• Effective alcohol policy controls needs development by health and education authorities 
• Based on scientific evidence, considering gender, age, ethnicity, income, and settings and contexts of use

The alcohol industry is strongly linked with the tourism industry in the country, and therefore, harm reduction approaches rather than total absenteeism, should be advocated by the health and education authorities as a realistic goal to promote population health. 

Moreover, effective alcohol control policies need to be developed based on scientific evidence considering gender, age, ethnicity and income variations of alcohol users in the country, and the settings and contexts in which drinkers use alcohol. 

These recommendations were made in an original research article on the ‘Epidemiology of alcohol use in Sri Lanka: Findings of the Sri Lanka Health and Ageing Study (SLHAS) national survey’ which was authored by B. Perera and H.M.M. Herath (both attached to the Ruhuna University's Medical Faculty), R.P. Rannan-Eliya, N. Wijemunige and S. Samarage (all three attached to the Institute for Health Policy), W. Jayawardene (attached to the US’ Southern Illinois University's Human Sciences School) and M.R. Torabi (attached to the US' Indian University's Public Health School), and published in the Global Epidemiology journal's 10th Volume.

Alcohol use causes substantial premature deaths and disabilities worldwide, with individuals in low- and middle-income countries (LMICs) disproportionately affected (K. Shield, J. Manthey, M. Rylett, C. Probst, A. Wettlaufer, C.D.H. Parry and J. Rehm's ‘National, regional, and global burdens of disease [GBDs] from 2000 to 2016 attributable to alcohol use: A comparative risk assessment study’, the WHO ‘Global status report on alcohol and health and the treatment of substance use disorders’, the Alcohol Collaborators' ‘Alcohol use and burden for 195 countries and territories, 1990-2016: A systematic analysis for the GBD study 2016’, and the GBD 2021 Risk Factors Collaborators' ‘GB and strength of evidence for 88 risk factors in 204 countries and 811 sub-national locations, 1990-2021: A systematic analysis for the GBD Study 2021’). 

In 2019, it accounted for 2.6 million deaths, or 4.7% of all global deaths. In 2021, high alcohol use was the 10th leading risk factor for the loss of disability-adjusted life years (DALY) in people, accounting for 2.5% of the global DALY burden. Additionally, alcohol use causes a range of socio-economic and environmental harms. Violence, property damage, absenteeism in work, and social misconduct are more common in heavy alcohol users and in communities where alcohol use is common (O. Waleewong, A.M. Laslett, R. Chenhall and R. Room's ‘Harm from others’ drinking-related aggression, violence and misconduct in five Asian countries and the implications’).

S. Ranaweera, H. Amarasinghe, N. Chandraratne, M. Thavorncharoensap, T. Ranasinghe, S. Karunaratna, D. Kumara, B. Santatiwongchai, U. Chaikledkaew, P. Abeykoon and A. De Silva's ‘Economic costs of alcohol use in Sri Lanka (a LMIC in South Asia [SA])’ (2015) estimated alcohol use costs at 1.1% of the gross domestic product (GDP), and 27.5% of the health expenditure. That year (2015), Sri Lanka generated $ 780 million in Excise tax revenue from alcohol, but, the estimated total economic cost of alcohol was $ 886 million. The socio-economic burden of alcohol use is severe in LMICs, and diminishes the economic and social wellbeing of people. Alcohol use also exacerbates disparities across income levels, age, gender, and urban and rural residences, reinforcing health and economic inequalities within and between countries. It also hinders progress toward several SDGs, particularly SDGs One, Three, Five and 10 which aim to reduce poverty, improve health, promote gender equality, and reduce inequalities.

