Background

Adolescence is not only a transitional stage between childhood and adulthood, it is also a critical phase in brain development and maturity during which the brain undergoes progressive and regressive changes [1, 2]. Excessive alcohol use at a young age can disturb this process and can seriously damage the development of the brain [3, 4]. Binge drinking in particularly is a harmful way to consume alcohol [5, 6]. During puberty, adolescents exhibit risk behaviour [7–9], such as drinking too much, [10–13], and studies have shown that excessive alcohol consumption in adolescents is strongly related to problem behaviour and an increased risk of suicidal behaviour [14, 15], with six or more glasses of alcohol weekly being associated with an increased risk of depression [16]. Moreover, boys with clinical depression start drinking alcohol at a younger age than their non-depressed peers [17, 18]. Subclinical mood changes are common in adolescence, ranging from "dips" in mood that usually last no longer than a few weeks to subclinical depression, which affects 17% of young people [19]. Depression and anxiety disorders have a high disease burden, even in a mild form [20] and teenagers with subclinical depression have a 6 times higher risk of developing clinical depression than teenagers without subclinical depression [19], and adolescents diagnosed with a depressive disorder are at higher risk of substance abuse, future depression, and suicidal behaviour [14, 21, 22]. Fifteen percent of adolescents report anxiety [23], with social phobia and generalized anxiety disorder in particular developing during childhood and adolescence [24]. These young people are at increased risk of developing other anxiety disorders, depression, and substance dependence [25]. Indeed, current anxiety is strongly associated with alcohol abuse in adolescents seen in primary care settings [26].

Although the number of binge drinking youngsters aged 12-14 years has decreased in recent years from 28% in 2003 to 19% in 2007, the proportion of binge drinking students aged 15-16 years has remained stable (57%). Compared with their peers in other European countries, Dutch students can be considered heavy drinkers [27–29]. In the Netherlands, the Youth Health Service provides the parents and guardians of children aged 4-18 years with guidance regarding the physical, mental, and social development of their children. One of the Service's primary tasks is to identify health risks at an early stage, which necessitates monitoring mental health and lifestyle risks, including alcohol consumption. Little is known about the drinking behaviour, and especially binge drinking, of secondary school students who have moderate or poor mental health but who have not been clinically diagnosed with depression or anxiety disorder. The aim of this study was to establish whether self-reported moderate or poor mental health is associated with binge drinking in boys and girls aged 12 to18 years.

Methods

Participants and procedures

The study was carried out by the Community Health Service (GGD) Brabant-Zuidoost, the Netherlands. Data for this study were obtained from "jeugdmonitor 12 t/m 18 jarigen 2007, GGD Brabant Zuidoost", the Provincial Youth Survey held in the south-east of the Netherlands in November 2007. This was a cross-sectional survey with self-administered questionnaires held among 19 386 youngsters aged 12 to 18 years on 1 October 2007 and living in the province of North Brabant, in the south-east part of the Netherlands. The adolescents were randomly selected, stratified by municipality, using the software application of "Statistical Package for Social Sciences (SPSS 16). The name and date of birth of the adolescents were obtained from the personal records database of the municipality.

All adolescents were sent, to their home address, a letter containing a user name and log-in code for filling in the Internet-based questionnaire, and a paper version of the questionnaire, for those without access to Internet, plus a stamp-addressed envelope for returning the completed questionnaire. The letter was addressed to the parents, who implicitly agreed to their child's participation if the child completed the questionnaire. After 6 and 12 weeks, non-responders received a reminder and a new user name, log-in code, and the paper version of the self-administered questionnaire. A raffle was held among participants, with prizes worth €15. Ethical review was not necessary for the secondary analysis of anonymous data.

Measures

Mental health

Mental health was measured with the Mental Health Inventory (MHI-5), a brief 5-item questionnaire [30]. The MHI-5 measures general mental health and can be used to screen for depressive symptoms and feelings of anxiety [31, 32]. It is part of the Short Form Health Survey (SF-36) [33, 34]. The MHI-5 compromises five questions (Table 1), each with six possible response, scored between 1 and 6 (total score ranges from 5 to 30). The total score is transformed into a variable ranging from 0-100 using a standard linear transformation, with a score of 100 representing optimal mental health. Because no formal cut-off point has been determined for the Dutch version of the MHI-5, we used a cut-off score of 60 to define moderate-to-poor mental health. This cut-off is widely used and provides the best sensitivity and specificity for detecting depressive symptoms [35].

