Predictors of Alcohol and Drug Dependence

Abstract

Objective:

Our study sought to identify sociodemographic, clinical, life perception, and service use characteristics that distinguish new cases of people dependent on substances from the general population; and to determine predictors of substance dependence over a 2-year period. Variables that differentiate people dependent on substances according to sex and age were also assessed.

Methods:

Among 2434 people who took part in an epidemiologic catchment area health survey at baseline, 2.2% were identified with substance dependence at the second measurement time only. Using a comprehensive framework, various aspects were considered as predictors for multivariate statistics.

Results:

Participants with substance dependence at time 2 only showed worse clinical conditions, life events, life and health perception, and neighbourhood characteristics than other participants, but only 2.5% used health care services. Male sex, younger age, stigmatization, and impulsiveness were predictors of substance dependence. Regarding sex, females with dependence were only more likely to suffer from social phobia than males. In terms of age categories, participants over 50 with substance dependence were more likely to have a lower household income and less social support than younger people.

Conclusion:

Stigmatization was the strongest predictor of substance dependence. Our study also confirmed that males and younger people were more likely to have substance dependence. Anti-stigmatization, prevention, and outreach programs are needed to overcome the reluctance of this clientele to use health care services. Health professionals should also pay more attention to life and health perception and neighbourhood characteristics of newly identified drug users.

According to the National Epidemiologic Survey on Alcohol and Related Conditions, the prevalence of lifetime and 12-month dependence on alcohol in the United States in 2001–2002 was 12.5% and 3.8%, respectively,^1,2 while the prevalence of dependence on drugs was 2.6% and 0.6%, respectively.^1,3 In Canada, the prevalence of 12-month dependence in 2002, according to the CCHS 1.2, was 2.2% for alcohol and 1.1% for drugs.⁴ Substance dependence is often associated with poor mental and physical health, poor treatment outcomes, and inferior social conditions, including poverty, lack of education, stigmatization, domestic violence, incarceration, and homelessness.^2,3,5

Cross-sectional studies have largely identified the determinants and risk factors of substance dependence. According to the results of the US National Comorbidity Survey, one-half of people with a lifetime SUD have at least 1 lifetime mental disorder.^6,7 Anxiety disorders,⁸ MDD,⁹ bipolar disorders,¹⁰ and personality disorders (mainly antisocial personality disorder)⁹ are all associated with substance dependence. Co-occurring substance dependence and mental disorders interact and influence the other’s course.¹¹ Mental disorders often lead to increased severity of SUDs.¹² There is also heterogeneity among co-occurring disorders, according to the different combinations of mental disorders and substance dependence (for example, depression and alcohol, severe mental disorders, and marijuana), as well as among patients’ demographic, socioeconomic, or service use characteristics.^11,13

Younger people are more likely to abuse substances, whether they have co-occurring disorders or not.^14,15 People with an SUD are less likely to have a post-secondary education¹⁶ or high household incomes.¹⁵ However, people whose sole affliction is substance dependence have lower psychological distress than those affected by co-occurring disorders.¹⁶ Use of health care services for mental reasons is significantly higher among people with co-occurring disorders.^14,17 The latter, however, are more likely to be unsatisfied with the care received.^16,18–20

Clinical Implications

• The strongest predictor of substance dependence was stigmatization.

• People with new substance dependence had bad self-perceptions of their QoL and of their physical health.

• People with new substance dependence were more likely to have a negative perception of the physical state of their neighbourhoods than people without dependence.

Limitations

• Serious mental disorders, such as schizophrenia and personality disorders, were not included in our study.

• A large number of participants with substance dependence were lost or excluded from follow-up at time 2.

• Sample size was too low to show a high volume of both new cases of substance dependence and users of services with substance dependence.

Several studies found differences between males and females in terms of SUDs. Females are more likely to have mental disorders only, while males are more likely to have substance dependence only or substance dependence with co-occurring disorders.¹⁴ Females with substance dependence are more likely to have co-occurring anxiety and mood disorders, whereas males are more likely to be affected by antisocial disorders.²¹ Psychological distress is more frequent among females with substance dependence, while dependence among males is more commonly associated with antisocial behaviour.¹¹ Males tend to be more often dependent on alcohol, marijuana, and hallucinogens, while females are more often dependent on amphetamines.¹¹

Most cross-sectional studies analyze variables associated with a specific substance dependence (for example, alcohol or cocaine). To our knowledge, none of those cross-sectional studies looked at variables associated with substance dependence in general (including alcohol and common illicit drugs, such as marijuana, cocaine, speed, and ecstasy), or at predictors of substance dependence (new cases at follow-up in longitudinal studies). Moreover, these studies considered only clinical variables and some specific demographic variables, such as sex and age. Several compelling variables, such as neighbourhood characteristics (criminality or physical state) and health care service use, are usually neglected.