In Sri Lanka, non-communicable diseases (NCDs) account for about 81% of deaths. There is a decreasing trend in the production of legal alcoholic beverages, from 51.7 million litres (l) of hard liquor in 2015 to 26.5 million l in 2022, a 51.3 % drop. The per capita consumption of alcohol, estimated as the sales of legal alcohol, has fallen from 4.3 l in 2016 to 2.8 l in 2019. But, alcohol revenues of the Government have steadily increased from Rs. 66 billion in Government Excise tax on liquor in 2016 to Rs. 165.2 billion in 2022, a 250% increase. Although price increases of legal alcoholic beverages are probably the major reason for the drop of legal liquor consumption and the increase in Excise tax, illegal, low cost liquor production has increased over the last few decades (H. Leifman's ‘Trends and patterns of alcohol consumption in Sri Lanka 1981-2017 - Analyses of sales and survey data’, and P. Katulanda, C. Ranasinghe, A. Rathnapala, N. Karunaratne, R. Sheriff and D. Matthews's ‘Prevalence, patterns and correlates of alcohol consumption and its’ association with tobacco smoking among Sri Lankan adults: A cross-sectional study’) and has probably led to rising alcohol consumption in the country. An estimated 65% of the total alcohol market in Sri Lanka is believed to be illicit today, and J.B. Sørensen, F. Konradsen, T. Agampodi, B.R. Sørensen, M. Pearson, S. Siribaddana and T. Rheinländer's ‘A qualitative exploration of rural and semi-urban Sri Lankan men's alcohol consumption’ suggests that the alcohol consumption rates are on an upward trend in the country. 

Methodology

Perera et al.'s study used baseline data from a national, longitudinal health survey to identify the prevalence, patterns and associated factors of alcohol use among adults in Sri Lanka. This analysis used data from the first wave of the SLHAS conducted from mid-November 2018 to mid-November 2019. The SLHAS is a national, longitudinal, cohort study approved by the Health Ministry. Stratified, multistage probability sampling was used to recruit a nationally representative sample of the non-institutionalised adult (18 years or above) population. Stratification involved two steps. In the first, all Grama Niladhari Divisions (GNDs) (14,104), which are the smallest administrative unit in Sri Lanka and primary sampling units (PSUs) in the SLHAS, were categorised by District (25) and the sector of residence (urban, rural, estate, rural/estate) into 57 preliminary strata. Then, PSUs (GNDs) within each preliminary stratum were further stratified into equally sized population quantiles after having ranked them using an index of the area socioeconomic status (SES), generating 157 strata. 

Within the strata, individuals were sampled using multistage, probability cluster sampling. In the first stage, a minimum of two PSUs was selected from each stratum using probability proportionate to the size of their adult population, with large PSUs picked with certainty. At the second stage, four to six widely spaced households were identified in each PSU by systematically sampling the electoral register, or in rural areas where the households were not distributed by street, by using satellite maps showing the distribution of buildings to pick dispersed geo-locations. Recruitment teams visited these households or the household nearest to the sampled geo-location, and then additional households at pre-set inter-household intervals of two to four by walking in a predefined track, with larger intervals in more densely populated or urban PSUs. 

If they gained entry, recruiters enumerated all household residents using a computer-assisted personal interviewing (CAPI) application running on computer tablets. If the household gave consent, the CAPI software randomly selected one adult (18 years or above) using weighted probabilities that targeted a final equal distribution of respondents by sex and by age up to 69 years with the oversampling of those aged 70 years or above. 

An interviewer administered pre-tested questionnaire was used to collect data. The study protocol excluded pregnant women, and adults unable to give informed consent, and if the selected individual declined participation, the whole household was excluded, with the CAPI system preventing recruiters from selecting another individual. The selected individuals were invited to attend a field clinic near their residence, where they were interviewed to assess their health and collect other individual and household information including substance use related data. Individuals with mobility limitations were interviewed at home. 