Table 1 Questions and response categories of the MHI-5.
Full size table

Binge drinking

In this study, binge drinking was defined as 5 or more alcoholic drinks consumed on one occasion, a definition commonly used in Europe [23, 36]. The National Institute on Alcohol Abuse and Alcoholism (NIAAA) defines binge drinking as an increase in blood alcohol level of 0.8 or higher [37]. For adults, this means drinking 5 or more "standard" drinks within 2 hours for men, and 4 or more within 2 hours for women. However, there is no international consensus on the number of drinks or on the time within which the alcohol should be consumed. Moreover, the amount of alcohol in a standard drink varies by country.

Participants in this study were asked to report the number of times they consumed 5 or more drinks on one occasion (one evening or at a party) in the past 4 weeks. To derive the frequency of alcohol consumption and the average quantity of alcohol consumed, participants were asked to indicate the usual number of weekdays and weekend days per week they drunk alcohol and the number of drinks usually consumed per drinking day. Participants also reported whether they sometimes used drugs and alcohol together. For this study, the answers were recoded into two categories (binge drinker yes/no).

Confounding factors

The following factors were considered to be potential confounders: age (recoded into the category '12 to 15' and '16 to 18 years'), sex, ethnicity, family situation (living or not living with both parents), the participant's relationship with their parents (below average), education level (recoded into lower general secondary education, and higher general secondary education), stressful life events, having ADHD, and frequency of drug use (cannabis).

Data analysis

A weighting procedure was used to enable generalization of the findings to the general adolescent population in the south-east Netherlands. Post-stratification weights were calculated by comparing sample distributions and known population distributions of sex, age, and town size (the national statistics were obtained from the Central Bureau for Statistics; see http://www.cbs.nl). Non-response to a particular question was treated as missing for all analyses using that variable. Descriptive statistics for the total sample, including frequencies, means, and standard deviations (SD), were calculated. Analyses were stratified by age (12-15 and 16-18 years) and sex. Chi-square tests were performed (2 × 2 contingency tables) with binge drinking and all potential confounders and with mental health and all potential confounders (e.g., binge drinking or mental health and drugs, binge drinking or mental health and ADHD, binge drinking or mental health and ethnicity etc.), to determine which factors were confounders. Differences were tested for significance (p < 0.05). Odds Ratios (OR) and Likelihood ratios were calculated.

To investigate the association between binge drinking and mental health, multivariate logistic regression analyses were conducted with binge drinking as dependent variable and mental health as independent variable, controlling for confounders. The Forced Entry method was used, so that all predictor variables were tested in one block to assess their predictive ability, while controlling the effects of other predictors in the model. Within the logistic regression interactions and significance are examined. As regression analysis indicated that age and sex were significant predictors of both binge drinking and mental health, interaction terms were calculated for these variables.

A-2 log likelihood test with maximally 20 iterations was performed. A variable was entered into the model if the probability of its score statistic was less than the Entry value (= 0.05) and was removed if the probability was greater than the removal value (= 0.10). Cases with predicted values that exceed the classification cutoff (= 0.5) are classified as positive, while those with predicted values smaller than the cutoff are classified as negative. All statistical analyses were performed using SPSS 16.0

Results

The response rate was 52%; 77% of the participants completed the paper version of the questionnaire and 23% the Internet version.

Data collected about the frequency of mental health problems and binge drinking, socioeconomic variables, and possible confounders are given in Table 2.

Table 2 Sample characteristics.*
Full size table

Girls reported poor or moderate mental health more often than did boys. The older boys and girls were, the more often they reported mental health problems, with the reporting rate increasing from 9.0% in 12-year-olds to 18.4% in 18-year-olds. Most of the adolescents aged 16-18 years had been binge drinking in the previous 4 weeks. The proportion of adolescents with mental health problems who were binge drinkers is given in Table 3.

Table 3 Chi-square tests: Association between mental health problems and binge drinking.*
Full size table

Girls and boys aged 12-15 years with mental health problems were more likely to be binge drinkers than were their peers without mental health problems. However, boys aged 16-18 years with mental health problems were less likely to be binge drinkers than their peers without mental health problems. Their was no such significant relationship in 16-to 18- year old girls.

Table 4 shows the results of the logistic regression analysis with binge drinking as dependent variable. Girls aged 12-15 were 2.4-times more likely to be binge drinkers if they reported moderate or poor mental health, even after correction for education, ethnicity and family situation. Boys of the same age were 1.6 times more likely to be binge drinkers if they reported mental health problems after correction for potential confounders. In contrast, boys aged 16-18 years reported binge drinking significantly less often in the past 4 weeks if they experienced moderate or poor mental health, they were 0.6 times less likely to be classified as binge drinkers if they experienced mental health problems. No such association was seen in girls of the same age.

