Association between age at first alcohol use and heavy episodic drinking: An analysis of Thailand's smoking and alcohol drinking behavior survey 2017.

According to evidence from developed countries, age at first alcohol use has been identified as a determinant of heavy episodic drinking (HED). This study aimed to investigate the association between age at first alcohol use and HED using data from the Smoking and Drinking Behavior Survey 2017, a Thai nationally representative survey. Binary logistic regression was used to examine the association. This study used data from 23,073 current drinkers in the survey. The survey participants were chosen to represent the Thai population aged 15 years and older. The prevalence of HED and frequent HED among Thai drinkers was 18.6% and 10.1%, respectively. Age at first drinking <20 years was associated with higher odds of HED (adjusted OR, 1.43; 95% CI, 1.26-1.62) and frequent HED (adjusted OR, 1.31; 95% CI, 1.12-1.53) relative to age at first drinking ≥25 years. Regular drinking, drinking at home, and exposure to alcohol advertising increased the odds of HED. Drinking at home was associated with frequent HED. There was a significant interaction between the effect of age at first alcohol use and sex on HED and frequent HED with a stronger effect of age at first alcohol use observed in females. This study provides evidence from a developing country that early onset of alcohol use is associated with HED. Effective measures such as tax and pricing policy should be enforced to delay the onset of drinking.

1. Introduction

Alcohol use is associated with a substantial burden on mortality, morbidity, and healthcare costs [1, 2]. Globally, alcohol use was the leading cause of death among females (3.8%) and males (12.2%) aged 15–49 years [2] and disability-adjusted life years among females (2.3%) and males (8.9%) in 2016 [2].

Heavy episodic drinking (HED) or binge drinking, defined as an episode of drinking five or more standard drinks of alcoholic beverages for males or four or more drinks for females in about 2 hours [3], is a high-risk drinking pattern associated with harm to drinkers and others [4]. HED is associated with alcohol-related injuries [5], violence [6], mortality [7], and high healthcare costs [1]. Drinkers’ quality of life is negatively associated with HED [8]. The prevalence of HED among drinkers varied between countries, for instance, 28% in the US [9], 38% in Canada [10], 65.1% in France [11], 36% in Australia [12], 13.7% in Singapore [13], and 11.9% in Thailand [14].

Age at first alcohol use is recognized as an important risk factor for HED, alcohol dependence, and alcohol-related consequences [15-19]. An analysis of the 2010 National Survey on Drug Use and Health in the US found that age at first alcohol use of <12, 12–14, and 15–17 years was associated with 3.0-, 2.6-, and 1.9-fold increase in the likelihood of HED compared to initiation of alcohol drinking at 18–24 years [15]. A prospective study of Australian students and parents reported that earlier age at first alcohol use is associated with binge drinking and a higher quantity of alcohol consumed after adjusting for parental and family factors [12]. Another prospective study from Canada found that early alcohol use was associated with current binge drinking and current alcohol use [10]. Other factors including exposure to alcohol advertising [20], drinking contexts [21], smoking [22], income [23], and sex [10] are associated with HED.

Most studies on the association between age at first alcohol use and HED have been conducted in developed countries where alcohol use and binge drinking have decreased since the early 2000s [24-28]. This trend is less obvious in developing countries [24]. In Thailand, the prevalence of current drinkers among individuals aged 15–19 years has slightly increased from 11.0% in 2001 to 13.6% in 2017 [14]. Nevertheless, the prevalence of youth drinking is generally higher in developed countries than in developing countries [29, 30]. Although the initially high prevalence of youth drinking in developed countries has been declining, no obvious decrease was noted in the initially low prevalence in developing countries.

This difference in the prevalence and trend of youth drinking between developed countries and developing countries may reflect the difference in contextual factors that are associated with the trend of youth drinking [24, 28, 31] and the relationship between age at first alcohol use and HED [16]. The association between age at first alcohol use and HED in developing countries may differ from what is observed in developed countries. However, evidence on the association from developing countries is lacking. With the trend of youth drinking observed in some developing countries such as Thailand, knowledge about the association can facilitate decision-making for implementing measures to delay youth drinking. This study aimed to investigate the association between age at first alcohol use and HED using data from a nationally representative survey in Thailand.

To provide context for this study, nearly one-third of Thais are current drinkers. Between 2007 and 2017, the prevalence of current drinkers ranged between 28.4% and 32.3% [14], whereas the prevalence was 59.9% in WHO’s European Region and 54.1% in WHO’s Region of the Americas in 2016 [29]. This distinction can be attributed in part to Thailand’s cultural background. Thailand is a predominantly Buddhist country. Thai people adhere to Buddhism’s teaching known as the Five Percepts, which include refraining from intoxication—including alcohol use. The Buddhist Lent Abstinence Campaign, which encourages people to abstain from drinking for three months, is Thailand’s major sobriety campaign. In 2016, almost six million Thai drinkers were estimated to abstain completely during the campaign period [32]. Alcohol regulations in Thailand are implemented primarily under the Alcohol Beverage Control Act, B.E. 2551 (2008). The National Alcohol strategy is a blueprint for alcohol policy. It comprises five strategies: controlling economic and physical access to alcohol, altering social norms toward alcohol and reducing drinking motivation, reducing harms from drinking, promoting community-based solutions, and creating policy supporting mechanisms [33].

