Principles of Social Psychiatry

COMMON MENTAL DISORDERS 233 mental health service provision for those affected. and professionals of secondary (mental) health care. Thus, CMDs are still undervalued when compared to Special emphasis should be given to the comorbidity other mental illnesses, underrecognized and conse- of CMDs and physical disorders. This of course would quently undertreated. As such, CMDs are not yet lead to more health service use and more treatment. considered as a ‘core business’ of social psychiatry However, the benefits for the wider society would by although many patients with CMDs develop severely far outweigh the costs, especially when taking into disabling conditions and are, thus, on the caseloads of account the considerable economic loss due to sick- secondary mental health services. Even worse, gen- ness and disability caused by CMDs. eral practitioners are faced with ‘unexplained symptoms’ that they do not recognize as CMDs. All this leads to the fact that peoplewith CMDs are at great 17.8.2 Poverty and food insecurity risk of not being recognized to have an illness, not being appropriately treated and therefore developing a Poverty and food insecurity is a topic that one might chronic and disabling condition. This might be espe- think is only relevant to developing countries. How- cially true for groups with increased risk of CMDs, ever, research evidence shows that the contrary is true; such as women and socially and/or economically the relationship between CMDs and poverty particu- disadvantaged people, e.g. ethnic minorities, unem- larly holds true for developed countries as well. ployed people or those with low education. This is a Regardless of the direction of the relationship, pro- finding irrespective of the overall state of development grammes to tackle poverty and food insecurity should of the society these people live in. consider incorporating promotion of mental health. In public health terms, there are a number of im- For instance, child nutrition programmes in Asia plications of these findings. They might vary depend- should promote maternal mental health, taking into ing on the cultural and social background, but can be consideration the fact that maternal CMDs and poor summarized in four topics. child nutrition status are closely related [30]. 17.8.1 Education 17.8.3 Special needs groups The most encouraging fact is that low education is a potentially preventable risk factor. Promotion of gen- Some groups that are more at risk of developing eral education is a key factor in tackling different CMDs than others. Women and ethnic minorities social and health concerns. This holds true for all are two paradigmatic populations. Programmes to societies irrespective of their levels of development. reduce the burden of CMD in women should target, Education has different levels. School education is among others, poverty, women exposed to violence certainly one level. This would increase social oppor- and those with chronic physical illness. Moreover, tunities and access to education, but also the quality the clinical assessment of CMD in women must and the completion rate of mandatory education. include exploration of violence and other forms of Educating and training health professionals is an gender disadvantage. CMDs in ethnic minorities intervention that could also substantially impact on should be better understood, particularly underre- the identification and the treatment of people with cognition or misinterpretation of symptoms by health CMDs. This includes both primary care physicians professionals. 17.9 FUTURE RESEARCH The issues to put on the research agenda are deducible their relationship to CMDs must be better understood. from the above. First of all, vulnerable groups and There is no clear model available on how these factors

234 SOCIAL DETERMINANTS are related and what the importance of each of these results from the WHO Collaborative Study on Psycho- variables is. Second, we need longitudinal studies logical Problems in General Health Care. Journal of the to determine causal relationships between CMDs American Medical Association, 272, 1741–1748. and factors known to be related to these disorders. 6. Ustun, T., Sartorius, N., Costa e Silva, J. et al. (1995) These would open the field for clearly tailored inter- Conclusions, in Mental Illness in General Health ventions. Third, research should clarify how CMDs Care: An International Study (eds T. B. Ustun and could be destigmatized. There is a stigma attached to N. Sartorius), John Wiley and Sons, Ltd, Chichester, pp. 371–375. CMDs in that they are not regarded as ‘proper ill- nesses’, especially among the general population. 7. Kessler, R. C., Chiu, W. T., Demler, O. et al. (2005) Prevalence, severity, and comorbidity of 12-month This might have an influence on the poor identification DSM-IV disorders in the National Comorbidity Survey in primary care and further on the poor treatment rates Replication. Archives of General Psychiatry, 62, seen in these disorders. A better understanding of 617–627. CMDs would not only help those affected but also 8. Singleton, N., Bumpstead, R., O’Brien, M. et al. (2003) those around them, carers, health services, public Psychiatric morbidity among adults living in private health and the wider society, by improving physical households 2000. Office for National Statistics. Inter- and mental health, psychosocial well-being and social national Review of Psychiatry, 15, 65–73. inclusion for a much marginalized group. 9. Kessler, R. C., McGonagle, K. A., Zhao, S. et al. (1994) To sum up, CMDs are highly prevalent, but under- Lifetime and 12-month prevalence of DSM-III-R psy- and misidentified. They are insufficiently treated and chiatric disorders in the United States. Results from the often lead to sickness and disability. This is a finding National Comorbidity Survey. Archives of General that holds true independent of the cultural back- Psychiatry, 51, 8–19. ground. Although there are some known potentially 10. Melzer, D., Buxton, J. and Villamil, E. (2004) Decline modifiable stressors, among them education, poverty in CMD prevalence in men during the sixth decade of and food insecurity, more research is needed to under- life. Evidence from the National Psychiatric Morbidity stand causal relationships as a basis for developing Survey. Social Psychiatry and Psychiatric Epidemiol- ogy, 39, 33–38. interventions. 11. Melzer, D., Fryers, T. and Jenkins, E. (eds) (2004) Social Inequalities andthe Distribution of the Common Mental Disorders: A Report to the Department of REFERENCES Health Policy Research Programme, Psychology Press, London. 1. Goldberg, D. and Huxley, P. (1992) Common Mental 12. Patel, V., Todd, C., Winston, M. et al. (1998) Outcome Disorders: A Biosocial Model, Tavistock/Routledge, of common mental disorders in Harare, Zimbabwe. London, pp. 7–8. British Journal of Psychiatry, 172, 53–57. 2. Koopmans, G. T., Donker, M. C. and Rutten, F. H. 13. Cooper, C., Bebbington, P. E., Meltzer, H. et al. (2005) Common mental disorders and use of general (2008) Depression and common mental disorders health disorders: a review of the literature on popula- in lone parents: results of the 2000 National tion-based studies. Acta Psychiatrica Scandinavica, Psychiatric Morbidity Survey. Psychological Medicine, 111 (5), 341–350. 38, 335–342. 3. Lahelma, E., Laaksonen, M., Martikainen, P. et al. 14. Patel, V., Kirkwood, B. R., Pednekar, S. et al. (2006) (2006) Multiple measures of socioeconomic circum- Risk factors for common mental disorders in women. stances and common mental disorders. Social Science Population-based longitudinal study. British Journal of and Medicine, 63, 1383–1399. Psychiatry, 189, 547–555. 4. Murray, C. J. and Lopez, A. D. (1997) Global mortality, 15. Patel, V., Kirkwood, B. R., Pednekar, S. et al. (2006) disability, and the contribution of risk factors: Global Gender disadvantage and reproductive health risk fac- Burden of Disease Study. Lancet, 349, 1436–1442. tors for common mental disorders in women: a com- 5. Ormel, J., Von Korff, M., Ustun, T. B. et al. (1994) munity survey in India. Archives of General Psychiatry, Common mental disorders and disability across cultures: 63, 404–413.

COMMON MENTAL DISORDERS 235 16. Horwath, E., Johnson, J. and Hornig, C. D. (1993) 25. Lopes, C. S., Faerstein, E., Chor, D. and Werneck, G. L. Epidemiology of panic disorders in African-Ameri- (2008) Higher risk of common mental disorders after cans. American Journal of Psychiatry, 150, 465–469. experiencing physicalviolenceinRiodeJaneiro,Brazil: 17. Araya, R., Montgomery, A., Rojas, G. et al. (2007) the Pro ´-Sau ´de Study. International Journal of Social Common mental disorders and the built environment in Psychiatry, 54, 112–117. Santiago, Chile. British Journal of Psychiatry, 190, 26. Ludermir, A. B. and Lewis, G. (2005) Is there a gender 394–401. difference on the association between informal work 18. Haynes, J. C., Farrell, M., Singleton, N. et al. (2005) and common mental disorders? Social Psychiatry and Alcohol consumption as a risk factor for anxiety and Psychiatric Epidemiology, 40, 622–627. depression: results from the longitudinal follow-up of 27. Shah, A. (2003) Ethnicity and the Common Mental the National Psychiatric Morbidity Survey. British Disorder. Social Inequalities and the Distribution of Journal of Psychiatry, 187, 544–551. the Common Mental Disorders: A Report to the De- 19. Das-Munshi, J., Goldberg, D., Bebbington, P. E. et al. partment of Health Policy Research Programme, Psy- (2008) Public health significance of mixed anxiety and chology Press, London. depression: beyond current classification. British Jour- 28. Bhui, K., Bhugra, D., Goldberg, D. et al. (2001) Cul- nal of Psychiatry, 192, 171–177. tural influences on the prevalence of common mental 20. McCrone, P., Dhanasiri, S., Patel, A. et al. (2008) disorders, general practitioners’ assessments and help- Paying the Price: The Cost of Mental Health Care in seeking among Punjabi and English people visiting England to 2026, King’s Fund, London. their general practitioner. Psychological Medicine, 21. Ludermir, A. B. and Lewis, G. (2001) Links between 31, 815–825. social class and common mental disorders in Northeast 29. Stewart, R. C., Umar, E., Kauye, F. et al. (2008) Brazil. Social Psychiatry and Psychiatric Epidemiol- Maternal common mental disorder and infant growth – ogy, 36, 101–107. a cross-sectional study from Malawi. Maternal and Child Nutrition, 4, 209–219. 22. Patel, V. and Kleinman, A. (2003) Poverty and common mental disorders in developing countries. Bulletin of 30. Hadley, C., Tegegn, A., Tessema, F. et al. (2008) the World Health Organization, 81, 609–615. Food insecurity, stressful life events and symptoms of anxiety and depression in East Africa: evidence from 23. Pothen, M., Kuruvilla, A., Philip, K. et al. (2003) the Gilgel Gibe growth and development study. Jour- Common mental disorders among primary care atten- nal of Epidemiology and Community Health, 62, ders in Vellore, South India: nature, prevalence and risk 980–986. factors. International Journal of Social Psychiatry, 49, 119–125. 31. Harpham,T.,Huttly, S.,DeSilva,M. J.and Abramsky, T. (2005) Maternal mental health and child 24. Ludermir, A. B., Schraiber, L. B., D’Oliveira, A. F. nutritional status in four developing countries. Jour- et al. (2008) Violence against women by their intimate nal of Epidemiology and Community Health, 59, partner and common mental disorders. Social Science 1060–1064. and Medicine, 66, 1008–1018.



18 Suicide Gwendolyn Portzky and Kees van Heeringen Department of Psychiatry and Medical Psychology, University Hospital, Gent, Belgium 18.1 INTRODUCTION An abundance of research has led to a greater under- factors, including social variables. Many studies have standing of the complex aetiology of suicidal beha- addressed the relationship between suicidal behaviour viour. There is clear evidence that suicidal behaviour and social factors, showing that such variables influ- is caused by multiple factors such as an underlying ence suicide risk. In addition, suicide rates are com- trait vulnerability, which includes biological and psy- monly used by sociologists as indicators of social chological characteristics as well as state-dependent disintegration, distress and quality of life [1]. 18.2 CLASSICAL SOCIOLOGICAL DETERMINANTS OF SUICIDAL BEHAVIOUR: DIVORCE, RELIGIOUS AFFILIATION AND UNEMPLOYMENT The classical sociological approach to the study of as social beings who need to believe in a community, suicidal behaviour is almost synonymous with the be part of it and have the norms and rules of the theories of Durkheim [2]. Durkheim was the first to community for support. Low social integration (the state that differences in suicide rates between extent to which people are joined together by social countries indicate a considerable influence of social networks) and low social regulation (the influence of organization and society on individual behaviour. tradition, norms and habits on behaviour) thus have a According to Durkheim’s vision individual suicidal negative impact on suicide rates. Social integration, as behaviour is the result of individual factors and cir- defined by Durkheim, was associated with marriage/ cumstances, while the variations in the prevalence of divorce, (un)employment and religious affiliation, suicide are explained by the moral and psychological which is possibly one of the reasons why Durkheim’s climate in the society, which affects the reactions to work has often been a target of criticism; aside from problems and pain [3]. Durkheim considered humans taking his work too literally, critics have replicated his Principles of Social Psychiatry, second edition Edited by Craig Morgan and Dinesh Bhugra Ó 2010 John Wiley & Sons, Ltd.

238 SOCIAL DETERMINANTS analyses using modern data, but divorce or religious and colleagues [16], which examined the association affiliation are obviously not the same in modern, between religious affiliation and suicide attempts in Western society as they were in Catholic nineteenth patients with depression, showed greater moral century France [3]. objections to suicide and lower aggression levels in Notwithstanding the criticism, modern-day studies religiously affiliated subjects. These may serve as still often focus on divorce, religious affiliation and protective factors against suicide attempts, whereas unemploymentasindicatorsofsocialintegration.Such religious affiliation may not. studies have often shown ambiguous results. Using a Another example of the complicated role of religion time series analysis to examine the effect of divorce on insuicidalbehaviourcanbefoundinAsianandChinese suicide in Finland, Stack [4] found a positive correla- societies. These societies have important differences in tion between divorce and suicide rates. However, a suicide patterns compared to most Western countries, similar analysis in Japan did not confirm this relation- includingahigherfemalesuicideratethanmalesuicide ship between divorce and suicide, although a general rate andvery high rural suicide rates. These differential association with family integration was found [5]. A suicide patterns have been partly linked to the complex review by Stack [6] showed that the majority (86.4%) role of religion and its effect on the traditions, norms of 37 cross-sectional studies supported the hypothesis andhabits inthe society.Chinaisnota religioussociety that the suicide rate among divorced persons is higher andgoingtochurchisnotpopular,whichcouldresultin than the rate among married persons, with the excep- alackofsocialsupportandcontactwiththecommunity. tion of younger age groups in which both marriage and However, traditional Chinese culture is founded by the divorce are uncommon. The study of Caces and threeimportantAsianreligions:Buddhism,Taoismand Harford [7] found no effect of the divorce rate on the Confucianism. These religions differ strongly from frequency of suicide while Gunnell and colleagues [8] Western religions with regard to the afterlife, rituals, identified the increasing divorce rate as one of the organization and the existence of one God, and as a explanations for the increasing suicide rate in young result have different attitudes regarding suicide. For males in England and Wales. According to Stack [9], instance, Buddhism considers self-sacrifice for reli- the association between divorce and suicide may be gious reasons as honourable, and although Confucian- explained by the fact that divorced people have higher ism does not allow suicide because it deplores dama- levels of suicidogenic conditions compared to the ging physical attractiveness, it does tolerate suicide as general population. These suicidogenic conditions an expression of high moral or emotional protest [17]. include: an increased risk of depression, increased Buddhism and Taoism also imply the reincarnation of financial pressures, increased risk of alcohol abuse thesoul,whichresultsintheChinesebeliefthatthenext during the divorce process, and increased risk of feel- life is formed by the previous life and that suicide is a ings of emotional hurt, shame and guilt. possibility to receive a new, better life. In Japan suicide The relation between religion and suicide has been is generally considered as a honourable act of self- a matter of debate. The classic Durkheimian premise sacrifice, which originates from Confucianism, stating states that religious affiliation is associated with high that being dutiful, obedient and loyal to the group is social integration, which protects against suicide. more important than the individual [18]. However, studies investigating such a positive influ- It is thus clear that religious beliefs are associated ence on suicidal behaviour have often shown conflict- with norms, values and attitudes in Asian societies as ing results, with several studies showing a positive these societies can be described as collectivistic socie- influence of religion on suicidal behaviour [10–12] ties, while Western nations are traditionally consid- and other studies showing no effect [13–15]. More ered as individualistic cultures. Individualism has recently, it has become clear that the relation between been associated with psychological disadvantages and religion and suicidal behaviour is complex and that mental health problems due to the evidence that other factors than mere social integration, or the sheer individualistic values are related to smaller and less number of beliefs and practices, play a role in this satisfying social support networks, less adaptive emo- complex association. For example, the study of Dervic tional coping, lower help-seeking behaviour and

SUICIDE 239 higher levels of hopelessness and suicidal idea- tions (sometimes only for specific sociodemographic tion [19]. Members of individualistic societies are subgroups), 2 reported negative findings and 15 known to emphasize values such as self-direction, reported nonsignificant findings (the number adds up autonomy and universalism while members of col- to more than 29 because several studies reported lectivistic societies strongly focus on tradition and multiple and differing associations arising from the conformity. Although China has traditionally been disaggregation of the data). More nonsignificant described as collectivistic and Western societies as associations were found in females. individualistic, research has shown that Chinese peo- The relationship between unemployment and sui- ple are indeed strongly collectivistic within family cidal behaviour is thus complex as overall suicide relations and clans but are more individualistic on rates are often found not to fluctuate with the unem- a community and society level compared to ployment rate, as would be expected. The ‘cry of pain Americans [20]. It can be suggested that Chinese model’ of Williams [23] may, however, explain some collectivism is primarily related to strong familial of the controversies regarding suicidal behaviour and relations and marriage but not to the community, unemployment. Although this model also considers which also implies certain risk factors may have a psychological and biological aspects, there is a cru- greater impact as community networks and support cial combination with social aspects including social are often limited. Although family and marriage are comparison. Williams states that individuals are highly regarded in Chinese society, Chinese suicides sensitive to signals from their social environment of may well be related to adverse familial and relational threats to their rank within the group and to their triggers [21]. The Chinese social importance of family acceptance as part of the group. Such events are is also related to limited social contacts outside the perceived as a sign that they are a loser and are family. This would especially apply to young rural defeated in an important aspect of their lives. women who, as a result, are entirely dependent on According to Williams the effects of social compar- resources provided by the family. ison are strong and have an effect on psychopathol- The findings regarding the association between ogy and suicidal behaviour, but it is the comparative unemployment and suicidal behaviour are more effect that determines the amount of stress or depres- homogeneous. The majority of these studies indicate sion and not the absolute effect. It is thus the dis- that unemployment is associated with an increased tribution of inequality that signals a threat to a risk of suicidal behaviour [22]. More ambiguous person’s rank within the group. As a result, unem- results are, however, found in aggregate longitudinal ployment has less capacity to cause depression and studies in which the effect of unemployment on suicidal behaviour when individuals perceive every- suicide rates is examined. Platt and Hawton [22] found one else to be in the same situation [24], which could that, of the 29 studies looking at male suicide rates as explain why high unemployment rates are not always the outcome variable, 22 reported positive associa- associated with high suicide rates. 18.3 MODERN-DAY SOCIAL DETERMINANTS OF SUICIDAL BEHAVIOUR: SOCIAL PERFECTIONISM, SOCIAL SUPPORT AND (IM)MIGRATION Another psychological model for suicidal behaviour Recent studies have therefore examined the asso- that conceptualizes suicidal behaviour as driven by a ciation between social perfectionism and suicidal need for social approval is Baumeister’s [25] ‘escape behaviour. Perfectionism has been described as hav- from self model’. This model states that a person’s ing three dimensions: self-oriented perfectionism failure to attain (either self or) socially imposed unrea- (the standards individuals set for themselves), listic standards induces a chainlike process including other-oriented perfectionism (high expectations self-blame, negative self-awareness, negative affect and standards for other people’s behaviour) and and a desire to escape via suicidal behaviour. socially prescribed perfectionism (beliefs about the

