Unveiling binge drinking trends and triggers among army personnel: a cross sectional study

Abstract

Background

Military populations are known to have higher prevalence and heavier alcohol use compared to the general population globally. This has serious negative implications to the military. The objective of this study was to describe the prevalence, patterns and associated factors of binge drinking among male military personnel in the Sri Lanka Army.

Methods

A cross sectional study was conducted among 1337 male Army personnel in active service using multistage sampling. A self-administered questionnaire and the interviewer-administered Alcohol Use Disorders Identification Test which is a 10-item screening tool were used. Prevalence of binge drinking was summarised as a proportion with 95% Confidence Intervals (CI). Age specific prevalence rates and the age standardized prevalence rate of binge drinking were calculated. The standard measure of one unit of alcohol being equivalent to 10 g of pure alcohol was used as a reference to calculate the units of alcohol consumption. Binary logistic regression analysis was used to determine the factors associated with binge drinking.

Results

The overall prevalence of binge drinking was 51.2% (95% CI 48.5%-54.0%). The age standardized prevalence of binge drinking was 28.3%. The majority binge drank once a month (50.4%). Those engaged in binge drinking used 5.6 median units of alcohol on a typical day, 84% consumed arrack, 69.3% have ever thought or attempted to quit and median age of first alcohol consumption was 18 years. When controlled for confounding, those who had mental distress (AOR 2.46, 95% CI = 1.72–3.53), had sex with a commercial sex worker (AOR 1.92, 95% CI = 1.21–3.06), ever smoking (AOR 1.69, 95% CI = 1.27–2.25), had serious consequences (AOR 1.58, 95% CI = 1.13–2.20), currently used cannabis (AOR 1.39, 95% CI = 1.02–1.89) and had combat exposure (AOR 1.37, 95% CI 1.00-1.87) had a higher likelihood of binge drinking.

Conclusions

The high prevalence of binge drinking warrants immediate advocacy to the highest level of command of the Sri Lanka Army for support to implement sustainable evidence-based alcohol prevention programmes.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13011-025-00655-9.

Background

Alcohol has long survived and thrived in human civilization, standing the test of time. “Alcohol is the oldest and still one of the most widely used drugs,” [1] woven into various customs and traditions worldwide, and even used for medicinal purposes. Historically, alcohol use was linked to military customs and traditions worldwide, helping to alleviate stress in war and fostering social bonding and comradeship [2].

Alcohol is causally associated with many diseases and injuries, and a systematic analysis of the Global Burden of Diseases study in 2016 led to the remarkable and explosive declaration that there are no safe levels of alcohol consumption for health [3]. Binge drinking, also known as Heavy Episodic Drinking (HED), is an important prevalence indicator that is often used to highlight the burden of alcohol use. The World Health Organization (WHO) defines HED as “60 or more grams of pure alcohol on at least one single occasion at least once per month” [4].

Despite the adverse effects of alcohol, its use among the military continues to be high and much higher than that reported among civilians. The 2018 Health Related Behaviors Survey (HRBS) conducted in the United States (U.S) military reported that 34% of service members were binge drinking compared to 26.5% of the U.S general population [5]. The Army was amongst the service branches in the U.S military reporting higher rates of heavy alcohol use compared to the Air Force. This may indicate subcultures within service branches regarding alcohol use. In addition, the 2018 survey demonstrated a widened gap between the binge drinking prevalence rates between the military and the general population compared to the 2015 HRBS.

Research conducted on military alcohol use in Sri Lanka has been minimal. A study in the Sri Lanka Navy (SLN) reported a prevalence of 71.2% for current drinking, 16.69% for hazardous drinking and 26.1% for binge drinking [6]. This study was conducted three months after the conclusion of the armed conflict in Sri Lanka in 2009. The Sri Lanka military was engaged in combat for 30 years, with intensified operations from 2006 to 2009, resulting in the deaths of 190 officers and 5,700 soldiers in the Army, along with 27,000 injured personnel [6]. An increase in hazardous drinking prevalence to 25.7% was reported in a follow-up study in the SLN [7], three and a half years after the conclusion of the armed conflict Increased access to alcohol due to the easing of restrictions after the end of the war and greater availability have been implicated as possible reasons for this increase in hazardous drinking. The prevalence of males who currently consumed alcohol in the past 30 days among the general population in Sri Lanka was 34.8% in 2015, rising to 43.3% by 2021 [8, 9]. This is of great concern, as high prevalence rates in the general population can influence alcohol use among military personnel [6]. Hence, determining the prevalence of binge drinking in the military is both crucial and timely. Moreover, the pattern of binge drinking and its associated factors is important for identifying the severity of binge drinking, and these factors may differ from those in the general population. This information is highly useful for guiding the development of preventive interventions.

There has been a considerable change in the role and tasks of the Sri Lanka Army (SLA) since the conclusion of the armed conflict in 2009. As opposed to the exclusive combat operations for which the Army personnel were originally trained and prepared, the current work environment has assumed a heterogeneous nature, constantly adapting to national-level requirements. It has become imperative to maintain high levels of both physical and mental health among Army personnel for the efficient and effective execution of their multitude of responsibilities. Furthermore, the consumption of alcohol among military personnel has been shown to have detrimental effects on their overall combat readiness, operational performance, and discipline standards which collectively have a deleterious overall effect on the organization.

Preventing alcohol misuse in the military requires policies and individual-level interventions. Military-led abstinence movements began in the British Army in India due to heavy drinking among military personnel in the mid-19th century, with policies evolving over time. By World War II, alcohol rations were restricted, and in the 1980s, the U.S. military raised the drinking age to 21, reducing alcohol-related treatment cases. Since then, interventions have expanded to include prevention and treatment programs including screening and brief interventions, deglamorization campaigns, and policy measures such as alcohol price regulations, and retail restrictions. Recent research highlights the effectiveness of these interventions including the use of digital tools. A trial reported that a mobile app delivering a brief intervention significantly reduced alcohol use among UK military veterans in the short term [10].

This study aims to determine the prevalence, patterns, and associated factors of binge drinking among military personnel in the Sri Lanka Army, which can help inform and strengthen existing policies and interventions by providing evidence-based insights for more targeted prevention strategies.

Methods

Study design and participants

A cross-sectional study was conducted on male military personnel in active service at battalions (camp deployments) of the SLA over a period of five months, from September 2022 to February 2023. In a national survey in Sri Lanka [8], the current drinking prevalence was only 0.5% among females and the majority of the females (96.5%) were lifetime abstainers of alcohol. A longitudinal analysis conducted in the U.S [11] demonstrated that women enlistees were less likely to drink than their civilian counterparts, which was contrary to research findings among male personnel. Therefore, it was decided to conduct the study only on male Army personnel. Personnel in active service of the SLA for more than one year and who were domestically deployed were included. This excluded those who had deserted the service or were on long term leave. Due to feasibility concerns and legal obligations, the Principal Investigator (PI) could not include these personnel to the current study.

