Alcohol Use Disorder and the Persistence/Recurrence of Major Depression: Le trouble de l’usage de l’alcool et la persistance/récurrence de la dépression majeure

Abstract

French

Objective

This study aims to determine the role of alcohol use disorder and other potential risk factors on persistence/recurrence of major depression in a Canadian population sample.

Methods

Data were drawn from the National Population Health Survey (1994/1995 to 2010/2011), a prospective epidemiologic survey of individuals 12 years and older, living in 10 Canadian provinces (N = 17,276). Participants were reinterviewed every 2 years for 9 cycles. This study population was a cohort of individuals who at baseline met the diagnosis of a major depressive episode (MDE) in the previous 12 months (n = 908). After the 6-year (cycle 4) and 16-year (cycle 9) follow-up period, 124 of 718 participants and 79 of 461 participants met the criteria for MDE, respectively. Persistence or recurrence of major depression was defined as meeting a diagnosis of MDE after 6 years and 16 years. Modified Poisson regression models were used to assess the role of alcohol dependence and other risk factors on the persistence/recurrence of major depression using Stata 14.

Results

Alcohol use disorder was significantly correlated with a 6-year (odds ratio [OR]: 3.03; 95% confidence interval [CI], 1.68 to 5.48; P < .0001) and 16-year (OR, 3.17; 95% CI, 1.15 to 8.77, P = 0.003) persistence/recurrence of major depression. Other factors associated with the persistence/recurrence of major depression include female sex, childhood traumatic events, chronic pain restricting activities, daily smoking, and low self-esteem.

Conclusions

Comorbid alcohol use disorder was found to be a strong risk factor for the persistence or recurrence of major depression.

Keywords

comorbidity major depression alcohol dependence alcohol use disorder

Introduction

Alcohol use disorder (AUD) and major depression are prevalent and frequently co-occur.^1

–4 Studies have shown that with depression, the risk of being addicted to alcohol is higher than in the general population,^1,5 and depression is more prevalent with alcohol dependence.^4,6 For example, the Collaborative Study on the Genetics of Alcoholism found a lifetime prevalence of major depressive disorder to be 42.2% among treatment-seeking individuals diagnosed with AUD, while the prevalence prior to alcohol dependence was 5.3% and prevalence occurring outside of the context of alcohol dependence was 11.5%.⁷

The major clinical problems associated with depression resulting in significant personal and public health consequences are its persistence and/or recurrence.^8,9 This persistence or recurrence is one of the reasons for depression’s considerable burden.^9

–12 Persistence occurs when there is a prolonged recovery time from a major depressive episode (MDE), while recurrence is the manifestation of a new MDE after a sustained period of remission.¹³ Previous studies showed that about half of those who recover from a first MDE will have at least one additional event in their lifetime and nearly 80% of those with a history of two events will have another recurrence.^9,14 Significantly, a population-based Canadian study using the same data set as our study found that 55.3% of the participants reported repeated episodes of major depression in the following 6 years.¹⁵

Previous studies suggest there may be differences in risk factors, neurobiological basis, and therapeutic response between major depression and recurrent or persistent major depression.^9,16,17 Using the “dynamic stress-vulnerability model,”¹⁸ Spijker et al. classified the many risk factors for persistent major depression identified in the literature¹⁹ as age,^20,21 female gender,^20,22 low socioeconomic status,²³ psychiatric illness and comorbidity,^23,24 and previous MDE episode.²⁰

Studies have been inconsistent in showing the relationship between AUD and persistent or recurrent depression. While some researchers found an association between substance use disorders and recurrent depression, suggesting a resolution of depressive symptoms and consequent loss of depressive diagnosis with abstinence from alcohol,^25
–27 others have not.^19,28,29 In Canada, there is a paucity of longitudinal data in mental health research. A significant number of studies come from cross-sectional data such as the Canadian Community Health Survey: Mental Health and Well-Being data. However, few studies have investigated the mental health research potential of longitudinal cohort studies by utilizing the nationally representative National Population Health Survey (NPHS) database.³⁰ In light of this, we explored the NPHS data to provide generalizable patterns of the recurrence/persistence of mental health and substance use disorders in Canada. This study aimed to determine whether concurrent AUD predicts persistence or recurrence of MDE after 6 years and 16 years of follow-up in a Canadian population sample. Furthermore, our study examined other factors that predict persistent or recurrent MDE after 6 years and 16 years of follow-up in this cohort.

Methods

Data Source and Study Cohort

Data came from the longitudinal NPHS (1994/1995 to 2010/2011), a nationally representative prospective epidemiologic survey of individuals 12 years and older, living in the 10 Canadian provinces in 1994/1995. Criteria for exclusion from the survey were living in certain remote areas, institutions, and reserves. In addition, full-time members of the Canadian armed forces were not included. The survey sample comprised of 17,276 participants who were reinterviewed every 2 years for 9 cycles (cycles 1 to 9). Detailed information on the NPHS methodology is available on the Statistics Canada web page.³¹ In accordance with Statistics Canada rules, controlled access to the NPHS Master data file took place at the Saskatchewan Research Data Centre.

This secondary analysis of the NPHS data included only individuals 18 years and older who met the criteria for a diagnosis of an MDE in the previous 12 months at baseline (n = 908; Figure 1). The age restriction was based on the structured instrument for the diagnosis of MDE, which was only administered to individuals 18 years and older. This study focused on cycles 4 (6 years from baseline) and 9 (16 years from baseline). The original follow-up period for the NPHS ended in cycle 9. Cycle 4 which had a reasonable time interval from baseline (6 years) and a good response rate was used in this analysis due to a significant loss to follow-up in cycle 9 (Figure 2). The estimated response rates reported by Statistics Canada were 84.9% and 69.7% for cycles 4 and 9, respectively.³¹

Figure 1.