2. Methods

2.1 Study design

This study analyzed data from the Smoking and Drinking Behavior Survey 2017 (SADBeS), a nationwide cross-sectional survey in the Thai population. Variables that were relevant to the research questions were selected for the analysis.

2.2 Data source

Data for this study were obtained from SADBeS. The National Statistical Office, Thailand, conducted the survey. A stratified two-stage random sampling technique was used in the survey. In the first stage, enumeration areas (a sampling frame for the national census) were randomly selected with the probability proportional to size from 77 strata (the survey employed provinces as strata). In the second stage, households were randomly chosen from the selected enumeration areas. All household members aged 15 years and above who were fluent in Thai were invited to participate in the survey. The response rate of the survey was 93.5%.

The survey included data from non-drinkers, former drinkers, and drinkers. The present study analyzed data from current drinkers aged 15 years and above. Participants with missing data for the required variable were excluded. Therefore, the analysis included only complete cases. Ultimately, this study included data from 23,073 participants. The flow diagram of participant selection is shown in Fig 1.

Fig 1

Participant flowchart.

2.3 HED

The outcome of interest was HED. The amount of drinking considered HED differs by sex: five or more standard drinks of alcohol for males and four or more standard drinks for females [3]. The item for HED in the survey followed the definition for males. Hence, male and female participants were asked to respond to the same item for HED. The item was: "How often had you drunk heavily (5 drinks or more) in a short time period in the past 12 months? " The response options were as follows: no, 7 days per week, 5–6 days per week, 3–4 days per week, 1–2 days per week, 1–3 days per month, 8–11 days per year, 4–7 days per year, and 1–3 days per year. Two dichotomous variables were created based on these responses. The first variable, HED, indicated the HED frequency of at least once a month (which corresponds to the option ‘1–3 days per month’ or more frequent). The frequency of at least once a month was chosen to make the estimate somewhat comparable with a 30-day time frame used by the World Health Organization in reporting HED prevalence among drinkers [34]. Nonetheless, as the survey item used a 12-month time frame, the prevalence was underestimated in this analysis. The other variable, frequent HED, indicated the HED frequency of at least once a week (which corresponds to the option ‘1–2 days per week’ or more frequent).

2.4 Age at first alcohol use

In this study, the explanatory variable of interest was age at first alcohol use defined as the age at which the individual first drank alcohol. The original question asked in the survey was, "How old were you when you drank alcoholic beverages for the first time?” The response was open-ended. As the legal purchasing age for alcoholic beverages in Thailand is 20 years and above, a new variable was created by categorizing the responses into three categories: <20 years old, 20–24 years old, and ≥25 years old.

2.5 Other covariates

Covariates related to demographic characteristics included sex, age, marital status (single, married, and widowed/divorced/separated), education level (primary school or lower, secondary school, and college or higher), household income (<5,000 Thai baht [THB], 5,000–9,999 THB, and ≥10,000 THB), and smoking (no, occasional, and regular). Drinking-related covariates included frequency of drinking (occasional [less frequent than once a week] and regular [once a week of more frequent]), most frequent drinking venue (own home, someone else’s home, and party and traditional ceremony venue), and exposure to alcohol advertising (yes, no, and unsure; the unsure response was coded as ‘no’). The item for alcohol advertising exposure was a single question, “During the past 30 days, have you seen or heard any alcoholic beverage advertisement?”. No further information was given to survey participants regarding alcohol advertising.

2.6 Statistical analysis

Percentage, mean, standard deviation, minimum, and maximum were used for describing continuous variables. Categorical variables were represented by count and percentage. The prevalence of HED and frequent HED was computed. Multivariate analysis was performed using a binary logistic regression model. The primary explanatory variable was age at first alcohol use. All other variables included as covariates in the multivariate model were based on existing evidence of association with HED. The dependent variables were HED and frequent HED. Crude and adjusted odds ratios (ORs) with 95% confidence interval (CI) and p-value were estimated using data from the entire sample. As participants aged 24 years or below were unable to be fully coded for the age at first alcohol use variable (e.g., a participant aged 22 years cannot state the age at first alcohol use of ≥25 years old), three separate regression models were run using data from each age group (i.e., 15–19 years, 20–24 years, and 25+ years) to address this issue. All three models had the identical set of covariates as the full sample model. The only difference was how the age at first alcohol use variable was coded. For the youngest age group, a continuous age at first use variable was used in the model. For the 20–24 age group, a dichotomous age at first use variable (<20 years old and 20–24 years old) was used. The model for the 25+ age group employed the three-level age at first use variable (20 years old, 20–24 years old, and ≥25 years old) as the full sample models.