Colonel and above ranks and those who were not present on the day of data collection were excluded. Colonel and higher ranks were excluded due to the logistical difficulties in reaching senior officers for data collection. Similar exclusion criteria for higher ranks were used in past research on the SLA [12]. The sample size was calculated using a standard formula [13]. With a prevalence of 71.2% for current drinking [6] and a precision of 0.035, a multistage sampling method was used, considering a battalion as a cluster with a calculated design effect of 1.98 and a 5% non-response rate, which resulted in the final sample size of 1337. The cluster size was predetermined to be 50 per battalion. Hence, the number of clusters was 27. A three-stage sampling design was used, and this is summarized in Fig. 1.

Fig. 1.

Flow chart of the selection of participants

Study instruments and measures

A Self-Administered Questionnaire (SAQ) was used to collect data due to the sensitive nature of the behavior being investigated. The SAQ was developed based on past literature with the guidance of a multidisciplinary panel of experts and included nine broad sections:

1. First section: Demographic, socio-economic, occupational and health information.

2. Second section: Level of alcohol consumption, pattern of drinking and expenditure on alcohol use.

3. Third section: Consequences of alcohol use.

4. Fourth section: Knowledge, attitudes and other associated factors.

5. Fifth section: Tobacco smoking.

6. Sixth section: Use of other substances.

7. Seventh section: Adverse childhood experiences during the first 18 years of life.

8. Eighth section: Five items from SF-36.

9. Ninth section: GHQ-12 questionnaire.

The Alcohol Use Disorders Identification Test (AUDIT), which was validated in Sri Lanka [14] was included in the final tenth section of the questionnaire. This was interviewer-administered. The AUDIT questionnaire consists of 10 questions with scores ranging from 0 to 40, screening for risk levels of alcohol use. According to the AUDIT validation study, a score of 1–6 indicates low risk drinking, a score of 7–15 represents hazardous drinking and a score of 16 or higher signifies alcohol use disorders. A person with a positive response for question three in AUDIT, endorsing any frequency of binge drinking within the past one year, was classified as engaging in binge drinking. Binge drinking was the consumption of six or more drinks on one occasion. In order to capture the full spectrum of binge drinking, it was decided to include all persons who gave an affirmative response to any frequency of binge drinking as it was presumed that there can be certain personnel who may consume alcohol during their leave due to restrictions to alcohol consumption in the camp. A similar method was also used to describe binge drinking in the Australian military [15].

The first section of the SAQ consisted of questions on demographic, socio-economic, occupational and health information. The presence of combat exposures in this section was based on a ‘yes’ response to any of the 13 statements regarding such experiences. The presence of chronic disease and the additional health-related question on the presence of medical disposal were included to gauge the health status of the participants as self-report questions.

The second section of the SAQ focused on alcohol consumption and drinking patterns. It included simple yes/no questions to assess whether participants had ever consumed alcohol and whether they were currently drinking alcohol. Those currently drinking were also asked where they most often purchased alcohol, with three response options provided. Questions regarding the main reasons for drinking and the types of alcoholic beverages consumed over the past year allowed participants to select multiple responses.

Questions on serious consequences from drinking, risk behaviours, productivity loss, and military culture in sections three and four of the SAQ were adapted from the alcohol use section of the 2015 HRBS [16]. Individuals were classified as experiencing serious consequences from drinking if they endorsed any of 19 related statements (e.g., “I had trouble on the job because of my drinking,” “I didn’t get promoted because of my drinking,” or “I spent time in prison/cell because of my drinking”). Risk behaviours were identified through endorsement of any of three statements (“I operated power tools or machinery when I had too much to drink”, “I drove a car or other vehicle when I had too much to drink” and “I rode in a car or other vehicle driven by someone who had too much to drink”). Past-year job-related productivity loss was based on endorsement of any of six statements, such as “I worked below my normal level of performance because of drinking, a hangover, or an illness caused by drinking.” Perception of military culture as supportive of drinking was based on agreement or strong agreement with any of eight related statements. In addition, section four of the SAQ assessed the usual driving habit, seatbelt use while driving, sex with commercial sex workers, multiple sexual partners, sleep disturbances, and suicide attempts through direct questioning.

Section five of the SAQ addressed tobacco smoking. Those who were ever smoking were defined as individuals who smoked 100 smoking tobacco products (i.e. cigarettes/sticks of bidi/cigars/times of pipe) in the lifetime. Those who were never smoking were defined as individuals that have never smoked or have smoked less than 100 smoking tobacco products in their entire life, including those currently smoking but who have smoked less than 100 smoking tobacco products in their entire life [17]. This section included a series of statements describing smoking habits, from which participants selected the statement that best described their own.

Section six of the SAQ focused on the use of other substances, specifically smokeless tobacco and cannabis. Current smokeless tobacco users were defined as those that are currently using smokeless tobacco since the month before, and current cannabis users were those who have used cannabis over the last 12 months.

Section seven addressed adverse childhood experiences during the first 18 years of the life and positive responses for childhood physical abuse, domestic violence, and sexual abuse were based on the endorsement of one of a series of four questions each on the adverse experiences [18]. For example, the question on physical abuse during childhood included statements such as, “did a household member sometimes, often, or very often push, grab, shove, or throw something at you?” to which participants responded with either ‘yes’ or ‘no’. Questions about the father’s drinking behavior and being induced to drinking during childhood were included as direct questions based on previous research [19]. Further details on the study instrument are available in a previous manuscript by the authors based on this research [20].

The multidisciplinary panel of experts confirmed the judgmental validity of the SAQ. The SAQ was translated to Sinhalese and then back translated to English. The General Health Questionnaire (GHQ)-12 validated for Sri Lanka [21] consisted of 12 statements which were rated using a four-point scale. The Sinhalese GHQ-12 [22] demonstrated good internal consistency (Cronbach’s alpha = 0.88). Factor analysis revealed a two-factor solution accounting for 45% of total variance (correlation coefficient = 0.65). Criterion validity was established [21] by comparing results against psychiatric diagnosis as the gold standard. The GHQ-12 was included in the SAQ to identify the presence of mental distress. As the standard scoring method 0-0-1-1 was used, the cutoff value was taken as a score of two or more as per the validation study [21]. Hence, a participant with a score of ≥ 2 was classified with the presence of mental distress whereas those with a score < 2 with the absence of mental distress.