Cohort sample derivation. *Excluded from the analysis. NPHS = National Population Health Survey.

Figure 2.

Follow-up chart. NPHS = National Population Health Survey.

Measures

Major depressive episode

The diagnosis of an MDE in the past 12 months was assessed at baseline (1994/1995) using the World Health Organization Composite International Diagnostic Interview Short Form (CIDI-SF) based on the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised (DSM-III-R) criteria.³² The past year’s MDE was evaluated on the same cohort every 2 years. The 6-year and 16-year persistence/recurrence of major depression were defined as meeting a diagnosis of MDE in the past 12 months at cycle 4 (2000/2001) and at cycle 9 (2010/2011) follow-up periods, respectively. No distinction is or could be made between persistence and recurrence major depression during the follow-up periods because the study time frame and data did not allow for such accuracy.

Alcohol use disorder

AUD was assessed in cycle 2 (1996/1997) as alcohol dependence using the World Health Organization CIDI-SF based on DSM-III-R criteria.³² Alcohol dependence was not assessed at baseline.

Although the term “alcohol use disorder” better reflects the more recent iteration of the Diagnostic and Statistical Manual of Mental Disorders, the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), the term “alcohol dependence” is used in this article in circumstances where appropriate to flag the use of the DSM-III-R criteria in the data. Studies assessing the differences in the DSM-III-R versus the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) and DSM-5 criteria for problematic alcohol use suggest that the transition from DSM-III-R to DSM-IV saw modest increasing rates of abuse reported in epidemiological studies but decreasing rates of dependence. DSM-5 criteria that set aside the abuse/dependence dichotomy may result in an inflation of the prevalence rates of AUD.^33

–36

Other measures

The covariates were determined using Spijker et al.’s classification of risk factors for persistent depression based on the “dynamic stress-vulnerability model.”^18,19 Demographic factors (age, gender) and social vulnerability factors (education, marital status, dwelling owned—used as a proxy for socioeconomic status). The following are the psychobiological vulnerability factors measured: childhood and adult stress index—this measured the number of traumatic events respondents have been exposed to during their childhood, adolescence, or adulthood. Higher values indicate more stressors/traumatic events respondent has been exposed to in childhood/adolescence³²; presence of a physical chronic illness; Mastery index; and Self-esteem index scores. Higher scores indicate better mastery or self-esteem.³² Sustaining factors measured were perceived social support, negative life events score (which included physical abuse, unwanted pregnancy, abortion or miscarriage, major financial difficulties, and serious problems at work or in school), and family stress index (a measure of ongoing difficulties/stressors such as activity overload, financial difficulties, and problems with relationships in day-to-day encounters).³²

Statistical Analyses

All statistical analyses were conducted using STATA version 14.0. First, we described the participant’s sociodemographic characteristics and risk factors for persistent or recurrent major depression in both the baseline and follow-up cohorts. Second, modified Poisson models were used to fit the data. The modified Poisson model is an effective way of estimating the relative risk (RR) of binary data in a cohort.³⁷

The RR for the persistence or recurrence of MDE, the main outcome of interest, was assessed for 6 years (model 1) and 16 years (model 2) of follow-up. The modeling was carried out in two stages for each model. First, all covariates were assessed independently as potential predictors of persistent or recurrent major depression. In stage 2, covariates that were independent risk factors or met a significance level of P < 0.25 were included in a multivariate model. The multivariate modified Poisson models were built using stepwise analyses with backward elimination of covariates. Covariates that were eliminated were assessed for confounding and re-entered into the model if found to confound the RR for alcohol dependence. Plausible interactions for gender and alcohol dependence, smoking, and alcohol dependence were tested for in each model but were not significant. Deviance and Pearson goodness-of-fit were tested on each model to justify the use of Poisson models. The postestimation command “fitstat” in Stata 14 was used to compute other measures of fit and to compare models with the interaction term and models without the interaction term.

Results

Participants’ Characteristics

At baseline, almost half of the cohort sample were aged between 25 and 44 years (48.7%); one-third were male (31.7%); almost half were married or common-law with a partner (42.7%); and 69.4% or almost two-thirds self-identified an existing chronic condition (Table 1). After 6 years (cycle 4) and 16 years (cycle 9), approximately 20% and 49% of participants were lost to follow-up, respectively (Figure 2).

Table 1.

Participant Characteristics of the NPHS and Cohort Sample at Baseline.

