Alcohol and Alcoholism Advance Access originally published online on May 2, 2008
Alcohol and Alcoholism 2008 43(4):477-480; doi:10.1093/alcalc/agn037
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Alcohol Consumption in The Netherlands in the Last Decade: Sharp Decreases in Binge Drinking, Especially among Youngsters
1 Tranzo, Tilburg University, Tilburg, The Netherlands
2 Addiction Research Institute, Erasmus University, Rotterdam, The Netherlands
* Corresponding author: Tranzo, Faculty of Social and Behavioural Sciences, Tilburg University, P.O. Box 90153, 5000 LE Tilburg, The Netherlands. E-mail: h.garretsen{at}uvt.nl
Received 14 September 2007; first review notified 15 November 2007; in revised form 9 April 2008; accepted 10 April 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT IntroductionMethods and MeasurementsResultsDiscussionReferences |
|---|
Aim: The aim of this study was to examine the development of alcohol consumption and ‘binge drinking’ in The Netherlands over the period 1997–2005. Methods: Data from three national population surveys commissioned by the Dutch Ministry of Health were compared. Results: Decreases in binge drinking have appeared, especially for youngsters, which are not explained by changes in the composition of the population according to religious orientation, and for which no other explanation can be given at present. Conclusion: Trends in binge drinking in The Netherlands can vary over relatively short spaces of time.
|
Introduction |
|---|
|
TOPABSTRACTIntroductionMethods and MeasurementsResultsDiscussionReferences |
|---|
National population surveys on the use of alcohol and illicit drugs have been carried out in The Netherlands in 1997, 2001 and 2005, commissioned by the Dutch Ministry of Health, Welfare and Sports. The three surveys are to a large extent comparable. In this article, results of the three surveys with regard to the quantity and frequency of drinking and binge drinking (defined as drinking six or more glasses of alcoholic beverages at least 1 day per week in the past 6 months) are discussed.
Compared to other European countries, the level of alcohol consumption in The Netherlands is moderate. The per capita consumption in the past decade fluctuates around the 8.0 l of pure alcohol per year, indicating that no major changes in consumption have taken place. (Netherlands National Drug Monitor, 2006
). Per capita figures, however, only give a restricted insight. Differences between population subgroups may exist and differences in drinking patterns may occur.
Among European youngsters, the use of alcohol is high and binge drinking occurs frequently. In 2003, a study in 35 European countries revealed that Dutch students (aged 15 and 16) are in the lead with regard to binge drinking, together with students from Ireland, the Isle of Man, the United Kingdom and Germany. (Hibell et al., 2004). It should be noted that binge drinking was more prevalent amongst girls (albeit on a lower cut-off) than boys in countries like the UK, but not in The Netherlands. It is not clear why this is the case.
In this following paper we explore the development of alcohol consumption and ‘binge drinking’ in The Netherlands over the period 1997–2005 for the whole population and for different subpopulations, defined by gender, age and urbanization degree.
|
Methods and Measurements |
|---|
|
TOPABSTRACTIntroductionMethods and MeasurementsResultsDiscussionReferences |
|---|
Sampling and data collection
Data from the National Prevalence Studies (NPS) of 1997, 2001 and 2005 are analysed. These three studies are based on random samples selected from the Dutch Municipal Populations Registry.
The National Prevalence Study 2005 was executed by the Addiction Research Institute Rotterdam (Rodenburg et al., 2007). The sampling frame of NPS 2005 was the general national population aged 15–64 years, as registered on 1 January 2005 in the Dutch Municipal Population Registry, excluding institutionalized populations, such as people in nursing homes, penitentiary and psychiatric clinics. Following a two-stage sampling design, first a representative sample of municipalities was selected, and next a representative sample of persons. From May 2005 until December 2005, data were gathered by CAPI (Computer Assisted Personal Interview) by interviewing respondents at home. Collaboration was established with Statistics Netherlands (CBS, a national governmental agency). The NPS on prevalence of substances was incorporated into the data-collection of the integrated survey on household living conditions (POLS) of Statistics Netherlands. The survey on household living conditions contains questions on alcohol use, smoking and the use of medications. Questions on other substances, such as cannabis, cocaine, other hard drugs and performance enhancing drugs, were added to the survey of the POLS 2005.
The NPS 1997 and 2001 was executed by Cedro, an addiction research institute based in Amsterdam (Abraham et al., 1999, 2002). The sampling frames were comparable to NPS 2005, only the age limits differed (12 years and older in 1997 and 2001). NPS 1997 and 2001 were not incorporated into the data collection of another survey.
In comparison to NPS 1997, NPS 2001 and NPS 2005 differed somewhat in the data collection modes. In NPS 1997, data collection consisted of personal interviews in the home situation (CAPI). However in NPS 2001, besides the CAPI method, four other interview methods were used, namely telephone interviews, postal questionnaires, questionnaires on disk (floppy disk by mail) and online questionnaires. In NPS 2005, besides the CAPI method, questionnaires were administered within an online panel study.
The aim of this paper is to detect trends. Therefore, with regard to the NPS 2001 and NPS 2005, only the CAPI-data are used. In addition, in all studies, participants were selected within the age-range of 15–64 years.