SF-36 was validated among Sri Lankan soldiers and matched healthy males using triangulation [23]. A multidisciplinary expert panel established judgmental validity. Convergent-discriminant validity was confirmed against the Nottingham Health Profile, with strong correlations for similar dimensions (highest r = 0.785) and weak for dissimilar ones (lowest r = 0.187). Construct validity assessment yielded 5–6 factors with eigen values > 1.0 across both groups, mirroring the tool’s dimensions. Reliability was satisfactory with Cronbach’s alpha > 0.8 across all dimensions.

Following precedent in military research [7, 24, 25], only the five items assessing functional impairment were included from the validated Short Form (SF)-36. This selective approach was appropriate as military personnel are typically physically healthy, making the questions about daily physical activity limitations less relevant to our study. Additionally, this targeted selection prevented redundancy, as the GHQ-12 already comprehensively assessed overall mental wellbeing. The questionnaire was pre-tested at the Army hospital, Colombo.

Data analysis

SPSS (version 21) software was used for data entry and data analysis. The prevalence of binge drinking with 95% CI and age-specific prevalence rates were calculated using 10-year intervals instead of 5-year intervals due to the limited number of individuals in the oldest and youngest age groups. Using the age weights of the WHO World Standard Population 2000–2025 [26], the age-standardized prevalence of binge drinking in SLA was calculated through direct standardization. This involved multiplying each age-specific prevalence rate by the corresponding age weight and summing the results to obtain the final value. The patterns of alcohol consumption among those engaged in binge drinking were summarized using frequency distribution tables. This included the median number of alcohol units consumed on a typical day. As there was no definition of a standard drink for Sri Lanka, the standard of 10 g alcohol being equivalent to one unit was taken as a reference for this calculation [27].

The bivariate analysis was conducted among those currently consuming alcohol during the past 12 months (n = 1088). The dependent variable was binge drinking versus non-binge drinking. Non binge drinking is defined as having consumed alcohol within the past 12 months without engaging in binge drinking during that period. Binge drinking and non-binge drinking were identified based on question three of the AUDIT. Individuals who reported binge drinking at any frequency—less than once a month, once a month, once a week, or daily/almost daily—over the past 12 months were classified as engaging in binge drinking. Those who had not engaged in binge drinking during this period were categorized as non-binge drinking. The independent variables potentially associated with binge drinking were categorized into two levels.

To identify potential confounders and independent variables for inclusion in the logistic regression analysis, a liberal significance threshold of p- ≤0.25 [28] was used. This approach captured variables that might have meaningful associations despite not reaching conventional statistical significance levels. Several factors known to have a complex interrelationship with binge drinking, including sleep disturbances, tobacco smoking, cannabis use, and mental distress were identified. All these variables were subsequently incorporated into the multiple logistic regression model to comprehensively assess factors associated with binge drinking while accounting for potential confounding effects.

All subgroups included in the analysis had individuals experiencing the outcome event. With 1088 participants and binary independent variables, the sample size was deemed adequate for logistic regression. The assumption of multicollinearity was not applicable due to all variables being binary.

Missing data analysis showed 99.4% completeness. Thus, multiple imputation, generating five imputed datasets, was used to address missing data. Binary logistic regression was performed on the multiply imputed dataset, adjusting for confounders and yielding pooled ORs. The Omnibus test was significant across all multiply imputed datasets (χ2 = 182.195-182.478, df = 28, p < 0.001), with Nagelkerke R² indicating that only 20.6–20.9% of the variance was explained by the model. However, the Hosmer-Lemeshow test indicated a good fit (p = 0.524–0.933) as the p-values across the five datasets were not statistically significant. With the confounding accounted for, adjusted ORs had a > 10% difference than the unadjusted ORs.

Results

Descriptive data

Out of the 1337 participants selected for the study, 1301 consented to participate giving a response rate of 97.3%. Thirty participants did not respond to the critical two questions on ever drinking and current drinking and of these, 16 left most of their questionnaire blank, failing to complete at least six of the nine sections in the SAQ. After excluding those 30 participants, the final sample taken for the analysis was 1271. This accounted for 95.1% of the calculated sample.

Mean age of the participants was 31.1 years (SD ± 7.3) with almost 90% of the participants 40 years or younger. The highest proportion of participants had studied between grade 9 up to ordinary levels (40.4%) and a majority were currently married (58.7%). Overall, other ranks accounted for 95.8% of the total sample with Privates and Lance Corporals/Corporals comprising 81.8% of the sample. The highest proportion of participants (68.0%) were from infantry regiments. Majority (83.2%) were living in their camps. Over half have served more than 10 years (55.6%). Nearly 89% of the participants received a total income of more than Rs. 25,000.00 and less than 75,000.00. A majority of 96.2% have never been deployed outside of Sri Lanka and 67.4% have not been exposed to combat. Only 15.6% reported being diagnosed or told by a doctor to have any type of chronic disease and 10.8% were low medically categorized.

Prevalence and patterns of binge drinking

Binge drinking was reported by 51.2% of the study population (n = 651, 95% CI = 48.5- 54.0%).

According to Fig. 2, the age standardized prevalence of binge drinking was 28.3%. The prevalence of binge drinking among younger service members (40 years and younger) was similar as seen in Fig. 2. This prevalence was lower than that of 41-50-year-olds and higher than that of members older than 50 years.

Fig. 2.

Age specific prevalence rates and age standardized prevalence rate of binge drinking

Those who binge drink consumed 5.6 median number of units of alcohol on a typical day [Inter Quartile Range (IQR) 3.9]. The median age of first consumption of alcohol was 18 years (IQR 3) and among them 69.3% (n = 451) have ever thought of quitting or have attempted to quit drinking alcohol. Other information on the pattern of binge drinking is demonstrated in Table 1.

Table 1.

Pattern of binge drinking

Pattern of binge drinking	Frequency	Percentage
Binge drinking frequency among those who binge drink
Daily or almost daily	6	0.9
Once a week	99	15.2
Once a month	328	50.4
Fewer than a month	218	33.5
*Place of alcohol purchase
Mainly from camp	108	16.6
Mainly outside the camp	421	64.8
Equally	121	18.6
Types of alcoholic drinks currently consumed over the past 1 year
Strong beer/stout	289	44.4
Beer	536	82.3
Toddy	225	34.6
Kassippu	101	15.5
Arrack	546	83.9
Whiskey	171	26.3
Gin	138	21.2
Wine	119	18.3
Other liquor	83	12.8
Main reason(s) for drinking
Stress in personal life	101	15.5
Work stress	87	13.4
For pleasure	467	71.7
Due to loneliness/boredom	100	15.4
To get along with seniors	101	15.5
Other reasons	127	19.5

*Denominator = 650

According to Table 1, binge drinking monthly or less was the most popular pattern and only very few were binge drinking daily or almost daily (0.9%, n = 6). A majority of individuals who engaged in binge drinking purchased alcohol outside the camp.