	Baseline Cohort Sample Frequency (%), N = 908	Baseline NPHS Sample Frequency (%), N = 14,713
Age
18 to 24 years	149 (16.4)	2258 (15.4)
25 to 44 years	442 (48.7)	5828 (39.6)
45 to 64 years	231 (25.4)	3887 (26.4)
65 years and above	86 (9.5)	2740 (18.6)
Sex
Male	288 (31.7)	6733 (45.8)
Female	620 (68.3)	7980 (54.2)
Marital status
Married	325 (35.8)	7254 (49.3)
Common-law/with a partner	63 (6.9)	968 (6.6)
Widowed	80 (8.8)	1380 (9.4)
Divorced or separated	173 (19.1)	1445 (9.8)
Single/never married	267 (29.4)	3662 (24.9)
Highest level of education
<Secondary	239 (26.3)	4721 (32.2)
Secondary grad	137 (15.1)	2184 (14.9)
Some postsecondary	269 (29.6)	3545 (24.2)
Postsecondary grad	263 (29.0)	4224 (28.8)
Chronic condition
No	279 (30.7)	6215 (42.3)
Yes	629 (69.3)	8475 (57.7)
Total household income (CDN$)
<15,000	220 (24.2)	2646 (18.9)
15,000-29,999	247 (27.2)	3572 (25.5)
30,000-49,999	210 (23.1)	3817 (27.2)
50,000-79,999	160 (17.6)	2762 (19.7)
80,000 or more	71 (7.8)	1241 (8.8)
Dwelling owned by a household member
No	387 (42.6)	4592 (31.2)
Yes	521 (57.4)	10,115 (68.8)
Self-esteem index
Self-esteem index <18	299 (32.9)	1826 (13.5)
Self-esteem index = 18	190 (20.9)	3506 (25.8)
Self-esteem index = 19 to 22	266 (29.3)	4434 (32.7)
Self-esteem index = 23 to 24	153 (16.9)	3804 (28.0)
Mastery index
Mastery index <17	407 (44.8)	2960 (21.9)
Mastery index = 17 to 19	212 (23.4)	3421 (25.3)
Mastery index = 20 to 21	157 (17.3)	3396 (25.2)
Mastery index = 22 to 28	132 (14.5)	3728 (27.6)
Type of smoker
Never smoked	256 (28.2)	5286 (37.1)
Daily	374 (41.2)	3851 (27.0)
Former daily/occasionally but former daily	201 (22.1)	3781 (26.5)
Always/former occasional	77 (8.5)	1331 (9.3)
Recent life events scores
No recent life events	397 (43.7)	8738 (66.7)
<3 recent life events	400 (44.1)	3866 (29.5)
3 or more recent life events	111 (12.2)	488 (3.7)
Childhood and adult stress index
No traumatic events	260 (28.6)	6696 (51.2)
<3 traumatic events	414 (45.6)	5060 (38.7)
3 or more traumatic events	234 (25.8)	1329 (10.2)
Perceived social support index
No social support	38 (4.2)	133 (1.0)
Minimal social support	37 (4.1)	256 (1.9)
Some social support	162 (17.8)	1914 (14.2)
Adequate social support	671 (73.9)	11,234 (83.0)
Number of weeks felt depressed
Did not feel depressed	X	12,341 (91.1)
Felt depressed for <6 weeks	418 (46.0)	573 (4.2)
Felt depressed for 6 to 52 weeks	490 (54.0)	626 (4.6)
Family stress index
No stress	584 (64.3)	9983 (76.3)
Some stress	267 (29.4)	2752 (21.0)
Stress overload	57 (6.3)	356 (2.7)
Activities prevented by pain
No pain or discomfort	618 (68.1)	11,647 (82.1)
Pain does not prevent activities	51 (5.6)	697 (4.9)
Pain prevents few activities	81 (8.9)	813 (5.7)
Pain prevents some activities	76 (8.4)	552 (3.9)
Pain prevents most activities	82 (9.0)	478 (3.4)
Alcohol dependence (assessed in cycle 2, N = 896)
No	867 (96.8)
Yes	29 (3.2)

Note. NPHS = National Population Health Survey; <Secondary = less than secondary education; Secondary grad = secondary education graduate; Some postsec = some postsecondary education; Postsecondary grad = postsecondary education graduate.

X = All participants in the cohort were depressed at baseline.

Six-Year Persistence or Recurrence of Major Depression

After 6 years (cycle 4), 718 participants of the cohort sample of 908 participants meeting a diagnosis of MDE in the past 12 months at baseline remained in the study (Figure 2). Of these 718 participants, 124 still met the criteria for MDE in the past 12 months at this time. This accounted for 17.3% of the participants who were still in the study at cycle 4. Overall, 13.7% of the baseline cohort reported MDE symptoms after 6 years. The presence of alcohol dependence gave a 3-fold increase in the risk of having persistent or recurrent depression (Table 2). Other factors associated with the risk of persistent or recurrent depression were being a female, a daily smoker, having pain that prevents activities, and experiences of 3 or more traumatic life events. A self-esteem index of 23 and above was correlated with a reduction in the risk of persistent or recurrent depression by 46%.

Table 2.

Alcohol Dependence and Other Risk Factors for 6-Year Persistent or Recurrent Major Depression.

