Participant characteristics
Altogether, 13,177 participants (51.8% female) were included in the analysis. The mean age of the participants was 46.7 years (SD: 15.7) and 70.5% were married or in a de facto relationship. Of all participants, 57.8% had no formal education or had attained only a primary school level of education, 55.6% were employed in labor type work while 34.1% were employed in private or government work, and 10.3% were unemployed; 56.5% were living in rural areas.
Based on the AUDIT, 10.3% (95% CI: 9.8, 10.8) and 1.9% (95% CI: 1.7, 2.1) were classified into hazardous drinkers and harmful-dependent drinkers, respectively while 2.5% (95% CI: 2.3, 2.7) met the criteria for MDE in the past 12 months before the survey. Approximately 20% (95% CI: 18.9, 20.3) were current smokers. Chronic diseases were found among 41.7% (95% CI: 41.2, 43.0) of participants, with 10.2% (95% CI: 9.6, 10.7) having diabetes mellitus, 29.2% (95% CI: 28.1, 30.4) hypertension, 11.6% (95% CI: 10.4, 12.7) cholesterolemia, 3.3% (95% CI: 3.0, 3.7) cardiovascular diseases and stroke, 0.8% (95% CI: 0.7, 0.9) cancers and 15.3% (95% CI: 14.7, 15.8) other diseases.
Prevalence of MDE and AUD by socio-economic status and area of residence
The prevalence of MDE varied by all three socio-economic status indicators but not by area of residence. The prevalence of MDE was lower in the groups with higher levels of wealth, education, and among those who were employed (Table 1).
AUD was differentially distributed in different levels of wealth index, education levels, and area of residence. The highest prevalence of harmful-dependent drinking was found among people in the lowest wealth tercile and those with a secondary or higher level of education (12-year schooling). The prevalence of hazardous drinking was significantly higher among those living in rural areas but there was a non-significant difference in the prevalence of harmful drinking between those in urban and rural areas. Furthermore, no significant difference in AUD prevalence was found for employment status (Table 1).
Table 1 Weighted prevalence (%, (95% Confidence Interval)) of major depressive episode and alcohol use disorder in the past 12 months by wealth index, education, employment status, and area of residence (N = 13,177)
Category
|
Total
(N=13,177)
|
Wealth index
|
Education
|
Employment status
|
Area of residence
|
Tercile 1
(N=4,196)
|
Tercile 2
(N=4,017)
|
Tercile 3
(N=4,964)
|
1’ school (N=8,608)
|
2’ school (N=4,434)
|
Employed
(N=7,888)
|
Unemployed
(N=866)
|
Urban
(N=7,068)
|
Rural
(N=6,109)
|
MDE
(N=406)
|
2.5
(2.3, 2.7)
|
3.1
(2.8,3.5)
|
2.4
(2.1,2.8)
|
1.9
(1.6,2.2)
|
3.1
(2.9,3.4)
|
1.7 (1.5,2.0)
|
2.1 (1.9,2.3)
|
2.7 (2.1,3.5)
|
2.6 (2.3,2.9)
|
2.4 (2.2,2.7)
|
P-value = <0.001
|
P-value = <0.001
|
P-value = 0.047
|
P-value = 0.124
|
Non-problem (N=11,983)
|
87.8
(87.4,88.3)
|
85.8
(84.9,86.7)
|
87.8
(87.1,88.5)
|
90.0
(89.1,90.6)
|
89.6
(89.1,90.0)
|
85.4
(84.8, 86.0)
|
85.6
(85.0,86.2)
|
88.5
(86.4,90.3)
|
88.4
(87.8,89.0)
|
87.3
(86.8,87.9)
|
Hazardous
(N=1,019)
|
10.3 (9.8,10.8)
|
11.3 (10.4,12.2)
|
11.0 (10.1,11.8)
|
8.7 (7.9,9.5)
|
8.9 (8.3,9.4)
|
12.3 (11.5,13.2)
|
12.2 (11.5,12.8)
|
9.9 (8.4,11.7)
|
9.5 (8.9,10.2)
|
10.9 (10.2,11.6)
|
Harmful (N=175)
|
1.9 (1.7,2.1)
|
2.9 (2.5,3.5)
|
1.2 (1.0,1.5)
|
1.4 (1.2,1.8)
|
1.6 (1.4,1.8)
|
2.3 (2.0,2.6)
|
2.3 (2.0,2.5)
|
1.6 (1.1,2.3)
|
2.1 (1.8,2.4)
|
1.8 (1.5,2.0)
|
|
|
P-value = <0.001
|
P-value = <0.001
|
P-value = 0.117
|
P-value = <0.001
|
P-values from Chi-squared test, 1’ school = primary school, 2’ school = secondary school or higher. Wealth index: Tercile 1 = lowest (poorest) and Tercile 3 = highest (wealthiest).