Since the Thai population’s drinking patterns differ considerably between males and females [30], we examined the effect of interaction between age at first alcohol use and sex on HED and frequent HED using a logistic regression model. In all multivariate analyses, only participants with complete data on the outcome variable and covariates were included. All analyses were unweighted and conducted using the R statistical language via RStudio version 4.0.3. A significance level of 0.05 was used for all analyses.

2.7 Ethics approval

This study was approved by the Human Research Ethics Committee of Walailak University, Thailand (WU-EC-MD-3-470-63). The data were analyzed anonymously.

3. Results

3.1 Characteristics of current drinkers

Characteristics of current drinkers are shown in Table 1. The majority were males. The average age was 44.3 years. More than 70.0% of current drinkers were married. Half had primary school education or lower. The mean monthly household income was 10,919.9 THB. The proportion of regular smokers was 38.7%. The average age of first alcohol use was 20.6 years. Almost half of the drinkers drank regularly. One-third of the drinkers were exposed to alcohol advertising. The prevalence of HED and frequent HED was 18.6% and 10.1%, respectively.

Table 1

Characteristics of current drinkers (n = 23,073).

Characteristics	n	%
Sex
Male	18,356	79.6
Female	4,717	20.4
Age (years)
Mean (SD)	44.3 (14.1)
15–19	734	3.2
20–30	3604	15.6
31–45	7753	33.6
46–60	8095	35.1
≥61	2887	12.5
Education
Primary school or lower	11,635	50.5
Secondary school	9,009	39.1
College or higher	2,391	10.4
Household income per month (Thai baht)
Mean (SD)	10,919.9 (11,160.3)
<5,000	5,440	25.0
5,000–9999	7,935	36.5
≥10,000	8,381	38.5
Marital status
Single	4,587	19.9
Married	16,221	70.3
Widowed/divorced/separated	2,261	9.8
Smoking
No	13,016	56.4
Occasional	1,113	4.8
Regular	8,944	38.7
Age at first alcohol use (years)
Mean (SD)	20.6 (6.2)
<20	11,177	48.4
20–24	7,631	33.1
≥25	4,265	18.5
Frequency of drinking
Occasional	13,057	56.6
Regular	10,016	43.4
Most frequent drinking venue
Own home	9,118	39.5
Someone else’s home	5,061	21.9
Party or traditional ceremony	8,894	38.6
Exposure to alcohol advertising
No	14,053	63.6
Yes	8,053	36.4
HED
No	18,792	81.4
Yes	4,283	18.6
Frequent HED
No	20,746	89.9
Yes	2,327	10.1

SD, standard deviation; HED, heavy episodic drinking.

3.2 Prevalence of HED and frequent HED among drinkers

The prevalence of HED and frequent HED by drinkers’ characteristics is shown in Table 2. The prevalence of HED and frequent HED was higher among drinkers with a lower age at first alcohol use. For example, the prevalence of HED was 22.1% among those who began drinking before the age of 20 years, but only 11.4% among those who began drinking at the age of 25 years or above. HED was far more common among regular drinkers than occasional drinkers. Logically, it is impossible for occasional drinkers to engage in frequent HED; hence, the prevalence of frequent HED in this group was 0%. One-fourth of regular drinkers had frequent HED. A high prevalence of HED and frequent HED was observed in drinkers who cited their own or other’s home as the most frequent drinking venue. Drinkers with exposure to alcohol advertising had slightly higher HED than those who were not exposed to such advertising. An association between alcohol advertising and frequent HED was not statistically significant. HED and frequent HED were far more common in males. A high prevalence of HED and frequent HED was observed in those who smoked more regularly.

Table 2

Prevalence of HED and frequent HED by drinkers’ characteristics (n = 23,073).