240 SOCIAL DETERMINANTS excessive expectations individuals perceive signifi- support to approximate social integration [1]. Social cant others have of them) [26]. Socially prescribed support provided by family and friends may constitute perfectionism appears to be the most robust correlate important resources for coping and may therefore be of suicidal behaviour as it has often been associated associated with suicidal behaviour. The higher suicide with suicidality. Donaldson, Spirito and Farnett [27] rates in males compared to females (which is examined socially prescribed perfectionism in ado- universal, with the exception of China) have also lescent suicide attempters and found that this dimen- been linked to social support issues. M€ oller- sionofperfectionism was highly correlated withhope- Leimb€ uhler [30] associates the male vulnerability for lessness. A study by Dean and Range [28] also showed premature death by suicide with the fact that males that socially prescribed perfectionism was the only tend to be less socially integrated, have less social dimension of perfectionism that showed a significant support and are emotionally more isolated. association with suicidal behaviour. It is, however, It is thus not surprising that many studies regarding important to note that the vast majority of studies social support and the influence on suicidal behaviour examiningtheassociationbetweensociallyprescribed have supported Durkheim’s social integration theory perfectionism and suicidal behaviour indicate that this of suicide. A study examining the role of social correlation is indirect and thus mediated through other relationships as predictors of depression and suicidal variables. The study of Rasmussen, O’Connor and ideation in older adults indicated that lower levels of Brodie [29] showed that socially prescribed perfec- social support were associated with higher levels of tionism interacted with overgeneral autobiographical depression and suicidal ideation. Interestingly, a sense memory of both positive and negative memories to of belonging to the community was not identified as a predict suicidal ideation and depression. The above- predictor of depression and suicidal ideation [31]. mentioned relationship between socially prescribed The results of a study examining the changing perfectionism and hopelessness found by Donaldson, suicide rates in Norway identified the weakening of Spirito and Farnett et al. [27] in young suicide attemp- family integration as a very important factor asso- ters was also attenuated after the effects of depressive ciated with the increasing post-war suicide rates [1]. cognitions were controlled for. There is also evidence A similar study in Cyprus examining the low indicating that there is an interactive effect of socially suicide rates in this country concluded that strong prescribed perfectionism and positive future thinking family relationships result in high social support and in relation to suicidal behaviour. O’Connor and low social isolation, which could be associated with colleagues[26]showedinrepetitivesuicidalself-harm the low suicide rates [32]. The study of Range and patients that the interaction between socially pre- colleagues [33] also found that factors associated scribed perfectionism and positive future thinking with social support and social networks can be iden- predicted psychological well-being two months fol- tified as a possible explanation for the different lowing a suicidal episode. The best outcome in terms suicide figures among the four most important ethnic of hopelessness and suicidal thinking was found in groups in the US. African Americans have a lower patients who reported low social perfectionism and suicide rate, with especially African American high positive future thinking following a suicidal women showing a very low suicide rate. The authors episode. Importantly, there was no such positive suggest that religion, extended family network and change in hopelessness and suicidal thinking in the important role of the elderly are among the pos- patients who reported high social perfectionism. sible sociocultural variables, which could be respon- sible for the low suicide rate in this group. Hispanic Americans also show lower suicide rates compared to 18.3.1 Social support as modern-day non-Hispanic Americans. Possible sociocultural pro- equation of social integration tective factors in Hispanic American society are the strong family networks along with the important More recent studies regarding social integration and ‘extended’ family connectedness and strong Catholic its relation to suicidal behaviour have often used social religion associated with the strong emphasis on

SUICIDE 241 ‘fatalismo’, which implies a lack of control when Notwithstanding this latter finding, which points at facing adverse events. In contrast, the suicide rate in an independent effect of social integration, it is impor- Native Americans is 1.6–4.2 times higher than the tant to also indicate that not all studies find a sig- general suicide rate in the US. It is assumed that nificant association between social support and disruption of the tribal unity, the absence of social suicidal behaviour. A study in a population sample integration and alcohol as a primary coping response of 2219 Chinese people in Hong Kong examined the are possible sociocultural factors associated with this contribution of hopelessness, depression and social elevated suicide rate. Asian Americans are also found factors to suicidal ideation and suicide attempts [36]. to have a lower suicide rate, which may be explained Multivariate modelling showed that about 40% of by the three important Asian religions (Confucianism, suicidal ideation and suicide attempts was attributable Buddhism and Taoism), de-emphasizing the indivi- to depression and about 20% was attributable to hope- dual compared to the group and the different attitudes lessness. This impact of depression and hopelessness towards suicide among Asian Americans, which have was, however, not affected by social support. There been described above. These studies thus indicate that was also a significant contribution – albeit to a lesser cultural differences in mental health, well-being and extent – of marital dissolution, along with drug abuse. suicidal behaviour are related to differences in social The authors conclude that hopelessness and depres- support and integration. sion are important contributors to suicidality and that Social support issues have also been identified as social support seems to have little effect as buffer. As it one of the factors distinguishing suicide attempters has been mentioned before that Asian and especially from suicide ideators. A study of Fairweather and Chinese people have important differences in suicide colleagues [34] in an Australian community sample patterns compared to Western countries, these results found that suicide ideators and attempters have com- have to be interpreted with caution and cannot simply parable levels of depression and anxiety, but also be adopted to Western individuals. showed that those who attempt suicide are more often unemployed, have more physical ill-health and experience more negative interactions with friends, 18.3.2 Migration resulting in less social support. This study thus shows that the differentiation between ideation and attempts As a consequence of the still increasing migration in is independent of mental disorders but related to many countries, many individuals are faced with physical health problems and social support issues. quickly changing social networks and the challenge It is indeed important to question whether social to integrate in new societies. Migration can be support and social integration have an independent described as an important interference in a person’s association with suicidal behaviour or whether they social network and integration system. It is known to are merely related through other factors such as disrupt important ties between the individual and his mental disorders. Studies on the relationship or her social network, including relationships with between social support or social integration and family, friends, colleagues and neighbours. Immigra- suicidal behaviour have often controlled inade- tion is known to add additional challenges such as quately for the effects of mental disorders, but the adapting to different habits, norms and values, lan- study of Duberstein et al. [35] showed some impor- guages, dress style and diets. According to the classic tant results regarding this issue. Their psychological Durkheimian view, these adversities in the social autopsy study showed that poor social integration is integration system should have a negative effect on associated with increased suicide risk over and above suicide rates and should result in increased suicide the effects of mood disorder and occupational status. rates in migrants. There are indeed several studies The authors conclude that the association between a pointing at an increased risk of suicide in migrants. lack of social interaction and suicide is robust and The results of a study examining the correlation largely independent of the presence of mental between migrant inpatients in Germany and nonfatal disorders. suicidal acts showed that Mediterranean immigrant

242 SOCIAL DETERMINANTS inpatients had a higher prevalence of suicide attempts risk can only be found in certain demographic (15.3%) compared to German inpatients (8.9%) [37]. groups. Young males until the age of 35 have often A highly elevated proportion of suicide by migrants is been described as the risk group most likely to be found in Saudi Arabia, with migrants comprising 77% affected by social change [9, 41]. Some of the of the suicide sample. Asian and Indian migrants variation in the strength of the effect of migration represented the majority of this group [38]. The on suicide can also be explained by the ethnic back- cross-cultural breakdown of Swedish suicide rates ground. Trovato and Jarvis [42] indicate that immi- also points to an overrepresentation of migrant sui- grant groups with Catholic backgrounds have greater cide [39]. Migrants from several countries such as abilities to integrate in new societies as they provide Russia, Finland, Germany, Denmark and Norway strong community ties for their members. showed elevated suicide rates compared with their Migrant studies are also an informative method for respective country of origin. The suicide risk for examining the genetics of suicide. Voracek and immigrants appears to be 1.5 times higher compared Loibl [43] recently conducted the first meta-analysis to native Swedes. Dorling and Gunnell [40] examined of the associations of migrant and country-of-birth social integration in Britain between 1980 and 2000 by suicide rates and found a strong positive association use of three components: migration, unemployment across 33 studies, containing data for nearly 50 and being single. This study identified migration as a migrant nationalities in 7 host countries located in significant and stable predictor of local suicide rates 3 continents. This findingindicates that migrants have, over the two decades. in their host country, a similar risk of suicide com- It can be concluded from these studies that pared to their homeland, which hints at possible migrants tend to have an elevated risk for suicide. underlying genetic vulnerabilities for suicide. However, some authors indicate that the increased 18.4 SOCIAL VARIABLES IN YOUNG PEOPLE The section above regarding migration indicated that trols [44–47]. The review of Evans, Hawton and young males are often identified as a risk group, and Rodham [48] indicated, however, that the evidence are particularly likely to be affected by social change. regarding the association between parents’ cohabita- It thus seems relevant to examine the role of social tional status and suicidal behaviour in youngsters is determinants of suicidal behaviour in this age group. inconclusive and that the possible relationship is indir- Epidemiological and psychological autopsy studies ect. Bridge, Goldstein and Brent [49] further elaborate have consistently shown that many social or environ- the association and conclude from their review that the mental variables are associated with youth suicide. relationship between divorce and suicidal behaviour These variables can be divided into: familial charac- may be explained by the increased prevalence of teristics, (mainly interpersonal) life events, social parental psychopathology, which could account for support by family and peers, and exposure to suicidal the increased divorce rate. Indeed, several psycholo- behaviour by others. gical autopsy studies of youth suicide have identified parental and familial psychopathology, in particular depression and substance abuse, as a risk factor for 18.4.1 Familial characteristics youth suicide [44–47]. The study of Brent, Perper and Goldstein [50] even identified parental psychopathol- Several familial characteristics are associated with ogy as an independent risk factor for suicide after suicide in young people. Regarding family structure controlling for psychopathology in the adolescent, but and living situation, there is evidence indicating that this could not be replicated by Gould et al. [45]. suicide cases come from broken families due to par- With regard to familial relationships, there are ental divorce more often than community con- many reports of a higher prevalence of parent–child

SUICIDE 243 relationship difficulties or conflicts with parents in Stressful life events can often be considered as young suicide victims [44, 45, 47, 49], not only in precipitants of adolescent suicidal behaviour, trig- comparison with community controls but also gering the behaviour instead of causing the suicidal when compared to suicidal psychiatric inpatient act. Cooper, Appleby and Amos [55] examined the controls [49]. According to Evans, Hawton and prevalence, timing and type of life events precipi- Rodham [48], the relationship between family discord tating suicide in young people with and without and suicidal phenomena is direct. It can be assumed mental disorders and found that youth suicide is that several family-related characteristics are related associated with life events in the last 3 months to suicidal behaviour in young people, but that the before thesuicide andmainlyinthe last week.In nature of the relationship still is a matter of debate. particular, interpersonal and forensic (arrest, con- Marttunen, Aro and L€ onnquist [47] concluded that viction) life events were more often found in family-related events such as divorce or death can be suicide victims. Comparison between suicide vic- considered more as general risk factors for the devel- tims with and without psychiatric disorder showed opment of mental health problems than as specific that more life events within the week prior to the risk factors for suicide. suicide were reported in suicide victims without a psychiatric disorder. 18.4.2 Life events 18.4.3 Social support by family and peers As vulnerability and sensitivity to certain stressful life events, indicating signals of defeat, are asso- As young people often experience negative life ciated with an increased risk of suicidal behaviour, it events, social support appears to be crucial at this is not surprising that stressful and traumatic life young age as a buffer against stress. Social support events are crucial in the pathway leading to suicidal can be provided by family or peers, but there is behaviour. Several controlled studies have indeed evidence indicating that there is a difference in the shown that suicide victims experiencemorestressful effect on suicidal behaviour according to the source life events than community controls [44, 51]. Other of social support. In adolescents, low support from controlled studies have provided more information the family is found to be most detrimental as it regarding the type of life events and indicated that predicts suicidality at follow-up after 6 months and interpersonal problems are most often identified in even into adulthood [56, 57]. Cumsille and young people. The studies of Houston, Hawton and Epstein [58] found that low support from the family Shepperd [52], Gould et al. [45] and Martunnen, Aro is more strongly related to depression than low and L€ onnquist [47] indicated that relationship diffi- support from other relationships. A study in Hong culties are a common problem in the lives of young Kong adolescents confirmed this finding as it showed suicide victims. However, the comparison between that suicidality, which was strongly predicted by suicide victims and suicidal psychiatric inpatients in depression, substance use and death attitude, was the study of Brent, Perper and Goldstein [50] could lowered by support from family and friends, but the not find any differences regarding conflicts with effect of family support was much stronger than partners. Legal problems and/or school-related pro- support from peers [59]. Similar findings were found blems have also often been related to youth sui- in the study by Kidd and colleagues [60], which also cide [45, 52] although the Brent, Perper and Gold- identified positive parent relations as the most con- stein study [50] identified more school problems in sistent protective factor for adolescent suicide suicidal control patientsthaninsuicide victims. attempts. Interestingly, for young males with a his- Traumatic experiences, such as the loss of an impor- tory of suicide attempts and poor peer relations, an tant person and sexual or physical abuse have also interactive effect of family relations and school rela- been related to suicidal behaviour in young peo- tions was found as the protective effect of parental ple [44, 46, 53, 54]. support was augmented by positive school relations.

244 SOCIAL DETERMINANTS There is also evidence indicating that in community influence rather than a general indication of family samples peer support may compensate for low family problems and psychopathology, as a family history of support, but this is not the case in more impaired suicidal behaviour has been identified as an indepen- youngsters [58, 61]. The review of Evans, Hawton dent risk factor after controlling for adverse parent– and Rodham [48] showed that having unsupportive child relationships and parental psychopathology parents is directly associated with suicidality in [45, 65]. Evans, Hawton and Rodham [48] differ- adolescents. Their findings regarding peer support entiate between completed suicide in the family and suggest that negative peer relationships can be a risk attempted suicide by family members as only a factor for suicidality but that positive peer relation- family history of attempted suicide is directly asso- ships are not necessarily protective. ciated with suicidal behaviour in young people, The findings of the study by Kerr, Preuss and whereas completed suicide in the family is not King [62] suggest that young males and females have independently associated, which may be related to different associations between social support and its relative rarity. De Leo and Heller [66] found suicidality and that these gender-specific associations similar results when data from four large studies have to be considered. The results showed that female were examined. With regard to the adolescent group, adolescents’ perceptions of low family support are they found that exposure to nonfatal suicidal beha- related to higher levels of hopelessness, depressive viour in family and friends was predictive of suicid- symptoms and suicidal ideation. It was, however, ality, even after controlling for exposure to fatal interesting that for adolescent males higher levels of suicidal behaviour. The reverse relation, in which peer support were associated with greater levels of fatal suicidal behaviour predicts suicidality when hopelessness, depressive symptoms and suicidal exposure to nonfatal suicidal behaviour is controlled ideation, which was independent of family support. for, could not be found. The authors conclude that peer support is sometimes Although less commonly examined, similar associated with more dysfunction in adolescents and results are found with regard to suicidal behaviour that support from peers as perceived by adolescents by friends, which has also been associated with may sometimes actually include shared antisocial suicidal behaviour in adolescents [52, 67]. A case- behaviour and maladaptive emotional coping. The control comparison of young suicide victims with findings also showed that for both genders low family matched psychiatric control patients showed that support was associated with more externalizing beha- suicide victims had been exposed more frequently viour and substance abuse, while more support from to suicidal behaviour by friends – and through media, peers was also associated with more externalizing which will be discussed below [68]. However, there behaviour problems. have also been reports indicating that exposure to suicide in a friend is not associated with an increased incidence of suicidal behaviour among young peo- 18.4.4 Exposure to suicidal behaviour ple [69, 70]. Evans, Hawton and Rodham [48] con- by others clude that there may be a possible direct association with attempted suicide but not with completed The study of Houston, Hawton and Shepperd [52] suicide. identified exposure to suicidal behaviour by family There is evidence indicating that exposure to members and/or friends in 41% of the young suicide suicidal behaviour through the media (newspaper, victims. In particular, suicidal behaviour in family fictional and nonfictional television reports) may members has often been related to suicidal behaviour influence the occurrence and characteristics of suici- in young people [44, 45]. A psychological autopsy dal behaviour. The above-mentioned psychological study in adolescents indicated that suicidal behaviour autopsy study in adolescents showed that 37% of the by family members alone could be identified in suicide victims had been exposed to suicidal beha- almost half of the sample [63]. According to Gould, viour through the media, which was very high com- Shaffer and Greenberg [64] this may reflect a genetic pared with the 5.3% in the group of control

SUICIDE 245 patients [68]. Hawton and colleagues [71] examined adolescents two weeks before and after the the effect of the broadcasting of a serious paracetamol broadcasting of four fictional movies regarding sui- overdose in a popular fictional television drama and cide. A significant increase in the number of both found an increase of 17% in the number of self- suicide attempts and completed suicides was found poisoning episodes presented to the hospital in the after the broadcasting. Young people appear to be first week after the television show and an increase especially vulnerable to the influence of media cover- of 9% in the second week. Gould and Shaffer [72] age of suicidality [73]. monitored the incidence of suicidal behaviour in 18.5 CONCLUSIONS It is clear that social factors play a role in the complex REFERENCES aetiology of suicidal behaviour. Although divorce, religious affiliation and unemployment can be 1. Barstad, A. (2008) Explaining changing suicide rates in described as more classical sociological determi- Norway 1948–2004: the role of social integration. nants of suicidal behaviour, which were already Social Indicators Research, 87, 47–64. identified by Durkheim in the nineteenth century, 2. Durkheim, E. (1897/1951) Suicide, Free Press, New they are still often a target in modern-day studies. York. The results are not always homogeneous, especially 3. Bille-Brahe, U. (2000) Sociology and suicidal with regard to religious affiliation, but the findings do behaviour, in The International Handbook of Sui- suggest an interaction, which seems to be complex cide and Attempted Suicide (eds K. Hawton and K. van Heeringen), John Wiley & Sons, Ltd, and influenced by other related factors. More Chichester. recently, there has been a focus on other social 4. Stack, S. (1992) The effect of divorce on suicide in determinants such as social perfectionism, social Finland – a time series analyses. Journal of Marriage support and migration, which all seem to be asso- and the Family, 54, 636–642. ciated with suicidal behaviour. Social support issues 5. Stack, S. (1992) The effect of divorce on suicide in even seem to be related to gender and cultural Japan – a time series analyses, 1950–1980. Journal of differences in the prevalence of suicide. There is Marriage and the Family, 54, 327–334. further evidence indicating that young people are 6. Stack, S. (1995) Divorce and suicide: a review of 132 especially vulnerable to adverse social situations and studies, 1880–1995. Paper presented at the annual social change. In particular, variables such as adverse meeting of the Michigan Association of Suicidology, familial characteristics, interpersonal life events, Holland, Michigan. social support by family and peers, and exposure to 7. Caces, F. and Harford, T. (1998) Time series analysis of suicidal behaviour by others have been identified alcohol consumption and suicide mortality in the Uni- as being related to suicidal behaviour in young ted States, 1934–1987. Journal of Studies on Alcohol, people. 59, 455–461. Notwithstanding the abundance of studies point- 8. Gunnell, D., Middleton, N., Whitley, E. et al. (2003) ing at an important relation between social factors Why are suicide rates rising in young men but falling in and suicidal behaviour, it has to be emphasized that the elderly? A time series analysis of trends in England social factors can only explain a part of the com- and Wales 1950–1998. Social Science and Medicine, plex aetiology. Other factors, such as biological and 57, 595–611. psychological determinants, are involved in making 9. Stack, S. (2000) Suicide: a 15-year review of the socio- an individual vulnerable to handling adverse logical literature. Part II: modernization and social social situations and social change via suicidal integrationperspectives.Suicide andLife – Threatening Behaviour, 30, 163–176. behaviour.