Association between MDE and AUD across levels of wealth index, education and area of residence
Table 2 shows crude odds ratios for the relationship between MDE and each of the SES and area of residence variables with AUD. MDE was significantly associated with AUD. The prevalence of MDE among non-drinkers, hazardous drinkers, and harmful drinkers was 2.6% (standard error; SE = 0.01), 1.3% (SE = 0.10) and 6.3% (SE = 0.9), respectively. Other variables significantly associated with AUD were wealth index, education, and area of residence while employment status was marginally significant.
Table 2 Crude odds ratios and 95% confidence intervals for the associations between major depressive episode, socio-economic status variables and area of residence with alcohol use disorder
Characteristic
|
Hazardous drinking
(N=1,019)
|
Harmful-dependent drinking (N=175)
|
P-value
|
Major depressive episode
|
|
|
|
No
|
1
|
1
|
<0.001
|
Yes
|
0.52 (0.43, 0.63)
|
2.55 (1.80, 3.62)
|
|
Wealth index
|
|
|
|
Tercile 1 (poorest)
|
1
|
1
|
<0.001
|
Tercile 2
|
0.95 (0.87, 1.05)
|
0.40 (0.32, 0.52)
|
|
Tercile 3 (wealthiest)
|
0.74 (0.67, 0.81)
|
0.47 (0.39, 0.56)
|
|
Education
|
|
|
|
1’ school
|
1
|
1
|
<0.001
|
2’ school
|
1.46 (1.39, 1.52)
|
1.51 (1.30, 1.76)
|
|
Employment status
|
|
|
|
Employed
|
1
|
1
|
0.05
|
Unemployed
|
0.79 (0.65, 0.96)
|
0.68 (0.47, 0.98)
|
|
Area of residence
|
|
|
|
Urban
|
1
|
1
|
<0.001
|
Rural
|
1.16 (1.09, 1.24)
|
0.87 (0.77, 0.98)
|
|
Significant interactions between MDE and wealth index (coefficient: 0.05, 95% CI: -0.00, 0.10 p=0.067 for tercile 2 and coefficient: 0.08, 95% CI: 0.04, 0.12 for tercile 3), education (coefficient: 0.49, 95% CI: 0.02, 0.80 p=0.008) and area of residence (coefficient: -0.06, 95% CI: -0.11, -0.01, p=0.027) were found, indicating that these factors modified the association between MDE and AUD. However, no significant interaction between MDE and employment status (coefficient: -0.02, 95% CI: -0.08, 0.03) in the association with AUD was seen.
Table 3 shows weighted percentages of AUD among those with and without MDE and adjusted odds ratios for the association between MDE and AUD across various socioeconomic levels. Adjusted for other variables, the associations between MDE and either hazardous or harmful-dependent drinking were strongest among those in the third tercile of wealth index (AOR = 2.23, 95% CI: 1.51,2.72 for hazardous drinking and AOR = 8.68, 95% CI: 5.34, 14.11 for harmful-dependent drinking). The AOR for the association between MDE and harmful-dependent drinking was also significant among those in the first tercile of wealth index (AOR = 7.14, 95% CI: 3.71, 13.73) but lower and not significant among those in the second tercile (AOR = 1.78, 95% CI: 0.74, 4.25). The association between MDE and either hazardous or harmful-dependent drinking was significant in people who had a secondary school level of education or above (AOR = 1.75, 95% CI: 1.33, 2.30 for hazardous drinking and AOR = 16.0, 95% CI: 10.3, 24.9 for harmful-dependent drinking) but not among those with a primary school level of education or lower, indicating a strong influence of educational level on the association. Finally, the association between MDE and harmful-dependent drinking was stronger among those living in urban areas (AOR = 8.5, 95% CI: 5.50, 13.13) than in rural areas (AOR = 4.7, 95% CI: 3.31, 6.77). Regarding employment status, there were significant associations between MDE and both hazardous (AOR = 1.26, 95% CI: 1.02, 1.56) and harmful-dependent drinking (AOR = 6.81, 95% CI: 4.71, 9.85 for harmful-dependent drinking) but the number of unemployed participants with MDE was too small for conducting a meaningful stratified analysis.