For the types of alcoholic beverages currently consumed within the past one year and for main reason(s) for drinking, the participants could choose multiple response options. Arrack and beer were the most popular alcoholic drinks, and pleasure was the main reason of alcohol use among those who binge drink.

Factors associated with binge drinking

The results of the bivariate analysis to determine the strength of the association between the outcome of binge drinking and the selected predictor variables, with the unadjusted odds ratio and the respective 95% CIs, are demonstrated in Table 2.

Table 2.

Association between binge drinking and socio-demographic, health related, occupational and other factors

Variable	Drinking status		OR (95% CI) P value
	Binge drinking n (%)	Non binge drinking n (%)
Age group (n = 1088)
≤ 30 years	318 (48.8)	212 (48.5)	OR = 1.01 (0.8–1.3) p = 0.963
≥ 31 years*	333 (51.2)	225 (51.5)
Educational status (n = 1088)
Up to O/L	337 (51.8)	203 (46.5)	OR = 1.24 (0.97–1.58) p = 0.098
Passed O/L & beyond*	314 (48.2)	234 (53.5)
Marital status (n = 1088)
Not currently married	286 (43.9)	168 (38.4)	OR = 1.26 (0.98–1.6) p = 0.082
Currently married*	365 (56.1)	269 (61.6)
Presence of chronic diseases (n = 1088)
Yes	104 (16.0)	58 (13.3)	OR = 1.24 (0.88–1.76) p = 0.254
No*	547 (84.0)	379 (86.7)
Low medically categorized (n = 1088)
Yes	80 (12.3)	32 (7.3)	OR = 1.77 (1.15–2.72) p = 0.01
No*	571 (87.7)	405 (92.7)
Number of children among those who are married/separated/divorced/widowed (n = 653)
No children	60 (15.8)	52 (19.0)	OR = 0.80 (0.53–1.20) p = 0.325
One child or more*	320 (84.2)	221 (81.0)
Rank (n = 1088)
Other ranks	619 (95.1)	418 (95.7)	OR = 0.88 (0.49–1.57) p = 0.773
Officers*	32 (4.9)	19 (4.3)
Within other ranks (n = 1037)
Corporal and below	391 (63.2)	251 (60.0)	OR = 1.14 (0.88–1.47) p = 0.343
Sergeant and above*	228 (36.8)	167 (40.0)
Enlistment force (n = 1088)
Regular	455 (69.9)	293 (67.0)	OR = 1.14 (0.88–1.48) p = 0.355
Volunteer*	196 (30.1)	144 (33.0)
Service arm (n = 1088)
Infantry	458 (70.4)	285 (65.2)	OR = 1.27 (0.98–1.64) p = 0.086
Support & Service*	193 (29.6)	152 (34.8)
Residence (n = 1088)
Living in	539 (82.8)	371 (84.9)	OR = 0.86 (0.61–1.19) p = 0.404
Living out*	112 (17.2)	66 (15.1)
Years worked away from home (n = 1081)
More than 10 years	278 (42.9)	194 (44.8)	OR = 0.93 (0.72–1.18) p = 0.579
10 Years or less*	370 (57.1)	239 (55.2)
Total military service (n = 1087)
More than 10 years	354 (54.5)	245 (56.1)	OR = 0.94 (0.73–1.20) p = 0.646
10 years or less*	296 (45.5)	192 (43.9)
Net salary (n = 1077)
Rs 50,000 or less	450 (69.8)	296 (68.5)	OR = 1.06 (0.82–1.38) p = 0.713
More than Rs. 50,000*	195 (30.2)	136 (31.5)
Additional income (n = 1077)
Has additional income	96 (14.9)	81 (18.8)	OR = 0.76 (0.55–1.05) p = 0.111
No additional income*	549 (85.1)	351 (81.2)
Deployment outside of Sri Lanka (n = 1088)
Yes	29 (4.5)	14 (3.2)	OR = 1.41 (0.74–2.70) p = 0.379
No*	622 (95.5)	423 (96.8)
Combat exposure (n = 1086)
Present	226 (34.7)	129 (29.7)	OR = 1.26 (0.97–1.64) p = 0.094
Absent*	425 (65.3)	306 (70.3)
Serious consequences (n = 1087)
Present	250 (38.4)	83 (19.0)	OR = 2.65 (1.99–3.53) p = 0.001
Absent*	401 (61.6)	353 (81.0)
Risk behaviours (n = 1087)
Present	414 (63.6)	199 (45.6)	OR = 2.08 (1.63–2.66) p = 0.001
Absent*	237 (36.4)	237 (54.4)
Alcohol related productivity losses (n = 1087)
Present	83 (12.7)	19 (4.4)	OR = 3.21 (1.92–5.36) p = 0.001
Absent*	568 (87.3)	417 (95.6)
Military culture (n = 1087)
Supportive of drinking	383 (58.9)	246 (56.3)	OR = 1.11 (0.87–1.42) p = 0.425
Not supportive*	267 (41.1)	191 (43.7)
Usual driving habit (n = 1087)
Sometimes/ always drive over the speed limit	205 (31.5)	101 (23.1)	OR = 1.53 (1.16–2.02) p = 0.003
Do not drive or always drive within the speed limit*	445 (68.5)	336 (76.9)
Seat belt use (n = 1082)
Sometimes/always do not wear the seatbelt when driving	180 (27.8)	86 (19.8)	OR = 1.56 (1.17–2.10) p = 0.003
Do not drive or always wear the seat belt when driving*	467 (72.2)	349 (80.2)
Ever had sex with a Commercial Sex Worker (CSW) n = 1088
Yes	103 (15.8)	34 (7.8)	OR = 2.23 (1.48–3.35) p = 0.001
No*	548 (84.2)	403 (92.2)
Multiple sexual partners (n = 1087)
Yes	66 (10.2)	25 (5.7)	OR = 1.86 (1.16- 3.00) p = 0.013
No*	584 (89.8)	412 (94.3)
Sleep disturbances (n = 1087)
Yes	107 (16.5)	47 (10.8)	OR = 1.64 (1.13–2.36) p = 0.011
No*	543 (83.5)	390 (89.2)
Ever tried to kill yourself (n = 1088)
Yes	50 (7.7)	10 (2.3)	OR = 3.55 (1.78–7.08) p = 0.001
No*	601 (92.3)	427 (97.7)
Smoking tobacco use status (n = 1084)
Ever smoking	456 (70.3)	229 (52.6)	OR = 2.13 (1.65–2.74) p = 0.001
Never smoking*	193 (29.7)	206 (47.4)
Smokeless tobacco use (n = 942)
Current user	139 (24.3)	61 (16.5)	OR = 1.63 (1.16–2.27) p = 0.005
Non user*	433 (75.7)	309 (83.5)
Cannabis use status (n = 1083)
Current user	306 (47.2)	117 (26.9)	OR = 2.43 (1.87–3.16) p = 0.001
Non user*	342 (52.8)	318 (73.1)
SF-36 item- Cut down the amount of time spent on work or other activities (n = 1087)
Yes	60 (9.2)	15 (3.4)	OR = 2.86 (1.60–5.11) p = 0.001
No*	590 (90.8)	422 (96.6)
SF-36 item- Accomplished less than would like (n = 1087)
Yes	95 (14.6)	28 (6.4)	OR = 2.5 (1.61–3.88) p = 0.001
No*	555 (85.4)	409 (93.6)
SF-36 item-Were limited in the kind of work or other activities (n = 1086)
Yes	120 (18.5)	46 (10.5)	OR = 1.93 (1.34–2.78) p = 0.001
No*	529 (81.5)	391 (89.5)
SF-36 item-Had difficulty performing the work or other activities (n = 1086)
Yes	137 (21.1)	48 (11.0)	OR = 2.17 (1.52–3.09) p = 0.001
No*	512 (78.9)	389 (89.0)
SF-36 item-Physical health or emotional problems interfered with social activities (n = 1087)
All or most of the time	51 (7.8)	27 (6.2)	OR = 1.29 (0.79–2.09) p = 0.364
Other*	600 (92.2)	409 (93.8)
Mental Distress (n = 1088)
Present	235 (36.1)	63 (14.4)	OR = 3.35 (2.46–4.58) p = 0.001
Absent*	416 (63.9)	374 (85.6)
Childhood physical abuse (n = 1088)
Positive	133 (20.4)	58 (13.3)	OR = 1.68 (1.19–2.35) p = 0.003
Negative*	518 (79.6)	379 (86.7)
Domestic violence during childhood (n = 1088)
Positive	61 (9.4)	28 (6.4)	OR = 1.51 (0.95–2.40) p = 0.102
Negative*	590 (90.6)	409 (93.6)
Childhood sexual abuse (n = 1087)
Positive	78 (12.0)	27 (6.2)	OR = 2.06 (1.31–3.25) p = 0.002
Negative*	573 (88.0)	409 (93.8)
Living with a problem drinker/alcoholic (n = 1088)
Yes	119 (18.3)	64 (14.6)	OR = 1.3 (0.94–1.82) p = 0.137
No*	532 (81.7)	373 (85.4)
Minimum number of drinks father consumed per week (n = 809)
4 times or more per week	117 (23.4)	67 (21.8)	OR = 1.10 (0.78–1.54) p = 0.659
3 times or less per week*	384 (76.6)	241 (78.2)
Induced to drinking during childhood (n = 1087)
Yes	52 (8.0)	15 (3.4)	OR = 2.44 (1.35–4.39) p = 0.003
No*	599 (92.0)	421 (96.6)

