Poverty and Mental Illness

Benedetto Saraceno, MD1, Corrado Barbui, MD2

Objective: To assess the relationship between poverty and mental illness in order to stimulate debate on future international cooperation programs in mental health.

Method: Epidemiological data in the international literature addressing the issue of material poverty as a risk factor for the development of mental illness and as a prognostic factor for the outcome of mental illness were reviewed.

Results: The international literature reviewed supports the notion that material poverty is a risk factor for a negative outcome among mentally ill people. In addition, preliminary epidemiological data suggest that service-related variables may be determinants of outcome of mental illnesses. In our view, cooperation with developing countries is a great opportunity to evaluate mental health services in a natural setting.

Conclusions: A new generation of programs for international cooperation in mental health is needed, in which knowledge and technology transfer is based on a service-research attitude. Attention should be focused on variables related to the poverty of services that might be linked to the course and outcome of mental illnesses.

(Can J Psychiatry 1997;42:285–290)

Key Words: material poverty, cultural poverty, service-delivery poverty, mental illness

Over the past decades, the relationship between poverty and mental illness has been a topic of considerable epidemiological research (1,2) aimed at clarifying the importance of material poverty as a risk factor for the development of mental illness (3) and as a prognostic factor for the outcome of mental illness (4,5). There is, however, a third way of approaching this relationship that has not been the objective of intensive research: this approach refers to cultural and service-delivery poverty as factors in the outcome of mental illness (6).

The present paper briefly discusses these fields of epidemiological investigation with a view to promoting debate on the need for a new generation of international cooperation programs in mental health. Although the term “mental illness” has been chosen to encourage a comprehensive attitude towards mental health, the term “schizophrenia” was used when data on other diagnostic groups were lacking.

Material Poverty as a Risk Factor for the Development of Mental Illness

Epidemiological data on this issue can be summarized as follows: people with the lowest socioeconomic status (SES) have 8 times more relative risk for schizophrenia than those of the highest SES (7). Schizophrenic people, in comparison with people without mental disorders, are 4 times more likely to be unemployed or partly employed (8), one-third more likely not to have graduated from high school, and 3 times more likely to be divorced (5). Poverty, from an epidemiological perspective, means low SES, unemployment, and low levels of scholarship and family standing (Figure 1).

Figure 1. The 3 different levels of the relationship between poverty
and mental illness: poverty as a risk factor for the development of
mental illness, as a prognostic factor for the outcome of mental illness,
and poverty of the mental health service as a determinant of outcome.

Prior to 1980, psychiatric epidemiology clearly showed that SES was inversely related to the prevalence rates of psychiatric disorders such as schizophrenia, major depression, and other subtypes of psychopathology (9,10). The higher rate of mental illnesses among people with the lowest SES suggested it might be a risk factor for the development of mental disorders. The underlying rationale was the association between low SES and greater environmental adversity: increased life event stressors, poor quality of maternal and obstetric care, and scarce social resources thus leading to a higher risk for the disorders (11). This is the so-called social causation explanation (see Figure 1) (12). The value of this relationship, however, has been biased by a nonetiological possibility, which argues that rates of mental illnesses are higher in lower SES groups because people with the disorders drift into lower SES groups (12). This is the so-called social selection explanation (see Figure 1).

A number of studies have tried to address the social causation–social selection question and have obtained contrasting answers. The discrepant findings are mainly due to practical and ethical difficulties in approaching the matter with experimental strategies. Nevertheless, quasiexperimental studies have provided evidence that, at least for schizophrenia, people with psychiatric disorders or personal characteristics predisposing to the disorder drift down to  or fail to rise out of lower SES (13). Although this evidence needs to be confirmed in different settings with different ethnic groups, it provides preliminary support for the social selection explanation.

Material Poverty as a Prognostic Factor for the Outcome of Mental Illness

Regardless of the social causation–social selection issue, most people with mental disorders have a low SES and share the consequences of the environment of poverty: homelessness and high levels of mortality and substance abuse (14,15). From an epidemiological as well as a clinical point of view, the question is therefore to understand whether this status affects the long-term course and outcome of the disorders (see Figure 1). This is a central point in the debate because an association between SES and outcome would imply that mental disorders cannot be managed without directly taking into account the environment of poverty.