Table 4 Logistic regression analysis of the relationship between binge drinking (dependent variable) and mental health problems,
Full size table

Discussion

The purpose of this study was to investigate whether self-reported moderate or poor mental health is associated with binge drinking in boys and girls aged 12-18 years. We found that boys and girls aged 12-15 years with mental health problems were at significantly higher risk of binge drinking than their mentally well peers and this association remained after correction for confounders. The majority of boys aged 16-18 years were binge drinkers (69.6%). In the Netherlands, the legal age for buying drinks containing less than 15% alcohol is 16 years (1964 Licensing and Catering Act). Most youngsters go to the pub once they become 16, and most drink at home with friends before going out [12]. Drinking is a social activity. While it has been reported that binge drinking boys exhibit externalizing behaviour [15], we found that the boys aged 16-18 years with mental health problems were significantly less likely to be binge drinkers. A possible explanation is that these boys tend to withdraw when experiencing social phobia or depressive feelings, so they are more likely to stay at home than to go out drinking with friends. We did not find mental health problems to be associated with binge drinking in girls aged 16-18 years. This is contrary to our expectation, because alcohol consumption is reported to increase in female students who experience daily sadness [38]. This discrepancy might be due to differences in study population or as a result of selection bias (relatively many mentally well girls might have completed the questionnaire). Further research is recommended.

Strengths and limitations

The large sample size (10 090 adolescents) of this study made it possible to stratify by age and sex. We chose to stratify by age (under 16 and 16 years and older) because 16 is the legal age for buying alcohol in the Netherlands and because binge drinking and mental health problems are strongly associated with age. We investigated the frequency of binge drinking in adolescents aged 12-18 years with or without mental health problems for each age group separately, but because few children aged 12-13 years were binge drinkers and had mental health problems, we could not make comparisons for each age group. However, we did find binge drinking to be associated with self-reported mental health problems in children younger than 16 years. In contrast, binge drinking was not associated with mental health problems in children older than 16 years, in the sense that in this age group it was typically children without mental health problems who were binge drinkers.

The administration of the questionnaires at home and the assurance that data would be processed anonymously might have encouraged the participants to respond truthfully, so that we collected reliable and valid data. It can be expected that peer influence is less when questionnaires are completed at home rather than at school. A potential limitation of this study is the reliance on self-report data, so that responses to sensitive questions about undesirable or illegal behaviour may have been biased. Another potential limitation is the seemingly low response rate of 52%. We did not perform a non-responder analysis, and so we do not know whether heavy drinkers or adolescents with poor mental health were under-represented or over-represented in our sample. We think that binge drinking among adolescents with mental health problems is more likely to have been underestimated than overestimated in our study, because both mental health problems and binge drinking increase with age and our study included more younger than older adolescents. Moreover, Knudsen et al. showed that the non-participants in their population-based health study typically had poorer health habits (including risky alcohol consumption) and poorer general somatic and mental health than the participants. These authors also showed that internal associations tended not to vary much with response rate [39]. In a review article published in 2007, Galea and Tracy also concluded that the available empirical findings showed "little evidence for substantial bias as a result of nonparticipation" [40].

There may also have been a selection bias because few immigrants and more adolescents who were following higher general secondary education than lower general secondary education participated in the study. Moreover, more girls than boys and more young adolescents than older adolescents participated; however, we attempted to correct for this by using a weighting procedure for sex and age. Lastly, the cross-sectional design limited the ability to draw conclusions about causality.

Conclusions

Anxiety and depression are major causes of morbidity and disability and constitute a major public health burden [20]. Binge drinking is a particularly harmful way to consume alcohol [6], and especially at a young age when the brain is still developing and is susceptible to alcohol-induced damage [4, 5]. Different definitions of binge drinking are used in the literature, but we defined binge drinking as the consumption of 5 or more alcoholic drinks on one occasion. As the blood alcohol level increases when alcohol is drunk over a short period, other definitions of binge drinking that specify the period over which the drinks are consumed (such as the NIAAA) may result in a stronger or weaker association between binge drinking and poor mental health.

The association between mental health and binge drinking emphasizes the importance of alertness on the part of primary care practitioners and youth health services to depressive and anxiety symptoms and binge drinking, especially among boys and girls aged 12-15 years. In the Netherlands, all youngsters in the second class of secondary school, when most children are 13-15 years old, are given a medical examination by the youth health services. If adolescents, especially girls, report moderate or poor mental health, they should be asked about their drinking habits and, vice versa, if they report binge drinking, they should be asked whether they are anxious or depressed.

Many professionals are not aware of the association between mental health problems and alcohol use [41]. Yet knowledge of the risk of comorbidity between anxiety or depressive feelings and binge drinking in adolescents, and early identification of those at risk, will lead to early and more effective intervention. Effective interventions for youths are described in the database of the Netherlands Youth Institute (NJI); see http://www.nji.nl.

A longitudinal study of the association between mental health and binge drinking is needed to understand the causal link between binge drinking and mental health problems, which may help to improve the early detection and treatment of both problems.