	Crude RR (95% CI)	Adjusted^a
	Crude RR (95% CI)	RR (95% CI)	Statistic (P value)
Alcohol dependence
No	1	1
Yes	2.54 (1.46 to 4.41)**	3.03 (1.68 to 5.48)**
Age			0.06
18 to 24 years	1	1
25 to 44 years	1.33 (0.85 to 2.08)	1.37 (0.82 to 2.28)
45 to 59 years	0.79 (0.45 to 1.38)	0.84 (0.44 to 1.61)
60 years and above	0.59 (0.23 to 1.49)	0.97 (0.35 to 2.68)
Sex
Male	1	1
Female	1.39 (0.95 to 2.04)	1.53 (1.02 to 2.29)*
Marital status
Married	1	1
Common-law/with a partner	1.29 (0.69 to 2.40)	0.81 (0.41 to 1.59)
Widowed	.053 (0.22 to 1.28)	0.62 (0.24 to 1.56)
Divorced or separated	1.48 (0.98 to 2.23)	1.01 (0.65 to 1.56)
Single/never married	1.07 (0.71 to 1.61)	0.86 (0.53 to 1.38)
Highest level of education			0.90
<Secondary	1	X
Secondary grad	0.77 (0.44 to 1.35)
Some postsecondary	0.97 (0.64 to 1.46)
Postsecondary grad	0.83 (0.54 to 1.28)
Dwelling owned by a household member
No	1	1
Yes	0.61 (0.44 to 0.83)**	0.74 (0.53 to 1.03)
History of chronic condition
No	1	X
Yes	1.13 (0.79 to 1.61)
Self-esteem index			<0.0001
Self-esteem index <18	1	1
Self-esteem index = 18	0.56 (0.36 to 0.87)**	0.73 (0.47 to 1.12)
Self-esteem index = 19 to 22	0.49 (0.32 to 0.74)**	0.59 (0.40 to 0.89)**
Self-esteem index = 23 to 24	0.43 (0.25 to 0.74)**	0.54 (0.32 to 0.92)*
Mastery index
Mastery index <17	1	X
Mastery index = 17 to 19	0.56 (0.36 to 0.86)**
Mastery index = 20 to 21	0.64 (0.40 to 1.01)
Mastery index = 22 to 28	0.43 (0.23 to 0.77)**
Type of smoker			0.02
Never smoked	1	1
Daily	2.05 (1.34 to 3.12)**	1.58 (1.03 to 2.43)*
Former daily/occasionally but former daily	1.09 (0.62 to 1.91)	1.09 (0.63 to 1.87)
Always/former occasional	1.50 (0.78 to 2.88)	1.24 (0.63 to 2.44)
Recent life event scores			0.43
No recent life events	1	1
<3 recent life events	1.59 (1.11 to 2.27)*	1.19 (0.81 to 1.76)
3 or more recent life events	1.47 (0.88 to 2.47)	0.92 (0.54 to 1.57)
Childhood and adult stress index			0.03
No traumatic events	1	1
<3 traumatic events	1.68 (1.05 to 2.69)*	1.38 (0.87 to 2.19)
3 or more traumatic events	2.67 (1.66 to 4.28)**	1.72 (1.04 to 2.83)*
Perceived social support index
No social support	1	X
Minimal social support	0.78 (0.19 to 3.14)
Some social support	1.54 (0.59 to 4.02)
Adequate social support	1.06 (0.42 to 2.66)
Number of weeks felt depressed
Felt depressed for <6 weeks	1	X
Felt depressed for 6 to 52 weeks	1.25 (0.90 to 1.74)
Family stress index			0.68
No stress	1	1
Some stress	1.40 (1.01 to 1.96)*	1.08 (0.76 to 1.52)
Stress overload	1.11 (0.57 to 2.16)	0.73 (0.37 to 1.46)
Activities prevented by pain			0.03
No pain or discomfort	1	1
Pain does not prevent activities	1.33 (0.69 to 2.55)	0.95 (0.54 to 1.68)
Pain prevents few activities	1.19 (0.68 to 2.07)	0.96 (0.59 to 1.56)
Pain prevents some activities	1.35 (0.77 to 2.39)	1.14 (0.63 to 2.06)
Pain prevents most activities	2.13 (1.37 to 3.31)**	1.70 (1.06 to 2.71)*

Note. Significant values are marked in boldface. CI = confidence interval; RR, relative risk; <Secondary = less than secondary education; Secondary grad = secondary education graduate; Some postsec = some postsecondary education; Postsecondary grad = postsecondary education graduate; X = lost in multivariate analysis.

^a Adjusted in a multivariate model.

*P value ≤ 0.05. **P value < 0.001.

Sixteen-Year Persistence or Recurrence of Major Depression

After 16 years of follow-up (cycle 9), 461 (51%) participants of the cohort sample of 908 participants who met the criteria for a MDE in the past 12 months at baseline remained in the study (Figure 2). Of these 461 participants, 79 (17.14%) still met the criteria for MDE in the past 12 months at this time. Alcohol dependence remained a risk factor for the persistence/recurrence of major depression with a 3-fold increase in the risk compared to those who were not alcohol dependent (Table 3). There was a 2-fold increase in the associated risk of persistent or recurrent depression if participants had pain preventing activities compared to no pain or discomfort.

Table 3.

Alcohol Dependence and Other Risk Factors for 16-Year Persistent or Recurrent Major Depression.