246 SOCIAL DETERMINANTS 10. Clarke, C., Bannon, F. and Denihan, A. (2003) Suicide 25. Baumeister, R. (1990) Suicide as escape from self. and religiosity – Masaryk’s theory revisited. Social Psychological Review, 97, 90–113. PsychiatryandPsychiatricEpidemiology,38,502–506. 26. O’Connor, R., Whyte, M., Fraser, L. et al. (2007) 11. Hilton, S., Fellingham, G. and Lyon, J. (2002) Suicide Predicting short-term outcome in well-being following rates and religious commitment in young adult males in suicidal behaviour: the conjoint effects of social per- Utah. American Journal of Epidemiology, 155, 413–419. fectionism and positive future thinking. Behaviour 12. Neeleman, J., Halpern, D., Leon, D. and Lewis, G. Research and Therapy, 45, 1543–1555. (1997) Tolerance of suicide, religion and suicide rates: 27. Donaldson, D., Spirito, A. and Farnett, E. (2000) The an ecological and individual study in 19 Western coun- role of perfectionism and depressive cognitions in tries. Psychological Medicine, 27, 1165–1171. understanding the hopelessness experienced by ado- 13. Malone, K., Oquendo, M., Haas, G. et al. (2000) lescenct suicide attempters. Child Psychiatry and Protective factors against suicidal acts in major Human Development, 31, 99–111. depression: reason for living. American Journal of 28. Dean, P. and Range, L. (1996) The escape theory of Psychiatry, 175, 1084–1088. suicide and perfectionism in college students. Death 14. Kelleher, M., Chambers, D. and Corcoran, P. (1999) Studies, 20, 415–424. Suicide and religion in Ireland: an investigation of 29. Rasmussen, S., O’Connor, R. and Brodie, D. (2008) Thomas Masaryk’s theory of suicide. Archives of The role of perfectionism and autobiographical mem- Suicide Research, 5, 173–180. ory in a sample of parasuicide patients – an exploratory 15. Lester, D. (1996) Religion and suicide. Psychological study. Crisis, 29, 64–72. Reports, 78, 1090–1090. 30. M€ oller-Leimk€ uhler, A. (2003) The gender gap in sui- 16. Dervic, K., Oquendo, M., Grunebaum, M. et al. (2004) cide and premature death or: why are men so vulner- Religious affiliation and suicide attempt. American able? European Archives of Psychiatry and Clinical Journal of Psychiatry, 161, 2303–2308. Neuroscience, 253, 1–8. 17. Zhang, J., Conwell, Y., Zhou, L. et al. (2004) Culture, 31. Vanderhorst, R. and McLaren, S. (2005) Social relation- risk factors and suicide in rural China: a psychological ships as predictors of depression and suicidal ideation in autopsy case control study. Acta Psychiatrica older adults. Aging and Mental Health, 9, 517–525. Scandinavica, 110, 430–437. 32. Zacharis, C., Hadjivassilis, V., Madianos, M. et al. 18. Young, J. (2002) Morals, suicide and psychiatry: a view (2005) Suicide in Cyprus 1988–1999. European from Japan. Bioethics, 16, 412–424. Psychiatry, 20, 110–114. 19. Scott, G., Ciarrochi, J. and Deane, F. (2004) Disadvan- 33. Range, L., Leach, M., McIntyre, D. et al. (1999) Multi- tages of being an individualist in an individualistic cultural perspectives on suicide. Agression and Violent culture: idiocentrism, emotional competence, stress Behaviour, 4, 413–430. and mental health. Australian Psychologist, 39, 34. Fairweather, A., Anstey, K., Rodgers, B. and 143–153. Butterworth, P. (2006) Factors distinguishing suicide 20. Zhang, J. (2000) Collectivism or individualism. An attempters from suicide ideators in a community sam- analysis of Chinese interactive culture. American ple: social issues and physical health problems. Psy- Review of China Studies, 1, 57–65. chological Medicine, 36, 1235–1245. 21. Zhang, J. (1996) Suicide in Bejing, China, 1992–1993. 35. Duberstein,P.,Conwell,Y.,Connor,K.etal.(2004)Poor Suicide and Life-Threatening Behaviour, 26, 175–180. social integration and suicide: fact or artifact? A case- control study. Psychological Medicine, 34, 1331–1337. 22. Platt, S. and Hawton, K. (2000) Suicidal behaviour and the labour market, in The International Handbook of 36. Cheung, Y., Law, C., Chan, B. et al. (2006) Suicidal Suicide and Attempted Suicide (eds K. Hawton and K. ideation and suicidal attempts in a population-based van Heeringen), John Wiley & Sons, Ltd, Chichester. study of Chinese people. Risk attributable to hope- lessness, depression and social factors. Journal of 23. Williams, M. (1997) A Cry of Pain. Understanding Suicide and Self-Harm, Penguin, Harmondsworth. Affective Disorders, 90, 193–199. 37. Grube, M. (2004) Nonfatal suicidal acts in a group of 24. Platt, S. (1984) Unemployment and suicidal behaviour: psychiatric inpatients. Situation of Mediterranean a review of the literature. Social Science and Medicine, immigrants. Nervenarzt, 75, 681–687. 19, 93–115.

SUICIDE 247 38. Elfawal, M. (1999) Cultural influence on the incidence Scotland: a case-control study using the method of and choice of method of suicide in Saudi Arabia. psychological autopsy. Social Psychiatry and Psychia- American Journal of Forensic Medicine and tric Epidemiology, 34, 645–650. Pathology, 20, 163–168. 52. Houston, K., Hawton, K. and Shepperd, R. (2001) 39. Ferradanoli, M. (1997) A cross-cultural break-down of Suicide in young people aged 15–24: a psychological Swedish suicide. Acta Psychiatrica Scandinavica, 96, autopsy study. Journal of Affective Disorders, 63, 108–116. 159–170. 40. Dorling, D. and Gunnell, D. (2003) Suicide: the spatial 53. De Wilde, E. J., Kienhorst, I., Diekstra, R. and Willems, and social components of despair in Britain 1980–2000. W. (1992) The relationship between adolescent suicidal Transactions of the Institute of British Geographers, behavior and life events in childhood and adolescence. 28, 442–460. American Journal of Psychiatry, 149, 45–51. 41. Trovato, F. (1986) The relationships between marital 54. Shafii, M., Carrigan, S., Whittinghill, J. R. and Derrick, dissolution and suicide: the Canadian case. Journal of A. (1985) Psychological autopsy of completed suicide Marriage and the Family, 48, 341–348. in children and adolescents. American Journal of Psy- 42. Trovato, F. and Jarvis, F. (1986) Immigrant suicide chiatry, 142, 1061–1064. in Canada: 1971 and 1981. Social Forces, 65, 55. Cooper, J., Appleby, L. and Amos, T. (2002) Life events 433–457. preceding suicide by young people. Social Psychiatry 43. Voracek, M. and Loibl, L. (2008) Consistency of and Psychiatric Epidemiology, 37, 271–275. immigrant and country-of-birth suicide rates: a meta- 56. King, C., Segal, H., Kaminski, K. et al. (1995) A analysis. Acta Psychiatrica Scandinavica, 118, prospective study of suicidal behaviour following 259–271. hospitalisation. Suicide and Life-Threatening Beha- 44. Beautrais, A. (2000) Risk factors for suicide and viour, 35, 327. attempted suicide among young people. Australian and 57. Lewinsohn, P., Rohde, P., Seeley, J. and Baldwin, C. New Zealand Journal of Psychiatry, 34, 420–433. (2001) Gender differences in suicide attempts from 45. Gould, M., Fisher, P., Parides, M. et al. (1996) Psy- adolescence to young adulthood. Journal of the chosocial risk factors of child and adolescent com- American Academy of Child and Adolescent Psychia- pleted suicide. Archives of General Psychiatry, 53, try, 40, 427–434. 1155–1162. 58. Cumsille, P. and Epstein, N. (1994) Family cohesion, 46. Brent, D., Perper, J., Moritz, G. et al. (1994) Familial family adaptability, social support, and adolescent risk factors for adolescent suicide: a case-control study. depressive symptoms in outpatient clinic families. Acta Psychiatrica Scandinavia, 89, 52–58. Journal of Family Psychology, 8, 202–214. 47. Marttunen, M., Aro, H. and L€ onnqvist, J. (1993) Ado- 59. Cheng, S. and Chan, A. (2007) Multiple pathways from lescence and suicide: a review of psychological autopsy stress to suicidality and the protective effect of social studies. European Child and Adolescent Psychiatry, 2, support in Hong Kong adolescents. Suicide and 10–18. Life-Threatening Behaviour, 37, 187–196. 48. Evans, E., Hawton, K. and Rodham, K. (2004) Factors 60. Kidd, S., Henrich, C., Brookmeyer, K. et al. (2006) The associated with suicidal phenomena in adolescents: a social context of adolescent suicide attempts: interac- systematic review of population-based studies. Clinical tive effects of parent, peer, and school social relations. Psychology Review, 24, 957–979. Suicide and Life-Threatening Behaviour, 34, 386–395. 49. Bridge, J., Goldstein, T. and Brent, D. (2006) Adoles- 61. Gauze, C., Bukowski, W., Aquan-Assee, J. and Sippola, cent suicide and suicidal behaviour. Journal Child L. (1996) Interactions between family environment and Psychology and Psychiatry, 47, 372–394. friendship and associations with self-perceived well- being during adolescence. Child Development, 67, 50. Brent, D., Perper, J. and Goldstein, C. (1988) Risk factors for adolescent suicide. A comparison of ado- 2201–2216. lescent suicide victims with suicidal inpatients. 62. Kerr, D., Preuss, L. and King, C. (2006) Suicidal Archives of General Psychiatry, 45, 581–588. adolescents’ social support from family and peers: gender-specific assocations with psychopathology. 51. Cavanagh, J., Owens, D. and Johnstone, E. (1999) Life Journal of Abnormal Child Psychology, 34, 103–114. events in suicide and undetermined death in south-east

248 SOCIAL DETERMINANTS 63. Portzky, G., Audenaert, K. and van Heeringen, K. autopsy study. Journal of Adolescence, 32, 849–862. (2005) Suicide among adolescents. A psychological DOI: 10.1016/j.adolescence.2008.10.007. autopsy study of psychiatric, psychosocial and person- 69. Brent, D. A.,Moritz,G., Bridge,J. et al. (1996) ality-related risk factors. Social Psychiatry and Long-term impact of exposure to suicide: a three- Psychiatric Epidemiology, 40, 922–930. year controlled follow-up. Journal of the American 64. Gould, M., Shaffer, D. and Greenberg, R. (2003) The Journal of Child and Adolescent Psychiatry, 35, epidemiology of youth suicide, in Suicide in Children 646–653. and Adolescents (eds R. King and A. Apter), 70. Hazell, P. and Lewin, T. (1993) Friends of adolescent Cambridge University Press, Cambridge. suicide attempters and completers. Journal of the 65. Brent, D., Bridge, J., Johnson, B. A. and Connolly, J. American Journal of Child and Adolescent Psychiatry, (1996) Suicidal behaviour runs in families: a controlled 32, 76–81. family study of adolescent suicide victims. Archives of 71. Hawton, K., Simkin, S., Deeks, J. et al. (1999) Effects General Psychiatry, 53, 1145–1152. of a drug overdose in a television drama on presenta- 66. de Leo, D. and Heller, T. (2008) Social modelling in the tions to hospital for self poisoning: time series and transmission of suicidality. Crisis, 29, 11–19. questionnaire study. British Medical Journal, 318, 67. Rey Gex, C., Narring, F., Ferron, C. and Michaud, P. 972–992. (1998) Suicide attempts among adolescents: preva- 72. Gould, M. and Shaffer, D. (1986) The impact of suicide lence, associated factors and comorbidity. Acta in television movies. Evidence of imitation. New Psychiatrica Scandinavica, 98, 28–33. England Journal of Medicine, 315, 690–694. 68. Portzky, G., Audenaert, K. and van Heeringen, K. 73. Gould, M., Jamieson, P. and Romer, D. (2003) Media (2009) Psychosocial and psychiatric factors associated contagion and suicide among the young. American with adolescent suicide: a case-control psychological Behavioural Scientiest, 46, 1269–1284.

19 Personality disorder Priya Bajaj and Mike Crawford Department of Psychological Medicine, Imperial College London, London, UK People differ in the way they view themselves and pational and social functioning. The presence of per- others. These characteristics influence the way that sonality disorder is also an important risk factor for people behave, including the way they interact with mental health problems such as depression, substance others. The term ‘personality disorder’ (PD) is used to misuse and suicidal behaviour [2]. Many people describe patterns of relating to self and others that are referred to mental health services have significant inflexible, maladaptive and lead to significant perso- personality problems and coexisting personality dis- nal and social problems [1]. Personality affects occu- order may reduce the effectiveness of treatments [3]. 19.1 A CONTROVERSIAL DIAGNOSIS While there is widespread acceptance that personality to justify withdrawal of treatment or exclusion from varies between individuals and that personality has an services [7]. important influence on people’s health and social Negative experiences of services among people functioning, the diagnosis of ‘personality disorder’ is with personality disorder are mirrored by those of a contentious one. Concerns have been expressed that some health care professionals who report finding it the term is a way of labelling socially undesirable difficult to work with people with PD and are pessi- behaviour [4]. Others have highlighted how the diag- mistic about the chances of someone with PD achiev- nosiscan be misusedto try to explaina person’s failure ing better mental health [8]. In a survey of 240 to respond to treatment for other mental disorders [5]. psychiatrists in England, respondents felt that those Many people with a diagnosis of personality disorder with a diagnosis of PD were less deserving of services are dissatisfied with the services they are offered and than thosewith other mental health problems [9]. Over feel that being given this diagnosis has a detrimental recent years, greater awareness of the emotional dis- effect on the care that they receive [6]. Some service tress that many people with personality disorder users feel they were given the label as a way of trying experience together with evidence that the prognosis Principles of Social Psychiatry, second edition Edited by Craig Morgan and Dinesh Bhugra Ó 2010 John Wiley & Sons, Ltd.

250 SOCIAL DETERMINANTS of some disorders is better than previously thought, logical treatments for some forms of PD further has led to a re-evaluation of the use of this diagnosis. emphasize the importance of recognition and assess- Research demonstrating the effectiveness of psycho- ment of personality in clinical practice. 19.2 CLASSIFICATION OF PERSONALITY DISORDER Two main approaches have been used to classify Eachoftheabovecategoriesisoperationallydefined personality-related problems; dimensional and cate- by between seven and nine specific criteria, a set gorical.Theuseofadimensionalapproachissupported numberbeingrequiredfordiagnosisofthedisorder[1]. by population-based research showing that problems The ICD-10 system defines a ‘specific personality associated with personality can be placed on a con- disorder’ as a ‘severe disturbance in the charactero- tinuumof normal toabnormaland that adiscretegroup logical constitution and behavioural tendencies of the of people with ‘disorder’ does not exist. Dimensional individual, usually involving several areas of the models of personality are also supported by a large personality, and nearly always associated with con- bodyofevidencethatsuggeststhatalimitednumberof siderable personal and social disruption’ [12]. The personality ‘traits’, such as neuroticism, extraversion general diagnostic guidelines applying to all person- and openness, provide a reliable measure of most ality disorders are: aspects of a person’s personality [10]. However, dimensionalapproachestoclassifyingmentaldisorder a. Markedly disharmonious attitudes and behaviour are cumbersome and current classification systems are are apparent, usually involving several areas of based on a categorical approach. This approach stems functioning, e.g. affectivity, arousal, style of relat- from the work of the German psychiatrist Schneider ing to others. who described personality disorder as abnormal per- sonalitieswheretheabnormalitycausessufferingtoan b. The abnormal behaviour pattern is enduring, of individual or the community [11]. long standing and not limited to episodes of mental The current DSM-IV classification system is based illness. on the concept of personality traits, i.e. ‘enduring patterns of perceiving, relating to, and thinking about c. The abnormal behaviour pattern is pervasive and the environment and oneself’. However, traits consti- clearly maladaptive to a broad range of personal tute personality disorder only when they are and social situations. ‘inflexible and maladaptive and cause significant functional impairment or subjective distress’ [1]. d. The above manifestations always appear during Recent editions of the DSM classify mental disorders childhood or adolescence and continue into (clinical syndromes) on Axis I and personality dis- adulthood. orders on Axis II. Ten patterns or categories of per- sonality disorder are identified and grouped into three e. The disorder leads to considerable personal distress clusters: but this may become apparent late in its course. . Cluster A (odd/eccentric) includes paranoid, schi- f. The disorder is usually, but not invariably, asso- zoid and schizotypal ciated with significant problems in occupational and social performance. . Cluster B (dramatic/erratic) includes antisocial, borderline, histrionic and narcissistic The subtypes include paranoid, schizoid, dissocial, emotionally unstable, histrionic, anankastic, anxious . Cluster C (anxious/fearful) includes avoidant, (avoidant) and dependent personality disorder. dependent and obsessive-compulsive. In order to qualify as a diagnosis, clear evidence of

PERSONALITY DISORDER 251 at least three of the traits or behaviours are required. Clinical practice suggests that many people with The ‘emotionally unstable’ subtype of PD is further personality disorder do not fit neatly into one of these broken down into impulsive and ‘borderline’ PD subtypes. Many people meet criteria for several per- (BPD). The latter is characterized by emotional labi- sonality disorders [13]. This finding has led to increas- lity, fear of abandonment, chronic sense of emptiness ing use of the three broad clusters of personality and recurrent suicidal ideation and behaviour. The disorder listed above. It has also given rise to the idea word ‘borderline’ was used to refer to the border that those whose personality disturbance meets between psychosis and neurosis in recognition of the criteria for disorders in more than one cluster transient psychotic symptoms that some people with should be regarded as having ‘severe personality this disorder experience at times of crisis. disorder’ [14]. 19.3 PREVALENCE OF PERSONALITY DISORDER The prevalence of personality disorder in community ality disorder, especially when it is based on a single samples has not been investigated with the same interview [21]. vigour as the prevalence of other psychiatric condi- Among the DSM-IV clusters, cluster C person- tions [15, 16]. The small numbers of studies that have ality disorders are more prevalent than those in been completed have all come from North America, clusters A and B. While the overall prevalence of Europe and Australasia. These studies suggest that the personality disorder is similar in men and women, lifetime prevalence of personality disorder is between there are marked differences in the prevalence of 5 and 15% [17, 18]. The prevalence of personality subtypes within the cluster B category. Levels of disorder is generally higher in urban than in rural borderline personality disorder are higher in women, populations [19]. and men are nine times more likely to meet diag- Cross-sectional surveys have shown that the point nostic criteria for antisocial PD [18]. Similarities in prevalence of personality disorders increases from 4 to some characteristics of borderline and antisocial 10% in the community, to about 20% in primary care, personality disorder, such as a low threshold for rising to over 50% in psychiatric outpatient popula- frustration, have led to the suggestion that these two tions. Among psychiatric patients, the highest preva- forms of personality disorder are the same but are lence of personality disorders is found among those manifested differently in men (who may tend to suffering from substance use disorders, eating disor- externalize aggression) and women (for whom inter- ders and those presenting to services following delib- nalized aggression is manifested through acts of erate self-harming behaviour. In these populations, deliberate self-harm) [22]. cluster B personality disorders predominate. Studies Borderline personality disorder is estimated to indicate that the highest prevalence overall of person- affect 0.5–2% of the population. High levels of emo- ality disorders is found among offenders; in the largest tional distress and recurrent acts of deliberate self- survey of psychiatric morbidity in prisoners in Eng- harm mean that the prevalence of borderline person- land and Wales, the prevalence of personality disorder ality disorder is considerably higher among those in among sentenced male prisoners was over 60% [20]. contact with mental health services. Between 0.6 and Differences between prevalence rates in different 3% of the general population of Westernized countries studies may be explained by sampling procedures, meet diagnostic criteria for antisocial personality diagnostic instruments and the number of disorder disorder. The prevalence is higher in inner-city areas categories included, rather than true differences and among thosewith substance misuse problems, and between populations. All studies in this field are higher still among those in contact with criminal limited by the poor diagnostic reliability of person- justice services [23].