Warner has analyzed most of the variables associated with recovery from schizophrenia by reviewing 85 long-term follow-up studies of schizophrenic people (16). As expected, these studies showed that recovery from psychosis is worse in the lower SES groups. The time spent in hospital, as well as the number of admissions, was significantly higher in lower-class patients; in addition, these patients had a worse work record and were more socially isolated and stigmatized than those from a higher class. A number of studies found a strong relation between SES and symptomatic outcome: lower-class patients had a worse pattern of psychotic symptoms than higher-class patients at follow-up.

If we shift from a strictly “clinical “ approach to a more comprehensive one—from psychiatry to mental health—the overlap between problems becomes more complex. As stated in the executive summary of the report “World Mental Health: Problems and Priorities in Low-Income Countries” (17), “Key social initiators of mental health problems include repressive genders practices, ethnic conflict, poverty, and local settings that place certain people at high risk” (17). In other words, it is artificial to isolate the notion of poverty from a wider context of problems. The model of overlapping clusters of problems is probably the only way to understand the 2 intriguing follow-up studies on the determinants of outcome of severe mental disorders (18,19). One of the main difficulties in comparing the results of follow-up studies is that they may not have the same criteria for selecting patients and measuring outcomes.

The World Health Organization (WHO) overcame this difficulty by introducing standardized methods of patient sampling and data collection in 10 different countries around the world. The well-known finding that people suffering from psychosis in developing countries had a better outcome than their counterparts in developed countries was surprising. Two-thirds of patients in developing countries but only one-third in developed ones had full remission at follow-up. How can we interpret these data according to the notion of a relationship between poverty and bad outcome? Some authors argued that, despite the standardized recruitment methods, selection bias could have arisen (20). Patients admitted to psychiatric centres in developing countries may not have been comparable to patients admitted in developed ones. Probably, only schizophrenic people with a mild form of disease were admitted for treatment in developing countries, and patients with severe disease may not have been able to reach the psychiatric services because of their inaccessibility in these areas. In developed countries, by contrast, where psychiatric facilities are much more accessible, schizophrenic patients admitted for treatment would have been representative of all the schizophrenic people. If this bias did indeed occur, the better outcome for Third World cases would be due to patient-related variables such as severity of schizophrenia and diagnosis.

Apart from this possibility, it can be hypothesized that not only material but also cultural and service-delivery poverty could represent determinants of outcome of mental illness.

Cultural Poverty

The WHO investigators proposed that cultural poverty affected the course of schizophrenia (19). They claim the need to investigate variables related to the cultural context with “culturally sensitive” research (21). Although this concept of culture refers to a wide variety of variables, such as way of life and traditional beliefs and practices (22), cultural poverty in this context revolves around employment, community involvement, and lack of stigma (16).

In developing countries, patients return to their productive role more easily than in developed countries (23–26)—for example, at follow-up in India, the WHO study found that the vast majority of patients were working in the fields or in domestic work. Unemployed patients showed a worse course and outcome of the disease than employed ones, suggesting a strong relationship between employment and outcome.

The concept of extended family structure is another of these variables (27). In developing countries, it is common for people in the same community to share emotional involvement and responsibility, allowing the nuclear family to provide a support in the management of a disease (28). WHO data showed a better outcome for patients managed in extended families.

Finally, scant stigma is attached to mental disorders in the Third World; there is no evidence strongly supporting an association between low stigma and good outcome, but it is reasonable to assume that this variable could also influence the course and outcome of mental disorders.

Among environmental and cultural factors under scrutiny for a possible role in the course and outcome of schizophrenia, the lack of attention to variables related to psychiatric facilities for patients is surprising.

Service-Delivery Poverty

An alternative explanation for the WHO results is that the treatment of schizophrenic people in developing countries was more effective than that in developed ones. In other words, service-related variables may have affected the course and outcome of schizophrenia (29,30). As a consequence, service-delivery poverty of psychiatric facilities, which are intended to provide adequate assistance and support to patients, might be one of the strongest prognostic factors related to the outcome of mental illnesses.