For alcohol use, the participants were asked whether they had consumed at least ‘one drink’ of any form of alcoholic beverage (arrack, kassippu, whisky, gin, brandy, beer, toddy, wine or other alcoholic beverages) at least once during the previous year. Those answering “no” were categorised as abstainers or former users, and those answering “yes” as current (yearly) users. Current users were further asked to describe their consumption frequency, categorised as daily, four or more times a week, two–three times a week, two-four times a month, once a month, once a year or more but not monthly, or less than once a year. The weekly alcohol consumption was calculated in units (one unit equals 8 g of alcohol) based on the average alcohol content of different alcoholic beverages: one drink (25-30 ml) of arrack/whisky, 50 ml of illicit alcohol, half a pint of beer/toddy and a small glass of wine [175 ml] were each considered to be equivalent to one unit (S. Eriksen's ‘Alcohol as a gender symbol’). Hazardous drinking was defined as consuming 21 or more units/week of alcohol for men and 14 or more units/week for women (the National Health Service's ‘Alcohol units’, and M.C. Reid, D.A. Fiellin and P.G. O’Connor's ‘Hazardous and harmful alcohol consumption in primary care’).

The data were weighted to ensure such being representative of Sri Lanka's adult population. 

Results

Of the 10,689 sampled households, 10,247 were contacted, and 10,062 agreed to participate, with 6,627 selected adults attending a field clinic, and 41 completing home interviews, giving an effective response rate of 65%. The Medical Officer of Health (MOH) Offices located in the survey areas were utilised for data collection. Some of the participants who agreed to participate in the research did not show up at the data collecting centres on their given date. 

However, substantial selection biases were not observed. Response rates were higher in women (69%), adults aged 45 years or above (74%), and rural residents (70%). Alcohol use data were collected from 6,559 participants, with 86 respondents (1.3%) excluded due to missing or incomplete data on alcohol use. This left 6,473 participants for analysis. Of the total, 52.5% were women, 25.2% were young adults (aged 18-29 years) and 19.8% were older adults (aged 60 plus years). The majority were Sinhalese (predominant ethnic group in Sri Lanka) (75.6%) followed by 12% of Sri Lankan Tamils and 9% of Sri Lankan Muslims. 

The prevalence of current alcohol use (any form of alcohol use in the past year) was 26.3% (25.5-27%), significantly higher in men (53.1%; 51.9-54.3%) than women (2%; 1.9-2.1%). Among the current users, men consumed a mean of 12.26 alcohol units/week (11.94-12.57) compared to 0.30 units/week (0.29-0.32) for women. A total of 275 of the total of 1,644 current male alcohol users were hazardous drinkers. The prevalence of hazardous drinking in current male alcohol users was 14.1%, with male hazardous drinkers consuming 62.56 units/week (61.21-63.91). The per capita consumption of alcohol was 1.48 l, and in men it was 3.12 l and in women it was 0.028 l.

Alcohol use peaked among middle-aged (30-49 years of age) adults. Overall, among middle-aged people, 61.8% (60.8-61.9%) of men and 2.5% (2.4-2.6%) of women were current alcohol users. The lowest prevalence rates were in older adults (60-plus years) with a sharp decrease with increasing age. The prevalence rate in young adults (aged 18-29) was 48.7% (48.6-49.8%) and 2.2% (2.2-2.3%) for men and women, respectively. A high prevalence rate of current alcohol use was observed among women aged 30-39 years (3.2%; 2.9-3.5%). Among the current male users, the prevalence of hazardous drinking was the highest among those aged 80-plus years (24%; 23.4-24.7%) followed by the 50-59 years age group (20.4%; 20.3-20.5%) and the 30-39 years age group (15.8%; 15.7-15.9%). Hazardous drinking rates were comparatively low among those men in the age groups of 40-49 years (12.9%; 12.8-13%) and 60-69 years (11.1%; 11-11.2%). None of the women surveyed reported hazardous drinking behaviour.

In men, current alcohol use was the highest among the Sinhalese (59.7%; 59.3-60.1%), while in women, the highest rate was among Indian Tamils (5.9%; 5.8-6%). Muslims of both sexes reported the lowest use rates. In men, the prevalence of hazardous drinking was the highest among Sri Lankan Tamil and Indian Tamil ethnic groups where the majority (80%) practice the Hindu religion.

