Social class and substance use disorders: The value of social class as distinct from socioeconomic status

https://doi.org/10.1016/S0277-9536(98)00011-2Get rights and content

Abstract

The relationship between social class and substance use disorders (SUDs) is explored and compared to the relationship between SES and SUDs. Social class and SES are two different conceptualizations of socioeconomic inequality (SEI) which emanate from two different theoretical orientations in sociology. SES is commonly used in epidemiological research and is usually measured in terms of education, income or occupational prestige. Social class is less known and less used. Here, following the work of Wright et al. (Wright, E. O., Hachen, D. and Costello, C. et al. (1982) The American class structure. American Sociological Review 47, 709–726) it is measured in terms of four types of control people have in their work place: ownership, control over budget decisions, control over other workers, and control over one's own work. Data are derived from an epidemiological survey, conducted in Israel, using a two stage sampling procedure for the identification of cases. In the first stage 4914 respondents were screened with the Psychiatric Epidemiological Research Interview (PERI). In the second stage (n=2741), those who screened positive (and a sample of the negatives) were diagnosed by psychiatrists using a structured interview that yielded diagnoses according to the Research Diagnostic Criteria (RDC). The results indicate that those who are advantaged in terms of ownership, i.e. self-employed, have higher rates of SUDs compared to employees. Furthermore, it appears that most disorders have an onset subsequent to entry into the current job, indicating that ownership plays a causal role in the onset of SUDs rather than the other way around. These results are contrasted with those of a previous report from the same study by Dohrenwend et al. (Dohrenwend, B. P., Levav, I. and Shrout, P. E. et al. (1992) Socioeconomic status and psychiatric disorders: the causation selection issue. Science 255, 946–952) which showed just the opposite association between SES and SUDs, i.e. those who are advantaged in terms of SES have lower rates of SUDs. As an explanation of these apparently conflicting results, the possibility is entertained that social class and SES represent independent causal pathways to the onset of SUDs with social class mainly related to primary SUDs and SES mainly to secondary SUDs.

Introduction

Socioeconomic inequality (SEI) is associated with a variety of psychiatric disorders, among which are substance use disorders (SUDs). The literature indicates that regular and heavy non-problem drinking are more common among the advantaged, while abstention, problematic use and alcohol use disorders are more common among the disadvantaged (Halldin, 1985; Bucholz and Robins, 1991). However, several studies reveal that the relationship between SEI and alcohol use disorders is modified by gender and ethnicity. For example, results from the Epidemiological Catchment Area (ECA) study show that the negative association between SEI and alcohol abuse and dependence is much stronger among males than females, probably because the SEI status definition of women depended as much on their spouse's as on their own occupation (Helzer et al., 1991). In addition, Jones-Webb et al. (1995), using a composite measure of SES that took into account income, education, and occupation, found a significant interaction between socioeconomic inequality and ethnicity in the prediction of the number of drinking problems and alcohol dependence symptoms. Among blacks, the inverse relationship with the SEI measure was much stronger than among whites. However, Herd (1994), whose report is based on the same study as that of Jones-Webb et al. (1995), and employs the same definitions of drinking problems and dependence symptoms, but a different indicator of SEI did not observe the two-way interaction between SEI and ethnicity reported by Jones-Webb et al. (1995). Thus, this examination of the literature reveals that the manner in which SEI is defined and measured has an impact on the nature of the relationships found with alcohol abuse and dependence.

The findings regarding the relationship between SEI and drug use disorders are less clear, with some studies indicating a generally negative association and others reporting overall positive associations (Warner et al., 1995; Robins, 1977). In addition to differences in gender, ethnicity and the definition of SEI which is employed, these differences are probably due to the different ways of distinguishing use from abuse and dependence and to differences in the relationship between SEI and different substances of abuse. For example, Antony (1991), using data from the Epidemiological catchment Area (ECA) study, showed that educational attainment is positively related to cannabis use disorders but negatively related to heroine and cocaine abuse and dependence. Moreover, unskilled labour jobs were over-represented among employed heroin abuse/dependent cases but under-represented among employed cocaine cases. In summary, it seems that the definition of SEI that is used also influences the nature of the relationship between SEI and drug use disorders.

