Definition Of Drug Abuse
Gender, occupational, and socioeconomic correlates of alcohol and drug abuse among U.S. rural, metropolitan, and urban residentsChamberlain C. Diala INTRODUCTION
Many studies have investigated alcohol and drug abuse and dependence separately for urban and rural areas (1-6), however, few studies have looked at the association between geographic contexts (i.e., urban, rural, metropolitan area) and alcohol and drug abuse and dependence (2,7-10). Hence the first objective of this study is to analyze separately the associations among rural, urban, and metropolitan residents' use of alcohol and drugs. Secondly, we assess the strength of association between demographic and socioeconomic status variables and alcohol and drug disorders after stratification by geographic context.
Substance abuse and dependence is a public health problem with far-reaching implications (11) and individuals dependent on illicit drugs experience higher rates of comorbid psychiatric syndromes (12). Many studies on drug use and dependence have focused on urban, nonwhite, low-income, adolescent males, excluding other drug-using groups. Examples include studies on prevalence and consequences (13), drug trafficking (14,15), risk factors (3,4), weapon carrying (16), psychosocial predictors (17), and social correlates (2,3). Until now, no study has employed nationally representative data to compare rural, urban, and metropolitan areas, in terms of the associations between sociodemographic (age, gender, race) and social class (education, household income, and wealth) factors. And the National Comorbidity Survey (NCS) is the only nationally representative dataset with proportionate rural representation to examine alcohol and drug-disorders-using psychiatric diagnostic criteria.
Results of studies comparing rural and urban alcohol and drug abuse and dependence are mixed and are often affected by the possibility of bias or limited generalizability by sample selections. For example, Warner and Leukefeld (10) studied incarcerated subjects and found rural drug abusers to have significantly higher rates of lifetime drug use than do abusers in urban areas. Donnermeyer (7) reviewed 65 reports of research on youth alcohol and drug use and found alcohol use to be similar for both rural and urban youth. Roundtree and Clayton (9) analyzed data from a stratified random subsample of the Kentucky Youth Survey Sample and found alcohol use in both urban and rural schools as well as among racially mixed and racially homogenous schools. Early reports found higher rates of alcoholism among African Americans (18,19). Anthony and Helzer (20), found true racial differences in the prevalence of drug abuse and dependence with Kessler et at. (8), reporting that African Americans in the NCS have significantly lower prevalence of substance-use disorders than whites. This is consistent with the Epidemiological Catchment Area (ECA) finding of higher prevalence of drug and alcohol abuse and dependence among young whites, 18-29 years of age, compared with that among young African Americans (20,21).
Following the aims of our analysis, we used the National Comorbidity Survey (NCS) because in addition to providing a representative sample of U.S. residents, it includes geographic area of residence, demographic and socioeconomic variables, and the prevalence of alcohol and drug abuse and dependence-related disorders.
METHODS
Lifetime Risks of Disorders in the NCS
We used the National Comorbidity Survey (NCS) for our study, because it included demographic and socioeconomic variables associated with the risks of mental disorders, which included alcohol and drug abuse and dependence. Access to (alcohol and drug use and dependence) services in rural areas and availability of rural populations to respond to surveys are two technical difficulties that traditionally burdened psychiatric epidemiology studies in rural areas. However, valid screening instruments in the NCS overcame these limitations because the NCS was conducted in 212 countries across the country with balanced rural representation, thus overcoming access barriers for these hard-to-reach populations (22). Structured interviews like the Diagnostic Interview Schedule (DIS) (23) are the preferred method of assessment in psychiatric epidemiology, and the NCS is the only large study of remote rural areas that incorporated structured interviews. In the present analysis, we carry this research one step further by providing data on the socioeconomic and demographic correlates of alcohol and drug abuse and dependence disorders, separately, in U.S. rural, urban, and metropolitan areas.
The NCS was designed to study the distribution, correlates, and consequences of psychiatric disorders in the United States (8) and was collected between 1990 and 1992 by the Institute for Social Research at the University of Michigan. As the first, and thus far only, national sample on which psychiatric disorders have been ascertained, the NCS data remain pertinent to our purposes. The DSM-III-R is the tool used to generate diagnoses for alcohol and substance abuse and dependence (24). The DSMIII-R is a modified version of the Composite International Diagnostic Interview (CIDI) (25), which was designed to be used by trained interviewers who are not clinicians (26). Interviewers went through a 7-day study-specific training in the use of the CIDI and were closely monitored throughout the data collection period. The CIDI, developed by the World Health Organization, is a reliable and valid epidemiological instrument suitable for use in conjunction with different diagnostic systems (25,26).
Lifetime risk of alcohol and drug (cannabis, cocaine, and opiates) abuse and dependence refers to the proportion of respondents who have ever experienced such disorders. The NCS is nationally representative, and focused on a younger age range than previous studies such as the Epidemiological Catchment Area (ECA). The NCS interview was administered in two parts to respondents aged 15-54 years, in the 48 contiguous states. A total of 8098 respondents were included in Part One, which comprised the core diagnostic interview, a brief risk factor battery, and an inventory of sociodemographic information. The NCS selection of respondents as young as 15 years (instead of the normal lower age limit of 18) was to minimize recall bias with regard to lifetime disorders. Kessler et al. (8), reported that the exclusion of respondents older than 54 years was based on evidence from the ECA study that active comorbidity between substance use disorder and nonsubstance psychiatric disorder is much lower among persons older than 54 years. The NCS response rate was 82.4%. Part Two (N = 5877) of the NCS, which served as the basis for this study, included a more detailed risk factor battery and secondary diagnoses. Kessler et al. (8), provides details for respondent selection processes and criteria. A nonresponse adjustment weight was constructed for the main survey data to compensate for systematic nonresponses. Households were selected at random. A second weight was constructed to adjust for probabilities of selection between and within households. The sample was then weighted (Table 1) to approximate the U.S. population distribution for age, sex, race/ethnicity, and education as defined by the 1992 National Health Interview Survey (8).
