Alcohol use is one of the top five leading risk factors for death and disability worldwide (World Health Organization, 2018). Research on alcohol use and alcohol-related problems has focused largely on how drinking harms the drinker. However, over the past few years, there has been a shift in research from alcohol-related problems suffered by the drinker to focus on the problems others suffer because of the drinker, i.e., alcohol’s harm to others, or AHTO (Laslett et al., 2023; Room et al., 2019).

In particular, research has evidenced the contribution of heavy or hazardous drinking to a range of significant harms to people other than the drinker (Quigg et al., 2019; Sundin et al., 2024). Previous research on AHTO has studied drink-driving (Dunn & Tefft, 2021), prenatal (Popova et al., 2023), and neonatal harms (Bandoli et al., 2023) overlooking broader harms to others’ health, quality of life, living conditions, and environmental resources which result in significant costs to individuals and society (Jiang et al., 2022). Most research on AHTO has been conducted with Western populations and explored various types of serious harm including family problems, financial difficulties, damage to property, and physical injuries (Kilian et al., 2024). While the degree of alcohol’s harm to others may differ between different countries and the types of harm defined (Laslett et al., 2020), certain sub-groups of victims are found to be strongly associated with reported harm across studies, namely women (Bloomfield et al., 2019; Laslett et al., 2020), young people (Beynon et al., 2019; Huhtanen & Tigerstedt, 2012), and the victim’s own hazardous (or binge) drinking (Nayak et al., 2019).

Yet, little is known about AHTO in the Southeast Asian region, where alcohol consumption is on the rise (World Health Organization, 2018). Understanding the effects of alcohol to others, and identifying the population groups most affected, will help to inform clinical practice and policies. This study aims to (i) analyze the prevalence of AHTO in Singapore's general population and evaluate the risk of alcohol-related harms to people other than the drinker, and (ii) examine the relationship of sociodemographic factors and an individual’s own alcohol consumption with experiences of AHTO or harm due to others’ drinking.

Methods

Sample and Design

This report utilized data from the Singapore Health and Lifestyle Survey, which was a nationwide, epidemiological cross-sectional study covering a range of physical and mental health conditions and risk factors (Central Narcotics Bureau, 2023) including questions on alcohol-related harm and hazardous alcohol use in the general population. A more detailed description of the study methodology has been reported elsewhere (Subramaniam et al., 2024). In brief, a disproportionate stratified sampling method was used to obtain a national probability sample of 6509 Singaporeans aged 15 to 65 years from a national registry. Respondents completed the survey in person or online, between April 2021 and July 2022. Since study recruitment had begun during the evolving COVID-19 situation, a hybrid approach was utilized to comply with the national public advisory guidelines to reduce direct contact with respondents in the study. The hybrid approach involved a brief door-to-door visit by a trained layperson, who conducted the verbal informed consent, and participants who expressed willingness to take part in the study were provided with a unique URL and password to complete the survey online. Eligibility for the study was limited to those currently residing in Singapore, who were not hospitalized or institutionalized, and able to be interviewed in one of the four official languages (English, Mandarin, Malay, or Tamil) of Singapore. This was an anonymous survey in which identifiers were not collected and/or linked to the respondents’ identity, i.e., data that can either directly identify the respondents or link the respondents to their identity were not collected in the study. Thus, verbal consent instead of written informed consent was obtained from all respondents. A maximum of ten attempts were made to reach respondents at their households. Varying visits to different times and days of the week maximized the likelihood of reaching eligible respondents. The study obtained a response rate of 73%. The institutional ethics board, the National Healthcare Group Domain Specific Review Board (NHG DSRB), approved all study materials, survey protocols, and research procedures.