*Reference category

The independent variables selected to the logistic regression model using a liberal p-value (≤ 0.25) can be viewed in supplementary file I.

Table 3 presents the pooled logistic regression coefficients (β), along with the adjusted odds ratios (ORs) and corresponding confidence intervals (CIs) for the factors associated with binge drinking identified in the final model. After controlling for confounding effects, six independent variables were found to be statistically significant predictors of binge drinking. The adjusted ORs demonstrate more than a 10% difference compared to their unadjusted counterparts, reflecting the consideration of confounding in the final model.

Table 3.

Statistically significant factors identified from logistic regression

Variable	Β	SE	P value	Adjusted Odds ratio	95% Confidence Interval
					Lower	Upper
Mental distress present	0.902	0.183	0.001	2.46	1.72	3.53
Had sex with a CSW	0.654	0.237	0.006	1.92	1.21	3.06
Ever smoking	0.527	0.146	0.001	1.69	1.27	2.25
Serious consequences	0.455	0.169	0.007	1.58	1.13	2.20
Current cannabis user	0.327	0.158	0.039	1.39	1.02	1.89
Combat exposure	0.314	0.158	0.047	1.37	1.00	1.87

After adjusting for all variables in the model, six factors were significantly associated with the odds of binge drinking as demonstrated in Table 3. Those with mental distress were 2.46 times as likely to binge drink than those without such distress. Those who had sex with a CSW within the past 12 months were 1.92 times as likely to binge drink compared to those who did not have sex with a CSW. Those ever smoking were 1.69 times as likely to binge drink compared to those never smoking. Those with serious consequences from drinking were 1.58 times as likely to binge drink compared to those who did not have serious consequences. Those currently using cannabis were 1.39 times as likely to binge drink than non-users. Those with combat exposure were 1.37 times as likely to binge drink than those without combat exposure.

Discussion

The prevalence of binge drinking in the SLA was high at 51.2% as frequently observed in past research on military alcohol use. Nevertheless, militaries in other countries seem to have higher prevalence rates and more severe patterns of binge drinking compared to the SLA. Moreover, a majority of personnel in the SLA engaged in binge drinking once a month or less frequently. Direct comparisons with other studies were challenging due to country-wise differences in the measure of the standard unit of alcohol and variations in the definition of binge drinking. For example, high binge drinking prevalence rates (61.4%) were observed among U.S. Navy male recruits [29]. An overall binge drinking prevalence of 30.5% was reported among Army personnel in the 2015 HRBS [16]. The definition of binge drinking in that study was the consumption of five or more drinks by males at least once over the past 30 days whereas the current study included all frequencies of binge drinking, including those who consumed alcohol less than once a month within the past year. Furthermore, the standard drink in the U.S. accounts for 14 g of pure alcohol [30]. Hence, binge drinking rates in the U.S military during the past 12 months could be much higher. Among male personnel of United Kingdom (UK) armed forces [31], 48% reported binge drinking on a weekly or daily basis. In the Royal Navy [32], only a small proportion of male respondents (4%), reported not binge drinking at all. Binge drinking was defined as the consumption of ≥ 9 units in a drinking session, with 48% engaging in binge drinking less than weekly, and a further 48% binge drinking at least weekly. The standard drink in the UK accounts for 8 g of alcohol [33] further complicating direct comparisons. A high prevalence of binge drinking (81.9%) was reported among the Australian Defense Forces [15], which used a binge drinking definition similar to the current study.