Geographic areas were used as stratification variables in this study, based on the U.S. Bureau of the Census (27) standard approach for defining "rural" populations as places or towns of less than 2500 inhabitants and in open country outside the closely settled suburbs of metropolitan cities. Several problems surround the definition of rurality (28), as such arbitrariness and many important variables are not taken into account (29). By contrast, urban areas consist of contiguous counties, which contain at least one city of 50,000 inhabitants or more. Metropolitan Statistical Areas (MSA) have at least 100,000 inhabitants, comprise one or more central cities with at least 50,000 inhabitants, and include adjoining areas that are socially and economically related to the central city. The term MSA was devised to account for the social and economic activity patterns of an area's population (27).
Social Stratification Variables
Several social stratification variables: race, gender, education, occupation, income, and wealth were used in this study. African Americans and whites were the two race categories compared, with whites as reference. Males were the reference to compare females and their risks of alcohol and drug use and dependence. Education is a common measure of social stratification (30) with possible influence over mental health risks and behavior change (31). Education is stable and reliable over adult life (32) and was measured as "years of formal education." In the analyses, four categories were used [less than high school (0-11 years), high school graduate (12 years), some college (13-15 years), and college graduate or more (16 or more years of education)]. Those with 16 or more years of education were used as the reference category.
Occupation in the NCS identified work hierarchy, job autonomy, and technical aspects of work based on respondents' current employment status and as such, is a major indicator of social classification associated with skills, prestige, wealth, and working conditions (33). Survey questions on occupation determined occupational or industry status using the 1996 Census Bureau Occupation Code (BOC) Index of Industries and Occupations. We used six categories: no occupation indicated (currently unemployed), professionals, technicians, service workers, craft workers, and laborers in the analyses with those not in the labor force as reference.
Household income represents economic resources that enable reduction in health risk and behavior change. Household income was grouped into four categories ($0-$19,999; $20,000-$34,999; $35,000-$69,999; and $70,000 or more). Households with $0-$19,999 annually were used as reference.
Household wealth is more unequally distributed than household income (34,35), and due to inheritance, wealth may not show a strong positive correlation with income, education, or occupation. A single question in the survey asked respondents "suppose you needed money, and you [and your (husband/wife/partner)] cashed in all your checking and savings accounts, stocks and bonds, real estate, sold your home and paid off your mortgage. If you added what you got, about how much would this amount to?" Respondents were then asked to identify their wealth from a list of nine possible wealth "categories." We collapsed this to four household wealth categories ($0-$9999; $10,000-$49,999; $50,000-$199,999; $200,000 or more), with those at the first category used for reference.
Health insurance is a critical enabling factor for health services utilization and was represented with three categories: no health insurance, public health insurance, and private health insurance, which was used as the reference category. Analysis also included four variables for age, 15-24 years 25-34 years, 35-44 years, and 45-54 years, with the youngest ages used for reference.
Statistical Analysis
Lifetime risks and correlates for both alcohol and drug abuse and dependence were estimated. Standard errors for proportions were estimated using the Taylor series linearization method to deal with the complex sample design and weighting of data (36). Next, we conducted bivariate logistic regressions for each variable and subsequent multiple logistic regression controlling for sociodemographic factors (age, gender, and race/ethnicity) as in previous NCS analyses (37).
Statistical Analysis System (SAS) multiple logistic regression was conducted and the SAS models were reanalyzed using SUDAAN (38) to account for the complex sample design and weights employed in the NCS and in order to obtain more accurate standard etimates generalizable to the U.S. population. Marginally significant SAS results were not usually statistically significant with SUDAAN analysis. The data were stratified by means of an iterative procedure to approximate the national population distributions of the cross classification of age, sex, race, ethnicity, education, household income and wealth, and health insurance status as defined by the 1992 U.S. National Health Interview Survey. Multivariate results are presented in the form of odds ratios (OR), obtained by exponentiating the regression coefficients from logistic models. The positive correlation (less than 0.5) of household income and wealth was not significant and unlikely to be problematic in the multivariate models.
RESULTS
Sample Description
Rural, urban, and metropolitan residents in the NCS showed similarity in age and gender (Table l), with half the study population being about 34 years or younger, and about half being males. Rural regions were comprised primarily of whites (85%), this was less so in urban (78%) and metropolitan (68%) regions.
Urban and metropolitan residents report more years of formal education compared to rural residents with about 13% of rural residents as compared to 23% of metropolitan residents having completed college or continued postgraduate training. The proportion of laborers was similar across the three settings and there were more technicians in urban and metropolitan areas and more craft workers in rural areas. One out of 10 rural residents earned $35,000 or more annually, compared to nearly one out of four metropolitan residents who earned $35,000 or more annually. Only 7% of rural households reported wealth of $200,000 or more, compared to 16% of metropolitan households.
Multivariate associations for alcohol and drug disorders were reported as odds ratios with 95% confidence intervals. In this study, age and education (data not presented here) were not significantly associated with alcohol and drug abuse and dependence.
Alcohol Abuse and Dependence Disorders
Bivariate Results
There were no significant rural age differences in alcohol disorders (Table 2). Urban (O.R. = 1.3, p<0.05) and metropolitan (O.R. = 1.5, p <0.05) residents aged between 25-34 years were more likely, while urban residents aged between 45-55 years were significantly less likely (OR = 0.7, p < 0.05) than their respective counterparts aged 15-24 years to abuse or depend on alcohol.
Women were less likely than men were to report alcohol disorders and African Americans were less likely than whites to report alcohol disorders. Except for those with college degrees or more, education was not associated with alcohol disorders. As in previous NCS papers, education shows weak association with mental disorders, including alcohol and drug abuse and dependence.
Urban and metropolitan occupation strata were positively associated with alcohol disorders compared to their counterparts not in the labor force. Also, among rural residents, professionals, services, and craft workers were positively associated with alcohol disorders compared with those not in the labor force.