Longitudinal analysis offers a stronger method than cross-sectional analyses for examining the relation between substance dependence and associated variables. Based on a longitudinal study and a comprehensive framework (Figure 1), our study aimed to identify sociodemographic, socioeconomic, clinical, life event, health, life perception, neighbourhood, and service use characteristics that distinguish new substance dependence cases from the general population; and to determine predictors of substance dependence over a 2-year period. Also assessed were variables that differentiate people dependent on substances according to sex and age.

Figure 1.

Comprehensive framework: predictors of substance (alcohol and drug) dependence

Methods

Study Design and Setting

Our study was based on an ECA in Montreal, Canada’s second-largest city, with a population of 3.6 million. The ECA had a population of 269 720 within 4 neighbourhoods, whose population ranged from 29 680 to 72 420. It included various health care and mental health care services. The latter services and the socioeconomic characteristics of the area are described in detail in other publications.^22,23 Substance use services are provided mostly outside the ECA at the regional level. The Montreal area has 2 public addiction rehabilitation centres offering specialized care (1 each for the French and English communities) and numerous private and voluntary groups, such as Alcoholics Anonymous and Narcotics Anonymous. However, there are few specialized services for co-occurring disorders—as is mostly the case everywhere else in the world.²⁴

Selection Criteria and Survey Sample

To be included in the survey, participants had to be aged between 15 and 65 and reside in the study ECA. The objective was to obtain a representative sample in geographical terms, that is, recruiting participants from all parts and in proportion to the population density, as well as in terms of socioeconomic status. Two data collections were conducted at a 2-year interval by specially trained interviewers.

At baseline (time 1: June 2007 to December 2008), 2434 people took part in the survey. All were contacted for a second interview (time 2) between June 2009 and December 2010. Only 611 were lost or excluded from follow-up because they had moved out of the catchment area or had died, and 1823 responded to time 2, for a retention rate of 74.9%. The attrition rate at time 2 (25.1%) included only 138 (5.7%) refusals to participate; 230 people (9.4%) had moved outside the catchment area, 231 (9.4%) were not reachable, and 12 (0.5%) had died. This attrition rate after 2 years was better than those observed in American ECAs after 1 year (20.4%, including 12.6% refusals).²³ The attrition rates were higher among youths, singles, the less-educated, low-income earners, and people with substance dependence. This is similar to the findings of other ECA studies.^25–27 Relevant ethics boards approved the research. The sampling strategy and data collection (especially at time 1) are described in detail in related publications.^22,23

Instruments

Many instruments were used to assess health, psychosocial, and service use parameters. Sociodemographic and economic data were collected using the CCHS 1.2.²⁸ Mental health diagnostics were based on the CIDI, an instrument created by a Word Health Organization working group. CIDI diagnoses, based on the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders, included anxiety disorders (that is, agoraphobia, social phobia, panic disorder, and PTSD), and mood disorders (MDD and mania). Several large-scale community epidemiologic surveys throughout the world have used the CIDI since its development in 1990.^17,29,30 SUDs (alcohol and drugs) were based on the CIDI Short Form, and other questions related to the World Mental Health CIDI and interference in life and activities, and on Canada’s Alcohol and Other Drugs Survey, respectively. Psychological distress was rated using the 10-question Kessler Psychological Distress Scale based on 5-point Likert scales.³¹ Impulsiveness was evaluated with the Barratt Impulsivity Scale, which assesses 30 items on a 4-point Likert scale.³² Social support was measured with the Social Provisions Scale, comprising 24 items in 6 areas (for example, emotional support and social integration) based on a 4-point-Likert scale.³³ QoL scores were determined with the Satisfaction with Life Domains Scale, using 7-point Likert scales to rate 20 items under 5 categories (for example, daily living, social relationships, and living environment).³⁴ Stigmatization was evaluated using the Devaluation Discrimination Scale, comprising 12 questions with a 6-point Likert scale.³⁵ Self-reported aggressive behaviour was assessed using the Modified Overt Aggression Scale, which rates 4 types of aggressive behaviour (verbal violence, physical violence, damage to property, and self-inflicted injuries) and their severity.³⁶ The physical state of neighbourhoods was evaluated using the Physical Conditions of the Neighbourhood Scale,³⁷ which measures 7 items, and criminality, using the Neighbourhood Disorder Scale, which measures 9 items.³⁸ Finally, the questionnaire on mental health care service use was adapted from the CCHS 1.2. Participants with mental disorders or substance dependence were asked which services (hospitalizations, rehabilitation centres, support, or help groups) and professionals (psychiatrist, family physician, psychologist, addictions counsellor, or other) they used, and how often. All these instruments were validated among the French-speaking population.

