Less drinking, less harm: declines in adolescent alcohol use are accompanied by declines in self-reported alcohol harm

ABSTRACT

Adolescent alcohol use has declined in many high-income countries, yet few studies have measured the impact on alcohol harm. We investigated whether declines in alcohol use among secondary school students in Aotearoa New Zealand were accompanied by declines in self-reported alcohol harm. We used data from the Youth2000 survey series (2007, N = 9098; 2012, N = 8487; 2019, N = 7311). We found alcohol use declined substantially on all indicators (lifetime, current, past month, and weekly alcohol use; past month binge drinking; typical consumption of 10 + drinks/session), with declines concentrated in the 2007–2012 period. Self-reported alcohol harm also declined markedly over the 2007–2012 period, e.g. the proportion reporting being injured declined from 22% to 16%; and/or doing something that could get them in serious trouble (e.g. stealing) declined from 19% to 12%. Despite declines, alcohol harm was common in 2012 with 33% of current drinkers reporting at least one indicator in the past 12 months. In conclusion, declines in adolescent drinking were accompanied by declines in self-reported alcohol harm. However, adolescent binge drinking remained prevalent in 2019 compared with similar countries. Addressing adolescent alcohol harm remains a pressing public health priority, and there is an urgent need to better monitor youth drinking and associated harms.

KEYWORDS:

• Alcohol
• harm reduction
• public health
• primary prevention
• adolescent

Introduction

The early twenty-first century saw a steep decline in adolescent alcohol use and other risk behaviours in most high-income countries (Vashishtha et al. 2020; Ball et al. 2023). For example, in Aotearoa New Zealand (NZ) the prevalence of past-month binge drinking (5 + drinks/session) declined from 42% in 2001 to 22% in 2019 (Fleming et al. 2022) mirroring similar declines in other English-speaking and Nordic countries (Vashishtha et al. 2020). The decline in adolescent drinking is generally seen as positive from a public health perspective since it would be expected to result in declines in both acute alcohol harm and long-term health and social problems (Holmes et al. 2022). Yet to date, few studies have measured the impact of youth drinking decline on alcohol-related harm.

An early report from Sweden (Hallgren et al. 2012) noted that alcohol-related hospitalisations among young people had increased during a period of youth drinking decline and suggested that polarisation of youth drinking patterns (i.e. decreasing alcohol use among the majority, but increasing use among heavy drinkers) could explain this paradoxical finding. Subsequent research has largely refuted the polarisation hypothesis, showing that, in most countries where youth drinking declines have been observed, alcohol use has declined across all demographic groups and all levels of consumption (though with some exceptions and sometimes to a greater extent among lighter drinkers) (Norstrom and Svensson 2014; Jackson et al. 2017; Oldham et al. 2020; Loy et al. 2021). Authors have argued that data quality or methodological problems may explain the early finding of apparently diverging trends in alcohol use and harm (Norstrom and Svensson 2014), yet studies of trends in adolescent alcohol harm using alternative methods are lacking.

Social and public health concerns about alcohol harm have been growing in NZ and similar countries in recent decades (Livingston et al. 2023). In NZ the landmark Law Commission Review of 2010 (Law Commission 2010) documented alcohol harms and recommended evidence-based measures aimed at curbing them, such as excise tax increases, and strict limits on alcohol advertising and sponsorship. Subsequent Government-commissioned reports such as the Ministry of Justice Report on Alcohol Pricing (White et al. 2014), the Ministerial Forum on Alcohol Advertising and Sponsorship (Lowe et al. 2014) and the Mental Health and Addiction Inquiry (The Government Inquiry into Mental Health and Addiction 2018) have showcased further evidence to support stronger alcohol policy reform. Yet successive Governments have failed to act decisively and to date only minor policy changes have been implemented. However, at the time of writing more substantive alcohol law reform is under debate, and in this context, it is vital that NZ-specific data on adolescent alcohol use and associated harm are available to inform effective policy and targeted interventions to reduce alcohol harm. An epidemiological description of trends in alcohol use and self-reported harm can help identify harm mitigation needs and potentially effective strategies.