Which variables thus belong to the term “poverty of services” and which experimental design is methodologically correct to test these variables? The notion that good procedures and adequate resources increase the likelihood of good outcome is widely accepted in every field of medicine (31). Good procedures provided in a context of adequate resources are those whose efficacy has been demonstrated scientifically, namely through the randomized, controlled trial. Hence services should provide a sort of  “sum of all good procedures” to achieve the best outcome for the patient (rational use of drugs, psychological and psychosocial support). This model can explain differences in the recovery rates from physical illness in developing and developed countries: the poverty of procedures and resources in poor countries produces the low recovery rate. In psychiatry, however,  it seems that this relationship between outcome and resources and procedures does not fit so easily. There is a vast body of literature showing that developing countries are often unable to provide the best resources and procedures for patients, from the scarcity of psychotropic drugs to the limited psychological and psychosocial management following precise technical approaches (32–34). Nevertheless, the recovery rates in these countries are no worse than in developed nations.

Thus the “poverty of services” in psychiatry presumably does not refer only to limited resources or unsophisticated procedures, but reflects a lack of some “other” service-related variables (see Figure 1). These variables are probably related to the organization of services dealing with mental health, their comprehensiveness in a given catchment area, their intersectoriality, and their integration with the community. Finally, what has been defined as “the style of work” of a service—affability, sympathetic sharing, knowledge of the patient’s context, flexibility of the professionals’ roles—can also be considered a service-related variable. As Strathdee and Thornicroft state, “Services should be local and accessible . . . comprehensive . . . flexible . . . consumer orientated . . . racially and culturally appropriate . . . should empower clients . . . focus on strengths . . . should be normalized and incorporate natural supports by being in the least restrictive, most natural setting possible . . . should meet special needs and be accountable to the consumers and carers” (35).

In developing countries, despite some shortages of good resources and procedures, a high level of many of these variables might explain the high recovery rate found by WHO. In other words, material poverty may be less important than richness of both cultural factors (for example, familial and nonfamilial social support and stigma) and service- delivery processes. It is, however, difficult to test this research hypothesis in experimental or quasiexperimental studies. From a methodological point of view, randomized, controlled trials would be ideal. Recently, 2 such trials were conducted in an inpatient setting in which the treatment under scrutiny included freedom to leave the ward at will, possibility for the patients to manage their affairs on their own, possibility to interact freely with the outside world, freedom to wear street clothes, and availability of group sessions to discuss psychiatric symptoms, prevention of relapse, and other problems (36,37). This “treatment” achieved significant improvement in the severity of negative and positive symptoms, as well as in the overall functioning, after both 6 months and one year of follow-up.

Although these data provide interesting information on the effect of health-related variables, some methodological difficulties remain, such as lack of randomization in one trial, problems with blind evaluation of outcome and, most of all, poor generalizability of the results to everyday patient care. The question of generalizability of results from randomized, controlled trials, widely discussed in the medical literature (38), seems to be a central problem also in this context, because we are not interested in evaluating the effect of, say, “freedom to wear street clothes” in an experimental setting, with a small, carefully selected group of patients; much more meaningful is whether the introduction of the “freedom to wear street clothes” into the organization of the mental health service improves the outcome of patients.

For this purpose, new epidemiological tools have to be developed. Alternatives to conventional trials have been proposed, including the “real world” randomized trial (39). Dunn stressed the need for introducing a controlled design to evaluate clinical services, the so-called health care trial (40). In health care trials, the unit of randomization is not the patient, but the health service: health services can be randomly allocated to competitive treatments, for example, “freedom to wear street clothes” in the experimental arm and standard treatment in the control arm. This approach would greatly increase the external validity of the results.

Regardless of the experimental design adopted, there is a pressing need to evaluate mental health services in a new way in order to provide information for mental health professionals on variables constituting the poverty of services, which could be closely linked to the course and outcome of mental illnesses.

Poverty and Cooperation with Developing Countries

It is interesting to notice that cooperation with developing countries can represent a great opportunity to evaluate mental health services in a natural setting.