High rural household incomes were protective ($35,000-$69,999, O.R. = 0.5, p < 0.05; and $70,000 or higher, O.R. = 0.05, p < 0.0001), while metropolitan household incomes were positively associated ($20,000-$34,999, O.R. = 1.4, p < 0.001; $35,000-$69,999, O.R. = 1.6, p < 0.0001; and $70,000 or higher, O.R. = 3.2, p < 0.0001) with alcohol disorders. Household wealth and lack of health insurance were positively associated with alcohol disorders.
Multivariate Results
Women in rural (O.R. = 0.2, p < 0.0001), urban (O.R. = 0.4, p < 0.0001), and metropolitan (O.R. = 0.4, p < 0.0001) areas were less likely to abuse or depend on alcohol compared to their respective male counterparts.
African Americans were protected from alcohol disorders in rural (O.R. = 0.2, p < 0.0001) and urban (O.R. = 0.3, p < 0.0001), but not in metropolitan areas.
Compared to those not in the labor force urban and metropolitan occupation strata were positively associated with alcohol disorders. Rural professionals (O.R. = 2.7, p < 0.0001), services (O.R. = 6.2, p < 0.0001), and craft workers (O.R. = 6.8, p < 0.0001) were significantly associated with alcohol disorders compared with those not in the labor force.
Rural ($20,000-$34,999, O.R. = 0.6, p < 0.05; $35,000-$69,999, O.R. = 0.3, p < 0.0001; and $70,000 or higher, O.R. = 0.04, p < 0.0001) and urban ($20,000-$34,999, O.R. = 0.6, p < 0.001; $35,000-$69,999, O.R. = 0.6, p < 0.001) household incomes were protective from alcohol disorders. However, metropolitan household income was positively associated ($35,000-$69,999, O.R. = 1.4, p < 0.05; and $70,000 or higher, O.R. = 2.0, p < 0.05) with alcohol disorders. In fact, greater rural household income meant greater protection from alcohol disorders, while the reverse relationship applied with metropolitan household incomes.
Rural residents with household wealth between $10,000-$49,000 (O.R. = 1.9, p < 0.0001) and $50,000-$199,000 (O.R. = 1.6, p < 0.0001) and urban residents with household wealth between $10,000-$49,000 (O.R. = 1.4, p < 0.0001) were more likely to present alcohol disorders. In addition, the lack of health insurance was positively associated with alcohol disorders in all three geographical contexts.
Drug Abuse and Dependence Disorders
Bivariate Results
With the exception of the oldest age (45-54 years) categories for rural (O.R. = 0.1, p < 0.0001) and urban (O.R. = 0.2, p < 0.0001) residents, there were no significant age differences in the prevalence of drug disorders by age. Females were less likely than men to report drug disorders. Rural and urban African Americans were less likely to report drug disorders compared to whites.
Except for rural (O.R. = 0.1, p < 0.0001) and metropolitan (O.R. = 1.6, p < 0.05) college graduates, education was not associated with drug disorders. All occupation strata within metropolitan areas were positively associated with drug disorders, while in rural and urban regions, only professionals, services, and craft workers were positively associated with drug disorders.
A high rural household income was protective, but rural and urban households reporting wealth of $10,000-$49,000 are positively associated with drug disorders. In addition, respondents without health insurance were more likely to have drug disorders across geographical contexts.
Multivariate Results
Table 3 describes the multivariate associations for substance abuse disorders. Those aged between 45-54 years in rural (O.R. = 0.1, p < 0.05) and urban (O.R. = 0.2, p < 0.001) areas were protected from drug abuse and dependence. While there were gender differences among urban (O.R. = 0.6, p < 0.05) and metropolitan (O.R. = 0.7, p < 0.01) respondents, there were no gender differences in drug disorders among rural residents. Rural (O.R. = 0.3, p < 0.05) and urban (O.R. = 0.2, p < 0.0001) African Americans were less likely to report drug abuse and dependence compared to their white counterparts; however, there were no race differences among metropolitan residents. Also, education showed no association with drug disorders in the NCS.
Only metropolitan occupations were most strongly associated with drug disorders. While rural (O.R. = 3.2, p <0.0001) and urban (O.R. = 2.8, p < 0.0001) craft workers were strongly associated with drug disorders, all metropolitan occupational strata were positively associated with drug disorders. Among metropolitan residents, there was no relationship between household income and drug disorders. However, rural (O.R. = 0.1, p < 0.001) and urban (O.R. = 0.5, p < 0.05) households earning $35,000-$69,000 and rural (O.R. = 0.06, p <0.0001) households earning >$70,000 per year were protected from drug disorders. Household wealth appeared to be weakly associated with drug disorders across geographic regions, however, rural and metropolitan residents without health insurance (or with public health insurance) were more likely to have drug disorders than those with private health insurance coverage.
DISCUSSION
The lack of rural gender differences in drug abuse and dependence should be interpreted within geographical and sociocultural contexts. Previous epidemiologicaI studies consistently found men (across the geographic regions) to report significantly more alcohol and drug abuse and dependence, as compared to women (39) (ECA, Epidemiological Catchment Area). However, the findings from our analysis in a representative sample of the U.S. population show no gender differences in drug disorders among rural residents. The absence of rural gender differences may be due to increased drug use by rural women in recent years. Recent studies (24,36) have reported increased availability, access to, and use of drugs by rural residents. Increased drug use among rural residents may be due to the downward economic mobility, resulting in the loss of social cohesion especially among whites (40,41) who comprise about 85% of rural residents.
Secondly, alcohol and drug abuse and dependence in the workplace appear to be comparable across geographic regions. Previous studies have reported the problems caused by alcohol (42-44) and drugs (45) in workplaces; and as our results indicate, this appears most predominantly among sales, craft, and services workers.