252 SOCIAL DETERMINANTS 19.4 COURSE OF PERSONALITY DISORDER Although the definition of personality disorder is an mental disorder, personality traits in the cluster B ‘enduring’ pattern of inner experience and behaviour, group became significantly less pronounced over which is inflexible, pervasive, stable and of long time. In contrast, those in the cluster A and C groups duration, research increasingly reveals that these con- generally became more pronounced [24]. ditions show major fluctuations. Epidemiological stu- These findings are supported by other studies, dies have shown that personality abnormality is at its which show that people with BPD and other cluster most marked in the late teens and early twenties, and B personality disorders generally improve over time. even short-term follow-up studies suggest improve- Zanarini and colleagues found that among 290 people ment in most areas of personality function within two with BPD who were recruited from inpatient mental years. Longer-term follow-up studies suggest that health units, over 70% no longer met criteria for this some, but not all, forms of personality disorder diagnosis when followed up six years later. This improve substantially over time. Seivewright, Tyrer change was reported to be due to a reduction in and Johnson reported that among 202 people recruited impulsivity, with affective symptoms changing far into a randomized trial of treatments for common less during this period [25]. 19.5 IMPACT OF PERSONALITY DISORDER Given that interpersonal problems are a central com- common disorders were anxious, anankastic and ponent of the definition of personality disorder, it is paranoid. self-evident that people with PD will have poor social functioning [26]. People with PD are less likely to be married and more likely to be single, separated or 19.5.1 Premature mortality divorced [27]. Those with cluster A PD appear to be the least likely to get married. People with PD report People with personality disorder have elevated mor- lower levels of contact with relatives, smaller numbers tality resulting from poorer physical health and higher of friends and smaller social networks [28]. levels of suicide and accidental death. A large com- Levels of Axis I mental disorder are higher among munity-based cohort study of 12 103 cases in Sweden peoplewho have a personality disorder than thosewho [33], which examined mortality among those with a do not, as are levels of deliberate self-harm. Border- range of mental disorders, showed that Standardized line personality disorder has been reported as the most Mortality Ratios for those with ‘external causes’ were common personality disorder in several studies of highest among those with a diagnosis of personality deliberate self-harm (DSH) patients [29–31]. In a disorder (21 for men and 29 for women). Standardized study by Haw and colleagues [32] patients aged 15 Mortality Ratios for suicide among people with PD years and over who presented to the district general were even higher (38 among men and 42 among hospital in Oxford, UK, following an episode of DSH women). between 10 February and 1 December 1997 were Psychological autopsy studies among people who screened. The definition of DSH that was used die by suicide suggest that as many as 40% have a included acts of self-poisoning and self-injury, but personality disorder, though this is generally in com- excluded acts of self-cutting that was part of a repe- bination with other mental disorders, especially titive pattern of self-mutilation. Personality disorders depression and alcohol misuse [34]. Levels of mor- were present in 51 cases (45.9%): 19 patients (17.1%) tality are especially high among people with cluster B met criteria for one personality disorder, 15 (13.5%) personality disorder. Paris and Zweig-Frank reported for two and 17 (15.3%) for three or more. The most that, among a group of 64 people aged under 35 who

PERSONALITY DISORDER 253 were admitted to inpatient units with evidence of on their victims, receiving injuries themselves and borderline PD, 18% died over the following 27 years, beinginvolvedinmultipleincidents,therebyincreasing with over half of all deaths resulting from suicide [35]. the burden of care upon health care services. Indivi- duals with antisocial personality disorder demonstrated strong associations with injuring victims, and their 19.5.2 Harm to others violence was repetitive. They victimized partners and family members as well as strangers, and were most The impact of personality disorder on the level of likely to be violent towards the police. They reported violence in society remains unclear. Coid and collea- violenceinall locations studied,and violencewas more gues [23] have explored this topic using data from the likely when intoxicated [23]. Psychiatric Morbidity Survey of UK adults in 2000. Comorbid PD also increases risk of violence among As part of this survey, over 8000 householders were those with other mental disorders. In a secondary asked questions about their violent behaviour over the analysis of data collected as part of a randomized preceding 5 years. Over a third of all incidents were controlled trial of intensive management compared reported by respondents with no disorder. Mental with standard case management for people with psy- disorder appeared to have an impact on reported chosis, Moran and colleagues [36] screened partici- violent behaviour, although this was overwhelmingly pants for comorbid personality disorder. Physical associated with substance use/dependence and per- assault was measured from multiple data sources over sonality disorder. This cross-sectional survey there- the next 2 years, and logistic regression was used to fore provides some evidence to show that personality assess whether the presence of comorbid personality disordered individuals, and particularly those with disorder independently predicted violence in the sam- antisocial personality disorder (APD), significantly ple; 186 patients (28%) were rated as having a comor- contribute to the overall level of violence in the UK bid personality disorder, and they were significantly population. Despite a relatively low prevalence, indi- more likely to behave violently over the two-year viduals with antisocial personality disorder made period of the trial, even after adjusting for the effects substantial contributions to self-reported violence in of substance misuse, previous violence and severity of the household population of Britain. psychotic symptoms. The authors concluded that Persons with antisocial personality disorder and comorbid personality disorder is independently asso- substance dependence were more likely to report ciated with an increased risk of violent behaviour in involvement inviolent incidents and inflicting injuries psychosis [36]. 19.6 RESEARCH INTO AETIOLOGY OF PD As with other mental disorders, research conducted to have relatives with schizophrenia or schizophrenia date suggests that personality disorder arises from an spectrum disorders. Third, individuals with avoidant, interplay of hereditary and environmental factors. dependent and compulsive personality disorders tend Studies concentrating upon the heritability of person- to have relatives with anxiety disorders [5]. ality disorder have pointed to three causal mechan- Torgersen and colleagues [38] published the first isms in the development of personality disorder [37]. twin study to investigate the complete range of DSM First, it has been found that antisocial personality personality disorders. The main part of their sample disorder, borderline personality and substance abuse was ascertained by matching the Norwegian Twin frequently occur together in family studies, resulting Register with the National Register for Mental Dis- in the hypothesis that they form a group of impulsive order in Norway. Heritability for each of the three spectrum disorders associated with a common tem- clusters was as follows: cluster A: 0.37, cluster B: perament. Second, individuals with schizoid, para- 0.62, cluster C: 0.62. As the sample for this study was noid and schizotypal personality disorders tend to drawn mainly from clinical settings, the authors

254 SOCIAL DETERMINANTS acknowledge that these estimates may not be general- violence begets violence, this study also highlighted izable to the wider population. Other studies have that the majority of abused and neglected children do suggested a greater contribution of early environmen- not become delinquent, criminal or violent. tal factors, especially for cluster B personality dis- Berenbaum et al. [43] examined the relationship order. Indeed, most studies examining links between between schizotypal symptoms and psychological early environments and personality disorder have trauma (which includes childhood maltreatment and been conducted among people with cluster B PD, the experience of an injury or life-threatening event) especially those with borderline personality disorder. and found a gender difference. Schizotypal symptoms These are explored in greater detail below. were found to be more strongly associated with child- hood maltreatment among men whereas schizotypal symptoms were more strongly associated with post- 19.6.1 Childhood trauma traumatic stress disorder (PTSD) criterion A among women. Several studies have examined the association between early childhood experiences and develop- ment of personality disorder. It has been found that many, but not all, individuals with borderline PD 19.6.2 Parenting and attachment (BPD) report a history of abuse, neglect or separation as young children. Zanarini et al. [39] found that 40 to Retrospective studies in which people with BPD are 71% of borderline PD patients report having been asked to describe their early childhood experiences sexually abused by a noncaregiver. Using a develop- show that relationships with caregivers are often mental interview, among several psychological risk unstable or disordered [45, 46]. Parental neglect and factors, i.e. childhood sexual abuse, childhood phy- overprotection have also been examined in BPD. In a sical abuse, early separation or loss, and abnormal cohort of borderline patients several aetiological parental bonding, women with BPD have reported hypotheses were tested by Soloff and Millward [47]. more severe and a greater frequency of childhood They found that relationship with parents was seen as sexual abuse, as well as more physical abuse [39]. Of conflicted and negative and mothers were seen as these factors, childhood sexual abuse emerged as overinvolved while at the same time less caring and the sole factor that discriminated between BPD and helpful. Reports of early parenting in a study by Parker non-BPD patients [40]. Despite this finding, a meta- et al. found that parents were rated as uncaring, analysis of published literature from 1980 to 1995 did overcontrolling and abusive [48]. Most consistent not support the hypothesis that childhood sexual abuse links between perceived dysfunctional parenting were is a major psychological risk factor or a causal ante- found with the cluster C (anxious) and cluster B cedent of BPD [41]. (dramatic) styles but were insignificant for the cluster Apart from BPD, the risk of development of anti- A (eccentric) style. Meeting criteria for an increasing social personality disorder following early childhood number of personality disorder clusters was asso- abuse or neglect has also been researched. A prospec- ciated with increasing levels of adverse parenting. tive study followed 699 subjects into early adulthood Overall in this study, disordered functioning was most and found that childhood victimization was a signifi- distinctly associated with paternal indifference and cant predictor of the number of lifetime symptoms of maternal overcontrol. antisocial personality disorder and of a diagnosis of However, most BPD patients do not have mothers antisocial personality disorder [42]. Long-term con- with BPD, and the estimated 10% that do may repre- sequences of childhood victimization were explored sent a particularly high-risk subgroup in which both in a cohort study by Widom [44]. Being abused or constitutional and environmental influences com- neglected as a child was found to increase risk for pound each other [49]. Interactional patterns of delinquency, adult criminal behaviour and violent mothers with BPD and their infants as well as parent- criminal behaviour. Despite the suggestion that ing perceptions of mothers with BPD were studied

PERSONALITY DISORDER 255 further by Newman et al. [50]. Mothers with BPD Furthermore, mothers with BPD reported being less were found to be less sensitive and demonstrated less satisfied, less competent and more distressed. These structuring in their interaction with their infants, and findings indicate that there is a role for early inter- their infants were found to be less attentive, less vention to promote maternal sensitivity and maternal interested and less eager to interact with their mother. perceptions of competence. 19.7 EVIDENCE FROM INTERVENTION STUDIES Retrospective studies examining the salience of social that it may be possible to reduce the likelihood of factors in the aetiology of personality disorder are interpersonal problems through improving the social susceptible to a variety of different forms of bias. As environments of children. For instance, Raine and with other health-related problems, adults with per- colleagues [51] have evaluated the impact of a two- sonality disorder may be more likely to recall adverse year programme of educational and nutritional inter- events in their background than those with no person- ventions and structured physical activity for three- ality disorder. year-old children on the island of Mauritius. The team Recentlyconductedexperimentalstudiesthataimto were able to follow up 75 (90%) of 83 children who study the impact of psychosocial interventions deliv- completed the programme and 288 (81%) of matched ered inchildhoodonsubsequent mental health provide controls and demonstrated lower levels of personality- a means of prospectively examining the impact of related problems and criminal behaviour twenty years improving a person’s social circumstances. Few stu- later. Thegeneralizability of such findings to resource- dies havebeenconductedsufficientlylongagoinorder rich settings have not been explored, but they support to study their impact on personality in adulthood. the notion that improving early childhood environments However, initial findings from some studies suggest can impact on adult personality status. 19.8 RELEVANCE OF AETIOLOGY FOR THE DEVELOPMENT OF TREATMENT APPROACHES Interventionsforpersonalitydisorderareconsideredin change [52]. Current interventions for personality detail in Chapter 34. While there are many gaps in our disorder do not extend beyond the first category, understanding of the aetiology of personality disorder usually involving tertiary referral services. Conse- our existing knowledge base provides an indication of quently, at present tertiary services that provide access areaswhereeffortsbothtopreventandtreatpersonality to peer support, help people with both social as well as disorder may be most effectively targeted. mental health problems, and help address social skills The high incidences of personality disorder in those deficits appear to offer the greatest potential to help who have been in local authority or institutional care, people with PD [53]. particularly in the cluster B group, and their subse- quent criminal convictions, suggest that preventive and treatment strategies in this population could have REFERENCES a major influence on public health [18]. Public health also has the advantage of linking the biological basis 1. American Psychiatric Association (1994) Diagnostic of health and disease with social and political pro- and Statistical Manual of Mental Disorders, 4th cesses of society. Rose divided preventive interven- edn, American Psychiatric Association, Washington, tions into ‘high-risk strategies’, which are targeted at DC. individuals identified as at high risk, and ‘population 2. Krueger, R. F., Caspi, A., Moffit, T. E. et al. (1996) strategies’, aimed at bringing about overall population Personality traits are differentially linked to mental

256 SOCIAL DETERMINANTS disorders: a mutitrait-multidiagnosis study of an ado- and Treatment (ed. W. Livesley), Guilford Press, lescent birth cohort. Journal of Abnormal Psychology, New York, pp. 107–123. 105, 299–312. 17. Casey, P. R. and Tyrer, P. J. (1986) Personality, func- 3. Dolan-Sewell, R. T., Krueger, R. F. and Shea, M. T. tioning and symptomatology. Journal of Psychiatric (2001) Co-occurrence with syndrome disorders, in Research, 20, 363–374. Handbook of Personality Disorders: Theory, Research, 18. Coid, J., Yang, M., Tyrer, P. et al. (2006) Prevalence and and Treatment (ed. W. J.Livesley),Guilford, New York, correlates of personality disorder in Great Britain. pp. 84–104. British Journal of Psychiatry, 188, 423–431. 4. Tyrer, P., Casey, P. and Ferguson, B. (1991) Personality 19. Keown, P., Holloway, F. and Kuipers, E. (2002) The disorder in perspective. British Journal of Psychiatry, prevalence of personalitydisorders, psychoticdisorders 159, 463–471. and affective disorders amongst the patients seen by a 5. Alwin, N., Blackburn, R., Davidson, K. et al. (2006) community mental health team in London. Social Understanding Personality Disorder: A Report by the Psychiatry and Psychiatric Epidemiology, 37, British Psychological Society, The British Psychologi- 225–229. cal Society, Leicester. 20. Moran, P. (1999) The epidemiology of antisocial per- 6. Ramon, S., Castillo, H. and Morant, N. (2001) Experi- sonality disorder. Social Psychiatry and Psychiatric encing personality disorder: a participative research. Epidemiology, 34, 231–242. International Journal of Social Psychiatry, 47, 1–15. 21. Tyrer, P. (2005) The stability of personality disorders. 7. Haigh, R. (2002) Services for People with Personality Psychiatry, 4 (3), 29–31. Disorder: The Thoughts of Service Users, National Insti- 22. Lowenstein, L. (2007) The Link between Borderline tute for Mental Health in England (NIMHE), London. Personality Disorder and Antisocial Personality Dis- 8. Markham, D. (2005) Attitudes towards patients with a order, Southern England Psychological Services, diagnosis of borderline personality disorder: social Eastleigh. rejection and dangerousness. Journal of Mental Health, 23. Coid, J., Yang, M., Roberst, A. et al. (2006) Violence 12, 595–612. and psychiatric morbidity in the national household 9. Lewis, G. and Appleby, L. (1988) Personality disorder: population of Britain: public health implications. Brit- the patients psychiatrists dislike. British Journal of ish Journal of Psychiatry, 189, 12–19. Psychiatry, 153, 44–49. 24. Seivewright, H., Tyrer, P. and Johnson, T. (2002) 10. CostaJr, P.T. and McCrae, R. R. (1985) The NEO Change in personality status in neurotic disorders. Personality Inventory Manual, Psychological Assess- Lancet, 359 (9325), 2253–2254. ment Resources, Odessa, Florida. 25. Zanarini, M. C., Frankenburg, F. R., Hennen, J. and 11. Schneider, K. (1950) Psychopathic Personalities, 9th Silk, K. R. (2003) The longitudinal course of borderline edn, Cassell, London (English translation, 1958). psychopathology: 6-year prospective follow-up of the 12. World Health Organization (1992) The ICD-10 Classi- phenomenology of borderline personality disorder. fication of Mental and Behavioral Disorders, WHO, American Journal of Psychiatry, 160, 274–283. Geneva. 26. Skodol, A. W., Johnson, J. G., Cohen, P. et al. (2007) 13. Fyer, M. R., Frances, A. J., Sullivan, T. et al. (1988) Co- Personality disorder and impaired functioning from morbidity of borderline personality disorder. Archives adolescence to adulthood. British Journal of Psychia- of General Psychiatry, 45, 348–352. try, 190, 415–420. 14. Tyrer, P. and Johnson, T. (1996) Establishing the sever- 27. Samuels, J., Eaton, W. W., Bienvenu, O. J. et al. (2002) ity of personality disorder. American Journal of Psy- Prevalence and correlates of personality disorders in a chiatry, 153, 1593–1597. community sample. British Journal of Psychiatry, 180, 536–542. 15. Casey, P. (2000) The epidemiology of personality dis- orders, in Personality Disorders: Diagnosis, Manage- 28. Tyrer, P., Merson, S., Onyett, S. et al. (1994) The effect ment and Course, 2nd edn (ed. P. Tyrer), Arnold, of personality disorder on clinical outcome, social London, pp. 71–79. networks and adjustment: a controlled clinical trial of psychiatric emergencies. Psychological Medicine, 24, 16. Mattia, J. and Zimmerman, M. (2001) Epidemiology, in 731–740. Handbook of Personality Disorders: Theory, Research,

PERSONALITY DISORDER 257 29. Ennis, J., Barnes, R. A., Kennedy, S. et al. (1989) meta-analytic study. Journal of Personality Disorder, Depression in self-harm patients. British Journal of 13, 268–280. Psychiatry, 154, 41–47. 42. Luntz, B. K. and Widom, C. S. (1994) Antisocial 30. Suominen, K., Henriksson, M., Suokas, J. et al. (1996) personality disorder in abused and neglected children Mental disorders and comorbidity in attempted suicide. grown up. American Journal of Psychiatry, 151, Acta Psychiatrica Scandinavica, 94, 234–240. 670–674. 31. Gupta, B. and Trzepacz, P. (1997) Serious overdosers 43. Berenbaum, H., Thompson, R. J., Milanek, M. E. et al. admitted to a general hospital: comparison with nono- (2008) Psychological trauma and schizotypal person- verdose self-injuries and medically ill patients with sui- ality disorder. Journal of Abnormal Psychology, cidalideation.GeneralHospitalPsychiatry,19,209–215. 117 (3), 502–519. 32. Haw, C., Hawton, K., Houston, K. et al. (2001) Psy- 44. Widom, C. S. (1989) The cycle of violence. Science, chiatricandpersonalitydisordersindeliberateself-harm 244, 160–166. patients. British Journal of Psychiatry, 178, 48–54. 45. Zanarini, M.C. (2000) Childhood experiences asso- 33. Tidemalm,D.,Waern,M.,Stefansson,C.-G.etal.(2008) ciated with the development of borderline personality Excessmortalityinpersonswithseverementaldisorderin disorder. Psychiatric Clinics of North America, 23 (1), Sweden: a cohort study of 12 103 individuals with and 89–101. without contact with psychiatric services. Clinical Prac- 46. Akiskal, H. S., Chen, E. S., Davis, G. C. et al. (1985) tice and Epidemiology in Mental Health, 4,23. Borderline: an adjective in search of a noun. Journal of 34. Foster, T., Gillespie, K., McClelland, R. and Patterson, Clinical Psychiatry, 46, 41–48. C. (1999) Risk factors for suicide independent of DSM- 47. Soloff, P. H. and Millward, J. W. (1983) Developmental III-R axis I disorder. Case-control psychological histories of borderline patients. Comprehensive Psy- autopsy study in Northern Ireland. British Journal of chiatry, 24, 574–588. Psychiatry, 175, 175–179. 48. Parker, G., Roy, K., Wilhelm, K. et al. (1999) An 35. Paris, J. and Zweig-Frank, H. (2001) A 27-year follow- exploration of links between early parenting experi- up of patients with borderline personality disorder. ences and personality disorder type and disordered Comprehensive Psychiatry, 42 (6), 482–487. personality functioning. Journal of Personality Disor- 36. Moran, P., Walsh, E., Tyrer, P. et al. (2003) The impact ders, 13 (4), 361–374. of co-morbid personality disorder on violence in psy- 49. Links, P. S., Steiner, M. and Huxley, G. (1988) The chosis – data from the UK700 trial. British Journal of occurrence of borderline personality disorder in the Psychiatry, 182, 129–134. families of borderline patients. Journal of Personality 37. Coid, J. W. (1999) Aetiological risk factors for person- Disorders, 2, 14–20. ality disorders. British Journal of Psychiatry, 174, 50. Newman, L. K., Stevenson, C.S., Bergman, L.R. and 530–538. Boyce, P. (2007) Borderline personality disorder, 38. Torgersen, S., Lygren, S., Oien, P. A. et al. (2000) A mother–infant interaction and parenting perceptions: twin study of personality disorders. Comprehensive preliminary findings. Australian and New Zealand Psychiatry, 41 (6), 416–425. Journal of Psychiatry, 41 (7), 598–605. 39. Zanarini, M. C., Gunderson, J. G., Marino, M. F. et al. 51. Raine, A., Mellingen, K., Liu, J. et al. (2003) Effects of (1989) Childhood experiences of borderline patients. environmental enrichment at ages 3–5 years on schi- Comprehensive Psychiatry, 30, 18–25. zotypal personality and antisocial behavior at ages 17 40. Paris, J., Zweig-Frank, H. and Guzder, J. (1994) Psy- and 23 years. American Journal of Psychiatry, 160, chological risk factors for borderline personality dis- 1627–1635. order in female patients. Comprehensive Psychiatry, 52. Rose, G. (1992) The Strategy of Preventive Medicine, 35 (4), 301–305. Oxford University Press, Oxford. 41. Fossati, A., Madeddu, F. and Maffei, C. (1999) Border- 53. Crawford, M. J. (2008) Services for people with per- line personality disorder and childhood sexual abuse: a sonality disorder. Psychiatry, 7, 124–128.