	Crude RR (95% CI)	Adjusted^a
	Crude RR (95% CI)	RR (95% CI)	Statistic (P Value)
Alcohol dependence
No	1	1
Yes	2.41 (1.09 to 5.29)*	3.17 (1.15 to 8.77)*
Age			0.07
18 to 24 years	1	1
25 to 44 years	1.29 (0.74 to 2.24)	1.16 (0.64 to 2.10)
45 to 59 years	0.67 (0.32 to 1.42)	0.49 (0.22 to 1.12)
60 years and above	0.60 (0.18 to 1.95)	0.48 (0.13 to 1.73)
Sex
Male	1	1
Female	1.23 (0.76 to 1.97)	1.39 (0.85 to 2.27)
Marital status			0.003
Married	1	1
Common-law/with a partner	0.97 (0.41 to 2.29)	0.86 (0.31 to 2.37)
Widowed	0.47 (0.12 to 1.83)	0.63 (0.15 to 2.85)
Divorced or separated	0.62 (0.32 to 1.21)	0.38 (0.19 to 0.74)
Single/never married	1.22 (0.79 to 1.90)	1.10 (0.66 to 1.84)
Highest level of education
<Secondary	1	X
Secondary grad	0.82 (0.34 to 1.96)
Some postsecondary	1.52 (0.82 to 2.80)
Postsecondary grad	1.28 (0.69 to 2.39)
Dwelling owned by a household member
No	1	1
Yes	0.63 (0.42 to 0.94)*	0.69 (0.44 to 1.07)
History of chronic condition
No	1	1
Yes	1.57 (0.98 to 2.51)	1.32 (0.80 to 2.18)
Self-esteem index
Self-esteem index <18	1	X
Self-esteem index = 18	0.81 (0.46 to 1.41)
Self-esteem index = 19 to 22	0.69 (0.41 to 1.14)
Self-esteem index = 23 to 24	0.72 (0.40 to 1.30)
Mastery index			0.01
Mastery index <17	1	1
Mastery index = 17 to 19	0.48 (0.25 to 0.91)*	0.47 (0.25 to 0.90)*
Mastery index = 20 to 21	0.66 (0.37 to 1.16)	0.68 (0.38 to 1.25)
Mastery index = 22 to 28	0.73 (0.41 to 1.28)	0.82 (0.48 to 1.40)
Type of smoker			0.03
Never smoked	1	1
Daily	1.35 (0.83 to 2.20)	1.15 (0.70 to 1.89)
Former daily/occasionally but former daily	1.29 (0.74 to 2.24)	1.28 (0.74 to 2.22)
Always/former occasional	0.33 (0.08 to 1.35)	0.23 (0.06 to 0.95)*
Recent life event scores			0.51
No recent life events	1	1
<3 recent life events	1.24 (0.79 to 1.95)	1.09 (0.69 to 1.72)
3 or more recent life events	1.53 (0.86 to 2.72)	0.90 (0.49 to 1.66)
Childhood and adult stress index			0.58
No traumatic events	1	1
<3 traumatic events	1.18 (0.69 to 2.00)	1.30 (0.78 to 2.19)
3 or more traumatic events	1.86 (1.09 to 3.19)*	1.32 (0.74 to 2.37)
Perceived social support index			0.002
No social support	1	1
Minimal social support	0.38 (0.08 to 1.70)	0.27 (0.05 to 1.32)
Some social support	0.24 (0.08 to 0.74)*	0.20 (0.07 to 0.57)*
Adequate social support	0.64 (0.30 to 1.38)	0.52 (0.24 to 1.14)
Number of weeks felt depressed
Felt depressed for <6 weeks	1	X
Felt depressed for 6 to 52 weeks	1.12 (0.74 to 1.68)
Family stress index			0.21
No stress	1	1
Some stress	1.70 (1.13 to 2.56)**	1.27 (0.82 to 1.98)
Stress overload	1.00 (0.39 to 2.58)	0.53 (0.20 to 1.45)
Activities prevented by pain			0.001
No pain or discomfort	1	1
Pain does not prevent activities	1.02 (0.39 to 2.64)	0.63 (0.22 to 1.77)
Pain prevents few activities	2.70 (1.67 to 4.39)**	2.24 (1.39 to 3.63)**
Pain prevents some activities	2.19 (1.15 to 4.16)*	1.89 (0.94 to 3.78)
Pain prevents most activities	1.85 (0.88 to 3.89)	2.23 (1.08 to 4.62)*

Note. Significant values are marked in boldface. CI = confidence interval; RR = relative risk; <Secondary = less than secondary education; Secondary grad = secondary education graduate; Some postsec = some postsecondary education; Postsecondary grad = postsecondary education graduate; X = lost in multivariate analysis.

^a Adjusted in a multivariate model.

*P value ≤ 0.05.

**P value < 0.001.

Discussion

Alcohol Use Disorder

This study showed that the associated risk for persistence/recurrence of major depression was 3 times higher with concurrent alcohol dependence. This supports the evidence that comorbidity is one of the strongest risk factors for persistent/recurrent major depression.^9,25,26,38 This also supports the evidence of the adverse effects of substance use disorders on the course of severe mental illness.^39
–41 Dependence on alcohol could be associated with an increased risk of persistent or recurrent depression by increasing the likelihood of symptom exacerbation, decreased social functioning, and treatment nonadherence.^9,39,42

Our findings could be explained by the “disorder-inducing” theory, which believes that one disorder causes the other, and the “self-medication” theory where major depression is being treated with alcohol to suppress symptoms.^43
–45 Another possible explanation could be the “overlapping predisposition” theory which postulates a common underlying vulnerability to AUD and major depression. This common vulnerability may be genetic and/or environmental^46,47 with a possible predisposition of individuals with significant vulnerability to have persistent or recurrent MDEs.^9,48
–50

Regardless of which pathway might be involved in causing the comorbid occurrence of major depression and alcohol dependence, once it is developed, a vicious cycle where one disorder exacerbates or maintains the other may have been triggered.^51,52 In other words, if one were to treat one disorder without addressing the other disorder, the risk of persistence or recurrence would be very high,^48,53,54 resulting in increased health-care utilization and costs.⁵⁵

Other Risk Factors

Childhood and adult stress index, a measure of the number of traumatic events the respondents have been exposed to, was associated with an increased risk of persistent/recurrent major depression in our study. This finding substantiates previous findings.^28,56
–58 Traumatic events such as physical and sexual abuse increase the prevalence, risk, and chronicity of depression,^56,57 which might be responsible for our finding. Physiologic dysregulation which results from stressful events during development might also lead to the increased risk of persistence/recurrence of depression^59
–61 and possibly treatment resistance.^56
–58

Consistent with previous findings, pain, smoking, lack of social support, low mastery, or low self-esteem were associated with persistent/recurrent depression.^{9,21,62
–64} Social vulnerability factors such as education and socioeconomic status were not associated with persistence or recurrence, consistent with findings from the Netherlands¹⁹ but inconsistent with findings from the United States.^20,23,65 This may reflect the income inequality in these countries, with the United States having more income inequality than the Netherlands and Canada.⁶⁶