In this study, occupational stratum was the class measure most powerfully associated with alcohol and drug disorders. The polarization of the U.S. socioeconomic structure in the last decades (46) has produced a shortage of previously well compensated blue collar jobs in the craft and precision production occupations accessible to workers with high school degrees. As rural America has disproportionately suffered as compared to other regions from a relative deficit in job growth (47), this may be related to both the mental health--increased levels of stress, anxiety and depression (48)--and substance abuse and dependence among rural populations reported here for that occupational group.
Rural and urban African Americans were less likely to report alcohol and drug disorders, as compared to whites, providing support for the "African American resiliency in deprived communities" hypothesis (49-51). Stanton and Galbraith (13) reported that 10% of male, urban, African American early adolescents engaged in drug trafficking; however, the strong association between drug trafficking and drug use (15) is not suggested by the NCS. Also of note is the lack of race differences in either alcohol or drug disorders among metropolitan residents. One explanation may be that race-related socioeconomic gaps are narrower in metropolitan areas than in either rural or urban regions, where socioeconomic resources for African Americans are often more limited.
Household income and wealth presented opposite but significant associations with both alcohol and drug disorders in rural areas. For example, rural households with $70,000 plus income are protected, while rural households with $50,000-$199,000 in accumulated wealth are at risk for alcohol disorders. This effect of income and wealth across geographical areas confirm that money flows and valuable assets have different implications for mental health (52). Also, household wealth is weakly associated with alcohol and drug disorders in both urban and metropolitan areas. As expected, respondents with no health insurance and those on public health insurance plans across the three regions are strongly associated with alcohol and drug disorders. Increased recognition of the severity of substance use disorders and levels of need among rural area residents in particular among women, craft, and precision occupations are important next steps in substance abuse research. Also increased appropriations and targeted resources to rural areas could result in more available and improved access to mental health services in rural America.
Earlier research has shown that rural areas have disparate access to mental health resources and service utilization patterns, compared to nonrural areas (53-58). The observation that sales, craft, and service workers are not considered at the bottom of the socioeconomic ladder underscores the analytical payoff of using multiple indicators (e.g., education, income, occupation, wealth) in studies of substance use in populations. Prior findings coupled with those from the current investigation also underscore the need for enhanced awareness around the mental health challenges of rural communities.
Table 1. Demographic characteristics of rural, urban, and
metropolitan residents in the National Comorbidity Survey (NCS).
Rural residents
(N = 1752)
Respondent characteristics % S. E.
Age
15-24 years 23.9 0.7
25-34 years 28.0 0.8
35-44 years 28.3 0.6
45-54 years 19.6 0.3
Gender
Males 51.5 1.2
Females 48.4 1.3
Race
Whites 84.6 2.3
African Americans 8.0 0.1
Other 7.3 0.5
Education
Less than high school 24.2 0.6
High school graduate 45.9 1.1
Some college experience 17.1 0.3
College degree or more 12.7 0.6
Occupation
Not in the labor force 30.0 0.6
Professional 18.4 0.5
Technician 12.7 0.2
Services 10.5 0.3
Craft 16.6 0.4
Laborer 11.6 0.3
Household income
$0-19,999 62.8 1.2
$20,000-34,999 26.6 1.6
$35,000-69,999 9.7 0.4
$70,000 or more 0.7 0.05
Household wealth
$0-9,999 44.3 0.9
$20,000-49,999 25.8 0.5
$50,000-199,999 23.2 0.5
$200,000 or more 6.5 0.2
Health insurance status
None 16.1 0.3
Public 5.1 0.1
Private 78.6 1.7
Urban residents
(N = 2690)
Respondent characteristics % S.E.
Age
15-24 years 25.5 1.3
25-34 years 30.8 1.1
35-44 years 25.5 0.7
45-54 years 18.1 0.9
Gender
Males 48.4 1.6
Females 51.5 1.7
Race
Whites 78.3 2.8
African Americans 8.8 0.6
Other 12.7 1.6
Education
Less than high school 24.2 0.9
High school graduate 38.6 1.6
Some college experience 20.8 0.7
College degree or more 16.3 0.6
Occupation
Not in the labor force 29.8 1.1
Professional 19.3 0.7
Technician 18.5 0.6
Services 8.9 0.4
Craft 11.6 0.4
Laborer 11.6 0.4
Household income
$0-19,999 56.4 1.8
$20,000-34,999 26.2 0.8
$35,000-69,999 16.1 0.5
$70,000 or more 1.1 0.1
Household wealth
$0-9,999 45.7 1.4
$20,000-49,999 22.6 0.7
$50,000-199,999 21.9 0.8
$200,000 or more 9.7 0.5
Health insurance status
None 16.4 0.7
Public 6.2 0.3
Private 77.2 2.4
Metropolitan residents
(N = 3654)
Respondent characteristics % S.E.
Age
15-24 years 24.4 0.9
25-34 years 30.5 0.9
35-44 years 27.8 0.8
45-54 years 17.1 0.6
Gender
Males 49.3 1.5
Females 50.6 1.2
Race
Whites 68.6 2.2
African Americans 15.0 0.7
Other 16.3 0.7
Education
Less than high school 19.8 0.7
High school graduate 32.3 1.1
Some college experience 24.6 0.6
College degree or more 23.1 0.8
Occupation
Not in the labor force 28.2 0.8
Professional 22.9 0.9
Technician 21.2 0.5
Services 9.0 0.3
Craft 8.3 0.4
Laborer 10.1 0.3
Household income
$0-19,999 46.9 1.1
$20,000-34,999 28.1 0.6
$35,000-69,999 22.1 0.8
$70,000 or more 2.7 0.2
Household wealth
$0-9,999 42.5 1.2
$20,000-49,999 19.6 0.5
$50,000-199,999 21.8 0.8
$200,000 or more 16.1 0.6
Health insurance status
None 12.4 0.4
Public 6.6 0.4
Private 80.9 2.2
Table 2. Lifetime prevalence and correlates by social class
of alcohol abuse and dependence among rural, urban, and
metropolitan residents in the National Comorbidity Survey (NCS).