Among the general population of Sri Lanka (18–69 years) the binge drinking prevalence was 16.8% [8]. However, this was reported only for the past 30 days whereas the current study included even those who consumed less than monthly. Moreover the current study consisted of a younger demographic with a mean age of 31 years which could have resulted in higher prevalence. The age standardized prevalence of binge drinking (past 30 days) for Sri Lankan males in 2020 was 18.9% [34] whereas the current study reported 28.3%. Thailand reported the highest age standardized binge drinking prevalence rates for the region at 31.2%. Nevertheless, when binge drinking within the past year is considered the age standardized binge drinking rates for the general population of Sri Lanka could be much higher making comparisons rather cumbersome. The most recent weighted prevalence of binge drinking (past 30 days) among the U.S military male personnel in 2018 was 35.2% [5]. This analysis applied analytic weights to ensure the sample was representative of the eligible service member population in the U.S which was much higher than the equivalent metric reported as 28.3% in the current study.

In the SLN the binge drinking prevalence was 26.1% [6] and binge drinking was defined as consumption of six or more drinks on one occasion at least once a month. Data collection in the study in this study commenced three months after combat operations ended. There were restrictions in alcohol use with limited availability and access to alcohol during that time which may have led to comparatively lower binge drinking prevalence in the Navy [6].

During the height of the COVID-19 pandemic, Army personnel were restricted to camps with limited access to alcohol, likely leading to lower consumption of alcohol. However, by the latter part of 2022, when data collection commenced, the Army had resumed its normal routine of duties. Therefore, this study most likely reflects typical patterns of alcohol use among Army personnel rather than temporary fluctuations due to pandemic-related restrictions.

Those currently drinking alcohol (within the past 30 days) among the general population of Sri Lanka consumed an average of 4.3 standard drinks per drinking occasion and in the latest survey this was 3.9 [9]. This is still higher than the South East Asia regional (26.3 g) and global (32.8 g) levels [4]. Alcohol use in the military can be significantly influenced by the general population levels [6] which may have resulted in a high median number of units of alcohol consumed on a typical day (5.6 units) in the SLA. This is consistent with past literature that has consistently demonstrated that military personnel who drink tend to drink more heavily [35].

Arrack was chosen by a majority engaged in binge drinking (84%). Similarly, spirits accounted for 87.9% of the total alcohol consumed in the South East Asia region [4] and almost 85% of males who current consumed alcohol in Gampaha district of Sri Lanka preferred to drink spirits and 52.9% preferred arrack [36].

The use of alcohol is described as a part of the medium of sociability to break down barriers serving as a social glue to bond people together, and the Army community being “close knit”, there is a high level of peer pressure, as large groups of men often work and live together, sharing accommodation [31]. Drinking alcohol was accepted in the military culture, with peer pressure to engage in heavy drinking [37]. A majority of those engaged in binge drinking in the current study consumed alcohol “for pleasure”. Among the third main reason for drinking was “to get along with seniors”. In the U.S Army, 57% of heavy drinkers also reported high pleasure seeking/enjoyment motive to consume alcohol [38].

The median age of first consumption of alcohol among those engaged in binge drinking was 18 years, which is around the time of recruitment to the Army. Among the general population in Sri Lanka, the mean age of onset of alcohol use was also 20.8 years [39]. Compared to the general population, those who currently drink alcohol in the SLA start consuming alcohol at a relatively early age. Moreover, it is evident from research that there is a propensity for drinkers to self-select to military service [40]. This poses an important gateway to strengthen preventive interventions at recruitment, at military training schools and academies. Over 60% of those engaged in binge drinking have ever thought of quitting or have attempted to quit drinking alcohol. This is of high importance, as a majority of drinkers could be more responsive to a preventive intervention.

There was no statistically significant association binge drinking and any of the other socio-demographic, health related, or occupational factors, except for combat exposure. This contrasts with past research which demonstrated that lower ranking personnel had a higher likelihood of drinking [15, 32, 35]. In the SLA, the majority (83.2%) of personnel reside in camps, where access to alcohol is more limited for lower ranks. Privates (34.6%) do not have access to bar facilities and only beer is available at the corporals’ club, which is the mess designated for corporals (47.2%). Alcoholic beverages can be purchased in the Army mess by officers and senior soldiers, but only during designated off-duty hours on specific days of the week. Therefore, access to alcohol within Army premises is more restricted compared to the general population. Additionally, alcohol is not sold at a concessionary rate, keeping its cost comparable to that outside military settings. However, once personnel proceed on leave they are free to consume alcohol without restrictions. This could be the reason for over 60% of drinkers to mostly purchase alcohol outside the camp, in contrast to the U.S Army, where a higher proportion of 43.3% purchased on base [16]. Moreover, challenges in recruitment encourage enlistment by communities already affected by heavy drinking [35]. Especially those from socially deprived backgrounds [32]. Individuals with risk taking characteristics often join the military and while those characteristics may make them successful in combat, can also lead to alcohol misuse [2]. Hence, there seem to be a propensity for higher levels of alcohol consumption in the Army overall. The median age of initiation of alcohol in the current study was also 18 years. The younger age of starting to drink may reflect a pre-enlistment vulnerability [31]. This predisposition towards drinking from enlistment onwards may be the reason for the lack of difference observed for binge drinking across ranks similar to age.

Unlike deployment outside Sri Lanka, exposure to combat had a statistically significant association with binge drinking in the current study. However, most past research indicates a significant association of alcohol use with deployment particularly in the presence of combat exposure [35, 41]. Moreover, specific deployment and combat experiences may impact alcohol use rather than deployment in general [37]. However deployment experiences could not be studied in depth in the current research, as only a small minority of 3.8% (n = 48) had ever been deployed outside of Sri Lanka. In addition, the SLA engages in mostly U.N peace keeping missions to Mali, Sudan and Lebanon rather than a direct combat role. A study on U.S national guard service members did not report a statistically significant increase in the post deployment estimated monthly drinks [42] further supporting the current study findings. Although there was an increase in alcohol consumption from baseline to post deployment first year, it dropped considerably by the second year post deployment, with levels even below the non-deployed personnel. This demonstrates that, though there can be an increase in alcohol use post deployment, it is not sustained and decreases with time, supporting the current study findings.