Questionnaires

Alcohol’s Harm to Others (AHTO) Measures

The set of items on harm caused by someone else’s drinking, comprised (i) being pushed, hit, or assaulted; (ii) having one’s property vandalized; (iii) having family problems or marriage difficulties; and (iv) having financial trouble. Participants were asked the questions as follows: (i) “Have you been pushed, hit, or assaulted by someone who had been drinking?”; (ii) “Has your property been vandalized by someone who had been drinking?”; (iii) “Have you had family problems or marriage difficulties due to someone else’s drinking?”; and (iv) “Have you had financial trouble because of someone else’s drinking?” These items have been similarly used in other large-scale, national population surveys (Eum & Choi, 2021; Laslett et al., 2023; Wilsnack et al., 2018). We created a collapsed set for an indicator of any AHTO present (none or at least 1 harm), and four specific AHTO type indicators as follows: (i) physical harm, (ii) vandalism, (iii) family problem, and (iv) financial trouble, caused by someone else’s drinking. Respondents who answered “NA” were coded as “No.” In the analysis, AHTO is treated as a dichotomous outcome variable (None = 0 vs. at least 1 harm present = 1) and the data was coded as “missing data” if all items in the questionnaire were unanswered.

Alcohol Use Disorders Identification Test (AUDIT)

The AUDIT is a ten-item instrument designed by the World Health Organization (WHO) with good validity in detecting current hazardous drinking patterns as well as more severe alcohol-related problems based on three aspects: alcohol intake, dependence, and past-year adverse consequences (Moehring et al., 2018; Reinert & Allen, 2007). While the AUDIT was developed in the West, it has been translated and validated for use in a variety of other populations including India (Endsley et al., 2017) and Taiwan (Wu et al., 2008), as well as Singapore (Tan et al., 2010). It has demonstrated good internal consistency (Cronbach’s α = 0.79). Possible scores range from 0 to 40, with scores of 1 to 7 indicating low-risk or light drinking and scores 8 or higher indicating hazardous alcohol use and the likelihood of alcohol dependence warranting further surveillance (Reinert & Allen, 2007). For own drinking status, those who endorsed not drinking in their lifetime were coded as Abstained, and those who endorsed that they did not drink in the past year and those who endorsed that they drank in the past 1 year and had an AUDIT score of less than 8 were coded as Light drinking. The cutoff score of 5 or higher was used to determine hazardous alcohol use among adolescents (aged 15–19 years) (Liskola et al., 2018).

Sociodemographic Characteristics

Sociodemographic data included age, gender, ethnicity, marital status, personal income, education level, and employment status.

Data Analysis

Sampling weights were used in all analyses to adjust for probability of selection and nonresponse, and post-stratified by age, gender, and ethnic distributions, thus allowing our sample estimates to be representative of the Singapore resident population at the time of data collection. Categorical variables were examined as weighted percentages and unweighted frequencies. Continuous variables were presented as weighted mean with standard deviation (SD). We estimated the population prevalence of any AHTO and of the five AHTO types separately for men and women among those who experienced harm from someone else’s drinking. Binary logistic regression analyses were used to examine how the prevalence (any vs. none) of AHTO differed by demographic characteristics. The relationship of the reporting of AHTO with the participant’s own drinking status was also examined. The results of the binary logistic regression analyses were presented as odd ratios (OR) with 95% confidence intervals (CI). Standard errors were estimated using Taylor’s linearization to adjust for the disproportionate stratified sampling design. All analyses were performed using IBM SPSS, version 23.0 and Stata/MP 17.0, using two-sided tests at a significance level of 5%.

Results

Prevalence of Alcohol’s Harm to Others

Table 1 shows the sample characteristics of participants and key prevalence estimates of alcohol’s harm from drinking by others. Of the 6465 participants, 36.4% were in the youngest age group (15–34 years) and about 51% of the sample were female. A majority of the victims were Chinese (73.6%), attained a degree or higher (36.7%), married (57.8%), and employed/self-employed (75.9%), with either no personal income or income below $2000 (41.9%). About 672 (9.6%) of the participants gave some indication of having experienced the harmful effects from drinking by others and 3.4% or 182 participants were found to have hazardous drinking use. Men had much higher rates of hazardous drinking (5.2% vs. 1.6%), and more men reported experiencing alcohol’s harm from drinking by others (10.6% vs. 8.6%). Men also reported physical harm more often than women (8.0% vs. 5.9%). Women reported harms related to family problems/marriage difficulties (3.9% vs. 2.5%) and financial trouble (1.8% vs. 1.1%) more commonly than men.