20 Drug use, drug problems and drug addiction: social influences and social responses John Strang, Michael Gossop and John Witton National Addiction Centre, Institute of Psychiatry, King's College London, London, UK 20.1 INTRODUCTION Substance abuse contributes to many of the social social environmental factors on the initiation of drug problems within society today. Over recent decades use, on the development of drug problems, on the there have been many changes in the prevalence of course of such problems and on their short-term and substance use disorders and in how these problems are long-term outcomes – both with and without treat- perceived as well as changes in the nature and delivery ment. In this chapter an examination is undertaken of of responses. These include an increased awareness of the social influences and social responses that may the interplay between genetic, developmental and influence drug use, drug problems and drug addiction. 20.2 DRUG USE, DRUG PROBLEMS AND DRUG ADDICTION TERMINOLOGY Drug addiction, sometimes also called dependence, is (iii) A relatively stereotyped drug-taking habit (i.e. marked by a clustering of cognitive, behavioural and a narrowing of the repertoire of the drug- physiological phenomena: taking behaviour) (i) Subjective awareness of compulsion to use a (iv) Evidence of neuroadaptation (tolerance and drug or drugs, usually during attempts to stop or withdrawal symptoms) moderate drug use (v) Use of the drug to relieve or avoid withdrawal (ii) Adesiretostopdruguseinthefaceofcontinueduse symptoms Principles of Social Psychiatry, second edition Edited by Craig Morgan and Dinesh Bhugra Ó 2010 John Wiley & Sons, Ltd.

260 SOCIAL DETERMINANTS (vi) The salience of drug-seeking behaviour rela- (i) Unsanctioned use: use of a drug that is not tive to other important priorities approved by a society or a group within that society. When the term is used, it should be (vii) Rapid reinstatement of the syndrome after a made clear who is responsible for the disap- period of abstinence. proval. The term implies that we accept disapproval as a fact in its own right without Attempts to separate physical dependence and having to determine or justify the basis of that psychological/psychic dependence are flawed be- disapproval. cause of the interrelationship between the two. The term ‘neuroadaptation’ covers the changes associated (ii) Hazardous use: use of a drug that will probably with physical and psychological withdrawal phenom- lead to harmful consequences for the user – ena and also with the development of tolerance. Thus either to dysfunction or to harm. This concept is neuroadaptation covers the cellular, metabolic and similar to the idea of risky behaviour. For behavioural adaptations that occur in response to instance, smoking 20 cigarettes each day may drug use, as drug users may move from occasional not be accompanied by any present or actual controlled use to the loss of control over drug harm but we know it to be hazardous. seeking and drug taking. The term may also be extended to the concept of reciprocal neuroadaptation (iii) Dysfunctional use: use of a drug that is leading to discover the phenomenon previously described as to impaired psychological or social function- cross-tolerance. ing (e.g. loss of job or marital problems). The World Health Organization [1] recommends the use of further terms to characterize the continuum (iv) Harmful use: use of a drug that is known to of drug-taking behaviour and effects: have caused tissue damage or mental illness in the particular person. 20.3 EPIDEMIOLOGY The majority of the commonly abused drugs (with the However, there are limitations to self-report house- striking exception of alcohol) are subject to legal hold surveys which, apart from the likelihood of restrictions on possession and use. This raises a num- deliberate withholding of information, may also miss ber of difficulties in attempting to reach valid esti- those who are socially disadvantaged. For example, mates of prevalence. Various measures do exist the 2007/8 BCS estimates that just 34 000 people used through self-report surveys or by using official noti- heroin in the last year – a number incompatible with fication figures. These include various morbidity and figures for those seeking treatment for their heroin use. mortality data such as drug-related deaths, hepatitis B The National Treatment Drug Monitoring System statistics, nonfatal drug emergencies and law enforce- (NDTMS) figures for England show that over ment data. 120 000 people in drug treatment reported heroin as In the UK the British Crime Survey (BCS) has their main misused drug [3]. The Office of Population charted the prevalence of drug use and trends among Censuses and Surveys (OPCS) with its National the general population in England and Wales since Psychiatric Morbidity Survey included the homeless 1996 [2]. Use of any illicit drug during the last year and institutional residents with psychiatric disorders declined from 12.1% of 16 to 59-year-olds in 1998 to as well as a general population sample in its survey. 9.3% in 2007–2008, a decline largely due to a fall in While 5% of the household sample were classified as the use of cannabis. While cannabis use accounts for alcohol dependent, over 21% of the homeless sample most of this estimate, nearly three-quarters of a mil- were recorded as alcohol dependent. Similarly lion people used cocaine powder in the last year [2]. tobacco and drug use were notably higher in the

DRUG USE, DRUG PROBLEMS AND DRUG ADDICTION: SOCIAL INFLUENCES AND SOCIAL RESPONSES 261 homeless group compared to the other two sam- month declined from 37% in 1979 to 16.4% in 1991, ples [4]. Prison inmates are also missed in general then rose to 26.2% by 1997. Rates have declined since population surveys. In a subsample of prisoners taken the 1990s peak to 21.9% of seniors reporting past from a national survey, just over two-fifths (42.5%) of month use in 2007 [6]. The use of most illicit drugs them were diagnosed as drug dependent using the showed a similar trend during this time, though the use Diagnostic Interview Schedule and 29% reached the of stimulant drugs differed from the overall pattern by criteria for ‘severe’ dependence for at least one showing an increase during the 1980s, especially for drug [5]. the use of cocaine by adults aged 26 years or older, In the US, national high school surveys on patterns with monthly cocaine use among high school seniors of drug use have been conducted annually since 1975, increasing in recent years despite a decline in the use as well as a national household survey of drug use of most other illicit drugs [6]. since 1972. These provide estimates of drug-using The 1988 US national survey was the first to collect trends. Based on information from these surveys, the information on items that are part of the ICD-10 and lifetime prevalence of cannabis use among adoles- DSM-III-R criteria for drug dependence and abuse. cents increased steadily during the 1970s, from 14% in The data on each individual in the survey were clas- 1972 to 31% in 1979. While this trend stabilized sified to yield categories of clear, probable, possible during the 1980s, rates rose again and peaked in the and unlikely need for treatment. In 2007, 3% of the mid-1990s. Among high school students, for instance, total population (7.5 million people) aged 12 or over the percentage of people smoking cannabis in the past needed treatment for an illicit drug use problem [7]. 20.4 SPECIAL STUDIES 20.4.1 The Vietnam study A number of points are vividly demonstrated by the Robins and colleagues’ papers. If the circumstances The story of heroin use by American forces in Vietnam are right (or to be more precise, if the circumstances is an extraordinary natural history experiment. Robins are wrong), the extent of use of prohibited drugs such and her colleagues charted heroin use by US army as heroin can be remarkably high. Winick [10] has personnel during and after the Vietnam War [8,9]. described how the extent of drug use within a society They examined two samples. One sample comprised may be determined by three factors: the availability of 500 army-enlisted men drawn randomly from those the drug, the degree of social proscription and the returning home from Vietnam during September extent of conflict or stress. In Vietnam, heroin was not 1971. While in Vietnam 43% had used an opiate and only readily available but was of high purity and at low nearly half of these men had used opiates on at least a cost relative to surplus income. As the use of heroin by weekly basis for six months or longer; a fifth of this smoking and injecting became more widespread general sample considered that they had been addicted among American soldiers, so the extent of moral to heroin while in Vietnam. The sample was followed censorship or social proscription against personal use up on their return home, and when interviewed eight to of such drugs became much less. ten months later, only 10% had used opiates since their The findings from Vietnam also provide clear evi- return to the US and only 1% had been addicted since dence of nondependent patterns of use of heroin. their return. The other sample studied by Robins was a Many users did not become dependent during their group of 500 ‘drug positive’ men (soldiers known to use of heroin. A separate study of this area has been be or to have been drug users). Three-quarters of these undertaken in different settings by Zinberg [11], who men had been addicted to opiates while in Vietnam, described the patterns of use and social structure of but on interview 8–12 months after return to the US, groups of heroin users who abide by strict internal only a third had used any opiates since Vietnam and rules (e.g. avoidance of dose escalation, avoidance only 7% had been addicted since their return. of use on consecutive days, preservation of outside

262 SOCIAL DETERMINANTS interests and friendship networks, etc.), which throughout England. The modalities were selected to Zinberg put forward as factors that were protective be representative of the main treatment modalities against the development of the more typical depen- within the UK. Residential modalities were specialist dent pattern of heroin use. inpatient treatment and rehabilitation programmes. A third area in which the Vietnam studies may The community treatments were methadone mainte- provide us with insight is with regard to the reversi- nance and methadone reduction programmes. The bility of the heroin-using and heroin-addicted status. NTORS provided detailed information about the Virtually all nondependent heroin users and the vast pre-treatment behaviours and problems of the cohort, majority of heroin addicts in Vietnam did not use the operational characteristics of treatment pro- heroin after they had returned to the US. While this is grammes and, particularly, the patient outcomes admittedly a profound piece of engineering of the across a range of measures. As with the American social environment, it nevertheless stands as a clear studies, a central feature of the NTORS was its con- demonstration of the evident reversibility of the her- cern with the impact of existing national treatment oin-using and heroin-addicted conditions. programmes delivered under day-to-day operating conditions. Clinical improvements were found in a wide range 20.4.2 The National Treatment Outcome of problem behaviours, including reductions in the use Research Study (NTORS) of heroin and other illicit drugs, reduced injecting and sharing of injecting equipment, improvements in psy- An important national development in the United chological health and reductions in crime. Frequency Kingdom during the 1990s was the government initia- of heroin use after one year, for example, was reduced tive to conduct a comprehensive survey of the clinical, to about half of the intake levels, and heroin use operational and cost effectiveness of existing services remained at this lower level throughout the full for drug misusers [12]. As a major part of its work, the 4–5 year follow-up period. The sharing of injecting Task Force set up by the Department of Health to equipment was more than halved among patients who oversee this process commissioned the National had been treated in both residential and community Treatment Outcome Research Study (NTORS). settings. Rates of abstinence from illicit drug use Because of cross-national differences, it was increased among the patients from both the residential unclear to what extent the US findings could be and the methadone programmes. Among the residen- generalized to different patient groups, with different tial patients, for instance, almost half (49%) were treatment systems and in different countries. Large- abstinent from heroin after 4–5 years, and the percen- scale, prospective, multisite treatment outcome stu- tage of residential patients whowereabstinent from all dies have played an important role in improving our six illicit target drugs had increased from 1% at intake understanding of treatment effectiveness. They pro- to 38% after 4–5 years. As in the American outcome vide valuable information about drug misusers, the studies, time in residential treatment was related to separate stages of their addiction careers, their various improved post-treatment outcomes. and complicated involvements with treatment ser- Many of the outcomes reported by the NTORS vices, and, of course, the changes that occur in their provided clear evidence of substantial and important drug use and other problem behaviours across reduction in illicit drug problems (reduced frequency extended periods of time after treatment. Such studies and quantity of drug use, increased rates of absti- are rare, however, because of the high costs in money, nence) and reduced injecting risk behaviours [13]. effort and organizational commitment necessary to These changes were accompanied by improved psy- implement, coordinate and sustain such data collec- chological and physical health, and by substantial tion systems over many years. reductions in criminal behaviour. However, not all The NTORS investigated a cohort of problem drug outcomes were so positive. Even in a treatment cohort users treated in four treatment modalities provided in that showed such major treatment gains, in the four either residential or community treatment settings years after intake, there was a continuing mortality

DRUG USE, DRUG PROBLEMS AND DRUG ADDICTION: SOCIAL INFLUENCES AND SOCIAL RESPONSES 263 rate of about 1% per year (six times higher than for an and identified the greater extent of penetration in age-matched group in the general population). In wards of greater socioeconomic deprivation, while addition, many clients were drinking heavily at intake also noting the subsequent spread of heroin problems and continued to drink heavily throughout the five- to adjacent local areas whose association with the year follow-up period. The NTORS recommended original heroin problem was only geographical. Over that drug treatment services should be modified to a similar period, Pearson and colleagues explored the address this continuing problem of alcohol abuse. Nor considerable variation in the extent of use in different can the NTORS results be seen as showing that any or local communities across the north of England, every type of treatment works. There was marked including variation in the distribution networks, the variability across treatment programmes both in treat- balance between amphetamines and heroin, and the ments provided and in outcomes achieved by clients. prevailing popularities of different routes of admin- The NTORS received a generally sympathetic and istration [16,17]. In particular, the strong association interested response from policy makers, and was able between the extent of spread of heroin use and the to inform the development of UK treatment policy existing levels of unemployment and other indices of responses in several ways. The main study findings socioeconomic disadvantage was noted [18]. and its conclusions were also generally well received by British treatment programmes. This is not surpris- ing since most results have been positive and have 20.4.4 Cities can mobilize a special been interpreted as supporting the ‘treatment works’ response message. The NTORS has also had an international influence, not least by encouraging the implementa- Just as towns and cities can be special insofar as a drug tion of similar projects in Australia, Scotland and problem may develop, so they can also be interesting Ireland. for the response to such a problem. The advent of HIV/ One unforeseen side-effect of the NTORS findings AIDS and emerging awareness of a potential/real link was to encourage a greater focus upon reduced crime with a growing heroin problem prompted a major as a goal of drug misuse treatment services. The review of public policy and practice, and several UK NTORS hoped that its results would contribute to the cities stand as interesting examples of innovative city- debate about how best to allocate scarce economic wide responses. Three UK city-wide examples are resources to tackle drug misuse problems, and the provided, in each of which a major ‘outbreak’ of new merits of retaining the current imbalance in resource heroin use was occurring at a point in time where the allocation whereby the greatest economic commit- HIVimplications were only gradually becoming clear. ment is to repression or other supply reduction mea- Inthefirstwaveofawareness,Edinburghfoundthat, sures rather than treatment. Subsequent to the among its young heroin-injecting population, HIV had NTORS, although increased resources have been already penetrated by the mid-1980s [19,20], and the directed towards treatment, there has also been a very entire orientation of the public response to the heroin marked change of focus whereby crime reduction has problem was a police-orientated control response, moved to the top of the list of political and social with virtually no treatment facilities whatsoever. priorities for drug misuse treatment. Awareness of the impending public health catastrophe prompted a major review and a consequent mobiliza- tion of provision of ready-access treatment [21], nee- 20.4.3 UK studies of neighbourhoods dle and syringe supply/exchange, and the introduction of a ‘harm reduction’ approach to public policy and Drug use in the Wirral was studied by Parker planning and treatment provision [22]. et al. [14,15] who described the rapid and extensive Similarly, in Liverpool and across Merseyside, spread of heroin use over the course of a year or two in concern about HIV prompted a major review even a community of variable social deprivation and in though HIV infection had not yet occurred to any which extensive heroin use was previously unknown, significant extent, and a major initiative was

264 SOCIAL DETERMINANTS established to provide ready access to treatment and treatment is less than 1 in 2000 and, for age-matched the widespread provision of needle and syringe general population, is probably about 1 in 20 000. exchange facilities [23]. Nevertheless, it is sobering This extreme mortality risk is probably a result of a to look back on these times to see how the best combination of lack of realization of loss of drug intentions can on occasions lead one down health- tolerance, deliberate celebratory post-release excess, conferring new avenues while on other occasions lead aggravated by co-consumption of alcohol and ben- one down distracting or even contrary directions [24]. zodiazepines. Studies to explore this further, and to As a final example of a UK city-wide response, the test preventive strategies, are currently in their early Glasgow response is striking for its high level of stages, but the example illustrates the importance of coordinated response to the introduction of treatment looking at the social and temporal aspects of both the in a city that was previously, to a considerable extent, drug problem itself and also of the associated without any formal prescribing component to any pre- sequelae. existing treatment response. Over a short period of time, a centrally coordinated city-wide provision of methadone maintenance was established [25,26] with 20.4.6 Addict doctors extensive involvement of local GPs [27,28] and with the establishment of a network of community phar- Medical practitioners themselves are an interesting macists who participated in a new scheme of super- group since they have higher than average rates of vision of consumption as well as dispensing of pre- alcohol and other drug addictions [36–38]. Certain scribed methadone [29,30]. In each of these city case forms of medical practice offer considerable access to studies, there has been only minor further spread of drugs: the anaesthetist or terminal care physician, or HIVand the spread of the heroin problem has probably GP collecting stocks of control drugs for transport been substantially checked, although here again, con- during home visits [39]. Additionally the degree of cerns have been raised that the rapid expansion of social proscription may be less for doctors who have provision of methadone maintenance and needle and seen the extensive analgesic use of morphine and syringe exchange has led to a perverse removal of heroin without instant emergence of addiction, so that incentivesto seek total abstinence as part of recovery – some of the protective taboos that apply to most of a debate that is active as this chapter is written. society may be eroded. Finally, the motivations behind the decisions to enter medicine may not sit comfor- tably with much of the reality of clinical practice for 20.4.5 Prison release as a time of many practitioners so that considerable dissonance special risk and role strain is generated. Study of the natural history of addiction in doctors also provides an oppor- The final area of special study to be considered is that tunity to look at how outcome may be influenced. The of release from prison. It might be thought that this treatment of addicted doctors is notable for the extent was a time when there would be celebration at release to which treatment is often linked to conditional and surely an absence of problems. However, for the possibility of return to a profession with high social former heroin user, the period following release from standing, and conditional return to work is often prison is fraught with real danger – former heroin preceded and coexists with intense monitoring with, users comprise approximately 40% of the UK prison for example, supervised administration of opiate population [31], among whom the vast majority antagonists, random call-up for physical examination are not using opiates while in prison (at least, not and supervised urine specimens; compulsory atten- regularly) [32]. For these former heroin users, there is dance at support groups; alerting family and collea- a 1 in 200 risk of dying in the first month post-release, gues so as to identify and abort relapses at an early a risk that relates almost entirely to risk of drug stage; and so on. These appear to be widely accepted overdose death [33–35]. This is difficult to compare by both the individual and the system and may go directly, but the risk among heroin addicts in some way to explain the good recovery rates.