This study examines trends in alcohol use for secondary school students in NZ (2007–2019), using a wide range of indicators many of which have not previously been reported. We also investigate changes in self-reported indicators of acute alcohol harm between 2007 and 2012, the period of steepest drinking decline. We hypothesise that declining alcohol use in the adolescent population will be associated with a reduction in self-reported alcohol harm. If correct, this would suggest that policy actions aimed at reducing adolescent alcohol use are likely to be effective at also reducing acute harm in this age group.

Materials and methods

This paper is based on a secondary analysis of data from the Youth 2000 survey series (2007, 2012 and 2019 waves) of health and well-being in secondary school students, predominantly aged 13–18 years. The Youth 2000 survey series are nationally representative random sample surveys with a complex survey design. Detailed information about the methods for these surveys, including participant characteristics, is available elsewhere (Adolescent Health Research Group 2008, 2013; Fleming et al. 2020; Rivera-Rodriguez et al. 2021).

Survey design

All survey waves used a two-stage clustered sampling design with randomly selected schools and, within these, randomly selected students. The 2017 and 2012 waves were national surveys. The most recent wave (2019) sampled secondary schools from three regions (Tai Tokerau, Auckland, and Waikato), an area that includes 47% of NZ's secondary school population. All surveys used stratification to improve the representation of key participant groups likely to be underrepresented such as Māori and students living in rural areas. There were four strata: kura kaupapa Māori schools (Tai Tokerau, Auckland and Waikato regions), mainstream-Auckland, mainstream-Tai Tokerau and mainstream-Waikato. Weighting and calibration were used to produce comparable national estimates for all waves (see ‘Analysis’) (Lumley et al. 2011; Lumley and Scott 2017).

Response rates and sample size

School response rates were 84% (2007), 73% (2012) and 57% (2019). Student response rates were 74%, 68% and 60% respectively. Sample sizes were 9,098 (2007), 8,487 (2012) and 7,311 (2019).

Measures

Alcohol use indicators were prevalence of (i) lifetime use (i.e. ever had more than a few sips), (ii) current use (i.e. students who continued to drink at the time of the survey), (iii) past month use, (iv) use weekly or more often, (iv) past month binge drinking (5 + drinks/session), and (among current drinkers) (v) consuming 10 or more drinks on a typical drinking occasion. The survey used a branching question design, and only those reporting lifetime use were asked subsequent alcohol questions.

Self-reported alcohol harm was based on the following questions: How many times in the last year have you: (i) had friends or family tell you to cut down your alcohol drinking? (ii) had your performance at school or work affected by your alcohol use? (iii) had unsafe sex (no condom) after you had been drinking alcohol? (iv) had unwanted sex after you had been drinking alcohol? (v) done things that could have got you into serious trouble (stealing, etc.)? (vi) been injured after you had been drinking alcohol? (vii) been injured after you had been drinking alcohol, requiring medical treatment? (viii) injured someone else after you had been drinking alcohol? (ix) had a car crash after you had been drinking alcohol? For each indicator a binary (yes/no) variable for past 12-month exposure was created.

For both alcohol use and harm indicators, question wording has been consistent across survey waves. Note that data on alcohol harm was not collected in the 2019 survey so analysis of alcohol harm is limited to 2007 and 2012.

Analysis

Analysis was conducted using R (The R Foundation for Statistical Computing Version 4.1.3). Data were initially weighted using inverse probability of selection weights. Generalised raking was used to correct for non-response and to calibrate the results of each survey wave to the national secondary school population in terms of school decile (a school-level measure of the socio-economic status of the student community), student age, gender, and ethnicity. This ensured that estimates were nationally representative and comparable across survey years despite variations in the demographic make-up of the samples. Further details about weighting and calibration are available elsewhere (Rivera-Rodriguez et al. 2021). The findings presented are national prevalence estimates and 95% confidence intervals (CI).

Results

All indicators show a decline in the prevalence of adolescent drinking between 2007 and 2019 (Table 1). The results show reduced prevalence and frequency of alcohol use among secondary school students over the study period. Most of the decline occurred in the 2007–2012 period, and changes between 2012 and 2019 were modest. Between 2007 and 2012, frequent drinking (i.e. weekly or more often), past month binge drinking (5 + drinks/session) and extreme binge drinking (10 + drinks/session) declined both in the population and among current drinkers.

Table 1. Trends in alcohol use, NZ secondary school students, 2007–2019.