Analysis

Analyses were conducted following the framework displayed in Figure 1, which includes substance dependence predictors. Analyses encompassed univariate, bivariate, and multivariate statistics. Univariate statistics comprised frequency distributions for categorical variables and mean values along with standard deviations for continuous variables. Bivariate analyses were used first for comparison analyses (at time 2, participants with, compared with without, substance dependence; males, compared with females, with substance dependence; participants with substance dependence aged 29 years and younger, compared with 30- to 49-year-olds, and aged 50 and older). For these comparisons, dependent variables were categorical. A chi-square test was applied when independent variables were categorical, and a Student t test, when independent variables were continuous. Bivariate analyses were also conducted in simple logistic regression analyses between independent variables and the dichotomous dependent variable “Having any substance dependence at time 2 only” with the alpha value set at P < 0.10. Variables that yielded significant association in bivariate analyses were introduced in the multiple logistic regression model through the backward elimination technique (alpha value set at P < 0.05). The goodness-of-fit of the model was estimated using the Hosmer–Lemeshow statistic, as was the proportion of the variance explained using Nagelkerke’s R².

Results

Figure 2 shows the flow chart of participants with substance dependence in the previous 12 months from the full study sample, from time 1 (2434) to time 2 (1823). The prevalence of substance dependence in the prior 12 months was 5.9% at time 1 and 3.7% at time 2. The number of persistent cases of substance dependence was 28 (1.5% of the total sample), or 41.7% of participants with substance dependence at time 2. The subsample with substance dependence at time 2 represents only 2.2% of the total sample.

Figure 2.

Flow chart of participants regarding substance (drug and alcohol) dependence in prior 12 months

To assess predictors of substance dependence among this cohort, 82 participants with substance dependence at time 1 were excluded from analyses, yielding a subsample of 1741 participants, who were reached at both times 1 and 2. At the end of the follow-up period, 39 new cases of substance dependence were reported.

Table 1 displays frequency distribution of the total sample of the 1741 participants at baseline (time 1), along with comparison analyses between the 39 participants with any substance dependence at time 2 only and the remaining 1702 people. The participant profile described for both groups is related to variables at baseline.

Table 1.

Characteristics of participants (total sample) with and without substance dependence (alcohol or illicit drugs)