Military personnel endure various forms of trauma during combat, including exposure to killing, which has been independently linked to problematic alcohol use, as well as increased frequency and quantity of alcohol consumption [43]. Killing in combat may serve as a unique risk factor for heavy drinking and alcohol-related problems among military personnel post-deployment. All possible combat exposures including killing in combat as identified in literature to screen for the presence of combat exposure were included to the SAQ and nearly one third of the SLA was found to be exposed to combat (32.6%). Although it has been 14 years since the end of the war during which the SLA occupied in an intense military offensive, the persisting effect of combat exposure on binge drinking resonates with significant associations demonstrated in past literature.

There was no statistically significant association between binge drinking and marital status, which is contradictory to most past research. However, the statistically significant likelihood of binge drinking among those with high-risk sexual behaviors observed in this study is consistent with findings from previous research conducted among male armed services personnel in Congo [44], in which most were not married but living with a partner (67%). Although a majority of participants (58.7%) in the current study were married, an overwhelming majority (83.2%) were living in the camp. Mobility, including overnight travel away from home, was associated with high risk sexual behaviours [45]. Hence, high risk sexual behaviours were highly probable in the current study, since the majority of personnel were living away from their spouse and the risk taking propensity is known to increase with alcohol use. Most individuals engaged in current drinking were only binge drinking once a month or less. Hence, it is plausible that there was no impact of the protective effect of a marriage on binge drinking, as many were residing in the camp and the frequency of binge drinking was much less anyway.

The statistically significant associations observed between binge drinking and serious consequences are consistent with findings from previous military research conducted in the U.S. Heavy drinking was associated with higher rates of alcohol problems, similar to the current study. A study on U.S military personnel in active duty [46] reported that those engaged in binge drinking were significantly more likely than those who do not binge drink to report a majority of alcohol related consequences when controlled for age and gender. Binge drinking individuals had a higher odds than those who do not binge drink to report serious consequences of drinking such as “got into a fight and hit someone” and received punishment, etc. Moreover, heavy alcohol users in U.S military on active duty showed nearly three times the rate of self-reported serious consequences than moderate/heavy drinkers [47].

Binge drinking was statistically significantly associated with smoking with ex-smokers and never smokers being protective factors among men in the UK armed forces [31]. The baseline survey (2001–2003) of the U.S. Military Millennium Cohort Study demonstrated that current smoking statistically significant increased the likelihood of new onset binge drinking [41]. These findings were maintained in the follow-up survey in 2004–2006. Among SLN personnel, a statistically significant association was observed between current smoking and hazardous drinking which was an AUDIT score of ≥ 8 [6]. Hence, past international and local research strongly supported the findings of the current study with regard to smoking. Likewise, a study that explored the use of cannabis among Navy personnel in Sri Lanka demonstrated a statistically significant association between cannabis use and hazardous alcohol use AUDIT ≥ 8 [48], similar to the findings of the current study.

There was also a statistically significant association between binge drinking status and four of the items in the SF-36 that assess functional impairment. In the Australian Defense Forces, higher AUDIT scores demonstrated clear declines in scores on all three subscales of SF-36; general health, role physical and social functioning [15]. However, these items that assessed functional impairment did not emerge as statistically significant factors that increased the likelihood of binge drinking in the final regression model. Similar findings were reported for binge drinking among SLN personnel [6]. This could be due to the low prevalence of binge drinking among the Navy personnel during the period of data collection, which was three months following the conclusion of the armed conflict. Conversely, in the presence of a high prevalence binge drinking (47%) among UK military personnel [24], the inability of the study to demonstrate any significant association between binge drinking and functional impairment through the SF-36 came as an unexpected finding. It was suggested that increasing the number of drinks per occasion to re-define binge drinking for future studies may be useful, as young military personnel in UK often drink six drinks on an occasion without experiencing any functional impairment at that level. The current study findings could be due to the fact that the severity of binge drinking was insufficient to result in functional impairment among Army personnel as most were binge drinking rather infrequently.

Mental distress, was found to significantly increase the likelihood of binge drinking in the current study. Similar findings were observed repeatedly in local and international studies. Among U.S military Millenium Cohort participants, baseline symptoms of depression, PTSD or both were found to statistically significantly increase the likelihood of the new onset of alcohol-related problems [41]. Similar findings were reported for the U.S military personnel in active duty [35]. Possible PTSD was significantly associated with binge drinking, as was depression. In the SLN, a statistically significant association between AUDIT scores ≥ 8 with GHQ case-ness was reported as well [6].

Military trauma accumulated during the military service predisposes to stress related mental health issues including PTSD, depression and problematic alcohol use [43]. Combat exposure involves highly traumatic events that threaten life or serious injury, overwhelming an individual’s ability to cope, increasing the risk of mental diseases [49]. Trauma triggers biological changes, including limbic system dysfunction, Hypothalamic-Pituitary-Adrenal axis dysregulation, and neurotransmitter imbalances affecting arousal and opioid systems in brain function, all of which are strongly linked to PTSD, mental illness, and substance use disorders [50]. Military medical professionals report that nearly all combat veterans experience stress, often leading to excessive alcohol use as a coping mechanism [35]. In a study of UK armed forces personnel [51], drinking to cope was more prevalent among individuals with alcohol misuse as well as those who binge drink. Social pressure was also linked to binge drinking. The association between coping motives and alcohol misuse may be explained by the self-medication hypothesis, which suggests that individuals consume alcohol to alleviate psychological symptoms.

However, underlying mental disorders were associated with a higher risk of alcohol misuse (AUDIT scores ≥ 8) when controlled for covariates including combat exposure even among the highly resilient U.S Special Forces personnel, who reported a low prevalence of PTSD post deployment [52]. A recent study reported that nearly a quarter of the SLA personnel were suffering from burnout (23.2%), and 14 correlates of burnout were identified, which included dysfunctional coping strategies [53]. In the absence of the stressors of frequent deployment in the SLA, the statistically significant association between binge drinking and mental distress could be partly explained by the persistent effects of past combat exposure in the current study. In addition, work-related stressors in the prevailing context could be explanatory, such as the recent COVID-19 control/prevention related tasks, handling of civilians during the ‘aragalaya’ protests, and during the fuel crisis for which the military men were not especially trained to perform.

Alcohol use is a sensitive behavior that may have serious negative repercussions on the careers of military personnel. Although, social desirability bias was inevitable, many measures were undertaken to minimize it. Due to the sensitive nature of questions, a SAQ was used as frequently employed in past military research in Sri Lanka [6, 12, 53]. During data collection, ethical procedures were strictly followed, ensuring voluntary participation, especially considering the fact that military is widely regarded as a captive population.