	2007		2012		2019
	n (N)	Weighted prevalence (95% CI)	n (N)	Weighted prevalence (95% CI)	n (N)	Weighted prevalence (95% CI)
Lifetime alcohol use	5938 (8295)	73.7% [72.0, 75.3]	4662 (8182)	61.2% [59.5, 62.8]	3278 (6841)	54.9% [53.4, 56.4]
Current alcohol use	5004 (8270)	63.3% [61.6, 65.0]	3696 (8163)	49.7% [48.1, 51.2]	2723 (6826)	47.2% [45.6, 48.7]
Past month drinking in population	3822 (8270)	48.6% [47.0, 50.3]	2555 (8163)	35.5% [33.9, 37.1]	1935 (6826)	33.9% [32.2, 35.7]
Drinking weekly or more often in population	1472 (8270)	18.7% [17.7, 19.7]	679 (8163)	9.5% [8.6, 10.4]	467 (6826)	9.2% [7.9, 10.4]
Past month binge drinking (5 + drinks) in population	2819 (8215)	36.2% [34.7, 37.8]	1838 (8140)	25.3% [23.8, 26.8]	1219 (6769)	21.7% [20.1, 23.4]
Past month drinking among current drinkers	3822 (5004)	76.8% [75.4, 78.2]	2555 (3696)	71.5% [69.9, 73.1]	1935 (2723)	72.0% [70.1, 73.9]
Drinking weekly or more often among current drinkers	1472 (5004)	29.5% [28.0, 31.0]	679 (3696)	19.2% [17.7, 20.7]	467 (2723)	19.4% [17.1, 21.7]
Past month binge drinking among current drinkers	2819 (4949)	57.5% [55.8, 59.1]	1838 (3673)	51.1% [49.0, 53.2]	1219 (2666)	46.6% [44.1, 49.2]
Consumes 10 or more drinks per typical session (current drinkers)	871 (4964)	17.0% [15.8, 18.2]	528 (3657)	14.6% [13.2, 15.9]	338 (2677)	13.2% [11.6, 14.7]

Self-reported alcohol harm among current drinkers also declined over the 2007–2012 period (Figure 1). The most commonly reported harm indicators in both periods were injuries (22% reported this indicator in 2007, declining to 16% in 2012), doing things that could get you in serious trouble e.g. stealing (19% in 2007, 12% in 2012), having friends or family tell you to cut down (16% in 2007, 11% in 2012), and having unsafe sex i.e. without a condom (14% in 2007, 12% in 2012).

Figure 1. Self-reported alcohol harm among current drinkers, past 12 months, 2007 and 2012.

Despite declines, alcohol harm remained prevalent among secondary school students in 2012, with 33% of current drinkers reporting at least one indicator of alcohol harm in the past 12 months, down from 43% in 2007 (data not shown).

Discussion

Our study provides new detail about adolescent drinking trends and associated alcohol harm in NZ. The findings confirm that, as expected, the sharp decline in adolescent drinking observed between 2007 and 2012 was accompanied by a marked decline in self-reported alcohol harm among current drinkers. Despite declines, the prevalence of potentially serious alcohol harms (including injuries, impact on schoolwork, unwanted sex, sex without a condom, criminal/anti-social behaviour, etc.) remained relatively common among adolescent drinkers in 2012, with the potential for cascading effects across the life course. The fact that more recent self-reported data on alcohol harm is unavailable highlights the urgent need for better monitoring of alcohol use and harm in this age group. However, as alcohol use changed little in the 2012–2019 period, it is likely that alcohol harm also changed little. Alcohol harm is strongly associated with binge drinking (Thor et al. 2017), which remained high among NZ adolescents in 2019 compared with similar countries such as Australia and the USA (Guerin and White 2020; Johnston et al. 2022). The Ministry of Health recommends that young people under 18 years do not drink alcohol at all, and if they do drink they should be supervised, drink infrequently and at levels never exceeding the adult daily limits: 3 drinks for males, 2 for females (Ministry of Health 2022a).