Characteristic	Participants with new cases of dependence, compared with without dependence
	Total sample n = 1741	With dependence at time 2 only n = 39	Without dependence n = 1702	P
Sociodemographic variables
Sex, n (%)				0.001a
Female	1106 (63.5)	15 (38.5)	1091 (64.1)
Male	635 (36.5)	24 (61.5)	611 (35.9)
Age, mean (SD)	42.9 (13.1)	36.8 (11.7)	43.0 (13.1)	0.004b
Socioeconomic variables
Household income, mean (SD)	54447.9 (54017.2)	43025.3 (46245.8)	54709.6 (54165.6)	0.18b
Principal source of income, n (%)				0.22c
Salary	920 (52.8)	19 (48.7)	901 (52.9)
Social welfare	107 (6.1)	4 (10.3)	103 (6.1)
Rent or pension benefits	115 (6.6)	0 (0)	115 (6.8)
Other	599 (34.4)	16 (41.0)	583 (34.3)
Social support score, mean (SD)	81.2 (9.4)	78.5 (9.8)	81.3 (9.3)	0.07b
Clinical variables
Mental health disorders in prior 12 months, n (%)
MDD	137 (7.9)	7 (17.9)	130 (7.6)	0.02a
Mania	21 (1.2)	1 (2.6)	20 (1.2)	0.43c
Panic disorder	28 (1.6)	1 (2.6)	27 (1.6)	0.47c
Social phobia	53 (3.0)	3 (7.7)	50 (2.9)	0.11c
Agoraphobia	23 (1.3)	0 (0.0)	23 (1.4)	0.10c
PTSD	12 (0.7)	0 (0.0)	12 (0.7)	0.10c
Psychological distress, mean (SD)	7.9 (6.3)	10.7 (7.8)	7.8 (6.3)	0.005b
Impulsiveness score, mean (SD)	56.2 (10.8)	65.2 (12.1)	56.0 (10.6)	<0.001b
Number of mental health disorders in prior 12 months excluding dependence, mean (SD)	0.2 (0.5)	0.3 (0.6)	0.2 (0.5)	0.045b
Life events, n (%)
Lifetime victimhood of violence	535 (30.7)	14 (35.9)	521 (30.6)	0.48a
Lifetime history of aggressive behaviour	486 (27.9)	14 (35.9)	472 (27.7)	0.26a
Victim of violence in previous 12 months	222 (12.8)	3 (7.7)	80 (4.7)	0.39c
Aggressive behaviour in previous 12 months	83 (4.8)	6 (15.4)	217 (12.7)	0.63a
Life and health perception
QoL, mean (SD)	109.5 (15.8)	100.5 (16.1)	109.7 (15.8)	<0.001b
Stigmatization score ≥2.5, n (%)	1386 (79.6)	37 (94.9)	1349 (79.3)	0.02a
Perception of one’s physical health, n (%)				0.02a
Excellent or very good health	808 (46.4)	11 (30.8)	796 (46.8)
Good	640 (36.8)	16 (41.0)	624 (36.7)
Fair or poor	293 (16.8)	11 (28.2)	282 (16.6)
Neighbourhood characteristics
Physical state of the neighbourhood, mean (SD)	37.0 (20.7)	45.8 (23.3)	36.8 (20.6)	0.007b
Criminality in the neighbourhood, mean (SD)	3.7 (1.3)	3.8 (1.4)	3.7 (1.3)	0.46b

^aPearson chi-square test;

^bIndependent sample t test;

^cExact Fisher test

Female participants made up most of the total sample but were the minority among people with substance dependence. Most participants in the total sample reported salary as their main income. The most prevalent mental health disorder was severe depression (8%). Almost one-half of the participants felt their physical health was good or excellent. Finally, 21% had a negative perception of the physical state of their neighbourhood.

Compared with the total sample, participants with substance dependence (n =39) were remarkable for their lower mean age. In terms of socioeconomic variables, they had marginally lower social support. Regarding clinical variables, those same participants showed significantly higher levels of MDD, and higher psychological distress and impulsiveness score. They also had more mental health disorders in the prior 12 months. Concerning life and health perception, QoL was lower among participants with substance dependence, as were the numbers of users perceiving their physical health as excellent or very good were poorer. The level of stigmatization was also higher among new cases of substance dependence. In terms of neighbourhood characteristics, participants with substance dependence tended to live in areas in better physical shape. Finally, only 1 of the 39 new cases of substance dependence used health care services for mental reasons at time 1, compared with 129 participants without substance dependence (not shown in the table).

Variables described in Table 1 were used in bivariate analyses to determine substance dependence predictors. The multiple logistic regression model displayed on Table 2 yielded 4 variables separately and significantly associated with substance dependence. Three were positively associated: sex (male), stigmatization score of 2.5 or more, and impulsiveness score. The fourth predictor, age, was negatively associated. This model explained 16% of the total variance, with an acceptable goodness-of-fit.

Table 2.

Predictors of drug and alcohol dependence (reference category: participants without dependence; index category: 39 participants with dependence at time 2 only) (n = 1741)^a

Predictor	β	Standard error	Wald	df	P	OR	95% CI
Sex, male	1.067	0.342	9.733	1	0.002	2.906	1.487–5.680
Self-perception of mental health	0.257	0.170	2.285	1	0.13	1.293	0.927–1.803
Age	−0.038	0.013	8.383	1	0.004	0.963	0.939–0.988
Household size	−0.166	0.104	2.512	1	0.11	0.847	0.691–1.040
Stigmatization score ≥2.5	1.661	0.749	4.923	1	0.03	5.267	1.214–22.855
Impulsiveness score	0.061	0.014	17.855	1	<0.001	1.063	1.033–1.094

R² = 16.1%; Hosmer–Lemeshow test: χ² = 6.069; df = 1, P = 0.64

^aVariables listed in Table 1 were used as independent variables to build this model using the backward elimination technique.