% S.E. O.R. (a) O.R. (b) 95% C.I.
Age (in years)
15-24 25.9 0.2 1.0 1.0 --
25-34 34.0 0.4 1.0 1.0 0.7-2.3
Sex
Males 77.5 0.6 1.0 1.0 --
Females 22.4 0.3 0.2 *** 0.2 0.1-0.4 ***
Race
Whites 87.9 0.8 1.0 1.0 --
Blacks 2.9 0.0 0.2 *** 0.2 0.1-0.5 ***
Occupation
Not in labor 23.0 0.1 1.0 1.0 --
force
Professionals 19.4 0.2 2.7 *** 2.7 1.5-4.7 ***
Sales 8.1 0.1 1.2 1.1 0.7-2.0
Services 18.3 0.2 6.2 *** 2.6 1.5-4.7 **
Crafts 23.8 0.2 6.8 *** 3.7 2.6-6.0 ***
Laborers 7.1 0.1 1.8 1.3 0.6-3.0
Income ($ in thousands)
0-19 62.2 0.5 1.0 1.0 --
20-34 28.1 0.3 1.0 0.6 0.4-0.9 *
35-69 9.5 0.1 0.5 * 0.3 0.2-0.5 ***
70+ 0.1 0.0 0.05 *** 0.04 0.01-0.2 ***
Wealth ($ in thousands)
0-9 42.9 0.3 1.0 1.0 --
10-49 28.6 0.3 2.6 *** 1.9 1.3-2.7 ***
50-199 23.2 0.2 2.0 *** 1.6 1.2-2.1 ***
200+ 5.2 0.1 0.8 0.9 0.4-1.9
Health insurance status
None 21.3 0.2 2.4 *** 1.9 1.3-27 ***
Public 5.3 0.1 1.3 2.4 1.2-4.8 **
Private 73.3 0.6 1.0 1.0 --
Age (in years) % S.E. O.R. (a) O.R. (b) 95% C.I.
15-24 23.8 0.4 1.0 1.0 --
25-34 37.0 0.5 1.3 * 1.5 1.0-2.1 *
Sex
Males 67.0 0.7 1.0 1.0 --
Females 32.9 0.6 0.4 *** 0.4 0.3-0.7 ***
Race
Whites 84.0 1.1 1.0 1.0 --
Blacks 3.7 0.1 0.2 *** 0.3 0.1-0.5 ***
Occupation
Not in labor 21.1 0.3 1.0 1.0 --
force
Professionals 23.9 0.3 3.7 *** 3.9 2.7-5.8 ***
Sales 14.3 0.2 2.2 *** 2.1 1.4-3.2 ***
Services 15.4 0.2 5.8 *** 3.5 2.0-5.8 ***
Crafts 15.9 0.2 4.8 *** 3.3 2.1-5.3 ***
Laborers 10.1 0.2 2.9 *** 2.3 1.5-3.6 ***
Income ($ in thousands)
0-19 57.4 0.7 1.0 1.0 --
20-34 25.5 0.3 0.9 0.6 0.5-0.8 **
35-69 18.4 0.2 0.8 0.6 0.3-0.9 **
70+ 10.1 0.0 0.8 0.6 0.3-1.1
Wealth ($ in thousands)
0-9 45.2 0.5 1.0 1.0 --
10-49 26.2 0.3 2.2 *** 1.4 1.1-1.8 **
50-199 18.4 0.2 1.5 * 0.9 0.6-1.4
200+ 10.1 0.1 1.5 * 1.2 0.7-2.0
Health insurance status
None 20.8 0.3 2.4 *** 2.0 1.4-2.7 ***
Public 3.4 0.1 0.8 1.2 0.7-2.0
Private 75.6 0.9 1.0 1.0 --
Age (in years) % S.E. O.R. (a) O.R. (b) 95% C.I.
15-24 21.2 0.2 1.0 1.0 --
25-34 37.7 0.5 1.5 * 1.3 0.9-1.9
Sex
Males 65.6 0.8 1.0 1.0 --
Females 34.3 0.4 0.4 *** 0.5 0.4-0.7 ***
Race
Whites 80.8 0.8 1.0 1.0 --
Blacks 7.2 0.2 0.5 * 0.6 0.3-1.0
Occupation
Not in labor 17.2 0.2 1.0 1.0 --
force
Professionals 26.0 0.4 4.8 *** 3.1 2.4-5.5 ***
Sales 20.9 0.3 4.3 *** 4.5 3.0-6.6 ***
Services 15.0 0.2 6.6 *** 4.4 2.8-6.9 ***
Crafts 12.0 0.2 4.0 *** 3.2 2.2-4.8 ***
Laborers 8.7 0.1 2.8 *** 2.6 1.6-4.4 ***
Income ($ in thousands)
0-19 45.7 0.5 1.0 1.0 --
20-34 28.8 0.3 1.4 ** 1.2 0.8-1.7
35-69 21.6 0.3 1.6 *** 1.4 1.0-2.0 *
70+ 3.7 0.1 3.2 *** 2.0 1.0-4.2 *
Wealth ($ in thousands)
0-9 40.8 0.4 1.0 1.0 --
10-49 21.9 0.3 2.0 *** 1.2 0.8-1.6
50-199 21.7 0.3 2.0 *** 1.1 0.9-1.5
200+ 15.4 0.3 2.8 *** 1.7 1.1-2.8
Health insurance status
None 15.1 0.2 1.5 ** 1.6 1.1-2.2 **
Public 6.0 0.1 1.4 ** 3.2 2.1-5.1 ***
Private 78.8 0.8 1.0 1.0 --
Results from bivariate and multiple logistic regression
using SUDAAN software.
(a) Bivariate odds ratio.