In this paper we focus on the impact of a different definition of SEI on the relationship between SEI and the prevalence and onset of SUDs, taking into account potential confounding and effect modification by gender and ethnicity. The most commonly used SEI measures in epidemiology are based on education, occupation, or income. These measures, referred to here as measures of socioeconomic status (SES), have a common underlying concept of a society that is stratified according to a socially desired commodity. Accordingly all these measures are gradational. A different concept and measure of SEI, referred to here as social class, relies on another view of society, one in which social classes are defined in terms of control over productive assets (Wright, 1985). For the empirical measure of social class we rely on the work of Wright and Perrone (1977)and Wright et al. (1982). The measure is based on four types of control people have in their work place: ownership, control over budget decisions, control over other workers, and control over one's own work. On the bases of these four types of control, eight class categories are defined. The work of Kohn et al. (1990)with occupational self-direction, Link et al. (1993)with occupational direction, control, and planning (DCP), and the work of Karasek (1979)with the control-demand model demonstrate that the element of control is important in the relationship of SEI with psychiatric disorders. We have previously shown that social class, defined according to Wright et al. (1982), has potential for etiological insights of psychiatric disorders by demonstrating that social class has a relationship with depression and SUDs that is independent of the relationship between SES and these disorders (Wohlfarth, 1997). The fact that social class has a relationship with SUDs independent from SES implies that the causal pathways responsible for these relationships are different as well.

As a first step in exploring these pathways, one needs to examine the nature and the direction of the relationship. Therefore, the specific aim of this paper is to examine the relationship between the different aspects of social class and SUDs (in the results section) and to compare these results with an earlier report from the same study by Dohrenwend et al. (1992)in which SES definition of SEI was employed (in the discussion section). Special attention will be paid to the causal role of the two SEI measures in the onset of SUDs.

Section snippets

Methods

The data come from a two-phase epidemiological study conducted in Israel. The study was designed to test the competing hypotheses of social-causation and social-selection, the two competing explanations for the inverse association between SES and most psychiatric disorders (Dohrenwend et al., 1992). The study is described in detail elsewhere (Dohrenwend et al., 1992; Levav et al., 1993). Here we provide only a summary of the sampling procedure followed by a more detailed description of the

The relationship between social class and substance use disorders

Table 1 presents rates of disorders in the different class categories. Only year prevalent disorders are presented here. The results for lifetime prevalence are quite similar and, therefore, do not add to the picture. Table 1 indicates that the main class difference in rates of alcohol and drug use disorders lies between the self-employed (Bourgeoisie, Petty Bourgeoisie, and Small Employers) and the employees (the other five classes), with self-employed having much higher rates than employees.

Discussion

Our aim in this paper is to examine the relationship between social class and SUDs and to compare this relationship to the one between SES and SUDs. Our results show a strong association between ownership and SUDs, with higher rates of SUDs among self-employed compared to employees. The causal test showed that the onset of most SUDs occurred after entry into the current job, indicating that being self-employed is likely to play a causal role in the onset of SUDs.

Furthermore, we found the

Acknowledgements

The authors like to thank Dr. Bruce P. Dohrenwend, Dr. Bruce G. Link and Dr. Sharon Schwartz for their valuable comments on earlier versions of this paper.

References (39)

  • T. Wohlfarth

    Socioeconomic Inequality and Psychopathology: Are Socioeconomic Status and Social Class Interchangeable?

    Social Science and Medicine

    (1997)
  • Antony, J. C. (1991) The epidemiology of drug addiction. In Comprehensive Handbook of Drug and Alcohol Addiction, ed....
  • Bucholz, K. K. and Robins, L. N. (1991) Recent epidemiologic alcohol research. In Annual Review of Addictions Research...
  • L. Buydens-Branchey et al.