In this study, we categorized individuals based on behavioral definitions established in past research, ensuring our approach was both methodologically sound and aligned with existing evidence. For instance, the classification of the smoking behaviour was based on a comprehensive Sri Lankan study [17] that adapted smoking definitions from the Centers for Disease Control and Prevention (CDC) guidelines. This ensured consistency with global standards while maintaining local relevance, further supported by qualitative input from experts. While research on adolescents [54, 55] suggests that nicotine-related changes can occur before the 100-cigarette threshold, a cross-sectional study on adults [56] highlights the need for longitudinal research to determine the most clinically meaningful smoking definitions. Given that our study focused on adult males (minimum age 19 years, mean age 31 years), the natural trajectory of smoking initiation may differ for those who start earlier or later in life [55] supporting the rationale for the pragmatic classification used.

The study population consisted only of male personnel in active service. Hence, findings can be generalized to all males in active service but not to females. As many international and local studies have consistently shown lower alcohol use prevalence rates among females compared to males, it was decided to conduct the study on males. Overall, women in the military were less likely to use alcohol than men, as frequently demonstrated in past research. According to the 2015 STEPS survey, there were only 0.5% females that currently consumed alcohol within the past 30 days in Sri Lanka [8], even when females were oversampled to consist of 61% of the sample. The vast majority (96.5%) of females were lifetime abstainers. In addition, there were no females in the high end or intermediate level of drinking, with only 0.5% females being in the lower- end level of drinking.

Although a cross-sectional study design was the most suitable approach to estimate the prevalence of alcohol use and identify factors associated with alcohol consumption, the nature of this design provides insights into associations rather than definitive causal relationships. Furthermore, the modest R-squared value in the logistic regression model reflects a limitation in the study. Important genetic factors like ADH1B, ALDH2, GABRA2, and CHRM2 variants, known to influence alcohol consumption patterns [57], beyond what family history alone captures could not be measured. These genetic markers, along with other unmeasured biological confounders, likely account for the unexplained variance in the results. Funding constraints prevented incorporating these biological determinants to the study, despite the comprehensive inclusion of key socio-demographic, occupation and behavioral factors.

Existing preventive measures to curb alcohol use among Army personnel in Sri Lanka include a dedicated Non-Communicable Diseases (NCD) Prevention campaign, which incorporates early detection and health education on NCDs and their risk factors, including alcohol use. Additionally, under the guidance of the Directorate of Army Preventive Medicine & Mental Health Services, Army health staff conduct health education sessions for troops at camps, training institutions, and academies. Furthermore, alcohol use is prohibited during basic training for recruits. Given the high prevalence of binge drinking among Army personnel in Sri Lanka, strengthening existing prevention strategies as well as implementing evidence-based prevention and treatment strategies, including the use of digital tools, should be considered. While cultural barriers persist, well-designed policies can help reduce alcohol-related harm in military settings.

Conclusion

Although the binge drinking prevalence seems to be high in the current study, in line with overall high rates often reported in militaries worldwide, binge drinking appears to be less frequent and less severe than in countries such as UK and U.S. The fact that the majority were binge drinking once a month or less frequently is coherent with the fact that the majority of Army personnel were of lower ranks with minimal/no access to alcohol within their camps, as bar facilities are considered a privilege of higher ranks in the military setting. In addition, majority were living in the camp, which makes access to increased amounts of alcohol rather challenging. Hence, it is highly likely that they may be consuming alcohol during their leave, which was evident from a higher proportion claiming to purchase alcohol mainly outside the camp. The increasing prevalence of alcohol use, along with the high amount of units of alcohol consumed by the general population of Sri Lankan males, could have influenced the alcohol use among male Army personnel. Furthermore, self-selection of individuals with higher risk taking tendencies in to the military, combined with military customs and traditions, may also have contributed to increased prevalence.

Given the high prevalence rates of binge drinking, it is essential to advocate to the SLA’s high command for the implementation of evidence-based alcohol control policies and prevention programs, as well as to further strengthen existing initiatives. Primary prevention strategies should be initiated from recruitment and basic training onwards. As a majority of those engaged in binge drinking have thought about quitting, outreach screening and treatment programmes should be implemented at the battalion level, by integration with the existing primary health care services of the Army, along with expansion of mental health services to improve secondary prevention of alcohol use. A health promotion program to prevent binge drinking in the SLA should be carried out alongside initiatives aimed at preventing smoking, cannabis use, high-risk sexual behaviors and promotion of mental well-being. Moreover, the evaluation of other substance use, past combat exposure, and serious consequences of drinking should be given high priority in the mental health assessment of those who binge drink. Whilst being clinically managed there should be a support group for quitters or those who want to quit consuming alcohol and maintain sobriety. Raising awareness at all levels is essential to reducing stigma and eliminating barriers to accessing mental health care. Future studies are recommended to assess the role of deployment on alcohol use among recently deployed cohorts of Army personnel.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

AUDIT

Alcohol Use Disorders Identification Test

CI

Confidence Intervals

CSW

Commercial Sex Workers

GHQ

General Health Questionnaire

HED

Heavy Episodic Drinking

HRBS

Health Related Behaviour Survey

IQR

Inter Quartile Range

SAQ

Self-Administered Questionnaire

SF

Short Form

SLA

Sri Lanka Army

SLN

Sri Lanka Navy

UK

United Kingdom

U.S

United States

WHO

World Health Organization

Author contributions

L.V.J. (principal investigator) played a role in designing the research, overseeing data collection, analyzing and interpreting the data, and writing the manuscript. S.P. was involved in the research design, data analysis, data interpretation and in manuscript writing. S.S. was involved in research design, data interpretation and manuscript writing. All authors have reviewed the final draft and agree with the manuscript’s content.

Funding

A self-funded research by the principal investigator.

Data availability

The datasets generated and/or analysed during the current study are not publicly available due to the sensitive nature. Data can be made available on reasonable request.

Declarations

Ethics approval and consent to participate

Ethical approval from the ethics review committee of the Postgraduate Institute of Medicine, University of Colombo (reference: ERC/PGIM/2022/039) and permission from Army Headquarters was sought prior to data collection. Confidentiality was ensured and written informed consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

L.V.J. is a physician in the Sri Lanka Army, currently undergoing post MD training in Community Medicine at the Postgraduate Institute of Medicine, University of Colombo attached to the Ministry of Health. S.S. was the former Director of the Directorate of Army Preventive Medicine and Mental Health Services. The Sri Lanka Army was not involved in the design, execution of the study, or the writing of the manuscript.