Additional analyses, not shown in tables, were computed to compare participants with substance dependence according to sex and age categories distribution, and in relation to variables identified in Figure 1. Regarding sex, only 1 variable was significant: females with substance dependence were more likely to suffer from social phobia than males (Fisher exact test, P = 0.05). However, males were marginally more likely to be victims of stigmatization (Student t test: P = 0.07). Two age category variables were significant: older participants (those 50 and older) with substance dependence had significantly lower income than participants aged 30 to 49 and those 29 and younger (Student t test: those 29 and younger, compared with those 50 and older, P = 0.03; 30 to 49 years old, compared with those 50 and older, P = 0.03). An inverse trend was observed regarding social support: younger participants with substance dependence had stronger social support than older participants (Student t test: those 29 and younger, compared with 30 to 49 years old, P = 0.01; those 29 and younger, compared with those 50 and older, P < 0.001; 30 to 49 years old, compared with those 50 and older, P = 0.03).

Discussion

At 5.9%, the 12-month prevalence of substance dependence at time 1 was higher than the Canadian level of 3% identified in the CCHS 1.2.⁴ This could be explained by the presence of a psychiatric hospital in the catchment area. At time 2, the 12-month prevalence (3.7%) was nearer to that of the CCHS 1.2, which could be the result of remission or people dropping out of the study at time 2. According to the literature, people with substance dependence often change residences, and are more likely to be jailed or hospitalized and to die prematurely.^25,27 Nonetheless, persistent substance dependence constituted 42% of the total substance dependence at time 2. International studies have revealed that almost one-half of people treated in specialized centres have a persistent substance dependence profile.^39,40

Using the multiple logistic regression model, stigmatization emerges as the major predictor of substance dependence. Several studies show that people with substance dependence, with or without co-occurring disorders, are stigmatized and discriminated against.^41–43 They are often seen as immoral or dangerous.⁴⁴ Some authors have concluded that stigmatization inflicts greater psychological pain than the mental disorder or addiction itself^45,46 and poses a key barrier to rehabilitation.⁴⁷ Sex has also been repeatedly reported as a strong predictor of substance dependence. Males are more likely than females to have substance dependence, mainly using alcohol and marijuana.^11,14 Previous studies^48,49 also found that co-occurring disorders were more prevalent among young people. It is logical that substance dependence would occur during young adulthood, considering that adolescence is usually the period of initiation to drugs.^50,51 Generally, mental disorders also appear during young adulthood.⁵² Impulsiveness is often a feature of co-occurring disorders,⁵³ especially personality disorders.⁵⁴ It is also a significant predictor of cocaine,⁵⁵ opioid,⁵⁶ and alcohol abuse or dependence.⁵⁷ Finally, note that no psychiatric diagnosis measured in our study was a predictor of substance dependence according to the multiple logistic regression model. An explanation could be that substance dependence precedes the appearance of a possible mental disorder—but other longitudinal studies are needed to validate this hypothesis.⁵⁸

However, in the bivariate analysis distinguishing recent substance dependence cases from nondependents, there is an association between substance dependence, MDD, and a greater number of mental health disorders. According to the literature, the presence of co-occurring mental disorders—mainly mood disorders—among the population with substance dependence is to be expected.^{5,6,8,10–12} The link between psychological distress and substance dependence is a little unexpected, considering that people with new dependence on substances were mostly males. Usually, psychological distress affects predominantly females with substance dependence.¹¹ The high level of psychological distress and meagre QoL among people with substance dependence could be the result of self-perceived poor physical health and inadequate social support. Finally, people were more likely to have a negative perception of the physical state of their neighbourhood, possibly because the high cost of drugs or alcohol forces them to seek cheap housing. People with substance dependence tend to live close to drug dealers for easy access to illicit drugs.⁵⁹