    Age of alcoholism onset. II Relationship to susceptibility to serotonin precursor availability

    Archives of General Psychiatry

    (1989)
  • R.J. Cadoret et al.

    Alcoholism and antisocial personality: interrelationships, genetic and environmental factors

    Archives of General Psychiatry

    (1985)
  • C.R. Cloninger

    Neurogenetic adaptive mechanisms in alcoholism

    Science

    (1987)
  • B.P. Dohrenwend et al.

    Socioeconomic status and psychiatric disorders: the causation selection issue

    Science

    (1992)
  • J. Endicott et al.

    A diagnostic interview: the schedule for affective disorders and schizophrenia

    Archives of General Psychiatry

    (1978)
  • W.A. Fuller

    Regression analysis for sample survey

    Sankhya C

    (1975)
  • Fuller, W. A., Kennedy, W. and Schnell, D., et al. (1986) PC Carp. Iowa State Univ, Statistical Library, Ames,...
  • Gerstley, L. J., Alterman, A. I., McLellan, A. T. and Woudy, G. E. (1990) Antisocial personality disorder in patients...
  • Goodwin, D. W. (1981) Psychiatric Diagnosis. Oxford University Press, New...
  • Goodwin D. W. (1983) Alcoholism. In The Child at Psychiatric Risk, ed. R. Tarter, pp. 195–213. Oxford University Press,...
  • J. Halldin

    Alcohol consumption and alcoholism in an urban population in central Sweden

    Acta Psychiatrica Scandinavica

    (1985)
  • Helzer, J. E., Burnam, A., McEvoy, L. T. (1991) Alcohol use and dependence. In Psychiatric Disorders in America: The...
  • D. Herd

    Predicting drinking problems among black and white men: results from a national survey

    Journal of Studies in Alcohol

    (1994)
  • Hidiroglou M. A., Fuller, W. A. and Hickman, R. D. (1980) SuperCarp. Iowa State University, Statistical Library, Ames,...
  • Hosmer D. W. and Lemeshow S. (1989) Applied Logistic Regression. John Wiley and Sons, New...
  • A.M. Hussong et al.

    The stress-negative affect model of adolescent alcohol use: disaggregating negative affect

    Journal of Studies on Alcohol

    (1994)
  • Cited by (46)

    • Opposite social gradient for alcohol use and misuse among French adolescents

      2013, International Journal of Drug Policy
      Citation Excerpt :

      There are several possible explanations for divergent findings regarding the relationship of F-SES to alcohol use. On one hand, the F-SES was defined by various means (level of diploma of the parents, occupational category, income, subjective SES, etc.) that sometimes may lead to different conclusions as it was shown for social class and socioeconomic status (Wohlfarth & van den Brink, 1998). On the other hand, most studies have focused only on a single alcohol use indicator (for example a given frequency of use during a period of time), which may be a strong limitation as it does not cover the full range of drinking patterns.

    • Employment relations, social class and health: A review and analysis of conceptual and measurement alternatives

      2010, Social Science and Medicine
      Citation Excerpt :

      A final group of 6 studies use the organizational control-criterion as a separate dimension. Health outcomes include substance use-related disorders (Wohlfarth and Van Den Brink, 1998), psychiatric disorders (Muntaner, Eaton, Diala, Kessler, & Sorlie, 1998; Parslow et al., 2004), job strain (D’Souza et al., 2003), CVD-related risk factors (Hirokawa, Tsutsumi, Kayaba, & Jichi Med Sch Cohort Grp, 2009), and general and cause-specific mortality (Macleod et al., 2005). Two out of the 6 studies are partially based on longitudinal data (D’Souza et al., 2003, Macleod et al., 2005).

    View all citing articles on Scopus
    View full text