Characteristic	HED			Frequent HED
	%	95% CI	p-valuea	%	95% CI	p-valuea
Drinking-related characteristics
Age at first alcohol use (years)			<0.001			<0.001
≥25	11.4	10.5–12.4		6.4	5.7–7.2
20–24	17.4	16.5–18.2		9.4	8.8–10.1
<20	22.1	21.3–22.9		12.0	11.4–12.6
Frequency of drinking			<0.001			N/A
Occasional	6.9	6.5–7.4		0.0	0.0
Regular	33.7	32.8–34.6		23.2	22.4–24.1
Most frequent drinking venue			<0.001			<0.001
Own home	24.7	23.8–25.6		15.6	14.9–16.3
Someone else’s home	22.4	21.3–23.6		11.2	10.4–12.1
Party or traditional ceremony	10.1	9.5–10.7		3.8	3.4–4.2
Exposure to alcohol advertising			<0.001			0.083
No	17.5	16.9–18.2		9.6	9.2–10.1
Yes	19.7	18.9–20.6		10.4	9.7–11.1
Demographic and socioeconomic characteristics
Sex			<0.001			<0.001
Male	21.5	20.9–22.1		11.8	11.3–12.3
Female	7.3	6.6–8.0		3.5	3.0–4.0
Age (years)			<0.001			<0.001
15–19	13.2	11.0–15.9		4.6	3.3–6.4
20–30	19.4	18.1–20.7		9.3	8.4–10.3
31–45	20.1	19.2–21.0		10.7	10.0–11.4
46–60	18.3	17.5–19.2		10.5	9.9–11.2
≥61	15.3	14.0–16.7		9.8	8.7–10.9
Education			<0.001			<0.001
Primary school or lower	18.4	17.7–19.1		10.8	10.3–11.4
Secondary school	19.6	18.8–20.5		10.1	9.5–10.7
College or higher	14.9	13.5–16.4		6.3	5.4–7.3
Household income (Thai baht)			<0.001			0.074
<5,000	16.6	15.6–17.6		9.3	8.5–10.1
5,000–9999	18.4	17.6–19.3		10.1	9.5–10.8
≥10,000	19.9	19.1–20.8		10.4	9.8–11.1
Marital status			<0.001			<0.001
Single	21.3	20.1–22.5		9.9	9.1–10.8
Married	17.2	16.7–17.8		9.6	9.2–10.1
Widowed/divorced/separated	22.5	20.8–24.2		13.6	12.3–15.1
Smoking			<0.001			<0.001
No	13.9	13.3–14.5		6.9	6.4–7.3
Occasional	16.8	14.7–19.1		7.1	5.7–8.8
Regular	25.6	24.7–26.5		15.1	14.4–15.9

a p-value computed using chi-squared test.

HED, heavy episodic drinking.

3.3 Factors associated with HED and frequent HED

The results of multivariate analysis are shown in Table 3. Drinkers who started drinking at <20 years had higher odds of HED (adjusted OR, 1.43; 95% CI, 1.26–1.62) and frequent HED (adjusted OR, 1.31; 95% CI, 1.12–1.53) than those who started drinking at ≥25 years. Regular drinkers had markedly higher odds of HED than occasional drinkers (adjusted OR, 5.47; 95% CI, 4.98–6.01). Compared to drinking at one’s own home, drinking at a party or traditional ceremony was associated with decreased odds of HED and frequent HED, whereas drinking at someone else’s home was associated with increased odds of HED but decreased odds of frequent HED. Exposure to alcohol advertising was associated with higher odds of HED but not frequent HED. Female sex was associated with lower odds of HED and frequent HED. Working age groups had higher odds of HED and frequent HED than teenagers. Drinkers with a college education or higher had lower odds of frequent HED than those with primary school education or lower. Having a monthly income ≥10,000 THB was associated with higher odds of HED and frequent HED.

Table 3

Factors associated with HED and frequent HED (multivariate logistic regression; n = 20,805).

Factor	HED					Frequent HED
	Crude OR	Adjusted OR	95% CI	p-value		Crude OR	Adjusted OR	95% CI	p-value
				Wald’s test	LR test				Wald’s test	LR test
Drinking-related characteristics
Age at first alcohol use (years)					<0.001					<0.001
≥25	1	1				1	1
20–24	1.59	1.24	1.09–1.42	<0.001		1.49	1.11	0.95–1.31	0.189
<20	2.22	1.43	1.26–1.62	<0.001		1.97	1.31	1.12–1.53	<0.001
Frequency of drinking					<0.001					N/A
Occasional	1	1				N/A	N/A	N/A	N/A
Regular	6.70	5.47	4.98–6.01	<0.001		N/A	N/A	N/A	N/A
Most frequent drinking venue					<0.001					<0.001
Own home	1	1				1	1
Someone else’s home	0.90	1.19	1.08–1.31	<0.001		0.70	0.70	0.62–0.78	<0.001
Party or traditional ceremony	0.35	0.78	0.70–0.86	<0.001		0.22	0.26	0.23–0.30	<0.001
Exposure to alcohol advertising					<0.001					0.078
No	1					1
Yes	1.14	1.16	1.07–1.25	<0.001		1.05	1.09	0.99–1.21	0.078
Demographic and socioeconomic characteristics
Sex					<0.001					<0.001
Male	1	1				1	1
Female	0.28	0.55	0.47–0.63	<0.001		0.25	0.39	0.32–0.47	<0.001
Age (years)					<0.001					<0.001
15–19	1	1				1	1
20–30	1.43	1.38	1.05–1.82	0.022		2.07	2.26	1.48–3.45	<0.001
31–45	1.42	1.41	1.07–1.86	0.016		2.31	2.64	1.73–4.02	<0.001
46–60	1.28	1.41	1.06–1.87	0.019		2.27	2.78	1.81–4.27	<0.001
≥61	1.03	1.09	0.80–1.47	0.582		2.10	2.40	1.54–3.75	<0.001
Education					0.191					0.010
Primary school or lower	1	1				1	1
Secondary school	1.12	1.09	0.99–1.19	0.080		0.96	1.02	0.91–1.14	0.756
College or higher	0.76	1.02	0.87–1.18	0.841		0.52	0.75	0.61–0.93	0.007
Monthly household income (Thai baht)					<0.001					<0.001
<5,000	1	1				1	1
5,000–9999	1.16	1.01	0.91–1.12	0.799		1.14	1.08	0.95–1.23	0.257
≥10,000	1.27	1.25	1.12–1.40	<0.001		1.16	1.28	1.11–1.46	<0.001
Marital status					<0.001					<0.001
Single	1	1				1	1
Married	0.74	0.70	0.63–0.78	<0.001		0.91	0.75	0.65–0.86	<0.001
Widowed/divorced/separated	1.04	1.11	0.95–1.30	0.172		1.37	1.34	1.11–1.61	0.002
Smoking					<0.001					<0.001
No	1	1				1	1
Occasional	1.26	1.01	0.84–1.23	0.884		1.08	0.86	0.66–1.12	0.257
Regular	2.11	1.19	1.10–1.30	<0.001		2.46	1.65	1.49–1.83	<0.001