Unfortunately, there are several fuzzy points in this apparently generous idea: first of all, the meaning of the word “cooperation” in this context is definitely vague. Does it mean financial donation, technology transfer, long-term loan, or does it have other implications, ranging from the most intelligent support to the most obtuse intrusion within a country. Second, the concept of “developing country” is itself ambigu- ous because a country can be underdeveloped in some areas, but developing or developed in others. Notions such as poverty and richness; developed, underdeveloped, and developing; or South and North, which are clear in terms of hard economic indicators, are less clear when applied to mental health services in a given country. If we assume that North means richness and South means poverty, then we should realize that every North has its own South, and it is quite common to see a North country providing a South-type service (41,42).

Although the concepts of both cooperation and developing country are fodder for continuous debate and discussion, experiences and initiatives conducted in several “developing countries” in the field of mental health care provided examples of innovative organization of services; traditional linkages between primary health care and psychiatry (33), the use of community resources, and the close contact with the community itself (41) are all components of the “Southern approach” to mental health. Although there are many such valuable innovations, the experience and knowledge of the developing countries are grossly underappreciated by the rest of the world. The crucial point is that these experiences should not only be of practical value but also should generate a theoretical knowledge. These experiences, in other words, should become the way to assess service-related variables as a factor in the outcome of mental illness.

We believe the best approach to obtain this goal is a scientific attitude (42). To be successful, the process of knowledge and technology transfer must be viewed as an intensive service-research project; the organizational and cultural constraints of less developed countries call for rigorous consideration of what is transferred. The model and instruments of psychiatry thus become an object of research through cooperation. From this perspective, the North–South relationship can best be seen in terms of the highly productive and revealing research contribution of the South to the understanding of psychiatry in the North. In other words, the encounter with psychiatry in poverty (South) reveals the poverty of psychiatry (North). The transfer results in the “discovery” that the package sometimes offered for export—the routine methods of thinking and practising psychiatry—could in many cases be a set of assumptions that is by no means scientifically validated.


The literature briefly reviewed here supports the notion that material poverty is a risk factor for bad outcome among mentally ill people; hence the concept of individual poverty depends to some extent on whether the patient lives in the North or South of the world. The same notion related to psychiatric services seems much more geographically independent, since rich or poor services, that is, services providing a high or low quality of assistance, may be located in either the North or the South. From this perspective, any program of international cooperation with developing countries in mental health should be seen as a service-research project that can shed light on service-delivery poverty as a determinant of outcome of mental illness.

Clinical Implications

  • Material poverty is a risk factor for a negative outcome among mentally ill people.
  • Service-delivery poverty might be a prognostic factor related to the outcome of mental illnesses.
  • Evaluative studies of mental health services in a natural setting are warranted.


  • Data on material poverty as a risk factor for the development of mental illnesses need to be confirmed.
  • Service-related variables as determinants of outcome of mental illness have not been clearly investigated.
  • Methodological problems remain in the evaluation of mental health services.


Corrado Barbui is the recipient of a CNR fellowship (National Research Council, Rome), grant No 201.12.78, “Progetto Finalizzato Invecchiamento.” The authors wish to thank J Baggott for editorial assistance.


1. Goldberg EM, Morrison SL. Schizophrenia and social class. Br J Psychiatry 1963;109:785–802.

2. Eaton WW. Epidemiology of schizophrenia. Epidemiol Rev 1985;7:105–26.

3. Dohrenwend BP. Socioeconomic status (SES) and psychiatric disorders: are the issues still compelling? Soc Psychiatry Psychiatr Epidemiol 1990;25:41–7.

4. Gift TE, Strauss JS, Ritzler BA, Kokes RF, Harder DW. Social class and psychiatric outcome. Am J Psychiatry 1986;143:222–5.

5. Cohen CI. Poverty and the course of schizophrenia: implications for research and policy. Hospital and Community Psychiatry 1993;44:951–8.

6. Saraceno B, Frattura L, Bertolote JM. Evaluation of psychiatric services: hard and soft indicators. In: World Health Organziation, editor. Innovative approaches in service evaluation. Geneva: WHO/MNH/MND; 1993. p 37–52.

