Prevalence and long-term change in alcohol consumption
In the present follow-up, 72.4% of the study participants from the baseline were included. There were no differences in alcohol consumption levels among participants and non-participants at baseline. The mean (SD) age at baseline and follow-up were 28.1 (1.2) and 45.9 (1.1) years respectively. The alcohol consumption prevalence at baseline was 54.5% (95% CI 51.6, 57.4)] and significantly decreased to 47.7% (95% CI 44.3, 51.1) during the follow-up (Z = 3.01, p < 0.01). In general, 34.3% were non-consumers of alcohol through the two decades, and 35.7% of individuals continued to drink in both the phases. Among the non-consumers at baseline, 25.9% began consuming alcohol during follow-up, whereas 33.6% of the baseline consumers quit drinking during follow-up. The median (IQR) alcohol consumption per week at baseline (n = 643) and follow-up(n = 402) are 4.53 (2.37, 7.23) SDU and 4.73 (4.73, 9.47) SDU, respectively. The maximum consumption during baseline was 187.13, and the follow-up was 110.46 SDU per week.
The type of alcohol consumption changed over time. Branded spirits largely replaced locally made spirits and made up the bulk, over 90%, of all reported alcohol consumption at follow-up. Locally made spirits reduced from 6.6 to 0.4%. Consumption of beer was reported by 37.5% at baseline and 3.2% at follow-up. Wine consumption was rare; 2.1% at baseline, dropping to 0.5% at follow-up.
Factors associated with long-term change in alcohol use
The long-term change among alcohol users was calculated as the absolute difference in alcohol consumption between two-time points (difference = follow-up—baseline). A positive difference signifies increased consumption at follow-up, while a negative difference indicates the opposite. The calculations do not include participants who were non-drinkers at both time points ( n = 289). The median(IQR) difference in alcohol consumption is 0.21(−2.49, 4.73) SDU per week with a minimum difference of -142.55 and a maximum difference of 84.42 for the alcohol consumers (n = 554). This difference in alcohol consumption was statistically significant (Z = −2.30,p = 0.021). An increased consumption was reported for participants who were unmarried, with lower education and lower SES, rural residents, smoking, non-smoking tobacco use and low levels of physical activity at the follow-up time point. These differences in level of consumption were statistically significant for marital status, place of residence, smoking and non-smoking tobacco use (Table 1).
A significant proportion (35.7%) of cohort participants continued to drink throughout both phases. A majority of the unmarried men(56%) were non-consumers in both phases, whereas the continuation of alcohol consumption was more common among married participants (36%). The non-consumption or discontinuation of alcohol at follow-up was reported higher for higher education (72.1%), skilled manual labourers (63.5%)/professional (70.6%) occupation and SES ≥ third quartile (58.0%). Discontinuation of alcohol consumption was associated with urban residents and non-smokers. Contrastingly, lower education (59.5%), unemployment (56.1%), and SES ≤ 2nd quartile (54.3%) were associated with continuing alcohol consumption throughout 18 years of follow-up or initiating alcohol consumption during the follow-up phase. Physical activity was not associated with a change in alcohol practice (Additional Table 1).
Factors associated with alcohol use
The association between between demographic factors and alcohol consumption was also studied. Alcohol consumption was associated with lower levels of education and socioeconomic status, with smoking and with low levels of physical activity. The associations remained the same for both time points (Table 2). The risk for alcohol consumption was significantly lower in semi-skilled labour compared to professionals/skilled labours at baseline [AOR(95% CI) 0.43 (0.23, 0.78)], whereas during follow-up semi-skilled labour [AOR (95% CI) 1.91 (1.01, 3.30)] had a higher risk compared to professionals/skilled labours. Similarly, the rural residents had lower risk for alcohol consumption at baseline [AOR (95% CI) 0.76 (0.56, 1.02)] with increased risk during follow-up [AOR (95% CI) 1.45 (1.06, 2.01)] compared to urban residents.
Factors related to an increase in the level of alcohol consumption were assessed by comparing the distribution of alcohol levels among the demographical variables. Alcohol levels were higher among those with low education, unemployed or unskilled labour, lower SES, and smoking and non-smoking tobacco use compared to their counterparts. The results were similar in baseline and follow-up (Table 3).
A significant decreasing trend in alcohol consumption levels was observed with higher education, skilled/professional occupation and higher SES. A significantly increased risk in alcohol levels was reported for smoking [Baseline: AOR (95% CI) 6.34 (4.91, 8.18); Follow-up: AOR (95% CI) 3.58 (2.04, 4.93)], and for non-smoking tobacco use [Baseline: AOR (95% CI) 2.79 (1.65, 4.72); follow-up: AOR (95% CI) 5.25 (2.04, 12.48)]. Contrastingly, rural residents had low levels of drinking at baseline [AOR (95% CI) 0.71 (0.54, 0.93)] where an increase was reported during follow-up [AOR (95% CI) 1.42 (1.06, 1.92)]. No significant trends were observed in marital status or physical activity (Additional Tables 2 & 3).
Alcohol use and chronic disease
Alcohol consumption was significantly more prevalent in groups with chronic disease (χ2 = 4.68,df = 1, p = 0.031). An increased level of alcohol consumption during the follow-up period was reported among the participants with chronic disease compared to the participants without chronic diseases (χ2 = 9.52, df = 3, p = 0.023). Furthermore, newly developed chronic disease during the follow-up period was more prevalent among the alcohol consumers compared to the non-consumers (46.02% vs 36.51%).
Over the decades, alcohol consumption was high in participants with chronic disease, whereas alcohol non-consumption or cessation was higher in participants without any chronic illnesses. Although a slight increase was observed in hazardous drinking and alcohol dependence among individuals with chronic disease, it was not statistically significant (Additional Table 4).
Drinking behaviour among the alcohol users
The AUDIT analysis was done for 380 men who were alcohol consumers at follow-up. The median AUDIT score was 6.0 (3.0, 10.0), with slightly higher values in rural areas [7.0 (4.0, 11.0)] compared to urban areas [6.0 (3.0, 10.0)]. The scores suggest hazardous drinking (Zone II) in 32.9%, harmful drinking (Zone III) in 4.7% and probable alcohol dependence (Zone IV) in 3.7% of the cohort participants.
Help-seeking behaviour
Problem awareness and help-seeking behaviour among the participants with AUDIT scores > 8 were studied (n = 114). The results showed that 25.4% were concerned about their high alcohol consumption, 12.3% considered seeking help, and 8.8% had sought help. A significant proportion of the individuals discussed the issue of high alcohol consumption with wife or family members (46.5%), while 13.2% did not discuss it with anyone. The participants with high SES were, to a more significant extent, aware of their consumption and were concerned. Among the heavy drinkers, there was a higher degree of concern about their drinking as well as a higher degree of discussing their drinking with family and friends and also of seeking help. Educational and occupational statuses were not associated with any help-seeking behaviour (Table 4).
Add Comment