The risk of persistent/recurrent depression was increased in females after 6 years of follow-up, consistent with previous findings.^20,22 In contrast with previous studies,^19
–21 age was not a determinant of persistent/recurrent depression in our study. There was no association between the duration of the most recent episode (number of weeks depressed) and persistent depression. This was consistent with previous findings of no associated risk of recurrence with the duration of the index episode.^9,67 This finding could be attributed to the fact that the index episode may not be the first episode, and results may simply be reflecting treatment response of participants.⁶⁸

Strengths and Limitations

The fairly large sample size based on a nationally representative sample of the Canadian population makes this study generalizable to the general Canadian population except individuals living in remotes areas, institutionalized, and are full-time members of the Canadian armed forces. This is one of the strengths of this study. The longevity of the study and the consistency of the relationships are other strengths of the study. Another strong feature of our study is that the diagnoses of alcohol dependence and MDE were derived using structured diagnostic instruments based on DSM-III-R criteria. However the transition from DSM-III-R to DSM-IV to DSM-5 diagnostic criteria for alcohol abuse and dependence poses a challenge for generalizing our findings to today’s diagnostic environment. Several articles have looked at the changing criteria for alcohol abuse and dependence and their potential impact of epidemiological estimates.^33

–36 As earlier mentioned, it is suggested that the transition from DSM-III-R to DSM-IV saw modest increasing rates of abuse reported in epidemiological studies but decreasing rates of dependence. DSM-5 criteria that set aside the abuse/dependence dichotomy may result in an inflation of the prevalence rates of AUD. However, these changes in diagnostic criteria do not distract from the main finding of our research that alcohol dependence is a strong risk factor for persistent or recurrent depression. Another limitation of this study was the loss of participants to follow-up. The inability to distinguish persistent depression from recurrent depression was a limitation. In addition, the diagnosis of MDE and alcohol dependence was interview-based, potentially introducing recall bias, and the severity of the index MDE which is an important determinant of prognosis was not assessed in the data.

Conclusion

There is a significant overlap in the mechanisms leading to AUD and major depression. The adverse effects of AUD on the long-term course of major depression and the increased health-care utilization caused by this comorbidity is a public health concern. It is imperative to treat both disorders simultaneously to improve treatment outcomes and reduce health-care costs.

Footnotes

Authors’ Note

VO conceived and designed the study, and analyzed the data. VO, TOF, and CD contributed to materials/analysis tools; VO, TOF, and CD wrote and edited the article.

Acknowledgments

The analysis presented in this article was conducted at the Saskatchewan Research Data Centre (SKY-RDC) which is part of the Canadian Research Data Centre Network (CRDCN). The services and activities provided by SKY-RDC are made possible by the financial or in-kind support of the SSHRC, the CIHR, the CFI, Statistics Canada, and the University of Saskatchewan. The views expressed in this article do not necessarily represent the CRDCN’s or that of its partners.’

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Vivian N. Onaemo, MBBS, MPH, PhD

References

Brière

Rohde

Seeley

Daniel Klein

. Comorbidity between major depression and alcohol use disorder from adolescent to adulthood. Compr Psychiatry. 2014;55(3):526–533.

Grant

Dawson

Stinson

Chou

Dufour

Pickering

. The 12-month prevalence and trends in DSM-IV alcohol abuse and dependence: United States, 1991-1992 and 2001-2002. Drug Alcohol Depend. 2004;74(2):223–234.

Grant

Goldstein

Saha

, et al. Epidemiology of DSM-5 alcohol use disorder. Results from the national epidemiologic survey on alcohol and related conditions III. JAMA Psychiatry. 2015;72(8):757–766.

Britton

Stephens

, et al. Comorbid depression and alcohol use disorders and prospective risk for suicide attempt in the year following inpatient hospitalization. J Affect Disord. 2015;187:151–155.

Kessler

Nelson

McGonagle

Liu

Swartz

Blazer

. Comorbidity of DSM-III-R major depressive disorder in the general population: results from the US National comorbidity survey. Br J Psychiatry Suppl. 1996;(30):17–30.

Kessler

Crum

Warner

Nelson

Schulenberg

Anthony

. Lifetime co-occurrence of DSM-III-R alcohol abuse and dependence with other psychiatric disorders in the national comorbidity survey. Arch Gen Psychiatry. 1997;54:313–321.

Schuckit

Tipp

Bucholz

Nurnberger

JIJ

Hesselbrock

Crowe

. The lifetime rates of three major mood disorders and four major anxiety disorders in alcoholics and controls. Addiction. 1997;92(10):1289–1304.

Judd

. The clinical course of unipolar major depressive disorders. Arch Gen Psychiatry. 1997;54(11):989–991.

Burcusa

Iacono

. Risk of recurrence in depression. Clin Psychol Rev. 2007;27(8):959–985.

10.

Balkrishnan

Joish

Yang

Jayawant

Mullins

. The economic burden associated with SSRI treatment failure in a managed care population. J Med Econ. 2008;11(2):601–610.

11.

Fostick

Silberman

Beckman

Spivak

Amital

. The economic impact of depression: resistance or severity? Eur Neuropsychopharmacol. 2010;20(3):671–675.

12.

Judd

Akiskal

Zeller

, et al. Psychosocial disability during the long-term course of unipolar major depressive disorder. Arch Gene Psychiatry. 2000;57(3):375–380.

13.