(b) Multivariate odds ratio with associated 95 per cent
confidence interval. Age cohorts 35-44 and 45-54
years, race category "other," education in years of
school (0-11, 12, 13-15) were not statistically
significant.
* p < 0.05.
* p < 0.001.
*** p < 0.0001.
Table 3. Lifetime prevalence and correlates of drug abuse and
dependence among rural, urban and metropolitan residents in the
National Comorbidity Survey (NCS).
Rural drug abuse and
dependence (N = 151)
% S.E O.R. (a) O.R. (b) 95% C.I.
Age (in years)
15-24 27.7 0.1 1.0 1.0 --
45-54 2.4 0.0 0.1 *** 0.1 0.06-0.6 *
Sex
Males 63.4 0.3 1.0 1.0 --
Females 36.5 0.2 0.5 * 0.6 0.3-1.1
Race
Whites 85.4 0.4 1.0 1.0 --
Blacks 5.5 0.0 0.4 * 0.3 0.1-0.8 ***
Occupation
Not in 28.5 0.1 1.0 1.0 --
Labor Force
Professionals 16.5 0.1 1.8 * 1.9 1.2-3.1 *
Sales 9.0 0.1 1.1 0.8 0.3-2.4
Services 12.6 0.1 3.1 * 1.6 0.6-4.1
Crafts 24.9 0.1 5.3 *** 3.2 2.0-5.1 ***
Laborers 8.2 0.1 1.7 0.9 0.3-2.9
Income ($ in thousands)
0-19 70.4 0.3 1.0 1.0 --
35-69 3.2 0.0 0.1 ** 0.1 0.06-0.5 **
70+ 0.1 0.0 0.04 *** 0.06 0.01-0.3 **
Wealth ($ in thousands)
0-9 52.3 0.2 1.0 1.0 --
10-49 25.5 0.1 1.8 * 1.6 1.0-2.4 *
Health insurance
None 31.3 0.1 4.4 *** 3.6 2.4-5.4 ***
Public 7.1 0.0 2.2 2.9 1.4-6.2 **
Private 61.0 0.3 1.0 1.0 --
Urban drug abuse and
dependence (N = 319)
% S.E O.R. (a) O.R. (b) 95% C.I.
Age (in years)
15-24 27.3 0.3 1.0 1.0 --
45-54 4.0 0.0 0.1 *** 0.2 0.08-0.6 **
Sex
Males 63.4 0.4 1.0 1.0 --
Females 36.5 0.3 0.5 *** 0.6 0.4-0.9 *
Race
Whites 86.6 0.6 1.0 1.0 -
Blacks 3.7 0.0 0.2 *** 0.2 0.1-0.4 ***
Occupation
Not in 27.1 0.3 1.0 1.0 --
Labor Force
Professionals 18.3 0.1 2.1 * 1.9 1.1-3.3 *
Sales 13.0 0.1 1.6 1.4 0.7-2.9
Services 14.8 0.1 4.0 *** 2.2 1.3-3.8 **
Crafts 18.9 0.1 4.3 *** 2.8 1.4-5.4 **
Laborers 7.7 0.1 1.7 * 1.1 0.6-2.0
Income ($ in thousands)
0-19 57.1 0.3 1.0 1.0 --
35-69 10.8 0.1 0.5 * 0.5 0.3-1.0 *
70+ 1.7 0.0 0.9 0.9 0.3-2.2
Wealth ($ in thousands)
0-9 47.7 0.3 1.0 1.0 --
10-49 22.4 0.1 1.7 ** 1.3 0.9-1.9
Health insurance
None 26.0 0.2 3.2 *** 2.9 1.9-4.3 ***
Public 4.9 0.0 1.3 2.0 0.9-4.4
Private 69.0 0.4 1.0 1.0 --
Metropolitan drug abuse
and dependence (N = 497)
% S.E O.R. (a) O.R. 95% C.I. (b)
Age (in years)
15-24 19.3 0.2 1.0 1.0 --
45-54 9.9 0.1 0.6 0.7 0.3-1.6
Sex
Males 57.9 0.4 1.0 1.0 --
Females 42.0 0.3 0.7 ** 0.7 0.5-0.9 *
Race
Whites 79.1 0.5 1.0 1.0 --
Blacks 8.6 0.1 0.7 0.7 0.4-1.2
Occupation
Not in 21.4 0.2 1.0 1.0 --
Labor Force
Professionals 25.4 0.2 3.7 *** 3.2 1.7-5.9 ***
Sales 21.5 0.2 3.5 *** 3.5 1.8-6.6 ***
Services 11.8 0.1 3.8 *** 3.2 1.5-6.7 ***
Crafts 10.0 0.1 2.6 *** 2.3 1.3-4.4 **
Laborers 9.8 0.1 2.6 *** 2.2 1.1-4.4 ***
Income ($ in thousands)
0-19 55.2 0.3 1.0 1.0 --
35-69 17.6 0.1 1.1 1.0 0.6-1.5
70+ 4.6 0.1 2.9 ** 2.2 0.8-6.1
Wealth ($ in thousands)
0-9 47.7 0.3 1.0 1.0 --
10-49 22.7 0.1 1.7 *** 1.1 0.8-1.5
Health insurance
None 20.3 0.1 2.3 ** 2.3 1.6-3.3 ***
Public 6.8 0.1 1.8 * 3.0 1.7-5.5 ***
Private 72.8 0.4 1.0 1.0 --
Results from bivariate and multiple logistic
regression using SUDAAN software.
(a) Bivariate odds ratio.
(b) Multivariate odds ratio with associated 95 per cent confidence
interval. Age categories 25-34 and 35-44 years, education in years
of school (0-11, 12, 13-15), household income in thousands (20-34)
and household wealth (50-199 and 200 plus) were not statistically
significant.
* p < 0.05.
** p < 0.001.
*** p < 0.0001.
ACKNOWLEDGMENT
This study was made possible in part by NIMH grant R03 MH63221.