DRUG USE, DRUG PROBLEMS AND DRUG ADDICTION: SOCIAL INFLUENCES AND SOCIAL RESPONSES 265 20.5 GENDER – WOMEN AND MEN AND DIFFERENT INVOLVEMENT WITH DRUG PROBLEMS AND WITH TREATMENT Although men outnumber women in both in-treatment tions and lack of antenatal care contribute to the health andout-of-treatmentsamplesofdrugmisusers[40,41], problems of infants born to drug-dependent women. women with drug misuse problems often have serious Morbidity in infants born to drug-dependent women is social,psychologicalandhealthproblems.Manyarein related to the amount of prenatal care as well as to the a relationship with partners who are drug dependent, types of drugs used, and increased prenatal care can and not infrequently with men who are physically reduce some maternal and neonatal adverse effects of violent towards them [42]. Many drug-dependent illicit drug use [48]. Services concerned with the care women have psychological problems. In a national of pregnant addicts should maintain good channels of clinical sample, nearly half of thewomen had received communication to optimize care during the antenatal previous treatment for a psychiatric problem; about period, but communication and liaison between addic- one-thirdhadbeenprescribeddrugsfordepressionand tion services, antenatal clinics and obstetric hospitals one-fifth had suicidal thoughts [43,44]. is often unsatisfactory. It is uncertain how much of what we know about Women drug misusers who are bringing up children treatment and recovery applies equally to men and face many special difficulties, especially when they women, and the extent to which special interventions are physically dependent on drugs. Powis et al. [42] and services are needed for women. Nonetheless, found that almost all of the opiate-dependent mothers treatment programmes that specifically address issues in her sample were living in poverty, and many in of particular relevance to women may be more effec- conditions of extreme poverty. Many women fear that tive for some women, or at least more attractive to their children will be taken into care, and often believe many women. Examples of this could include treat- that social services regard them as unfit mothers ment services that offer child care facilities, assess- purely on the basis of their drug use, regardless of ment (and treatment where required) for psychiatric their parenting capabilities [49]. Children are often disorder and work with partners and family. Women present when their parents are using drugs and are able have similar treatment outcomes to men despite the to describe their parents, drug activities in detail, may fact that some of their pre-treatment problems may be show symptoms of neglect and are often left without more severe than their male counterparts [45,46]. adult supervision, even at very young ages [50]. Pregnancy among female drug misusers may pre- Women who are severely dependent on heroin may sent difficult clinical and other problems. Drug misuse feel that if they seek treatment, this might expose their can adversely affect the mother, the foetus and even- drug problems and possible failings and increase the tually the baby [47]. Obstetric and medical complica- risk of their children being taken from them. 20.6 DRUG CAREERS OR THE NATURAL HISTORY OF ADDICTION ? These two perspectives relate to fundamentally dif- and the position of choice over behaviour (voluntar- ferent models of understanding the phenomenon. ism). Individuals may move in and out of drug use The concept of a drug-using career considers drug during a ‘career’. Parker et al. [15] refer to the use as a ‘career’during which the individual enters and ‘incidence’ and ‘outcidence’ of the behaviour that leaves various activities. Thus prostitution or drug use between them dictate the prevalence. may be considered as an ‘occupation’ within a deviant A different perspective is obtained from study of career. This ‘career’ model involves consideration of the literature on the natural history of addiction. external influences on the drug taking (contingency) Workers from various backgrounds have referred to

266 SOCIAL DETERMINANTS ‘epidemics’, with the linked presumption of greater perspective, for different degrees of involvement risk of ‘infection’ in the disadvantaged or vulner- with the drug at different points in time and able. Edwards [51] has described a tighter concept certainly represents a more tangible concept for of natural history as being ‘the sequential develop- consideration and debate. The two concepts ment of designated biological processes within the (‘career’ and ‘natural history’) may not be mutually individual’. This approach allows for a longer-term exclusive. 20.7 TREATMENT – POLICY AND CLINICAL APPROACHES Formal treatment procedures are only one set of syringe exchange schemes or wide provision of hepa- variables that affect any changes that might occur. titis B vaccination, or may alternatively be at the level Outcome is also affected by demographic and psy- ofindividual treatment (suchas management of detox- chological characteristics, and by a range of social ification).Oftenthereisoverlap,asinthepublicbenefit and environmental factors. seen with methadone and buprenorphine treatment for Treatment and prevention initiatives may be along opiate addicts and the resulting reduced levels of the lines of public health initiatives, such as needle and acquisitive crime and much lower risk of overdose. 20.8 COMMUNITY-BASED TREATMENT AND OTHER INITIATIVES Three major changes to the provision of treatment and report that they would not be willing to prescribe other responses in the UK are described in this section – either for detoxification or maintenance reasons to an the greater role of primary care, the introduction of opiate user and 10% report not being willing to see the harm reduction orientation and specific needle and patient or treat any aspect of their condition including syringe exchange schemes, and the deliberate roll-out general medical needs or, for example, hepatitis B of wider availability of opiate substitution treatments vaccination [53]. There are undoubtedly good aspects (methadone and buprenorphine maintenance). to this much wider primary care provision, but all is General practitioners (GPs) and their primary care not yet clear about further progress toward proper colleagues are an increasingly important and promi- availability at time of need. nent part of the contact with, and provision of treat- Needle and syringe exchange schemes were estab- ment to, those individuals involved in drug abuse. lished across the UK in the late 1980s [58,59] – some Back in 1985, only 19% of GPs had seen a drug freestanding, some linked to pre-existing nonstatutory misuser in their practice in the preceding month, but drug agencies and some linked to NHS drug treatment this figure had expanded to 50% by 2001 [52,53]; centres. A further variant has been the pharmacy- furthermore, this had led to the prescribing of based needle and syringe scheme, which has become treatment in only 31% of occasions compared with increasingly widely provided [60,61]. These are an 50% in 2001. GPs now probably treat about half of the example of an approach to public health planning and opiate addicts in treatment in the UK. Nevertheless, individual and public health response which is often alongside this enthusiasm for the greater involvement termed ‘harm reduction’. By 2005, it was estimated of GPs and presumed consequent greater access to that there were over 1700 exchange programmes in treatment [54,55], a cautionary note needs to be England, with the majority (over 70%) provided by sounded [56,57]. The integrated approach also has pharmacies [62]. An estimated 27 million syringes inherent problems, with clear evidence of continued were distributed in the UK during 1997 [63]. Such discrimination on grounds of diagnosis – 52% of GPs schemes have been successful in making contact with

DRUG USE, DRUG PROBLEMS AND DRUG ADDICTION: SOCIAL INFLUENCES AND SOCIAL RESPONSES 267 drug users whowere otherwise not currently in contact accumulating robust evidence base for methadone with drug treatment services. More recently data from maintenance [65], the potential constructive contribu- schemes in the North West of England found that tionofmaintenanceprescribingwasidentified[66,67], anabolic steroid users were the largest group of new since which time the extent of provision of methadone users [64]. Needle exchange schemes are now seen as maintenance has increased greatly, with more than a key to containing the spread of the hepatitis C virus quadrupling in the number of patients in maintenance among injecting drug users. treatmentthroughthe 15yearsuptothetimeofwriting. The third major development has been the delib- However, despite more provision of this important erate expansion of oral methadone maintenance (and treatment, there is a persistent disturbing trend for this more recently of similar buprenorphine maintenance). treatment to be provided subtherapeutically (or at least Although maintenance had always been part of the suboptimally), with more than half of all maintenance treatment system in the UK, it had become less doses still being prescribed at daily doses below the frequently prescribed through the 1980s. With the lower limit of the recommended therapeutic dose review of the response to HIV and with the steadily range despite a decade of advice to the contrary [68]. 20.9 NARCOTICS ANONYMOUS AND TWELVE-STEP TREATMENTS Since its inception in 1935, Alcoholics Anonymous Brown et al. [69] identified several essential char- (AA) has influenced the treatment of alcoholism acteristics of initiatives such as NA that differentiate and has gained increasing international popularity. them from other interventions. First, they involve the Narcotics Anonymous (NA) is a direct descendant of individual acting in collaboration with others who Alcoholics Anonymous. Both AA and NA have flour- share the same problems, to provide mutual support ished in many countries throughout the world and AA/ for each other’s recovery. NA suggested that ‘a NA now probably has a larger population of drug meeting happens when two or more addicts gather abusers involved in its programme than any other drug to help each other stay clean’ [71]. This element of recovery initiative [69]. mutual support is seen as a key dynamic for change. NA/AA is a fellowship of people who want to do It occurs both through the involvement of members something about their drug/drink problems, and who of the Fellowship who share common problems meet on equal and friendly terms. The primary pur- and through the specific support that can be offered pose of members is to stay sober and help others to by the sponsor. achieve sobriety. The programme consists of studying The NA/AA philosophy sees addiction as an ill- and following the Twelve Steps. These are the essen- ness that permeates all aspects of the individual’s life, tial principles and ingredients of the recovery pro- and which can only be controlled by life-long absti- cess [70]. The Steps emphasize two general themes: nence. This vast and daunting project is broken down belief in a ‘Higher Power’, which each individual into manageable parts and taken ‘one day at a time’. defines for themselves and which represents faith and All of the Twelve-Step programmes see recovery hope for recovery, and pragmatism – doing ‘whatever from substance abuse as a continuing process, with works’ for the individual. Group meetings are one of every day involving an effort to remain free from the best known aspects of NA. When individuals join drugs. The NA/AA disease concept is used to empha- NA, they are usually encouraged to attend more than size the need for the addict to take responsibility for one meeting a week, and a target of attending 90 their own behaviour and to participate actively in meetings in 90 days is often set. As members achieve their own recovery [72]. Although the addict is not sustained abstinence they may attend meetings less personally responsible for their illness, they are often, although it is recommended that they continue certainly seen as 100% responsible for their own to attend meetings at least once aweek, and more often recovery. Family, social and socioeconomic factors if they feel vulnerable to relapse. are recognized as possibly having contributed to the

268 SOCIAL DETERMINANTS addict’s plight, but it is the individual who must take finding new ways of using their leisure time. The NA responsibility for the ways in which they respond to group, with its evening meetings, provides an activity these conditions. during a high-risk time of the day that helps to support NA offers a peer group that can support efforts to continuing abstinence. achieve and maintain abstinence and can be a power- The influence of Twelve-Step programmes now ful asset for anyone seeking to recover from drug extends beyond the NA/AA meetings themselves, and addiction. The role modelling function of NA can be many treatment programmes provide interventions further assisted by the support, mentoring and policing that are substantially based upon the Twelve Steps. offered by the sponsor. A further socially supportive Many other services supplement their programmes by function is that NA provides a structure for the mem- recommending NA/AA attendance as an aftercare ber’s free time. Many drug addicts have difficulties in resource. 20.10 REHABILITATION HOUSES/THERAPEUTIC COMMUNITIES Residential drug-free rehabilitation houses have tra- Christian houses form a smaller body within UK ditionally been described as falling into three cate- rehabilitation houses. Two clear subcategories appear gories – concept-based therapeutic communities, to exist – one relates to houses in which the Christian Christian houses and community-based hostels. To faith is part of the motivation of the staff but in which this list needs to be added the more recently estab- faith is not a prerequisite for entry by a resident lished shorter-stay Twelve-Step or AA/NA-orientated whereas with the other group the commitment to residential houses. Christianity on the part of the resident is deemed Concept-based therapeutic communities are essential. derived from the original Synanon in California in Community houses represent the smallest group 1958. Typically such houses are staffed by a mixture within this section and are the least well defined. The of professionally trained personnel working along- emphasis is placed on reintegration back into the side former addicts who have usually graduated from community during this period of recovery, with early the treatment programme. Mahon [73] has described encouragement to develop links with the local com- the structure of the Concept House as rigidly hier- munity. Typically these houses contain a less rigid archical. Initially the new resident absolves them- structure with less reliance on confrontational and selves of all responsibility, which is taken over by group psychotherapeutic techniques. the community, but gradually the resident assumes The final type of residential rehab is orientated responsibility for self, for an extended network of around Twelve-Step or AA/NA work (see section residents and gradually to the wider community. This above) and is based on the Minnesota Model centres is often extended to include voluntary work outside in the US, such as those at Hazelden and, more the house as well as occupying a quasi-staff position recently, the Betty Ford clinics [75–77]. A distinctive within the community. Further components include characteristic of such centres is the explicit emphasis halfway houses and bridging accommodation for on helping the client to affiliate with the relevant graduates so as to ease the passage back into the Twelve-Step Fellowship (usually either AA or NA, general community. Length of stay correlated posi- but also Cocaine Anonymous (CA) and other fellow- tively with subsequent drug-free status. However, as ships). These Twelve-Step-orientated rehab houses with virtually all drug treatment programmes, ther- typically involve shorter durations of stay (e.g. 4 to apeutic communities are dogged with the problem of 8 weeks, in contrast to the several months of most early self-discharge: a study by Ogborne and other rehab houses), and this is probably more achiev- Melotte [74] reported that a third left within the able precisely because of the strong aftercare structure first month and only a third stayed for more than offered through the increasingly extensive network six months. of Twelve-Step Fellowship meetings in the wider

DRUG USE, DRUG PROBLEMS AND DRUG ADDICTION: SOCIAL INFLUENCES AND SOCIAL RESPONSES 269 community. This orientation is increasingly being tion treatment units [78,79] and this may increase ease incorporated, at least in part, in NHS inpatient addic- of flow between such agencies in future. 20.11 THE PROBLEM OF RELAPSE Relapse is one of the most central clinical problems The two central pillars of the relapse prevention in the addictions. Even for the heavily dependent approach are the identification of high-risk situations drug-taker it is often relatively easy to stop taking and the development and strengthening of coping drugs, but it is then very difficult to remain drug free. strategies. In Litman’s model of survival [83], the The early studies of Hunt, Barnett and Branch chances of avoiding relapse are seen as depending on demonstrated nearly identical patterns of relapse in an interaction between: groups of heroin addicts, alcoholics and cigarette smokers [80], and this finding has been repeatedly 1. Situations perceived to be dangerous for the indi- confirmed. vidual in that they may precipitate relapse. In a prospective study of relapse among heroin addicts after treatment, Gossop et al. [81] found that 2. The coping strategies available within the the critical period for relapse was immediately after individual’s repertoire to deal with these situations. leaving treatment. A large proportion of their subjects used opiates within a matter of only days or a few 3. The perceived effectiveness of these coping weeks of leaving an intensive inpatient treatment behaviours. programme: 42% of the sample had used heroin at least oncewithin oneweek of leaving treatment, and at 4. The individual’s self-perception and self-esteem six-week follow-up 71% had lapsed. This finding and the degree of learned helplessness with which would appear to confirm the most pessimistic views they view their situation. about the prognosis of heroin addicts. However, these apparently depressing findings must be seen in the Among the high-risk situations that are a common broader context of subsequent events. It is clear from problem for ex-addicts attempting to maintain their the results cited by Gossop et al. [81] that the initial abstinence from heroin are negative mood states (such lapse to opiate use did not herald a full-blown relapse as sadness, boredom and anxiety), external events and to an addictive pattern of use. There was a clear cognitive factors [84,85]. There are also strong risks ‘recovery-after-lapse’ effect during the six-month attached to being exposed to drug-related cues. follow-up period, which showed increasing numbers An important first task for the client is to develop an of subjects moving towards abstinence. By the six- understanding of the way in which internal and exter- month point, 45% of the sample were abstinent from nal cues may create a high-risk situation. Structured opiates (as confirmed by urine analysis) and were problem-solving techniques are employed alongside living in the community. rehearsal/role play. The client is warned of the dangers There has been much effort to develop a model for of the way in which covert planning may lead to understanding and preventing relapse. Marlatt and his relapse, and of apparently irrelevant decisions that colleagues have developed a self-management pro- may ‘by chance’ lead the client to happen to find gramme that combines behavioural skill training, himself outside his old drug dealer’s house. The client cognitive interventions and life-style change proce- must learn to spot early warning signals for these dures. It is designed to help support the maintenance potential relapse situations. stage of change, and its goal is to teach individuals The importance of the wider aspects of recovery who are trying to change behaviour now to anticipate and aftercare have come to the fore with the emer- and copewith the problems that will increase their risk gence of differing opinions on the appropriateness of of relapse [82]. maintenance treatments as techniques to achieve

270 SOCIAL DETERMINANTS improvement in health and social situations. It is UK Drug Policy Commission (UKDPC) has put for- clearly mistaken and offensive to regard mere reten- ward a wider more-inclusive perspective on recovery tion in a maintenance programme as recovery, and it is with a proposed overarching description that: ‘The similarly flawed to regard mere achievement of absti- process of recovery from problematic substance use is nence as recovery. On their own, neither ‘bums on characterised by voluntarily-sustained control over seats’ nor ‘clean piss in pots’ are valid measures of substance use which maximises health and wellbeing outcome, but they are legitimately part of the bigger and participation in the rights, roles and responsibil- picture that must be considered [86]. In this regard, the ities of society’ [87]. 20.12 `SPONTANEOUS' OR `NATURAL' RECOVERY Little attention has been paid to apparently sponta- regular drug dealer. Waldolf and Biernacki under- neous or natural patterns of recovery. Reference is took a series of studies of natural recovery among San occasionally made to Winick’s early description of Francisco heroin addicts, and in their review of the the ‘maturing out’ process in which recovery was literature [93] proposed that the process of natural seen as a spontaneous result of age and length of recovery concerns change in a person’s identity, with addiction [88]. Subsequent analyses have considered the replacement of the addict identity by some more the factors (including age and length of addiction) ordinary identity. In this regard it is interesting to which might influence the timing of this seemingly consider the intermediate state between these iden- spontaneous recovery. Anglin, Brecht and Wood- tities that might be achieved by successful recruit- ward [89] reported on their large-scale follow-up ment into methadone maintenance programmes, study and noted that natural recovery appears to be which Preble and Miller described as the ‘inter- influenced by addiction-related or contextual factors world’ [94]. This may perhaps link with Biernacki’s (e.g. property crime and drug dealing), which may later proposal [95] that for the entrenched addict retard the rate of any natural recovery. Stimson and it may be necessary to provide this new identity in Oppenheimer [90] reported on their London study of order to facilitate natural recovery (i.e. in contrast to heroin addicts recruited from treatment centres in relying on some pre-existing perceived role set). 1968. Over the course of a decade, 35% became Likewise Jorquez [96] suggests that recovery stably drug-free, but had travelled down many dif- involves two concurrent processes – extrication ferent pathways to reach this status. Attributions from from the social ties and dependencies of the drug the subjects themselves included the achievement of world, and accommodation to the new ordinary employment, finding new friends and the giving up of life-style. former addict friends; for many a geographical move It is also important to remember that many efforts had been central. Wille [91] looked at data from the at detoxification are actually self-detox efforts, and same cohort and identified two basic patterns of that the short-term success of these efforts may be recovery – a planned, internally motivated, voluntary similar to the success rate of more formal treatment way of becoming abstinent and a separate, externally approaches [97,98]. It is probable that our perspec- enforced way (both of which had been successful tives on ‘treatment’ and ‘recovery’ are too strongly with different individuals). Specific study of sponta- determined by the changes that are more easily neous recovery from heroin addiction was under- observed in the hospital or outpatient environments taken by Schasre [92], who reported on findings from and we may thus fail to give proper consideration 40 subjects who identified the following key factors – to strategies that might support individuals imple- negativeexperienceswithpeers,pressurefrompartner/ menting self-initiated change, just as might be encour- spouse, increased awareness of the stigma of addic- aged and supported in the alcohol and smoking cessa- tion, geographical move and the disappearance of the tion fields.