LR test, likelihood ratio test; HED, heavy episodic drinking; OR, odds ratio.

Table 4 shows the results for each age group from the regression model. Among participants aged 15–19 years, earlier onset of drinking was not significantly associated with HED and frequent HED (Model A). The finding was similar for the 20–24 years age group (Model B). In participants aged 25 years or above, alcohol use onset before the age of 20 years was associated with increased odds of HED and frequent HED (Model C). This is consistent with the result of the full sample model presented in Table 3.

Table 4

Effect of age at first alcohol use on HED and frequent HED in each age group.

Model	HEDa				Frequent HEDb
	Adjusted OR	95% CI	p-value		Adjusted OR	95% CI	p-value
			Wald’s test	LR test			Wald’s test	LR test
Model A: 15–19 years old (n = 547)				0.399				0.146
Continuous	1.09	0.89–1.34	0.399		0.81	0.61–1.07	0.146
Model B: 20–24 years old (n = 1,197)				0.093				0.305
20–25	1				1
<20	1.43	0.94–2.18	0.093		1.33	0.76–2.33	0.305
Model C: 25+ years old (n = 19,061)				<0.001				<0.001
≥25	1				1
20–24	1.25	1.10–1.42	<0.001		1.11	0.95–1.31	0.199
<20	1.42	1.25–1.61	<0.001		1.31	1.12–1.53	<0.001

a Adjusted for most frequent drinking venue, exposure to alcohol advertising, sex, age, education, marital status, household income, smoking, and frequency of drinking.

b Adjusted for most frequent drinking venue, exposure to alcohol advertising, sex, age, education, marital status, household income, and smoking.

LR test, likelihood ratio test; HED, heavy episodic drinking; OR, odds ratio.

An interaction between age at first alcohol use and sex in terms of their effect on HED and frequent HED was observed, indicating that the effect of age at first alcohol use differed by sex. Table 5 demonstrates that age at first alcohol use had a stronger association with HED and frequent HED in females by the factor of adjusted OR of the interaction term. For example, the adjusted ORs of age at first alcohol use <20 years old were 1.31 and 1.20 for HED and frequent HED, respectively, in males. The corresponding ORs in females were 2.17 (calculated from 1.31*1.66) and 2.44 (calculated from 1.20*2.03), respectively.

Table 5

Interaction between age at first alcohol use and sex (n = 20,805).

Factors	HEDa				Frequent HEDb
	Adjusted OR	95% CI	p-value		Adjusted OR	95% CI	p-value
			Wald’s test	LR test			Wald’s test	LR test
Interaction term (age at first alcohol use and sex)				0.009				0.009
≥25 and male	1				1
20–24 and female	1.43	1.02–2.00	0.036		1.40	0.87–2.25	0.164
<20 and female	1.66	1.19–2.32	0.003		2.03	1.29–3.20	0.002
Age at first alcohol use (years)				<0.001				0.004
≥25	1				1
20–24	1.14	0.99–1.32	0.065		1.03	0.87–1.23	0.700
<20	1.31	1.14–1.50	<0.001		1.20	1.02–1.41	0.030

a Adjusted for most frequent drinking venue, exposure to alcohol advertising, sex, age, education, marital status, household income, smoking, and frequency of drinking.

b Adjusted for most frequent drinking venue, exposure to alcohol advertising, sex, age, education, marital status, household income, and smoking.

LR test, likelihood ratio test; HED, heavy episodic drinking; OR, odds ratio.

4. Discussion

One-fifth of Thai drinkers engaged in HED, whereas one-tenth engaged in frequent HED. Lower age at first alcohol use, regular drinking, drinking at home, and exposure to alcohol advertising increased the likelihood of HED. Lower age at first alcohol use and drinking at home were associated with frequent HED. Demographic and socioeconomic factors were also associated with HED and frequent HED.