Table 1 Prevalence of alcohol’s harm from others’ drinking (N = 6465) by gender
Full size table

Risk Factors for Alcohol’s Harm

Table 2 presents the various relationships between AHTO and sociodemographic characteristics of the participant. The univariable analysis shows that gender was associated with harm experienced from the drinking of others. Similarly, age, ethnicity, education, and marital status, as well as, own drinking status were associated with experiencing AHTO. In the fully adjusted analysis, participants in the oldest age group (50–65 years) were less likely to experience AHTO (adjusted OR 0.42, p = 0.002) compared to those from the youngest age group (15–34 years). AHTO was more likely to be experienced by Malays (adjusted OR 1.50, p = 0.014), and Indians (adjusted OR 1.83, p < 0.001) as compared to Chinese, and those who were separated, widowed, or divorced (vs. married; adjusted OR 1.91, p = 0.022). Additionally, participants with alcohol use measured as “light drinking” (adjusted OR 2.50, p < 0.001) and “hazardous drinking” (adjusted OR 4.31, p < 0.001) were more likely to report experiencing AHTO than abstainers.

Table 2 Relative odds of reporting any harm from the drinking of others by sociodemographic groups
Full size table

Discussion

This study is the first large-scale effort to explore the various kinds of harm that can occur due to another’s drinking in the Singapore general population. The study has further enumerated the different relationships that exist for risk factors surrounding the person harmed. In general, our findings indicate that the experiences of AHTO varied significantly by age, ethnicity, marital status, and own drinking status.

The current study found a relatively low rate of AHTO, with just under one in ten (9.6%) Singaporeans reporting at least one or more harmful consequences—physical assaults, family problems, property vandalized, and financial trouble—experienced as a result of someone else’s drinking. Our study’s prevalence estimate is consistent with previous findings from several nationally representative population studies in countries such as Denmark (Seid et al., 2015) and USA (Greenfield et al., 2015). However, it is important to note that these studies are relatively dated and focus on different types of harm. For example, Seid et al. (2015) found that among 2357 Danish residents, approximately 8% of respondents reported that they had experienced any of four areas of harm (work, finance, injury, and family). While Greenfield’s trend analysis of U.S. adults across four national surveys indicated that 11.2% of the general population reported experiencing at least one of three harm types from others’ drinking, with prevalence rates between 4.4% for family or financial harms, 4.9% for physical/aggression-related harm, and 5.3% for driving-related harm. One potential explanation for our study finding could be attributed to the considerably low rate of per capita alcohol consumption in Singapore, which is among the lowest in the Asia–Pacific region, partly due to the heavy taxation of alcoholic beverages in the country (World Health Organization, 2018). Restricting public consumption of alcohol and reducing the hours for retail sales of alcoholic beverages are among the legislative measures under the Liquor Control (Supply and Consumption) Act (Ministry of Home Affairs, 2015), which came into force since 2015 in Singapore, and could explain to a certain extent the low prevalence rates for AHTO in Singapore.

In contrast, most other empirical studies on harm from others’ drinking have consistently demonstrated a higher prevalence of AHTOs across many countries. In the United States, recent data suggests about 43% of adults had experienced one or more of five harm types in their lifetime caused by another’s drinking (Nayak et al., 2019). Similarly, a population-based cross-sectional survey revealed that about one in five or 20% of Japanese residents suffered from AHTO, which includes physical and emotional abuse, being forced to drink, sexual harassment, cleaning up drinker’s problematic behavior, and financial problems (Kinjo et al., 2023). In both Australia and New Zealand, studies consistently estimate that between 48 and 70% of their population are victims of harm from others’ drinking in the form of abuse, threat, or property damage (Casswell et al., 2011; Laslett et al., 2023). Other recent studies in England (20.1%; with 18 harms) (Beynon et al., 2019), Denmark (50%; with 8 harms) (Beckhoff et al., 2022), Ireland (44–51%; with 14 harms) (Hope et al., 2018), and China (18.2%; with 4 harms) (Yu et al., 2022) also reported high levels of AHTO. While most of the studies are not directly comparable as each examined different types of harm, together they demonstrate that the prevalence and societal burden of alcohol-related harms are substantive and widespread.