DRUG USE, DRUG PROBLEMS AND DRUG ADDICTION: SOCIAL INFLUENCES AND SOCIAL RESPONSES 271 20.13 MANIPULATING THE ENVIRONMENT The psychosocial elements of treatment provision can designed to reward and increase the quantity and be every bit as influential as the medication elements robustness of health and social improvements in the that usually attract most of the media, scientific and drug user’s former circumstances [100]. Systematic clinical attention. The National Institute for Health reviews of the area have identified an extensive and Clinical Excellence [99] has recently reviewed the increase in health and social gains from contingency area of Psychosocial Aspects of Treatment in the Drug management approaches across a wide range of types Misuse field, and has identified the major enhanced of drug problem – with opiates such as heroin, sti- gains that can be achieved by contingency manage- mulants such as cocaine, as well as with alcohol and ment strategies and targeted incentives for health, with smoking [101,102]. 20.14 MANIPULATING THE HOME ENVIRONMENT Azrin and colleagues [103–105] have studied various tion system and group counselling [104], and in a community reinforcement strategies in the treatment later study by an additional component built around of recovering alcoholics (although the data have yet self-administration of disulfiram (Antabuse) [105]. to be replicated with subjects dependent on other The authors report on the strikingly better progress drugs). Their approach is based on principles of of alcoholics assigned to the community reinforce- operant reinforcement and involves the systematic ment programme, with the study group spending rearrangement of existing community reinforcers only 2% of their time drinking compared with such as job, family, social and recreational oppor- 55% for the control group, and with similar differ- tunities. These are rearranged in such a way that ences in outcome measures of employment, time further drinking results in time out from these rein- with families and time out of institutions. The effects forcers. The model is further developed by the were also remarkably durable, and were in large part incorporation of a buddy system, a telephone maintained at this level of magnitude for the two- report-in system, an early warning relapse notifica- year follow-up period. 20.15 STUDIES OF THE NEIGHBOURHOOD AND ITS INFLUENCE Study of the social network of drug users may be treatment were now in social relationships in which expected to be a productive line of enquiry. As yet, this the drug use and criminal behaviour were valued area has been minimally studied as a possible area for positively by their peers who saw little reward asso- intervention, although the current enthusiasm for ciated with more socially acceptable activities. In an detached social work and for outreach work represents extension of this study they looked at the post-treat- possible avenues into this field of enquiry and inter- ment progress of former heroin addicts leaving a vention. Hawkins and Fraser [106] found strong cor- residential treatment programme; here again Hawkins relations between frequency of opiate use and the and Fraser found that high levels of continued contact degree of engagement in drug-using networks (and with drug users was associated with higher levels also the degree of disengagement from non-using of relapse back to heroin use [106]. The evidence networks). This included the number of contacts from from these authors accords with the findings from work, school or other organizations, and the authors the naturalistic studies described above (e.g. see concluded that many such opiate addicts entering Reference [93]) – that social network support from

272 SOCIAL DETERMINANTS key individuals appears to be influential on both drug siderable scope for adaptation of the immediate social use and subsequent recovery, and there may be con- environment of addicts during and after treatment. 20.16 CONCLUSION Social and environmental factors have a major influ- Survey. International Review of Psychiatry, 15 ence on the extent of drug use both within the com- (1–2), 43–49. munity and within the individual. New epigenetic and 5. Farrell, M., Boys, A., Bebbington, P. et al. (2002) epigenomic work in this area will require exploration Psychosis and drug dependence: results from a of the interplay of these factors with genetic markers national survey of prisoners. British Journal of Psy- that are beginning to be identified. Furthermore, the chiatry, 181, 393–398. time-course and nature of continued drug use within 6. Johnston, L. D., O’Malley, P. M., Bachman, J. G. and individuals and within groups appears susceptible to Schulenberg, J. E. (2008) Monitoring the Future. National Survey Results on Drug Use, 1975–2007. influence in ways that produce strikingly different Volume I: Secondary School Students, National Insti- outcome data in different studies. As yet, efforts to tute on Drug Abuse, Bethesda, Maryland. tease out the active factors in bringing about these 7. Substance Abuse and Mental Health Services Admin- different outcomes have been disappointingly impre- istration (2008) Results from the 2007 National Survey cise: certainly one of the major areas of continued on Drug Use and Health: National Findings, research must be in the area of factors influencing the SAMHSA, Rockville, Maryland. natural history or career of drug use at the individual 8. Robins, L., Davis, D. H. and Goodwin, D. W. (1974) and group level, until such time as the essential Drug users in Vietnam: a follow-up on return to USA. therapeutic components can be isolated and subse- American Journal of Epidemiology, 99 (4), 235–249. quently administered in a deliberate and controlled 9. Robins, L., Helzer, J. E. and Davis, D. H. (1975) approach. There will then exist the task of identifying Narcotic use in South East Asia and afterwards. different matchings ofindividuals and treatments so as Archives of General Psychiatry, 32 (8), 955–961. to ensure maximum compatibility and acceptability of 10. Winick, C. (1980) A theory of drug dependence based different approaches. The study of drug use and on role, access to, and attitudes towards drugs, in addiction in its social context offers a suitable labora- Theories on Drug Abuse (eds D. J. Lettieri, M. Sayers tory for the development of tomorrow’s preventive and H. W. Pearson), National Institute on Drug Abuse, and treatment techniques. Rockville, Maryland, pp. 225–235. 11. Zinberg, N. (1984) Drug Set and Setting, The Basis for Controlled Intoxicant Use, Yale University Press, REFERENCES New Haven, Connecticut. 12. Polkinghorne, J., Gossop, M. and Strang, J. (2005) The 1. World Health Organization (WHO) (1994) WHO Government Task Force and its review of drug treat- Lexicon of Alcohol and Drug Terms, WHO, Geneva. ment services: the promotion of an evidence-based 2. Hoare, J. and Flatley, J. (2008) Drug Misuse Declared: approach, Heroin Addiction and the British system. Findings from the 2007/08 British Crime Survey. Volume II. Treatment and Policy Responses (eds England and Wales, Home Office, London. J. Strang and M. Gossop), Routledge, Abingdon, 3. National Treatment Agency (2008) Statistics from the pp. 198–205. National Drug Treatment Monitoring System 13. Gossop, M., Marsden, J., Stewart, D. and Kidd, T. (NDTMS), 1 April–31 March 2007, National Treat- (2003) The National Treatment Outcome Research ment Agency, London. Study (NTORS): 4–5 year follow-up results. Addic- 4. Farrell, M., Howes, S., Taylor, C. et al. (2003) Sub- tion, 98 (3), 291–303. stance misuse and psychiatric comorbidity: an over- 14. Parker, H., Newcombe, R. and Bakx, K. (1987) The view of the OPCS National Psychiatric Morbidity new heroin users: prevalence and characteristics in

DRUG USE, DRUG PROBLEMS AND DRUG ADDICTION: SOCIAL INFLUENCES AND SOCIAL RESPONSES 273 Wirral, Merseyside. British Journal of Addiction, 82 28. Scott, R. T. A., Gruer, L. D., Wilson, P. and Hinshel- (2), 147–157. wood, S. (1995) Methadone maintenance treatment – 15. Parker, H., Bakx, K. and Newcombe, R. (1988) Living Glasgow has an innovative scheme for encouraging with Heroin: The Impact of a Drugs ‘Epidemic’ on an GPs to manage drug misusers. British Medical Jour- English Community, Open University Press, Milton nal, 310, 464–465. Keynes. 29. Roberts, K., McNulty, H., Gruer, L. et al. (1998) The 16. Pearson, G., Gilman, M. and McIver, S. (1986) Heroin role of Glasgow pharmacists in the management of use in the north of England. Health Education Journal, drug misuse. International Journal of Drug Policy, 9 45 (3), 186–189. (3), 187–194. 17. Pearson, G. (1987) The New Heroin Users, Basil 30. Gruer, L. and Roberts, K. (2003) New approaches to Blackwell, Oxford. dispensing controlled drugs: supervised consumption of methadone, in Drug Misuse and Community Phar- 18. Pearson, G. (1987) Heroin and unemployment, in A macy (eds J. Sheridan and J. Strang), Taylor & Francis, Land Fit for Heroin: Drugs in Britain in the 1980s (eds London, pp. 137–148. N. Dorn and N. South), Macmillan, London. 31. Brooke, D., Taylor, C., Gunn, J. and Maden, A. (2002) 19. Robertson, J. R. (1985) Drug users in contact with Substance misuse as a marker of vulnerability among general practice. British Medical Journal, 290, 34–35. male prisoners on remand. British Journal of Psychiatry, 20. Robertson, J. R., Bucknall, A. B. V., Welsby, P. D. et al. 177 (3), 248-251. (1986) Epidemic of AIDS related virus (HTLV-III; LAV) infection among intravenous drug abusers. Brit- 32. Strang, J., Gossop, M., Heuston, J. et al. (2006) ish Medical Journal, 292, 527–529. Persistence of drug use during imprisonment: relation- ship of drug type, recency of use and severity of 21. Greenwood, J. (1990) Creating a new drug service in dependence to use of heroin, cocaine and ampheta- Edinburgh. British Medical Journal, 300, 587–589. mine in prison. Addiction, 101, 1125–1132. 22. Robertson, J. R. (2005) The arrival of HIV, in Heroin 33. Seaman, S. R., Brettle, R. P. and Gore, S. M. (1998) Addiction and the British System. Volume I. Origins Mortality from overdose among injecting drug users and Evolution (eds J. Strang and M. Gossop), Routle- recently released from prison: database linkage study. dge, Abingdon. British Medical Journal, 316, 426–428. 23. Stimson, G. V., Alldritt, L., Dolan, K. and Donoghoe, 34. Bird, S. M. and Hutchinson, S. J. (2003) Male drugs- M. (1988) Syringe exchange schemes for drug users in related deaths in the fortnight after release from prison: England and Scotland. British Medical Journal, 29, Scotland, 1996–99. Addiction, 98 (2), 185–190. 1717–1719. 35. Farrell, M. and Marsden, J. (2008) Acute risk of drug- 24. McDermott, P. (2005) The great Mersey experiment: related death among newly released prisoners in Eng- the birth of harm reduction, in Heroin Addiction and land and Wales. Addiction, 103 (2), 251–255. the British System. Volume I. Origins and Evolution 36. Brooke, D. (1995) The addicted doctor – caring pro- (eds J. Strang and M. Gossop), Routledge, Abingdon, fessionals? British Journal of Psychiatry, 166, pp. 139–156. 149–153. 25. Gruer, L., Wilson, P., Scott, R. et al. (1997) General 37. Gossop, M., Stephens, S., Stewart, D. et al. (2001) practitioner centred scheme for treatment of opiate dependent drug injectors in Glasgow. British Medical Health care professionals referred for treatment of alcohol and drug problems. Alcohol and Alcoholism, Journal, 314, 1730–1735. 36 (2), 160–164. 26. Gruer, L. (2005) The emergence of city-wide public health responses to the drugs problem, in Heroin 38. Fowlie, D. G. (1999) The misuse of alcohol and Addiction and the British System. Volume II. Treat- other drugs by doctors: a UK report and one ment and Policy Responses (eds J. Strang and M. region’s response. Alcohol and Alcoholism, 34 (5), Gossop), Routledge, Abingdon, pp. 156–166. 666–671. 39. Strang, J., Wilks, M., Wells, B. and Marshall, J. 27. Wilson, P., Watson, R. and Ralston, G. E. (1994) Methadone maintenance in general practice: patients, (1998) Missed problems and missed opportunities for addicted doctors. British Medical Journal, 316, workload, and outcomes. British Medical Journal, 405–406. 309, 641–644.

274 SOCIAL DETERMINANTS 40. Powis, B., Griffiths, P., Gossop, M. and Strang, J. treatment of opiate misuse. British Medical Journal, (1996) The differences between male and female drug 293, 427–430. users: community samples of heroin and cocaine users 53. Strang, J., Sheridan, J., Hunt, C. et al. (2005) The compared. Substance Use and Misuse, 31 (5), prescribing of methadone and other opioids to addicts: 529–543. national survey of GPs in England and Wales. British 41. Chatham, L. R., Hiller, M. L., Rowan-Szal, G. A. et al. Journal of General Practice, 55, 444–451. (1999) Gender differences at admission and follow-up 54. Gerada, C. (2000) Drug misuse and primary care in the in a sample of methadone maintenance clients. Sub- new NHS. Drugs, Education Prevention and Policy, 7 stance Use and Misuse, 34 (4), 1137–1165. (3), 213–221. 42. Powis, B., Gossop, M., Bury, C. et al. (2000) Drug- 55. Gerada, C. (2005) RCGP Guide to the Management of using mothers: social, psychological and substance use Substance Misuse in Primary Care, Royal College of problems of women opiate users with children. Drug General Practitioners, London. and Alcohol Review, 19 (2), 171–180. 56. Merrill, J. and Ruben, S. (2000) Treating drug depen- 43. Gossop, M., Marsden, J., Steward, D. et al. (1998) dence in primary care: worthy ambition but flawed Substance use, health and social problems of service policy? Drugs, Education Prevention and Policy, 7 users at 54 drug treatment agencies. British Journal of (3), 203–212. Psychiatry, 173, 166–171. 57. Merrill, J. and Ruben, S. (2005) Treating drug depen- 44. Marsden, J., Gossop, M., Stewart, D. et al. (2000) dence in primary care: reflecting on the problems as Psychiatric symptoms among clients seeking treat- well as the potential, in Heroin Addiction and the ment for drug dependence – intake data from the British System. Volume II. Treatment and Policy National Treatment Outcome Research Study. British Responses (eds J. Strang and M. Gossop), Routledge, Journal of Psychiatry, 176, 285–289. Abingdon, pp. 81–93. 45. Fiorentine, R., Anglin, M. D., Gil-Rivas, V. and 58. Stimson, G., Donoghoe, M., Lart, R. and Dolan, K. Taylor, E. (1997) Drug treatment: explaining the (1990) Distributing sterile needles and syringes to gender paradox. Substance Use and Misuse, 32 (6), people who inject drugs: the syringe exchange experi- 653–678. ment, in AIDS and Drug Misuse: The Challenge for 46. Stewart, D., Gossop, M., Marsden, J. et al. (2003) Policy and Practice in the 1990s (eds J. Strang and G. Similarities inoutcomes for men and women after drug Stimson), Routledge, London, pp. 222–231. misuse treatment: results from the National Treatment 59. Lart, R. and Stimson, G. (1990) National survey of Outcome Research Study (NTORS). Drug and Alco- syringe exchange schemes in England. British Journal hol Review, 22, 35–41. of Addiction, 85 (11), 1433–1444. 47. Johnson, K., Gerada, C. and Greenough, A. (2003) 60. Sheridan, J., Lovell, S., Turnbull, P. et al. (2000) Substance misuse during pregnancy. British Journal of Pharmacy-based needle exchange (PBNX) schemes Psychiatry, 183, 187–189. in South East England: a survey of service providers. 48. Jansson, L. M., Svikis, D., Lee, J. et al. (1996) Preg- Addiction, 95 (10), 1551–1560. nancy and addiction. Journal of Substance Abuse 61. Shorrock, T. (2003) The history and operation of Treatment, 13 (4), 321–329. pharmacy needle exchanges, in Drug Misuse and 49. Taylor, A. (1993) Women Drug Users: An Ethnogra- Community Pharmacy (eds J. Sheridan and J. Strang), phy of a Female Injecting Community, Clarendon Taylor & Francis, London, pp. 113–124. Press, Oxford. 62. Abdulrahim, D., Gordon, D. and Best, D. (2006) 50. Davis, S. K. (1990) Chemical dependency in women: a Findings of a Survey of Needle Exchanges in England, description of its effects and outcome on adequate National Treatment Agency, London. parenting. Journal of Substance Abuse Treatment, 7 63. Parsons, J., Hickman, M., Turnbull, P. J. et al. (2002) (4), 225–232. Over a decade of syringe exchange: results from 1997 51. Edwards, G. (1984) Addiction: A Challenge to UK survey. Addiction, 97 (7), 845–850. Society. New Society, 70 (1140) pp. 133–135. 64. McVeigh, J., Beynon, C. and Bellis, M. A. (2003) New 52. Glanz, A. and Taylor, C. (1986) Findings of a national challenges for agency based syringe exchange survey of the role of general practitioners in the schemes: analysis of 11 years of data (1991–2001) in

DRUG USE, DRUG PROBLEMS AND DRUG ADDICTION: SOCIAL INFLUENCES AND SOCIAL RESPONSES 275 Merseyside and Cheshire, United Kingdom. Interna- (eds J. Strang and M. Gossop), Routledge, Abingdon, tional Journal of Drug Policy, 14 (5–6), 399–405. pp. 175–186. 65. Ward, J., Mattick, R. P. and Hall, W. (1998) Methadone 78. Best, D. W., Harris, J. C., Gossop, M. et al. (2001) Maintenance Treatment and Other Opioid Replace- Are the twelve steps more acceptable to drug users ment Therapies, Harwood Academic Publishers, than to drinkers? A comparison of experiences of and Amsterdam. attitudes to Alcoholics Anonymous (AA) and Narco- 66. Advisory Council on the Misuse of Drugs (1993) Aids tics Anonymous (NA) among 200 substance misusers and Drug Misuse Update, HMSO, London. attending inpatient detoxification. European Addic- tion Research, 7 (2), 69–77. 67. Farrell, M., Ward, J., Mattick, R. et al. (1994) Metha- done maintenance treatment in opiate dependence: a 79. Harris, J., Best, D., Gossop, M. et al. (2003) Prior review. British Medical Journal, 309, 997–1001. Alcoholics Anonymous (AA) affiliation and the acceptability of the twelve steps to patients entering 68. Strang, J., Manning, V., Mayet, S. et al. (2007) Does UK statutory addiction treatment. Journal of Studies prescribing for opiate addiction change after national on Alcohol, 64 (2), 257–261. guidelines?: methadone and buprenorphine prescrib- ing to opiate addicts by general practitioners and 80. Hunt, W. A., Barnett, L. W. and Branch, L. G. (1971) hospital doctors in England, 1995–2005. Addiction, Relapse rates in addiction programmes. Journal of 102 (5), 761–770. Clinical Psychology, 27 (4), 455–456. 69. Brown, B. S., O’Grady, K. E., Farrell, E. V. et al. 81. Gossop, M., Green, L., Phillips, G. and Bradley, B. (2001) Factors associated with frequency of 12-step (1989) Lapse, relapse and survival among opiate attendance by drug abuse clients. American Journal of addicts after treatment: a prospective follow-up study. Drug and Alcohol Abuse, 27 (1), 147–160. British Journal of Psychiatry, 154, 348–353. 70. Emrick, C. D. (1999) Alcoholics Anonymous and 82. Marlatt, G. and Gordon, J. (eds) (1985) Relapse Pre- other twelve-step groups, in Textbook of Substance vention, Guilford Press, New York. Abuse Treatment (eds M. Galanter and H. D. Kleber), 83. Litman, G. (1986) Alcohol survival: the prevention of American Psychiatric Association, Washington, DC. relapse, in Treating Addictive Behaviors: Processes of 71. Narcotics Anonymous (1988) Twelve Steps and Change (eds W. R. Miller and N. Heather), Plenum, Twelve Traditions, Narcotics Anonymous, Chats- New York, pp. 391–405. worth, California. 84. Bradley, B., Phillips, G., Green, L. and Gossop, M. 72. Hill, S. Y. (1985) The disease concept of alcoholism: a (1989) Circumstances surrounding the initial lapse to review. Drug and Alcohol Dependence, 16 (3), opiate use following detoxification. British Journal of 193–214. Psychiatry, 154, 354–359. 73. Mahon, T. (1973) Therapy or brain washing? Drugs 85. Marlatt, G. A. and George, W. (1984) Relapse pre- and Society, 2 (5), 7–10. vention: introduction and overview of the model. British Journal of Addiction, 79 (3), 261–273. 74. Ogborne, A. C. and Melotte, C. (1977) An evaluation of the therapeutic community for former drug users. 86. Ashton, M. (2008) The New Abstentionists. Druglink British Journal of Addiction, 72 (1), 75–82. 23, January–February suppl. 75. Cook, C. (1988) The Minnesota Model in the manage- 87. UK Drug Policy Commission (2008) Recovery Con- ment of drug and alcohol dependency: miracle or sensus Statement: Developing a Vision of Recovery, myth? Part 1. The philosophy and the programme. UKDPC, London. British Journal of Addiction, 83 (6), 625–634. 88. Winick, C. (1962) Maturing out of narcotic addiction. 76. Cook, C. (1988) The Minnesota Model in the manage- UN Bulletin of Narcotics, 14 (1), 1–7. ment of drug and alcohol dependency: miracle, 89. Anglin, M. D., Brecht, M. L. and Woodward, J. A. method or myth? Part 2. Evidence and conclusion. (1986) Anempirical study of maturing out: conditional British Journal of Addiction, 83 (7), 735–748. factors. International Journal of the Addictions, 21 77. Leighton, T. and Barton, N. (2005) The origins, arrival (2), 233–246. and spread of residential Minnesota Model centres 90. Stimson, G. and Oppenheimer, E. (1982) Heroin across the UK, in Heroin Addiction and the British Addiction: Treatment and Control in Britain, system. Volume II. Treatment and Policy Responses Tavistock, London.