REFERENCES
(1.) Costello EJ, Agnold A, Burns B J, Stangle DK, Tweed DL. The great smoky mountains study of youth. Goals, designs, methods, and the prevalence of DSM-III-R disorders. Arch Gen Psychiatry 1997; 53(12): 1129-1136.
(2.) Dawkins MP. The social context of substance use among african american youth: rural, urban and suburban comparisons. J Alcohol Drug Educ 1996; 41(3):68-85.
(3.) Dawkins MP. Social correlates of alcohol and other drug use among youthful Blacks in an urban setting. J Alcohol Drug Educ 1986; 32(1): 15-28.
(4.) Farrell AD. Risk factors for drug use in urban adolescents--a three-wave longitudinal study. J Drug Issues 1993; 23(3):443-462.
(5.) Farrell AD, Danish SJ, Howard CW. Risk factors for drug use in urban adolescents: identification and cross-validation. American Journal of Community Psychology 1992; 20(3):263-286.
(6.) Silverman WH. Prevalence of substance use in a rural teenage population. J Adolesc Chem Depend 1991; 2(1):107-117.
(7.) Donnermeyer JF. The use of alcohol, marijuana, and hard drugs by rural adolescents: a review of recent research. Drugs Soc 1992; 7(1-2):31-75.
(8.) Kessler RL, McGonagle KA, Zhao S, Nelson CB, Hughes M, Escheman S, Wittchen HU, Kendler KS. Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States; results from the National Co-Morbidity Survey. Arch Gen Psychiatry 1994; 51:8-19.
(9.) Roundtree PW, Clayton RR. A contextual model of adolescent alcohol use across the rural--urban continuum. Subst Use Misuse 1999; 34(4 & 5):495-519.
(10.) Warner BD, Leukefeld CG. Rural--urban differences in substance use and treatment utilization among prisoners. Am J Drug Alcohol Abuse 2001; 27(2):265-280.
(11.) McRae AL, Brady KT, Sonne SC. Alcohol and substance abuse. Med Clin North Am 2001; 85(3):779-801.
(12.) Kandel DB, Huang F, Davies M. Comorbidity between patterns of substance use dependence and psychiatric syndromes. Drug Alcohol Depend 2001; 64(2):233-241.
(13.) Stanton B, Galbraith J. Drug trafficking among African American early adolescents: prevalence, consequences, and associated behaviors and beliefs. Pediatrics 1994; 93(6 Pt 2):1039-1043.
(14.) Li X, Stanton B, Feigelman S, Black MM, Galbraith J, Huang W. Drug trafficking and drug use among urban African American early adolescents. J Early Adolesc 1994; 14(4):491-508.
(15.) Li X, Feigelman S. Recent and intended drug trafficking among male and female African American early adolescents. Pediatrics 1994; 93 (6 Pt 2):1044-1049.
(16.) Black MM, Ricardo IB. Drug use, drug trafficking, and weapon carrying among low-income, African American, early adolescent boys. Pediatrics 1994; 93(6 Pt 2):1065-1072.
(17.) Matton KI, Zimmerman MA. Psychosocial predictors of substance use among urban Black male adolescents. Drug Soc 1992; 6(1-2):79-113.
(18.) Baskin D, Bluestone H, Nelson M. Ethnicity and psychiatry diagnosis. J Clin Psychiatry 1981; 37:529-537.
(19.) Jones BE, Gray BA. Problems in diagnosing schizophrenia and affective disorders among blacks. Hospital and Community Psychiatry: 1986; 37:61-65.
(20.) Anthony JC, Helzer JE. Syndrome of drug use and dependence. In: Robins LN, Regier DA, eds. Psychiatric Disorders in America: The Epidemiological Catchment Area Study. New York, NY: Free Press, 1991:116-154.
(21.) Helzer JE, Burnam A, McEvoy LT. Alcohol abuse and dependence. In: Robins LN, Regier DA, eds. Psychiatric Disorders in America: The Epidemiological Catchment Area Study. New York, NY: Free Press, 1991:81-115.
(22.) Dana MR, Tielsch JM, Enger C, Joyce E, Santoli JM, Taylor HR. Visual impairment in a rural Appalachian community. J Am Med Assoc 1990; 264:2400-2405.
(23.) Robins LN, Helzer JE, Croughan J, et al. National Institute of Mental Health diagnostic interview schedule: its history, characteristics and validity. Arch Gen Psychiatry 1981; 38:381-389.
(24.) American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 3rd ed. Washington DC: American Psychiatric Press, Inc., 1987.
(25.) World Health Organization. Composite International Diagnostic Interview (CIDI) Version 1.0. Geneva, Switzerland: World Health Organization, 1990.
(26.) Robins LN, Wing J, Wittchen H-U, Helzer JE, Croughan J, Ratcliff KS. The composite diagnostic interview: an epidemiological instrument suitable for use in conjunction with different diagnostic systems and in different cultures. Arch Gen Psychiatry 1998; 45:1069-1077.
(27.) U.S. Bureau of the Census. Statistical Abstract of the United States. 104th ed. Washington DC: U.S. Government Printing Office, 1989.
(28.) Kletke PR, Marder WD, Willke RJ. A projection of the primary care physician population in metropolitan and non-metropolitan areas. In: Primary Care Research: Theory and Methods. AHCPR Conference Proceedings, Washington DC: Agency for Health Care Policy Research, 1993:261-269.
(29.) Murray JD, Keller PA. Psychology in rural America: Current status and future directions. Am Psychol 1991; 46(3):220-231.
(30.) Liberatos P, Link BG, Kelsey JL. The measurement of social class in epidemiology. Epidemiol Rev 1988; 10:87-121.
(31.) Blane D. Editorial: social determinants of health--socioeconomic status, social class and ethnicity. Am J Public Health 1995; 85:903-905.