This is one of the few studies to investigate whether the recent shift in adolescent drinking had a measurable impact on alcohol harm. The study data were obtained from nationally representative samples of youth utilising a complex survey design. This allowed us to calculate national prevalence estimates representing all major sociodemographic groups within NZ. Our findings align with a Swedish study that found a consistent relationship over time between self-reported alcohol harm (using similar indicators) and alcohol use, i.e. when adolescent drinking decreased in the 2000s, self-reported alcohol harm also decreased (Thor et al. 2017). Self-reported measures have obvious limitations e.g. recall bias, misattribution of outcomes to alcohol use, failure to capture ‘invisible’ long-term impacts of alcohol use such as depression, suicide or cancer risk. Nevertheless, such studies provide quantification of changes in acute alcohol harms using measures relevant to young people, many of which are not captured by objective data such as hospitalisations (Crossin, Cleland, Rychert, et al. 2022). Triangulation of our findings using objective measures such as long-term trends in alcohol-involved emergency department presentations, and/or alcohol-involved vehicle crashes by age group is an area for future research. Unfortunately, such data series are not available in the public domain for comparison with self-reported trends and therefore analysis is required. Further monitoring and research are needed to better understand the relationship between trends in adolescent drinking and alcohol-related harm.

Research suggests that sociocultural changes, rather than alcohol control policies, are behind the international decline in adolescent drinking (Vashishtha et al. 2019; Ball et al. 2023). This shift is welcome from a public health perspective but there is no room for complacency in alcohol control policy. Alcohol remains a leading cause of health and social harm and contributes to ethnic and socioeconomic health inequities in NZ and globally (Global Burden of Disease 2018). For example, new research from NZ shows that alcohol is a key factor in suicide deaths, with Māori and Rainbow (i.e. lesbian, gay, bisexual, transgender, intersex and other sexual and gender minority) young people grossly over-represented in NZ’s appalling suicide statistics (Crossin, Cleland, Beautrais, et al. 2022). Recent reports have highlighted the failure of the Crown to protect Māori from alcohol harm (Walker 2019) and its obligation to implement Tiriti o Waitangi-informed alcohol legislation that enables the achievement of equitable health and social outcomes for Māori and empowers Māori to participate meaningfully in alcohol decisions that affect their communities (Maynard 2022).

It is notable that most of the decline in adolescent drinking occurred between 2007 and 2012, with little progress in the decade since. In fact, findings from the NZ Health Survey showed a sharp increase in hazardous drinking among 15- to 17-year olds between 2018/19 and 2020/21 (Ministry of Health 2021). The most recent findings (2021/22) appear to show a return to previous levels, but must be treated with caution due to the small sample size of 15–17 year olds in 2021/22 survey, which was affected by the Covid19 pandemic (Ministry of Health 2022b).

The findings of this study are encouraging, showing that alcohol harm in adolescents can be reduced by reducing alcohol use. However, the limited data available highlights the urgent need to monitor trends in adolescent alcohol use and alcohol harm, and conduct research to better understand contemporary determinants of alcohol harm, particularly for priority groups (e.g. Māori, Pacific and Rainbow youth). Evidence-based measures for reducing alcohol use, and thereby alcohol harm, are well established, but many have yet to be implemented in NZ, e.g. increased taxation on alcohol and stricter limits on alcohol marketing and availability (World Health Organization 2017). With alcohol policy reform currently under debate, NZ has an opportunity to implement impactful reforms. It is vital that alcohol policy in NZ have a focus on adolescent health and prevention, with an aim of continuing downward trends in adolescent alcohol use and eliminating inequity in alcohol-related harm.

Conclusions

The steep decline in adolescent binge drinking and frequency of alcohol use between 2007 and 2012 in NZ was accompanied by a significant decline in self-reported alcohol harm among current drinkers. However, adolescent binge drinking in NZ remains high relative to comparable countries and addressing adolescent alcohol harm remains an important public health priority. Ongoing monitoring of both alcohol use and related harm in this age group should be a public health priority, along with policy reforms to protect children and adolescents from alcohol harm.

Acknowledgements

The authors wish to thank the schools and students who participated in the Youth 2000 surveys and the data owners, the Adolescent Health Research Group, for access to the Youth2000 data. This research was part of a wider project on adolescent alcohol harm undertaken as a partnership between the Adolescent Health Research Group and Alcohol Healthwatch. They would like to thank the funders, the advisory group who guided the project, and Nicki Jackson, Director of Alcohol Healthwatch.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

This study used data from the Youth 2000 Survey series. Data are available on application to the Adolescent Health Research Group.

Additional information

Funding

This work was supported by the nib foundation through a 2022 Health Smart Grant; and by the Health Promotion Agency through a 2022 Alcohol Research Funding grant.