In men, the prevalence of current alcohol use was relatively higher (about 50% of the sample subjects were current alcohol users) among those with secondary school education (Grade Six up to having passed the General Certificate of Education Advanced Level [A/L]), compared to the others. Among the women however, a noteworthy overall higher prevalence of current alcohol use was observed among those who have the tertiary level of educational qualifications (3.8%; 3.6-4%). In men, hazardous drinking was more prevalent among those who have had no education (26.9%; 26.6-27.2%) and among those who studied up to the primary level (26.2%; 26-26.3%). In men, people in the middle wealth quartile have the highest prevalence rate (57.6%; 57.5-57.7%). In women, the prevalence rate was the highest in the richest wealth quintile (4.2%; 4.1-4.2%). In men, the prevalence of hazardous drinking in current drinkers was the highest among those in the poorest wealth quintile (21.3%; 21.2-21.4%).

Those who were currently married had a higher likelihood of being current alcohol users (55.1%; 54.7-55.5%) compared to those who were not married (46.7%; 46.6-46.8%) and ‘others’ (46.8%; 46.5-47%) where the ‘other’ category included those who are separated, widowed, divorced, and cohabitating.

Results indicate that gender is the strongest determinant of alcohol use in Sri Lankan adults, with only 2% of women reporting current alcohol use. Men aged 30-59 years had the higher odds of being current alcohol users, while those aged 60-plus years had lower odds, compared to men aged 18-29 years. However, hazardous drinking was more prevalent in older age groups: men aged 80-plus years are having about 3.6 times higher odds of hazardous drinking compared to young adults (18-29 years). Smoking was strongly positively associated with the current and hazardous drinking behaviour of men. Smokers were having 5.8 times higher odds of being current alcohol users than non-smokers, and among current users, smokers were having 2.2 higher odds of being hazardous drinkers.

Compared to non-hypertensives (no high blood pressure), hypertensives were having 1.12 times higher odds of being current alcohol users and 1.19 times higher odds of being hazardous drinkers. Having diabetes was associated with a lower likelihood of being an alcohol user. Obese, overweight and underweight men had lower odds of being current alcohol users than men with a normal body mass index (BMI). Among current male alcohol users, the likelihood of being a hazardous drinker was higher in those who were obese and underweight. A significant percentage of heavy drinkers underreport their amount of consumption, so, they are likely to be misclassified as moderate or mild drinkers, leading to the overestimation of the health consequences of moderate and mild drinking.

Those reporting lower self-related health (SRH) had higher odds of consuming alcohol, with hazardous drinkers having higher odds to report poor SRH. Hazardous drinking was also strongly associated with moderate and severe depression, though current alcohol users showed a lower likelihood of depression. Results indicated that those who have moderate and severe depression had higher odds than those with ‘no depression’ of current drinking. However, among current alcohol users, those who were identified as having severe, moderate or mild depression had lower odds of hazardous drinking than those with ‘no depression’.

In women, the prevalence of current alcohol use was comparatively higher in the Gampaha (Western Province [WP]), Kegalle (Sabaragamuwa Province [SP]), Puttalam (North Western Province [NWP]), and Moneragala (Uva Province [UP]) Districts. The Districts of Colombo (WP), Nuwara-Eliya (Central Province [CP]), Matale (CP), Kurunegala (NWP), Polonnaruwa (North Central Province [NCP]) and Anuradhapura (NCP) also have a higher prevalence of current alcohol use.

In men, the prevalence of current alcohol use is higher in the Gampaha, Matale, Polonnaruwa, Badulla (UP), Nuwara-Eliya, Ratnapura (SP), and Hambanthota (Southern Province) Districts. 

The Districts of Matale, Badulla, Nuwara-Eliya and Ratnapura have higher proportions of Indian Tamils. Gampaha has a significantly larger Roman Catholic and Christian population than the other Districts. Hambanthota has seen rapid development and increased tourism promotion in the last decade, mainly due to political initiatives. The Trincomalee (Eastern Province [EP]), Puttalam, Batticaloa (EP) and Ampara (EP) Districts have substantial Muslim populations, while Jaffna (Northern Province [NP]), Killinochchi (NP), and Vavuniya (NP) are predominantly Sri Lankan Tamil.