(32.) Kaplan GA, Keil JE. Socioeconomic factors and cardiovascular diseases: a review of the literature. Circulation 1993; 88:1973-1998.
(33.) Muntaner CM, Eaton WW, Diala CC. Social inequities in mental health: a review of concept and underlying assumptions. Health 2000; 4(1):89-113.
(34.) Oliver MO, Shapiro TM. Black Wealth/White Wealth. A New Perspective on Racial Inequity. New York: Routledge, 1997.
(35.) Wolff EN. Top Heavy: A Study of Wealth Inequity in America. New York: Twentieth Century Fund, 1996.
(36.) Woodruff RS, Causey BD. Computerized method for approximating the variance of complicated estimates. J Am Stat Assoc 1976; 71:315-321.
(37.) Blazer D, Kessler RC, McGonagle KA, Swartz MS. The prevalence and distribution of major depression in a national community sample: the National Co-Morbidity Survey. Am J Psychiatry 1994; 151:979-986.
(38.) Shah B, Barnell B, Bieler G. SUDAAN User's Manual: Software for Analysis of Correlated Data. Release 7.0. Research Triangle Park, NC: Research Triangle Institute, 1996.
(39.) Weissman MM, Bruce ML, Leaf PJ, Florio LP, Holzer C III. Affective disorders. In: Robins LN, Regier DA, eds. Psychiatric Disorders in America: The Epidemiologic Catchment Area Study. New York, NY: Free Press, 1991:53-80.
(40.) Kawachi I, Kennedy BP, Lochner K, Protbrow-Stith D. Social capital, income inequality and mortality. Am J Public Health 1997; 87(9): 1491-1498.
(41.) Danzinger S, Gottschalk P. America Unequal. Boston, MA: Harvard University Press, 1995.
(42.) Roman PM, Blum TC. The workplace and alcohol problem prevention. Alcohol Res Health 2002; 26(1):49-57.
(43.) McFarlin SK, Fals-Stewart W, Major DA, Justice EM. Alcohol use and workplace aggression: an examination of perpetration and victimization. J Subs Abuse 2001; 13(3):303-321.
(44.) Veazie MA, Smith GS. Heavy drinking, alcohol dependence, and injuries at work among young workers in the United States labor force. Alcohol Clin Exp Res 2000; 24(12): 1811-1819.
(45.) Bush DM, Autry JH 3rd. Substance abuse in the workplace: epidemiology, effects and industry response. Occup Med 2002; 17(1): 13-25.
(46.) McFate K, Lawson R, Wilson WJ. Poverty, Inequality, and the Future of Social Policy. Russell Sage Foundation, 1995.
(47.) Dilger RJ, Witt TS. West Virginia in the 1990s. Opportunities for Economic Progress. Morgantown, WV: West Virginia University Press, 1994.
(48.) Diala CC, Muntaner CM. Mood and anxiety disorders among rural, urban and metropolitan residents in the United States. Community Merit Health J 2003: 39(3):239-252.
(49.) Wandersman A, Nation M. Urban neighborhoods and mental health. Psychological contributions to understanding toxicity, resilience, and intervention. Am Psychol 1998; 53(6):647-656.
(50.) Calvert WJ. Protective factors within the family and their role in fostering resiliency in African American adolescents. J Cult Divers 1997; 4(4):110-117.
(51.) Mattis JS, Bell CC, Jagers RJ, Jenkins E. A critical approach to stress-related disorders in African Americans. J Natl Med Assoc 1999; 91(2): 80-85.
(52.) Cohen P, Hesselbart CS. Demographic factors in the use of children's mental health services. Am J Public Health 1993; 83(1):49-52.
(53.) Goldsmith HF, Wagenfeld MO, Manderscheid RW, Stiles D. Specialty mental health services in metropolitan and non-metropolitan areas: 1983 and 1990. Adm Policy Merit Health 1997; 24(6):475-488.
(54.) Merwin EI, Goldsmith HF, Manderscheid RW. Human resource issues in rural mental health services. Commun Ment Health J 1995; 31(6): 525-537.
(55.) Muntaner C, Eaton WW, Diala CC, Kessler RC, Sorlie P. Social class, assets, organizational control and the prevalence of common groups of psychiatric disorders: results from two U.S. epidemiologic surveys. Soc Sci Med 1998; 47(12):243-253.
(56.) Zahner GE, Daskalakis C. Factors associated with mental health, general health, and school-based service use for child psychopathology. Am J Public Health 1997:87(9):1440-1448.
(57.) Zahner GE, Jacobs JH, Freeman DH Jr, Trainor KF. Rural-urban child psychopathology in a northeastern U.S. state: 1986-1989. J Am Acad Child Adolesc Psych 1993; 32(2):378-387.
(58.) Zahner GE, Pawelkiewicz W, DeFrancesco JJ, Adnopoz J. Children's mental health service needs and utilization patterns in an urban community: an epidemiological assessment. J Am Acad Child Adolesc Psych 1992; 31(5):951-960.
Chamberlain C. Diala, Ph.D., M.P.H., (1,4) * Carles Muntaner, M.D., Ph.D., (2) and Christine Walrath, Ph.D.,M.H.S. (3)
(1) Africa Youth Development, Takoma Park, Maryland, USA
(2) Department of Behavior and Community Health Nursing and Department of Epidemiology and Preventive Medicine, University of Maryland-Baltimore, Baltimore, Maryland, USA
(3) ORC Macro, New York, New York, USA
(4) Academy for Educational Development, Health Communication Partnership (at GHPN), Washington, District of Columbia, USA
* Correspondence: Chamberlain C. Diala, Ph.D., M.P.H., Africa Youth Development, 7908 Garland Ave., Suite 2, Takoma Park, MD 20912, USA; Fax: (301) 718-3103; E-mail: africayouthdevelopment@yahoo.com.
COPYRIGHT 2004 Taylor & Francis Ltd.
COPYRIGHT 2004 Gale Group
|