Broadly consistent with previous research, our results also show that women reported higher rates of family problems (Hope et al., 2018) and financial trouble (Laslett et al., 2020) from others’ drinking, while more men reported suffering physical harm (Quigg et al., 2019). From a public health perspective, some of the AHTOs are more evident such as physical assaults and damage or vandalism to property. However, family and financial harms are much less visible, albeit with serious and chronic consequences for the well-being of those affected (Callinan et al., 2019; Jones et al., 2019) and thus warrant further attention in strategic policy and community interventional initiatives.

A closer look at the likelihood of victims at risk of one or more harms from others’ drinking reveals that young people, ethnic minorities (Indians and Malays), individuals with marital transitions (separated, divorced, or widowed), and those who themselves are drinkers (either light drinkers or hazardous alcohol use) were more likely to report suffering AHTO. In our study, harm from others’ drinking was considerably lower for older adults than for young people. Young people have consistently been reported to be at higher risks of experiencing AHTO (Nayak et al., 2019; Ramstedt et al., 2016; Yu et al., 2022). Greater exposure to drinking milieus oriented towards young people where AHTOs may likely occur (e.g., nightclubs, parties, raves) (Moan & Brunborg, 2021) and risky drinking behavior (e.g., binge drinking) (Lam et al., 2019; Lee et al., 2019) among certain subgroups may place young people at greater risks of experiencing AHTOs. It has also been suggested that the phenomenon could be explained, in part, by generational differences, such that younger generations—greatly sensitized to violence and other alcohol-related harms—are more likely to perceive and report AHTO compared to older adults who are probably less exposed to such drinking environments and more tolerant of negative consequences from someone else’s drinking (Laslett et al., 2023; Lund et al., 2016).

Our results, consistent with previous studies, demonstrate that married people tend to experience a lower risk of harm from another’s drinking, and the dissolution of marriage due to separation, divorce, or widowhood is associated with comparatively greater risks of experiencing AHTO (Nayak et al., 2019). On one hand, experiencing AHTO would have had detrimental impacts on a marriage and could have already resulted in marital separation and divorce (Leone et al., 2022; Wilson et al., 2020). On the other hand, the risk of alcohol consumption (and risky drinking), the restructuring of the social network, and thus changes to patterns of socializing and exposure to harmful drinking environments may increase following marital transitions, as separated, divorced, or widowed individuals may view drinking as a means to cope with their loss and reduce stress (Kendler et al., 2017).

Our study findings further reveal both low-risk or safelight drinking and individuals with hazardous drinking patterns were significantly associated with experiencing AHTO. In contrast, previous studies conducted on alcohol-related problems have demonstrated that light drinkers who do not drink alcohol episodically and non-drinkers are less likely to experience harms from others’ drinking (Nayak et al., 2019; Seid et al., 2015). While further research is needed to better understand the relationships between personal drinking patterns and experiencing harm from another’s drinking, our result broadly suggests that there is no threshold drinking for experiencing harm; regular alcohol consumption habits, even if not problematic, may put an individual at high risk of harm from another’s drinking.

Our results indicate that individuals from the Indian and Malay ethnic minority groups were at greater risk of one or more harms from others’ drinking. In line with previous research primarily in Western populations, differential experiences of alcohol-related harms have been evidenced with different ethnic groups. However, the relationship of ethnicity to experiences of alcohol-related harm remains unclear. Partially attributed to the different rates and patterns of drinking across ethnicities as reported in the National Population Health Survey 2022 (NPHS 2022) (Ministry of Health & Health Promotion Board, 2022), ethnic minority groups such as Indians have been found to have higher rates of hazardous alcohol consumption patterns putting them at greater risk of experiencing AHTOs, given that drinkers are more likely to socialize with and hence exposed to other drinkers (Moan & Brunborg, 2021). However, other ethnic minorities such as the Malays are found to experience harms from drinking that are disproportionate to their own consumption (Room et al., 2019). Future studies can include and investigate the reasons that lead to greater risk of harm experienced by ethnic minority populations. Nevertheless, our present findings indicate targeted public health policies and programs need to account for ethnicity-related vulnerabilities, particularly their relationship with the drinker causing harm.