In our study, only 2.5% of participants newly diagnosed with substance dependence used health care services for mental reasons in the past 12 months. According to the literature, 10 years or more can pass before a person with an SUD seeks help for their addiction.⁶⁰ Lower motivation, negative attitudes regarding treatment, or inadequate services could explain the underuse of health care services or support groups by people with substance dependence.¹⁴ According to previous studies,^14,61 the main reason cited by people for not seeking help for their alcohol dependence was that they were not ready to stop drinking. People with substance dependence rarely use health care services until their problem requires intensive treatment.^62–64 Epidemiologic studies have shown that few people with substance dependence use health care services. For example, according to the CCHS 1.2, only 14% of people dependent on substances (who are not necessarily new cases, as in our study) sought help for their disorder.⁶⁵

In terms of sex, note that the only diagnosis that significantly differentiated males and females with substance dependence was social phobia. According to the literature, substance dependence is more likely to be associated with anxiety and mood disorders among females.²¹ Our study might have arrived at a different conclusion as it included both alcohol and all other drugs in substance dependence. Some substances have a more direct connection with a specific mental disorder, for example, panic disorder and alcohol dependence.⁶⁶ The small number of participants with a specific mental disorder (for example, panic disorder or PTSD) could also explain the lack of any significant link. Further, our study did not include serious mental disorders generally found in males, such as schizophrenia,^67,68 and personality disorders⁶⁹ (particularly antisocial disorder⁹) that have a significant link with substance dependence. Social phobia is more common among females,⁵² and it could be that young females suffering from this anxiety disorder at adolescence would be more susceptible to peer pressure, and thus more at risk for developing an substance dependence.⁷⁰ Males with substance dependence were more likely to be stigmatized than previously reported.⁴⁴ Our study also hints at a tendency toward a higher ratio of stigmatization among females. Some cultures are more intolerant of women who abuse alcohol or drugs.⁷¹

Finally, in terms of age categories, participants aged 50 and older with substance dependence were more likely to have a lower household income and weaker social support than younger participants. Possible explanations for the low income would be that people 50 and older were more often single, divorced, or widowed; had no salary; or suffered from physical diseases. Further, several studies have shown lower social support for older people, which explains why they are not put in touch more often with health care services.^72,73

Limitations

Our study presents some limitations. First, it did not include the full spectrum of mental health disorders. Several studies have reported high prevalence of schizophrenia and personality disorders with substance dependence.^9,50,66,74 Second, a large number of participants, including some with substance dependence, were lost to follow-up at time 2. According to the literature, people with substance dependence often change their place of residence, which complicates follow-up.^25,27 However, retention was a key concern in most studies investigating long-term treatments or follow-up of such consumers. Our study was thus comparable with others in terms of retention rates.^75–77 Finally, the sample size was too low to establish a high volume of both substance dependence (new cases) and users of services for mental health including substance dependence, and especially at follow-up.

Conclusions

Our study was innovative in conducting a longitudinal analysis using sociodemographic, clinical, life perception, and service use variables (in a comprehensive framework) as possible predictors of substance dependence. It was also one of the few investigations (especially in Canada) based on a catchment area study with a representative sample of a population. Moreover, our study broke new ground by assessing variables that distinguish new people dependent on substances according to sex and age categories. Overall, our study found that stigmatization is the strongest predictor of substance dependence. A systematic effort is required to implement programs to reduce stigmatization related to substance dependence. Our study found that males with substance dependence are often victims of stigmatization. Further studies could be conducted to determine which subgroups of males are more likely to be stigmatized. Moreover, outreach and preventive programs are needed to overcome the reluctance of people newly diagnosed with substance dependence to use health care services. Another key issue is to provide wider access to specialized services for people with impulsiveness problems. Meanwhile, bivariate analyses have shown that participants with substance dependence were significantly distinct from the total sample in terms of QoL, perception of physical health, and neighbourhood conditions. Finally, among new cases of substance dependence, our study found significant differences between males and females concerning social phobia, and between people 50 and older compared with younger ones concerning household income and social support. These latter findings suggest that more attention must be paid to people who are socially isolated or living in disadvantaged neighbourhoods.

Abbreviations

CCHS 1.2

Canadian Community Health Survey: Mental Health and Well-Being

CIDI

Composite International Diagnostic Interview

ECA

epidemiologic catchment area

MDD

major depressive disorder

PTSD

posttraumatic stress disorder

QoL

quality of life

SUD

substance use disorder