Skodol

Grilo

Keyes

Geier

Grant

Hasin

. Relationship of personality disorders to the course of major depressive disorder in a nationally representative sample. Am J Psychiatry. 2011;168(3):257–264.

14.

Kupfer

Frank

Wamhoff

Mood disorders: update on prevention of recurrence. In: Mundt

Goldstein

, editors. Interpersonal factors in the origin and course of affective disorders. London, England: Gaskell/Royal College of Psychiatrists; 1996. p. 289–302.

15.

Colman

Naicker

Zeng

Ataullahjan

Senthilselvan

Patten

. Predictors of long-term prognosis of depression. CMAJ. 2011;183(17):1969–1976.

16.

Holzel

Harter

Reese

Kriston

. Risk factors of chronic depression - a systematic review. J Affect Disord. 2011;129(1):1–13.

17.

Szádóczky

Fazekas

Rihmer

Arató

. The role of psychosocial and biological variables in separating chronic and non-chronic major depression and early-late-onset dysthymia. J Affect Disord. 1994;32(4):1–11.

18.

Ormel

Neeleman

Towards a dynamic stress-vulnerability model of depression. the role of neuroticism, life events and gender. In: Harris

, editor. Where inner and outer Worlds meet. London: Routledge; 2000. p. 151–169.

19.

Spijker

de Graaf

Bijl

Beekman

ATF

Ormel

Nolen

. Determinants of persistence of major depressive episodes in the general population. Results from the Netherlands mental health survey and incidence study (NEMESIS). J Affect Disord. 2004;81(3):231–240.

20.

Sargeant

Bruce

Florio

Weismann

. Factors associated with 1-year outcome major depression in the community. Arch Gen Psychiatry. 1990;47(6):519–526.

21.

Spijker

Bijl

De Graaf

Nolen

. Determinants of poor 1-year outcome of DSM-III-R major depression in the general population: results from the Netherlands mental health survey and incidence study (NEMESIS). Acta Psychiatr Scand. 2000;103(5):122–130.

22.

Keitner

Ryan

Miller

Norman

. Recovery and major depression: factors associated with twelve month outcome. Am J Psychiatry. 1992;149:93–99.

23.

Keller

. Depression: a long term illness. Br J Psychiatry. 1994;165(Suppl 26):9–15.

24.

Ormel

Oldehinkel

Brilman

van den Brink

. Outcome of depression and anxiety in primary care. a three-wave study of psychopathology and disability. Arch Gen Psychiatry. 1993;50(2):759–766.

25.

Mueller

Lavori

Keller

, et al. Prognostic effect of the variable course of alcoholism on the 10-year course of depression. Am J Psychiatry. 1994;151(5):701–706.

26.

Barkow

Maier

Ustun

Gansicke

Wittchen

Heun

. Risk factors for depression at 12-month follow-up in adult primary health care patients with major depression: an international prospective study. J Affect Disord. 2003;76(1-3):157–169.

27.

Brown

Schuckit

. Changes in depression among abstinent alcoholics. J Stud Alcohol. 1988;49(5):412–417.

28.

Garcia-Toro

Rubio

Gili

, et al. Persistence of chronic major depression: a national prospective study. J Affect Disord. 2013;151(1):306–312.

29.

Alpert

Maddocks

Rosenbaum

Fava

. Childhood psychopathology retrospectively assessed among adults with early onset major depression. J Affect Disord. 1994;31(3):165–171.

30.

Tambay

Catlin

. Sample design of the national population health survey - its longitudinal nature. Health Rep. 1995;7(1):29–38, 31–42.

31.

Statistics Canada. National Population Health Survey: Household Component, Longitudinal (NPHS); 2012 [Accessed 2017 November 1]. http://www23.statcan.gc.ca/imdb/p2SV.pl?Function=getSurvey&SDDS=3225.

32.

Statistics Canada. National Population Health Survey Household Component Cycle 1 to 9 (1994/1995 to 2006/2007) Longitudinal Documentation. Ottawa, ON: Statistics Canada; 2009.

33.

Schuckit

Hesselbrock

Tipp

Anthenelli

Bucholz

Radziminski

. A comparison of DSM-III-R, DSM-IV and ICD-10 substance use disorders diagnoses in 1922 men and women subjects in the COGA study. Collaborative study on the genetics of alcoholism. Addiction. 1994;89(12):1629–1638.

34.

Rounsaville

Bryant

Babor

Kranzler

Kadden

. Cross system agreement for substance use disorders: DSM-III-R, DSM-IV and ICD-10. Addiction. 1993;88(4):337–348.

35.

Bartoli

Carrà

Crocamo

Clerici

. From DSM-IV to DSM 5 alcohol use disorder: an overview of epidemiological data. Addict Behav. 2015;41(5):46–50.

36.

Stoller

King

Kidorf

Neufeld

Brooner

. DSM-III-R Versus DSM-IV substance use disorders. Concordance in drug users homogenous for chronic and severe problems. Addict Disord Treat. 2006;5(4):165–171.

37.

Zou

. A modified Poisson regression approach to prospective studies with binary data. Am J Epidemiol. 2004;159(6):702–706.

38.

Segal

Pearson

Thase

. Challenges in preventing relapse in major depression: report of a national institute of mental health workshop on state science of relapse prevention in major depression. J Affect Disord. 2003;77(2):97–108.

39.

Owen

Fischer

Booth

Cuffel

. Medication noncompliance and substance abuse among patients with schizophrenia. Psychiatr Serv. 1996;47(1):853–858.

40.

Volkow

Baler

Compton

Weiss

. Adverse health effects of marijuana use. N Engl J Med. 2014;370(23):2219–2227.

41.