The study has several limitations. Due to the cross-sectional nature of the study, there may be limited capacity for drawing causal inferences. Rossow (2016) explains that cross-sectional data based on general population samples is still an important first step in elucidating how common these problems are in different populations. Nonetheless, longitudinal analysis may hold the potential to uncover the trajectory and contexts of alcohol’s harm to others. It is possible that some harms like physical violence perpetrated particularly by intoxicated intimate partners may have likely been under-reported by victims due to fears of being judged or endangered (Chen et al., 2023; Kanougiya et al., 2022). Furthermore, in this study, we could not identify the frequency or level of severity of exposure to AHTO as well as the perpetrator’s relationship (e.g., friend, relative, or stranger) to the harmed individual, which should be included in future studies. Lastly, the policies related to restrictions of public alcohol consumption were gradually relaxed from 26 April 2022, during the COVID-19 pandemic (Ministry of Health, 2022). It is possible that our results may have been influenced by increased alcohol use during this period. While studies on the prevalence of alcohol consumption during the pandemic in Singapore are scarce, the NPHS 2022 reported that the prevalence of regular drinking may not have changed significantly during the pandemic from 2020 to 2022. In future, more sophisticated research techniques might be able to reveal fresh information about the impact of the pandemic on trends surrounding alcohol consumption and the occurrence of alcohol’s harms to others.

Further Implications for Stakeholders and the Field

The present study makes a significant contribution to the global understanding of AHTO by highlighting the sociodemographic factors and victim’s own drinking patterns associated with these harms in a diverse, multicultural, multi-ethnic population. While the study is based in Singapore, its findings have broader implications that can inform international public health strategies. The identification of specific demographic subgroups—such as young people, ethnic minorities, and individuals undergoing marital dissolution—who are at greater risk of experiencing AHTO can guide targeted interventions and policies in other countries with similar vulnerable populations, thereby improving the effectiveness of public health initiatives globally. Moreover, our findings advocate for a holistic approach to public health that not only addresses individual alcohol consumption but also considers the broader social and familial impacts of drinking. Health professionals and public health practitioners around the world can leverage these insights to develop integrated strategies that encompass prevention, early intervention, and support for affected individuals and communities.

Furthermore, the examination of AHTO in a multi-ethnic context in our study highlights how cultural and ethnic differences can influence the experience and reporting of AHTO, offering valuable lessons for multicultural societies globally. Understanding these nuances will help in designing culturally sensitive public health campaigns and interventions in other regions. Additionally, this study underscores the potential role of stringent alcohol control measures, such as heavy taxation and restricted sale hours, in reducing the prevalence of AHTO. These insights can provide valuable lessons for policymakers worldwide, supporting the implementation or reinforcement of similar regulations in other regions. Singapore’s approach can therefore serve as a case study for analyzing the impact of such measures, offering helpful insights for other countries aiming to mitigate alcohol-related harms.

Conclusion

Results of the present study suggest that harms from others’ drinking are a common threat to people’s health, well-being, and their environment. Preventing harm to others from drinking is as cogent and urgent a public health goal as preventing harm to the drinker from his or her own drinking. Effective policy with tackling other kinds of second-hand harms (e.g., passive smoking) or alcohol-related harms (e.g., drunk driving) should be closely studied, and the implications for dealing with other kinds of harm to victims highlighted in the present study should be explored, paying particular attention to the sociodemographic groups found to be more susceptible to AHTO based on our results. Necessary legislation must be effectively communicated and enforced while closely working with community-led efforts to better improve the social and physical environments to protect against AHTO for public health.