276 SOCIAL DETERMINANTS 91. Wille, R. (1981) Ten-year follow-up of a representa- 100. Pilling, S., Strang, J. and Gerada, C. (2007) Psycho- tive sample of London heroin addicts. Clinic atten- social interventions for drug misuse and opiate detox- dance, abstinence and morality. British Journal of ification: Summary of NICE clinical guidelines. Brit- Addiction, 76 (3), 259–266. ish Medical Journal, 335, 203–205. 92. Schasre, R. (1966) Cessation patterns among neophyte 101. Lussier, J.P., Heil,S. H.,Mongeon, J.A. et al.(2006) A heroin users. International Journal of the Addictions, meta-analysis of voucher-based reinforcement ther- 1 (2), 23–32. apy for substance use disorders. Addiction, 101 (2), 93. Waldorf, D. and Biernacki, P. (1979) Natural recovery 192–203. from heroin addiction: a review of the literature. 102. Prendergast, M., Podus, D., Finney, J. et al. (2006) Journal of Drug Issues, 9 (2), 281–289. Contingency management for treatment of substance 94. Preble, E. and Miller, T. (1977) Methadone, wine and use disorders: A meta-analysis. Addiction, 101 (11), welfare, in Street Ethnography (ed. R. S. Weppner), 1546–1560. Sage, Beverley Hills, California. 103. Hunt, G. M. and Azrin, N. H. (1973) A community- 95. Biernicki, P. (1986) Pathways from Heroin Addiction: reinforcement approach to alcoholism. Behaviour Recovery without Treatment, Temple University Research and Therapy, 11 (1), 91–104. Press, Philadelphia, Pennsylvania. 104. Azrin, N. H. (1976) Improvements in the community 96. Jorquez, J. S. (1983) The retirement phase of heroin reinforcement approach to alcoholism. Behaviour using careers. Journal of Drug Issues, 13 (3), 343–365. Research and Therapy, 14 (5), 339–348. 97. Gossop, M., Battersby, M. and Strang, J. (1991) Self- 105. Azrin, N. H., Sisson, R. W., Meyers, R. and Godley, M. detoxification by opiate addicts. British Journal of (1982) Alcoholism treatment by disulfiram and com- Psychiatry, 159, 208–212. munity reintorcement therapy. Journal of Behavior Therapy and Experimental Psychiatry, 13 (2), 98. Noble, A., Best, D., Man, L. H. et al. (2002) Self- detoxification attempts among methadone mainte- 105–112. nance patients: What methods and what success? 106. Hawkins, J. and Fraser, M. W. (1981) The social Addictive Behaviors, 27 (4), 575–584. networks of drug abusers before and after treatment. Paper presented at the International Congress of 99. National Collaborating Centre for Mental Health Drugs and Alcohol, Jerusalem, Israel. Center for (2008) Drug Misuse: Psychosocial Interventions – Social Welfare Research, University of Washington, The NICE Guideline, British Psychological Society Seattle. and the Royal College of Psychiatrists, London.

21 Eating disorders Mervat Nasser Health Service and Population Research Department, Institute of Psychiatry, King's College London, London, UK 21.1 FOOD, CULTURE AND SOCIETY Health, foodand bodyare uniquely individualconcerns Changes in food consumption and their effects on that are governed by a vast scope of social judgement, health have been an important theme in much of the where the bodycan neverbedivided fromthe self orthe recent writing on food and eating practices. The new selffrom the social. Food, therefore, is never ‘just food’ concept of ‘food normality’ focuses on the idea of a and eating practices have little to do with hunger or ‘healthy diet’, a dietary pattern based on restricting health needs and more to do with conditions of sub- and/or promoting the consumption of certain food jective identity and social relationships. The behaviour items and nutrients. It also advises what, how much ofeatingisintimatelyboundupwithsocialrelationsand and when to eat [5]. The active promotion of specific cultural ideas, where food becomes analogues to lan- models of eating patterns aims mainly towards reg- guage through which the human body communicates ulating and standardizing the health and care of indi- with the outside world [1,2]. viduals in any given society. Sociological research has recently been concerned Self-regulation in relation to food is encouraged by with the relationship between eating behaviour and the rapid increase in the diet industry and the prolif- variables such as gender, socioeconomic class and the eration of commercial establishments concerned with impact of ethnicity on shaping appetites and deter- weight loss. Several studies showed dieting behaviour mining food choices [3]. In the majority of societies to be prevalent in 50–80% of younger women, who ‘feeding of the family’ is traditionally considered repeatedly reported concerns over weight and a desire women’s work and the nature and quantity of food is to be slimmer [6]. This heightened concern with food historically determined by states of affluence and pov- and weight took place, however, against an increase in erty, where food scarcity and abundanceare often at the population weight norms, particularly for women [7]. mercyofpoliticalandeconomicforces.Also,voluntary This means that the link between ‘body’ and ‘eating starvation or fasting is often connected with an ideo- practices’ cannot be properly evaluated except within logicalstanceandfrequentlyrespondstoabeliefsystem a framework of reconciled biomedical instruction and or a symbolic structure within the culture itself [4]. the sociocultural script. Principles of Social Psychiatry, second edition Edited by Craig Morgan and Dinesh Bhugra Ó 2010 John Wiley & Sons, Ltd.

278 SOCIAL DETERMINANTS 21.2 BEHAVIOURAL DISORDERS OF EATING: THE SPECTRUM OF MORBIDITY Fear of fatness and the pursuit of thinness are modern effects of overeating. Both anorexia and bulimia terms that are now used interchangeably to refer to nervosa were later subsumed under ‘eating dis- food restraint and pathological dieting behaviour. orders’ – a term that was introduced to acknowledge Anorexia nervosa is considered the parent syndrome. the full spectrum of eating psychopathology. The It was first reported in the latter part of the nineteenth concept has been incorporated in both the revised century by William Gull [8] in Britain and Charles American Diagnostic Manual (DSM-IV-R) [12] and Lasegue [9] in France. Both described a distinct state the International Classification of Mental and Beha- of self-starvation peculiar to young women and likely vioural Disorders (ICD-10) [13]. to be caused by a host of emotional factors. However, However, a degree of uncertainty about the classi- the syndrome of anorexia nervosa did not achieve its fication of eating disorders remained, as reflected in status as a specific psychiatric entity until it was categories such as ‘eating disorder not otherwise described by Gerald Russell in 1970. It was Russell, specified’, ‘binge eating disorder’ and obesity. ‘Eating too, who coined the term ‘bulimia nervosa’ and disorder not otherwise specified’ is a ‘catch all’ cate- regarded it initially as a serious variant of anorexia gorythatincludespeopleintheentryorresolvingphase nervosa [10,11]. Bulimia nervosa is described as a of either anorexia or bulimia nervosa and it has all the condition of periodic overeating coupled with exces- attitudesandbehavioursthatcauseclinicalimpairment sive concerns about weight and shape, leading to self- and yet does not meet the frequency, severity or dura- induced vomiting or purging to compensate for the tion criteria to be classified as a full syndrome [14]. 21.3 EATING DISORDERS: THE SOCIOCULTURAL PERSPECTIVES 21.3.1 The thinness ideal and certain subcultural groups where the demand for the spectrum hypothesis thinness is considered a licence to success, such as dancers, models and athletes [7,17,18]. The phenomenon of eating disorders was understand- ably linked to the cultivation of the ‘thin body ideal’ in Western culture and the promotion of thinness through 21.3.2 The gender bias the media, fashion and diet industries. The nature of eating disorder syndromes and the fact that they merge Epidemiological research has also found a clear gen- with more prevalent and the culturally acceptable der bias, with rates being consistently higher in dieting behaviour calls for an interpretation that is women. The female-to-male ratio for anorexia and more clearly grounded in the culture we live in. Also, bulimia nervosa is reported to be on average 10 : 1. prevalencestudiesestimatesubclinicalcases(orpartial This skewed distribution has led to some feminist syndromes)tobefivetimesmorecommonthanthefull- theorists positing that disordered eating is a kind of blown syndromes, suggesting that eating morbidity answer to the dilemma that Western women face in lies on a continuum of severity, with dieting at one end the modern world today, torn between traditional and the extreme forms of disordered eating at the notions of attractiveness and fashionability and new other[15].Thisfindingranparalleltoareportedsteady modern values of autonomy, achievement and self- increaseintherateoftheoccurrenceofeatingdisorders control [19]. Even in the small percentage of men who in the latter half of the twentieth century [16]. develop such disorders, a disproportionate number of The spectrum hypothesis was further supported them have gender identity problems, with concerns by the apparent increased risk for individuals in about body shape [20].

EATING DISORDERS 279 21.3.3 Urbanization rendered them immune from the challenges of mod- ernity that women in the West face [26,27]. Apart from this gender bias, epidemiological findings The issue of the culture-boundedness of eating have also shown an apparent increase in the preva- disorders was nonetheless challenged based on lence of eating disorders in proportion to the level of research emanating from the Middle East, the Far urbanization in any given society [21,22]. This has East, South America and Africa. This research sug- been explained on the basis of increased social gested that eating disorders were increasingly becom- mobility and subsequent changes within family struc- ing a global phenomenon and that rates of these tures, notably a tendency towards nuclearization. disorders in non-Western societies were similar to or Also, as cities became more densely populated and even higher than those reported in the West. diverse, eating patterns, food preferences and meal Most of these transcultural studies were structured times seem to change, with an inevitable rise in rates around a recurring binary hypothesis aiming to find out of obesity followed by heightened weight conscious- the prevalence rate of these disorders in non-Western ness and the development of disordered patterns societies and how these rates correlate with the level of of eating (for a further discussion refer to exposuretoWesternculturalvalues.TheEatingAttitude Reference [23]). Test Questionnaire (EAT) was commonly the main research instrument and some doubts were understand- ablyraisedaboutthevalidityofusingsuchaninstrument 21.3.4 Cross-cultural risks – the global in different cultural settings Nonetheless, the EAT was trend certainly helpful in allowing for preliminary cross- culturalcomparisonstobemadeinthisrespect[28–30]. Based on the notion of the cultural specificity of eating Thedeductionfromthemajorityofthesestudieswas disorders, minority groups in Western cultures were that identification with Western cultural norms was generally considered to be less likely to develop such responsible for this new psychopathology. There was disorders. However, when studies were carried out on noclear distinctionthen between ‘Westernization’and ethnic groups living in the UK and US the results ‘modernization’ and few studies actually attempted to showed the presence of these disorders, particularly measure this cultural change [26]. among those with high aspirations, achievement The issue of confused ‘racial identity’ that was orientation and conflict over ‘racial identity’. A cor- studied earlier among black populations in both the relation was also found between their morbid UK and US was explored again in South Africa after concerns over weight and the level of their the fall of the apartheid regime. A higher rate of eating acculturation [24,25]. disorders was detected in this country following this Across cultures, eating disorders were initially political change, and was attributed to a sense of thought to be rare or even absent, to the extent that shifting identity and altered self-conception among the phenomenon was first seen as exclusively bound to South African women [31,32]. Western culture. This was supported by the perceived The combined exposure to both ‘Western’ and differences in aesthetic standards between West and ‘religious’ values was considered responsible for the non-West. In contrast to the Western ideal of thinness, higher rates of disordered eating behaviour among non-Western societies were seen to favour plumpness Muslim women in the United Arab Emirates and associate with it positive attributes of wealth, and Iran [33,34]. Muslim women were the subject of fertility and femininity. Also, the fact that the majority earlier studies carried out in Egypt [35–37], Tur- of these societies belong to Third World economies key [38] and Israel, where Arab Muslims had among made them appear unlikely to develop a disorder the highest EAT scores [39]. This was again consistent commonly associated with wealth and affluence. with Mumford and Whitehouse’s [40] finding that Added to this, the role of women in non-Western Asian school girls from traditional Islamic back- societies continued to be viewed within a restricted grounds in the UK ran a higher risk of developing framework of the stereotyped and the traditional that eating disorders than their white counterparts.

280 SOCIAL DETERMINANTS Most of the research carried out on Asian groups in interpreted on the basis of the ambiguities women in the UK was initially part and parcel of studies carried this country feel about their identity and how the out on ‘Asian’ migrants. However, in the US, compar- ‘gender’ and the ‘national’ appear to interact to pro- able studies were more concerned with East Asian and duce such conflict and confusion [47]. South Asian cultures and countries. Asian women However, one of the most interesting findings of this attending American schools in the US were shown research has come from Eastern Europe following the to have increased vulnerability to abnormal eating political economic changes. High rates of abnormal behaviours, compared with those in Hong Kong [25]. eating attitudes were reported in Hungary, Poland and It was argued, however, that the overreliance on the the Czech Republic [48–51]. The liberalization of the ‘fear of fatness’ as a diagnostic criterion in eating economy and the undermining of the socialist collec- psychopathology could have been responsible for the tivist structure in former communist Europe, as well as overall underestimation of the magnitude of these in countries that experimented with socialism, problems in societies like China, India and Japan [41]. appeared to have contributed towards an increase in In the case of Japan, an increased tendency towards women’s ambiguity and ambivalence over gender anorexic behaviours was attributed to changes in roles in those societies. When socialist policies pre- traditional family structures in the post-war period, vailed for a time, women by and large felt protected in and was also linked to an increase in the level of their education, employment and child care and their urbanization with higher rates of eating disorders in sense of value seemed to derive from taking part in an cities than in rural areas of Japan [42–44]. overall social philosophy. However, with the political A similar situation was found in Latin America, and economic changes, greater disparity began to with eating disorders emerging as a significant pro- emerge between what women expected of themselves blem in urban areas in particular. Cases of anorexia and what they thought society expected of nervosa were reported in Chile and Brazil [45,46] and them [27,52] (for detailed reviews of all of these Argentina. In Buenos Aries the prevalence of eating studies refer to References [26] and [53]) (see disorders reached epidemic proportions, and was Figure 21.1). Percentage of Dieting/Abnormal Eating Attitudes Worldwide 35.00% 30.00% 25.00% % Eat+ive 20.00% 15.00% 10.00% 5.00% 0.00% Austria Germany Hungary China Poland Italy U.K. Egypt Israel (A) S.Africa Israel (K) Argentina India Japan Figure 21.1 Sociocultural risks/vulnerabil- ities Countries

EATING DISORDERS 281 21.4 EATING DISORDERS, TRANSITIONAL CULTURAL FORCES AND THE ISSUE OF CULTURAL IDENTITY What appears to be common in all the studies outlined clear-cut geographic entity in the same way as in the above is the direct relationship between disordered past. As they type and tap into a shared global envir- eating behaviours and changing social patterns, sug- onment they may in fact be travelling beyond their gesting that the underlying sociocultural dynamics of familiar nexus to ideas and life patterns never con- eating psychopathology lies in ‘cultural change’ and sidered by their homes or home country. However, the not in ‘culture’ per se [54]. Perhaps the most important ‘online culture’ clearly has its advantages and carries of all these cultural changes is the domination of a with it the promise of unlimited choices and oppor- ‘market economy’ in the majority of today’s societies. tunities, and yet the individual has to learn how to deal The liberalization of the economy is likely to produce with the influx of visual information and images to disturbance in the differentials of wealth, an increase negotiate a new cultural identity [57]. in consumerism and standardization of food prefer- The cyber culture, on top of other forces of urba- ences, and body ideals. The latter is bound to be nization, global markets and dietary transitions is marketed through the media, which is also currently bound to produce a threat to national identity and deregulated. create in individuals a sense of disconnection from Worldwide consumerism has introduced in many their more familiar pasts. This is usually the timewhen societies American fast/fat food chains, with the the ‘body’ is called for to express the anguish of the inevitable changes in traditional local diets. The soul and rearticulate through body language a new impact of the fast food culture could lead to a rise in cultural identity. The reformulated ‘corporeal the rates of obesity worldwide, prompting in turn identity’ becomes therefore a kind of coping mechan- greater weight consciousness and the possible emer- ism or a problem-solving tactic to handle the forces of gence of eating psychopathology [26,55]. This kind cultural transition. In the case of eating disorder this is of consumerism is commonly referred to as obviously done through manipulations of weight, diet ‘consumerism without development’, meaning that and food. these societies could in fact possess sufficient wealth The ‘body’, in its attempt to be more in tune with the to buy the life-style that makes them appear modern psychological needs of the individual, engages con- while their real economies remained weak. Subse- stantly with the ever-changing conditions of the ‘lived quently the gap between rich and poor will inevitably in’ society, merging the private and the public domains increase, creating a sense of social imbalance, inse- of human experiences and blurring the boundaries curity and resentment [56]. between what is normal and abnormal, social and Also, with the rapid advent of an ‘online’ culture medical [58] (see Table 21.1). people may no longer relate to their own nation as a 21.5 IMPLICATIONS FOR PREVENTION AND INTERVENTION As one examines the movement of eating disorders issue of self-determination, control and connection. from individual neurosis to cultural marker of dis- Nasser and Katzman [23] suggested that prevention tress, caused by transitional and conflicting cultural and intervention in eating problems could be forces, it becomes increasingly important to identify enhanced by the provision of new social supports ways of operationalizing treatment and prevention and new ways of belonging at the work and school, strategies. This could be done by organizing research possibly taking advantage of existing information and clinical questions around ways of assisting the technology.

282 SOCIAL DETERMINANTS Table 21.1 The sociocultural model of eating disorders 1. Specific to the Western cultural value system – cultivation of the thinness ideal 2. Spectrum of severity – symptoms merge with normal and culturally acceptable ‘dieting’ behaviour 3. Steady increase over the past 50 years 4. Subcultural/vocational risk – more prevalent in groups where thinness is a licence to success 5. Gender risk – 10 times more common in women (gender role confusion and ambivalence) 6. Intracultural, urban risk – increased risk in urban areas – changes in family structure and eating patterns 7. Cross-cultural risks Risk in minority ethnic groups in the West . Confused/disturbed racial identity . A strong need to correct a negative and traditionally stereotyped racial image . High level of acculturation and assimilation of the dominant cultural aesthetic . Desire for acceptance/approval Non-Western cultures and worldwide transitional forces . Deregulation of media and economy . Global standardization of beauty and commodification of the human body . Gender ambivalence following revision of gender roles . Revision of the traditional family structures . Revision of traditional national boundaries through universal media and cyber culture Electronic connections may indeed provide new 3. Caplan, P. (1997) Food, Health and Identity, Routle- ways of achieving this ‘connectedness’. Linked by dge, London and New York. computer technology, individual sufferers may be able 4. Mintz, S. (1994) Eating and being: what food means, in to overcome their social and political isolation and Food: Multidisciplinary Perspectives (eds B. Harris gain insights into new formulae for success and sur- and R. Hoffenberg), Blackwell, Oxford. vival. Several techniques are currently being used in 5. Cannon,G.(1992)FoodandHealth:TheExpertsAgree. the management of eating disorders, focusing on An Analysis of One Hundred Authoritative Scientific psychoeducation and self-help cognitive strategies. ReportsonFood,NutritionandPublicHealthPublished ThroughouttheWorldinThirtyYears,between1961and These interactive web-based multimedia programmes 1991, Consumers’ Association Ltd, London. are likely to make specialist therapies available to 6. Rand, C. and Kuldau, J. M. (1991) Restrained eating many more people who traditionally would have been (weight concerns) in the general population and among unable to access such help [59]. students. International Journal of Eating Disorders, The recognition of these new mechanisms is likely 10, 699–708. to stimulate research devoted to a transcultural and 7. Garner, D. M. and Garfinkel, P. E. (1980) Sociocultural transnational perspective for the prevention and man- factors in the development of anorexia nervosa. Psy- agement of eating disorders. chological Medicine, 10, 483–491. 8. Gull, W. W. (1874) Anorexia nervosa. Clinical Soci- REFERENCES ety’s Transactions, vii, 22. 9. Lasague, C. (1873) De 1’Anorexie Hystorique, 1. Barths, R. (1975) Towards a psychosociology of con- Reprinted in Evolution of Psychosomatic Concepts: temporary food consumption, in European Diet from Anorexia Nervosa and Paradigm (eds R. M. Kaufman Pre-industrial to Modern Times (eds E. Foster and R. and M. Heinman), 1964, International University Press, Foster), Harper Row, New York. New York. 2. Bourdieu, F. (1984) Distinction: A Social Critique of 10. Russell, G. (1979) Bulimia nervosa: an ominous variant the Judgment of Taste, Routledge & Keegan Paul, of anorexia nervosa. Psychological Medicine, 9, London. 429–448.