4.1 Association between age at first alcohol use and HED

The association between age at first alcohol use and HED among drinkers in a developing country was investigated using data from the SADBeS 2017. This study demonstrates that, after adjusting for covariates including drinking-related contextual factors (drinking venue and exposure to alcohol advertising), earlier alcohol initiation was associated with engagement in HED and frequent HED among drinkers. Drinkers who started alcohol drinking at <20 years had higher odds of HED and frequent HED, and those who started drinking at 20–24 years had higher odds of HED than drinkers who started alcohol drinking at ≥ 25 years. Nonetheless, this association was not observed when the analysis was limited to the age group of 15–19 years and 20–24 years. This may be due to less variability in drinking behavior in persons from a narrow age range. The analysis of the 25+ age group yielded results similar to those for the full sample model. This finding supports that the association between age at first alcohol use and HED in developing countries is similar to that in developed countries. For example, an analysis of the national survey in the US found that age at first use of alcohol was associated with increased HED in the last 30 days among drinkers [15]. Prospective studies from Australia and Canada confirmed this association [10, 12]. Moreover, earlier drinking initiation was associated with alcohol dependence and alcohol use disorder as well as substance use, violence, and employment problems [17-19]. Apart from the direct consequences of alcohol consumption, the initiation of a substance during adolescence may result in future use of other substances [35]. This issue should be examined in individuals with early drinking onset.

Drinking initiation is a result of a combination of contextual factors including parental drinking, parental supervision, friends’ drinking, and school bonding [16]. Apart from those who are likely to drink heavily themselves, parents with early onset of drinking perceived providing alcohol to children as acceptable [36]. Hence, an effect of earlier drinking initiation in one generation may lead to HED in the next generation. This underscores the importance of implementing measures to delay drinking onset. A recent study on Chilean youths reported the effect of pricing policy on the delayed initiation of alcohol use. The study estimated that 6.6 months delay in alcohol initiation could be brought about by a 10% increase in the price of alcoholic beverages [37]. This warrants the implementation of tax and pricing policies for delaying alcohol initiation and preventing engagement in HED.

The finding of lower odds of HED in females in this study was consistent with the finding from the International Alcohol Control Study, which reported four-fold chance of HED in Thai males than Thai females [30]. This ratio was in the same range as for other developing countries including Mongolia, South Africa, and Brazil [30, 38]. Although females were considerably less likely to engage in HED than males, our interaction model showed that the effect of age at first alcohol use on HED and frequent HED was significantly stronger in females. Hence, tax and pricing policies need to be implemented across the board including alcoholic beverages preferred by young females such as alcopops and cider [39, 40]. The interaction between sex and another determinant of heavy drinking was observed in a study from Brazil: higher education was protective against heavy drinking in males, but increased the chance of heavy drinking in females [38]. Sex likely modifies the effect of several factors on heavy drinking in the context of developing countries. This reflects different underpinning casual models of heavy drinking between males and females.

4.2 Effects of other covariates

Our study demonstrated that factors including frequency of drinking, most frequent drinking venue, exposure to alcohol advertising, age, smoking, and marital status are correlated with HED. These factors have been shown to be associated with HED in previous studies on the effects of drinking frequency [22], drinking location [21], alcohol marketing in youths [20, 41], age [21, 42], and smoking [43, 44]. The direction of the associations for most factors is consistent with previous research. The relationship between drinking location and HED, on the other hand, was complicated.

According to a study based on data from a Canadian survey, participants consumed the least amount of alcohol at a restaurant and the most at a bar/disco/nightclub. The amount consumed at home was moderate [45]. In the current study, participants indicating home as the most frequent drinking venue had a higher chance of HED and frequent HED than those mentioning party or traditional ceremony (which includes restaurants). Hence, this seemingly inconsistent finding may be due, in part, to the grouping of drinking locations in our study. Moreover, it should be noted that the drinking venue variable in our study does not directly refer to the location where HED occurred; it refers to the drinking location where participants visited the most for drinking in general.

A study conducted in the United States investigated heavy drinking in both public and private settings. It demonstrated that the other environmental factor, the presence of many intoxicated people at the drinking venue, was linked to heavy drinking in both public and private settings [46]. The current study did not include variables related to the immediate environmental factors of an HED occasion. Further investigation of HED in Thailand should include specific information about each HED occasion to fill this gap.

4.3 Limitations

This study has some limitations. Data used in the analysis are from a cross-sectional survey. This may lead to recall bias, especially concerning age at first alcohol use. The survey participants were limited to the Thai-speaking population, whereas some Thai citizens, such as hill tribes in the north, are unable to communicate in Thai. The findings of this study may not be applicable to these populations. The definition of HED used in this study was not the same as the one used by WHO regarding the time frame of the measurement. The item for HED in the survey used a 12-month time frame, whereas the WHO definition of HED uses a 30-day time frame. The HED variable used in the analysis was created by classifying heavy drinking at least once a month as HED. As a result, the prevalence of HED reported in this study may have been underestimated. Furthermore, the item for HED in the survey was based on the definition of HED for males, whereas the amount of drinking considered HED for females is slightly lower than that of males. As a result, the prevalence of HED in females was also underestimated in this study.