Table 3 Weighted prevalence of alcohol use disorder among those with and without major depressive episode and adjusted odds ratios for the association between major depressive episode and alcohol use disorder across levels of wealth index, education, employment status, and area of residence among the Thai adult population (N = 13,177)
Effect modifying variable
|
Depressive episode
|
Alcohol use disorder
|
Non-problem drinking
|
Hazardous drinking
|
Harmful-dependent drinking
|
Variable
|
Level
|
Status
|
% (95% CI)
|
AOR
|
% (95% CI)
|
AOR (95% CI)
|
% (95% CI)
|
AOR (95% CI)
|
Wealth index
|
Tercile 1
|
No
|
85.7 (84.8,86.6)
|
1
|
11.5 (10.8,12.3)
|
1
|
2.8 (2.5,3.1)
|
1
|
Yes
|
90.0 (85.1,93.4)
|
|
3.1 (2.3,4.2)
|
0.69 (0.44,1.07)
|
6.9 (4.0,11.6)
|
7.14 (3.71,13.73)
|
Tercile 2
|
No
|
87.7 (87.0,88.4)
|
1
|
11.1 (10.5,11.6)
|
1
|
1.2 (1.0,1.5)
|
1
|
Yes
|
92.6 (89.1,95.0)
|
|
6.6 (4.5,9.7)
|
1.53 (1.02, 2.31)
|
0.8 (0.4,1.7)
|
1.78 (0.74,4.25)
|
Tercile 3
|
No
|
90.0 (89.2,90.7)
|
1
|
8.7 (8.1,9.3)
|
1
|
1.4 (1.2,1.6)
|
1
|
Yes
|
85.9 (81.6,89.3)
|
|
8.4 (6.1,11.4)
|
2.23 (1.51,2.72)
|
5.8 (3.7,8.8)
|
8.68 (5.34,14.11)
|
Education
|
Primary school
|
No
|
89.4 (88.9,89.9)
|
1
|
9.0 (8.6,9.4)
|
1
|
1.6 (1.5,1.8)
|
1
|
Yes
|
94.9 (93.6,95.9)
|
|
4.1 (3.4,5.0)
|
1.04 (0.82, 1.31)
|
1.0 (0.5,2.0
|
1.44 (0.63, 3.33)
|
Secondary school or higher
|
No
|
85.6 (84.9,86.2)
|
1
|
12.4 (11.8,12.9)
|
1
|
2.1 (1.9,2.4)
|
1
|
Yes
|
76.7 (72.3,80.6)
|
|
9.1 (6.9,12.0)
|
1.75 (1.33, 2.30)
|
14.2 (10.4,19.1)
|
16.0 (10.3, 24.9)
|
Employment status
|
Employed
|
No
|
85.7 (85.0,86.3)
|
1
|
12.2 (11.7,12.7)
|
1
|
2.1 (2.0,2.3)
|
1
|
Yes
|
82.9 (79.5,85.7)
|
|
8.9 (7.3,10.8)
|
1.26 (1.02, 1.56)
|
8.3 (6.0,11.3)
|
6.81 (4.71, 9.85)
|
Unemployed
|
No
|
88.3 (86.2,90.2)
|
1
|
10.0 (8.4,11.9)
|
1
|
1.6 (1.2,2.3)
|
1
|
Yes
|
94.7 (91.9,96.6)
|
|
5.3 (3.4,8.2)
|
undetermined
|
0
|
undetermined
|
Area of residence
|
Urban
|
No
|
88.5 (87.9,89.1)
|
1
|
9.6 (9.2,10.1)
|
1
|
1.9 (1.7,2.1)
|
1
|
Yes
|
86.9 (84.9,88.6)
|
|
5.9 (4.5,7.5)
|
1.25 (1.05, 1.50)
|
7.3 (5.2,10.3)
|
8.50 (5.50, 13.13)
|
Rural
|
No
|
87.2 (86.6,87.8)
|
1
|
11.0 (10.6,11.6)
|
1
|
1.7 (1.6,1.9)
|
1
|
Yes
|
92.2 (89.7,94.1)
|
|
5.2 (4.1,6.7)
|
1.24 (0.90, 1.71)
|
2.6 (1.7,4.0)
|
4.73 (3.31, 6.77)
|
Reference category for outcome = non-drinking or non-problem drinking based on Alcohol Use Disorder Identification Test (AUDIT) score: 0-7, hazardous drinkers = AUDIT score: 8-15 and harmful-dependent drinkers = AUDIT score: 16-40.
Reference level for exposure = no major depressive episode
Wealth index: Tercile 1 = lowest (poorest) and Tercile 3 = highest socio-economic group.
AOR = Adjusted Odds Ratio, 95% CI = 95% Confidence Interval, all percentages are weighted for stratified sampling survey.
All odds ratios are adjusted for other socio-economic status variables, area of residence, age group, sex, marital status, religion, smoking status and presence of chronic illness.