Hobbs

JDJ

Kushner

Lee

Reardon

Maurer

. Meta-analysis of supplemental treatment for depressive and anxiety disorders in patients being treated for alcohol dependence. Am J Addict. 2011;20(4):319–329.

42.

Mueser

Bellack

Blanchard

. Comorbidity of schizophrenia and substance abuse: implications for treatment. J Consult Clin Psychol. 1992;60(6):845–856.

43.

Khantzian

. The self-medication hypothesis of addictive disorders: focus on heroin and cocaine dependence. Am J Psychiatry. 1985;142(3):1259–1264.

44.

Maremmani

Perugi

Pacini

Akiskal

. Toward a unitary perspective on the bipolar spectrum and substance abuse: opiate addiction as a paradigm. J Affect Disord. 2006;93(1-3):1–12.

45.

MacDonald

Baker

Stewart

Skinner

. Effects of alcohol on the response to hyperventilation of participants high and low in anxiety sensitivity. Alcohol Clin Exp Res. 2000;24(11):1656–1665.

46.

Goldman

Orozi

Ducci

. The genetics of addictions: uncovering the genes. Nat Rev Genet. 2005;6:521–532.

47.

Kendler

Prescott

Myers

Neale

. The structure of genetic and environmental risk factors for common psychiatric and substance use disorders in men and women. Arch Gen Psychiatry. 2003; 60(9):929–937.

48.

Canadian Centre on Substance Abuse. Substance abuse in Canada: Concurrent disorders. Ottawa, ON: Canadian Centre on Substance Abuse; 2009.

49.

Agrawal

Lynskey

. Cannabis controversies: how genetics can inform the study of comorbidity. Addiction. 2014;109(5):360–370.

50.

Burmeister

. New insights into the genetics of addiction. Nat Rev Genet. 2009;10(4):225–231.

51.

Stewart

Conrod

. Anxiety and substance use disorders: The vicious cycle of comorbidity. NY: Springer; 2008.

52.

Stewart

Conrod

. Anxiety disorder and substance use disorder comorbidity: common themes and future directions. In: Stewart

Conrod

, editors. Anxiety and substance use disorder: The vicious cycle of co-morbidity. NY: Springer; 2008. p. 239–257.

53.

Stewart

O’Connor

. Treating anxiety disorders in the context of concurrent substance misuse. In: Sookman

Leahy

, editors. Treatment of resistant anxiety disorders. NY: Routledge; 2009. p. 291–323.

54.

Bruce

Yonkers

Otto

, et al. Influence of psychiatric comorbidity on recovery and recurrence in generalized anxiety disorder, social phobia, and panic disorder: a 12-year prospective study. Am J Psychiatry. 2005;162(6):1179–1187.

55.

Willinger

Lenzinger

Hornik

, et al. Anxiety as a predictor of relapse in detoxified alcohol-dependent patients. Alcohol Alcohol. 2002;37(6):609–612.

56.

Klein

Arnow

Barkin

, et al. Early adversity in chronic depression: clinical correlates and response to pharmacotherapy. Depress Anxiety. 2009;26(8):701–710.

57.

Perez-Fuentes

Olfson

Villegas

Morcillo

Wang

Blanco

. Prevalence and correlates of child sexual abuse: a national study. Compr Psychiatry. 2013;54(10):16–27.

58.

Sugaya

Hasin

Olfson

Lin

Grant

Blanc

. Child physical abuse and adult mental health: a national study. J Trauma Stress. 2012;25(3):384–392.

59.

Gonzalez

Boyle

Kyu

Georgiades

Duncan

MacMilan

. Childhood and family influences on depression, chronic physical conditions, and their comorbidity: findings from the Ontario child health study. J Psychiatr Res. 2012;46(2):1475–1482.

60.

Choi

DiNitto

Marti

Choi

. Association of adverse childhood experiences with lifetime mental and substance use disorders among men and women aged 50+ years. Int Psychogeriatr. 2017;29(3):359–372.

61.

Merrick

Ports

Ford

Afifi

Gershoff

Grogan-Kaylor

. Unpacking the impact of adverse childhood experiences on adult mental health. Child Abuse Negl. 2017;69(6):10–19.

62.

Patten

Wang

Williams

JVA

Lavorato

Khaled

Bulloch

AGM

. Predictors of the longitudinal course of major depression in a Canadian population sample. Can J Psychiatry. 2010;55(10):669–675.

63.

Bottomley

Nazareth

Torres-Gonzalez

, et al. Comparison of risk factors for the onset and maintenance of depression. Br J Psychiatry. 2010;196(1):13–17.

64.

Meng

D’Arcy

. The projected effect of risk factor reduction on major depression incidence: a 16-year longitudinal Canadian cohort of the national population health survey. J Affect Disord. 2014;158(2):56–61.

65.

Mueller

Keller

Leon

, et al. Recovery after 5 years of unremitting major depressive disorder. Arch Gen Psychiatry. 1996;53(1):794–799.

66.

The Conference Board of Canada. Income Inequality—Society Provincial Rankings—How Canada Performs. [Online]; 2018 [Accessed 2018 April 28]. http://www.conferenceboard.ca/hcp/provincial/society/income-inequality.aspx.

67.

O’Leary

Costello

Gormley

Webb

. Remission onset and relapse in depression: an 18-month prospective study of course for 100 first admission patients. J Affect Disord. 2000;57:159–171.

68.

Gonzales

Lewinsohn

Clarke

. Longitudinal follow-up of unipolar depressives: an investigation of predictors to relapse. J Consul Clin Psychol. 1985;53(4):461–469.