14 Jul 2021

PONE-D-21-16712

Association between age at first alcohol use and heavy episodic drinking: An analysis of the Smoking and Alcohol Drinking Behavior Survey 2017

PLOS ONE

Dear Dr. Saengow,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Aug 28 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: http://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Zila M Sanchez, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

Reviewer #3: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

Reviewer #3: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors have undertaken an important subject, that of examining if associations between age of first drink and later HED are observed in a middle-income country (i.e., Thailand), as have been observed in data from several higher-income countries. The dataset is strong, and the topic is important. I have some concerns (some minor, some not) that may help improve the analyses and their contribution to the literature.

Major

1. My major concern is that the analytic sample needs to be revised in order to accurately evaluate cross-sectional associations with age of first drink as defined. The use of the full sample (ages 15 and up) is completely accurate for estimating prevalence of HID and frequent HED among Thai-speakers in Thailand. However, I do not feel it is appropriate when evaluating associations between age at first drink and HED to include cases in the models that have not had the opportunity to be coded fully using the defined age of first drink measure: <20, 20-25, >25. Respondents aged 15-19 have to have an age of first drink at <20; those aged 20-25 do not have the ability to be coded as >25. This skews the analyses and resulting estimates. Since age of first drink was asked as an open-ended measure, I recommend that a series of models possibly be run. One possible approach would be that for respondents aged 15-19, include age at first drink as a continuous measure with controls that are appropriate for the age range (e.g., not including marital status). For respondents aged 20-25, possibly include age at first drink as a dichotomy of <20 vs. 20+. For respondents aged 26+, use the initial trichotomy of <20, 20-25, >25. This would also help address the “telescoping” that happens when attempting to remember age at first drink as one gets older.

2. An additional limitation that should be noted is that only Thai speakers were included; as the population in Thailand includes many hill tribes in various locations (particularly the North), findings may not generalize to these populations.

3. In 2.3 HED, the wording of the item for HED reads, “How often have you drunk heavily (5 drinks)…”. Was this the actual wording used? Were females asked about the prevalence of 4 drinks?

4. No mention of what was done to address missing data on covariates or outcome measures is available. Please state the level of missingness present (or at least a range), and what was done to address this, or if complete case data were used.

5. I find it surprising that a yes/no item could be asked of exposure to alcohol advertising. Were respondents given any further information on what constitutes alcohol advertising?

6. Analyses should have included methods to address the complex sample (i.e., strata, clustering, weights). Please describe.

7. Please provide statistical testing of bivariate associations in Table 1.

Minor

1. The authors sometimes use “developed” and “developing” versus “higher-income” and “middle-income”. Please be consistent.

2. The authors should cite publications from national surveys for their prevalence estimates of nation-specific HID. Citation of a study looking only at HED among the Hispanic population at the US-Mexico border is not really appropriate for national HED prevalence, which can be found in sources such as the National Household Survey on Drug Use and the Monitoring the Future Survey.

3. The data presented on trends in current drinking among Thai adolescents is interesting, but confusing (p. 4). Please clarify.

4. The 2nd to last sentence in the Introduction is not clear.

5. In 2.2 Data source, the text reads 23,070 cases. However, in the sample flowchart, the total is 23,073.

6. On p. 12, I believe the use of the word “attenuated” is incorrect, as this indicates the association was smaller for females than males.

Reviewer #2: The study used national data in Thailand to examine the association between age at first alcohol use and heavy episodic drinking. This is an important issue less focused in research out of the developed countries and is worth exploration. The literature review laid a clear background for the focused topic, and the methods are suitable for the concerning questions and are clearly presented. The discussion interprets the results in the context of existing literature and provides policy implications. Here are some suggestions for the authors’ consideration:

1. The background introduction may be further strengthened by informing readers a little about the social/cultural background in Thailand and how that may contribute to the discrepancy in heavy drinking prevalence between Thailand and developed countries.

2. In the methods section

a. A little more specifics in the introduction of the database and sample. In particular, what is the proportion of the missing?

b. Is the age category of >=25 relative to the other two younger groups too broad?

3. In the result presentation section

a. It is better to provide specific statistics when presenting the results in the texts to facilitate reading

b. Given that this is a cross-sectional study, it is better to say correlates rather than “Determinants of HED and frequent HED”

c. The statement regarding the interaction between gender and initial alcohol use doesn’t seem to match the findings in Table 4 and is in contrast with its following statement: “however, female sex attenuated the effect of age at first alcohol use on HED and frequent HED. For example, in females, the effect of age at first alcohol use <20 years on HED and frequent HED was increased by 66% and 103%, respectively. ” Female earlier onset seems to increase the risk of HED instead.