7. Holzer CE, Shea BM, Swanson JW. The increased risk for specific psychiatric disorders among persons of low socioeconomic status. Am J Soc Psychiatry 1986;4:259–71.

8. Robins LN, Locke BZ, Regier DA. An overview of psychiatric disorders in America. New York: Free Press; 1991.

9. Dohrenwend BP, Dohrenwend BS. Social status and psychological disorders: a causal inquiry. New York: J Wiley; 1969.

10. Dohrenwend BP, Dohrenwend BS. Social and cultural influences on psychopathology. Annu Rev Psychol 1974;25:417–52.

11. Mueller D. Social networks: a promising direction for research on the relationship of the social environment to psychiatric disorder. Soc Sci Med 1980;14:147–61.

12. Dohrenwend BP. Socioeconomic status and psychiatric disorders: an update on the social causation–social selection issue. Epidemiologia e Psichiatria Sociale 1993;2:71–4.

13. Dohrenwend BP, Levav I, Shrout PE, Schwartz S, Naveh G, Link BG, and others. Socioeconomic status and psychiatric disorders: the causation-selection issue. Science 1992;255:946–52.

14. Eckholm E. Schizophrenic victims include strained families. New York Times 1986 Mar 17;A1–B11.

15. Regier DA, Farmer ME, Rae DS. Co-morbidity of mental disorders with alcohol and other drug abuse: results from the Epidemiologic Catchment Area (ECA) study. JAMA 1990;264:2511–8.

16. Warner R. Recovery from schizophrenia. New York: Routledge; 1994.

17. Executive summary: a call for a United Nations year of mental health, and a World Health Organization decade of mental health. In: Desjarlais R, Eisenberg L, editors. World mental health: problems and priorities in low-income countries. New York: Oxford University Press; 1995.

18. World Health Organization. Schizophrenia: an international follow-up study. Chichester (UK): J Wiley; 1979.

19. Jablensky A, Sartorius N, Ernberg G. Schizophrenia, manifestations, incidence and course in different cultures: a World Health Organization ten-country study. Psychol Med 1992; 20 Suppl.

20. Edgerton RB, Cohen A. Culture and schizophrenia: the DOSMD challenge. Br J Psychiatry 1994;164:222–31.

21. Jablensky A, Sartorius N, Cooper JE, Anker N, Korten A, Bertelsen A. Culture and schizophrenia: criticisms of WHO studies are answered. Br J Psychiatry 1994;165:434–6.

22. Lambo T. The importance of cultural factors in psychiatric treatments. In: Al-Issa I, Dennis W, editors. Cross cultural studies of behaviour. New York: Holt, Rinehart & Winston; 1970. p 548–52.

23. Westermeyer J. Psychosis in a peasant society: social outcomes. Am J Psychiatry 1980;137:390–4.

24. Kulhara P, Wig NN. The chronicity of schizophrenia in North West India: results of a follow-up study. Br J Psychiatry 1978;132:186–90.

25. Murphy HBM, Raman AC. The chronicity of schizophrenia in indigenous tropical people. Br J Psychiatry 1971;118:489–97.

26. Waxler NE. Is outcome for schizophrenia better in non industrial societies? J Nerv Ment Dis 1979;167:144–58.

27. Wig NN, Menon DK, Bedi H. Coping with schizophrenic patients in developing countries: a study of expressed emotions in the relatives. In: Leff J, editor. Seventh World Congress of Psychiatry; 1983 Jul 11–16; Vienna. p 157.

28. Rogler LH, Hollingshead AB. Trapped: families and schizophrenia. New York: J Wiley; 1965.

29. Gofin J, Levav I, Saraceno B. Vigilancia y evaluaciùn en los programas de salud mental comunitaria. In: Levav I, editor. Temas de salud mental en la comunidad. Washington (DC): Paltex; 1992. p 48–55.

30 Levav I, Saraceno B. La evaluaciùn de salud mental en la comunidad.  In: Levav I, editor. Temas de salud mental en la comunidad. Washington (DC): Paltex; 1992. p 56–77.