4. The discussion can be further enriched by more closely relating the findings to relevant literature. For example, evidence concerning gateway theory in substance use in general may help strengthen the discussion from early drinking onset to HED (e.g., Zhang, S., Wu, S., Wu, Q., Durkin, D. W., & Marsiglia, F. F. (2021). Adolescent drug use initiation and transition into other drugs: A retrospective longitudinal examination across race/ethnicity. Addictive Behaviors, 113, 106679.)

Reviewer #3: Major comments

The study is relevant for bringing results about HED in amiddle-income Asian country.

The results are interesting and in line withprevious literature on the subject.

Outcome variables have limitations that arewell described.

As contributions to the improvement of the report, I suggest:

1 - Review the use of the term gender, as apparently thevariable used was sex. For this, please consider the SAGER guidelines (https://www.equator-network.org/reporting-guidelines/sager-guidelines/)

2 - Emphasize, in the discussion, the main differences in theresults observed in Thailand in relation to other countries, in order toreinforce the importance of carrying out the study. Also inform about public policies to combat the harmful use of alcohol in force in Thailand, and how the results of the study can contribute to the improvement of such policies.

Minor comments and suggestions

Title:

- The title should include the name of the country where the survey was conducted.

Abstract:

- Please, use the same objective, as stated at the end ofthe introduction: “This study aimed to investigate the association between ageat first alcohol use and HED using data from a nationally representative surveyin Thailand.”

- The abstract should show effect sizes and confidence intervals. “Regular drinking, drinking at home, and exposure to alcohol advertising increased the likelihood of HED.” Also, it is suggested to replace “likelihood”by “chance”, which is more adequate to the type of analysis used (logisticregression).

- “There was a significant interaction between the effect ofage at first alcohol use and gender on HED and frequent HED.” Please, indicate the direction of the associations, eg: lower or higher age? Male or female?

- Please, review the use of the word “gender”.

- “This study provides evidence from middle-income countries that early onset of alcohol use is associated with HED.” – a middle income country

Keywords:

- Please include a keyword related to study design(cross-sectional studies)

Introduction:

- It should include data on HED prevalence in Thailand. Data on theburden of disease caused by alcohol could also be included.

Methods:

- At the “study design” subsection, please indicate clearly the study design (cross-sectional)

- Please, review the use of the word “gender”.

- Please, explain how “exposure to alcohol advertising” was measured and what was considered as “alcohol advertising”.

- Statistical analyses should include the criteria used for including and maintaining variables in the adjusted regression model.

- Ethics: please, indicate how participants gave consent.

Results:

- The results are clearly presented.

- The higher prevalence of HED while drinking at home couldbe emphasised in the text.

Tables 3 and 4: please, indicate as footnote the name of thestatistical tests used to calculate the p-values.

Discussion:

At the begining of this section, sex differences could also be mentioned. Also, sex differences could be more discussed. The discussion could be deepened and include dialogue with studies carried out with other middle-income countries, including Asian countries, and countries in other continents, such as Brazil, Mexico, South Africa.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Saijun Zhang

Reviewer #3: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

28 Aug 2021

The response has been uploaded as a separate file.

Submitted filename: Response to reviewer.docx

Click here for additional data file.

14 Sep 2021

PONE-D-21-16712R1Association between age at first alcohol use and heavy episodic drinking: An analysis of the Thailand's Smoking and Alcohol Drinking Behavior Survey 2017PLOS ONE

Dear Dr. Saengow,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Oct 29 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Zila M Sanchez, PhD

Academic Editor

PLOS ONE

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments:

The authors had answered most of the reviewers suggestions.

However some aspects should be revised:

- English edit of the manuscript

- the p values on the tables are sometimes not well placed. A p value for each category of answer is expected when when you have a variable with 3 or more categories (for example: age: 15-19, 20-24 and 25+ - here you should have 2 p values: 15-19 is the reference and 20-24 has one p value and 25+ has another p value). Most of the tables presents at least one error on the report of the p values).

- include in the methods section that missing data was excluded and that the analysis refers to complete cases only.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

30 Sep 2021

Please see the response to reviewer file upload along with the manuscript file.

Submitted filename: Response to reviewer.docx

Click here for additional data file.

22 Oct 2021

Association between age at first alcohol use and heavy episodic drinking: An analysis of Thailand's Smoking and Alcohol Drinking Behavior Survey 2017

PONE-D-21-16712R2

Dear Dr. Saengow,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Zila M Sanchez, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

28 Oct 2021

PONE-D-21-16712R2

Association between age at first alcohol use and heavy episodic drinking: An analysis of Thailand’s Smoking and Alcohol Drinking Behavior Survey 2017

Dear Dr. Saengow:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Zila M Sanchez

Academic Editor

PLOS ONE