31. Donabedian A. The quality of care: how can it be assessed? JAMA 1988;260:1743–8.

32. Saraceno B, Terzian E, Montero Barquero F, Tognoni G. Salud mental en el nivel de atenciùn primaria: estudio multicéntrico en seis países centroamericanos. Psiquiatría Pública 1993;2:76–94.

33. Saraceno B, Terzian E, Montero Barquero F, Tognoni G. Mental health care in the primary health care setting: a collaborative study in six countries of Central America. Health Policy and Planning 1995;10:133–43.

34. Saraceno B, Coen D, Tognoni G. Uso de los psicofármacos en la clínica. Washington: OMS/OPS, Cuadernos Técnicos; 1990.

35. Strathdee G, Thornicroft G. The principles of setting up mental health services in the community. In: Bhugra D, Leff J, editors. Principles of social psychiatry. Oxford: Blackwell Scientific; 1993. p 473–89.

36. Fan Z, Huang J, Wu O, Jiang S. Comparison of standard locked-ward treatment versus open-ward rehabilitation treatment for chronic schizophrenic patients: a one-year controlled trial in Canton. Br J Psychiatry 1994;165 Suppl 23:45–51.

37. Jin Z. Effect of an open-door policy combined with a structured activity programme on the residual symptoms of schizophrenic in-patients: a six-month randomized controlled trial in Yanbian, Jilin. Br J Psychiatry 1994;165:52–7.

38. Holmberg L, Baum M. Can results from clinical trials be generalized? Nature Med 1:734–6.

39. Simon G, Wagner E, Vonkorff M. Cost-effectiveness comparisons using “real world” randomized trials: the case of new antidepressant drugs. J Clin Epidemiol 1995;48:363–73.

40. Dunn G. Statistical methods for measuring outcomes. Soc Psychiatry Psychiatr Epidemiol 1994;29:198–204.

41. Saraceno B, Aguilar Briceno R, Asioli F, Liberati A, Tognoni G. Cooperation in mental health: an Italian project in Nicaragua. Soc Sci Med 1990;31:1067–71.

42. Saraceno B, Asioli F, Tognoni G, Flores Ortiz M. Laying foundations for improved care of the mentally ill. World Health Forum 1988;9:542–5.


Objectif : Évaluer la relation entre la pauvreté et la maladie mentale afin d’attiser le débat sur les futurs programmes de coopération internationale en matière de santé mentale.

Méthode : On a analysé des données épidémiologiques tirées de la littérature internationale et portant sur la question de la pauvreté comme facteur de risque d’évolution de la maladie mentale et comme facteur pronostique de l’issue de la maladie mentale.  On discute de la question à savoir comment la pauvreté de la culture et de la prestation des services influent sur l’issue de la maladie mentale.

Résultats : La littérature internationale étudiée appuie la notion voulant que la pauvreté matérielle soit un facteur de risque qui prédispose à un pronostic négatif chez les personnes ayant une maladie mentale. De plus, les données épidémiologiques préliminaires laissent entendre que les variables liées aux services pourraient être des facteurs déterminants de l’issue des maladies mentales. À notre avis, les programmes de coopération avec des pays en développement constituent une occasion exceptionnelle d’évaluer les services de santé mentale dans un cadre naturel.

Conclusions : Une nouvelle génération de programmes de coopération internationale s’impose en matière de santé mentale où le transfert des connaissances et des technologies est fondé sur une attitude de service et de recherche.  Il faudrait insister sur les variables liées à la pauvreté des services pouvant être associées à l’évolution et à l’issue des maladies mentales.

Manuscript received May 1996, revised September 1996.

1Head, Laboratory of Epidemiology and Social Psychiatry, WHO Collaborating Centre, “Mario Negri” Institute for Pharmacological Research, Milan, Italy.

2Research Fellow, Laboratory of Epidemiology and Social Psychiatry, WHO Collaborating Centre, “Mario Negri” Institute for Pharmacological Research, Milan, Italy.

Address for correspondence: Dr B Saraceno, Laboratory of Epidemiology and Social Psychiatry, “Mario Negri” Institute for Pharmacological Research, Via Eritrea 62, 20157 Milan, Italy

e-mail: barbui@irfmn.mnegri.it

Can J Psychiatry, Vol 42, April 1997