After the fieldwork, all data were statistically weighted on demographic variables that showed differential distributions, in comparison to the distribution of these variables in the general population, as registered by Statistics Netherlands. These were gender, age, marital status, family size, address density, part of the country and month of data collection.
Measurements
Participants’ alcohol use was assessed by asking participants whether they had drunk alcohol in the past 12 months. If so, participants were asked whether they had drunk in the past month as well. In addition they were asked whether they had drunk six or more glasses of alcoholic beverages in one day in the past six months. Also the frequency of drinking six or more glasses in one day in the past six months was asked. A dichotomous variable ‘binge drinking’ was constructed, differentiating between respondents who do and do not drink at least once a week six or more glasses of alcoholic beverages. The average glass of alcoholic beverage in The Netherlands contains 10-g ethanol.
In addition, information on demographic characteristics such as gender (2 levels; male, female), age (3 levels; 15–24, 25–44, 45–64 years) and address density (5 levels ranging from very low to very high) were obtained.
Response rates
The net sample sizes (aged 15–64) and the overall response rates are presented in Table 1.
|
As stated above, in 2001 other methods of data collection were introduced. The response rate of 37.4% in 2001 is the response rate for CAPI-respondents only. The total response in 2001 was 43.5%.
Analyses
Differences between respondents on demographic variables (age, gender and address density) were explored by computing cross tabs and Pearson's chi-square tests.
|
Results |
|---|
|
TOPABSTRACTIntroductionMethods and MeasurementsResultsDiscussionReferences |
|---|
Table 2 presents data on the development of alcohol consumption and ‘binge drinking’ over the period 1997–2005 (research question 1).
|
No major changes in the past year consumption and the past month consumption have taken place. However, a significant decrease has taken place between 2001 and 2005, both with regard to the drinking of six glasses or more in 1 day once or more in the past half year as well as with regard to binge drinking on a weekly basis. This percentage of people drinking at least one day a week six glasses or more decreased from 14.6% in 2001 to 10.3% in 2005.
Results differ somewhat for population subgroups defined by gender, age and urbanization degree.
As we can see in Table 3, with regard to the past year alcohol consumption, hardly any significant changes have taken place between 1997 and 2005, except for a significant decrease for the age group 25–44: from 87.3% to 83.9% (P < 0.001) and the same is true for past month drinking (data not shown), showing a significant decrease (P < 0.01) for the same age group.
|
Significant changes for different population subgroups have been found for binge drinking however.
A significant decrease in binge drinking is found for both men and women (see Table 4). It is striking that the proportion of binge drinking young men aged 15–24 appears to have halved between 2001 and 2005: from 39.9% to 19.5% (see Table 5). Sharp decreases are also found for women aged 15–24 and for men and women aged 25–44. Sharp decreases have also been found for people living in highly populated regions (see Table 4).
|
|
|
Discussion |
|---|
|
TOPABSTRACTIntroductionMethods and MeasurementsResultsDiscussionReferences |
|---|
The study has some limitations
The response rates in the different studies vary. Especially in 2001, the response rate was relatively low. However, this response rate is not low compared to other Dutch (alcohol) surveys (Dotinga, 2005
; Lahaut, 2006).
It is important to know whether bias might have resulted because of non-response. Problems could arise because a smaller sample size could lead to a loss in accuracy in population estimates and because it could lead to a possible selection bias. The first problem is unlikely: the sample sizes are big enough to allow for the analyses undertaken. With regard to the second, we analyzed the effect in 2001 and in 2005 of non-response on alcohol variables. In 2001 a slight under-representation of the alcohol use was found, and in 2005 a slight over-representation. In a previous study our research group studied in detail non-response and data quality in mail surveys on alcohol consumption (Lahaut, 2006). Strong evidence was found for a possible over-representation of non-responders among abstainers. More important for the conclusions in the present study, weak evidence was found for over-representation of non-responders among frequent excessive drinkers. This should be kept in mind when interpreting the present results, perhaps the group ‘frequent excessive drinkers’ is a little under-represented in the present study too. The high response of 2005 is noticeable. This possible under-representation of frequent excessive drinkers could be less in this year.
This high response of 2005 raises the question whether the incorporation of questions on substance use within an already existing survey (in this case a survey on household living conditions) has a positive effect on the response rate in itself. This could be the case because, for instance, non-drinkers may not respond to alcohol surveys because they assume it does not apply to them. However, earlier research of our research group with regard to the ‘labelling’ of a survey as an alcohol survey or a health survey did not show differences in response rates (Lahaut, 2006). So the picture remains confusing and it is questionable whether the incorporation within the existing survey in 2005 has led to the relative high response rate.
Moreover, the method of data collection used, the CAPI method, could lead to selective non-response. The one population subgroup could be reached better by this method than the other. It is expected that this mechanism will not have influenced the results concerning the differences in consumption found over the years, since in all three studies the same data collection mode was used.
Another limitation of the present studies is that the same definition of binge drinking is used for men and women. This influences the results. More men than women tend to drink six glasses or more in one occasion. It would be interesting to know what the result would have been if binge drinking among women had been defined as drinking four or more glasses on one occasion. This is an important issue for further research.
Sharper decreases in binge drinking were found for younger people living in highly populated regions, compared to low populated regions. We checked whether this could partly be related to the fact that people with an Islamic religious orientation (who are not allowed to drink alcohol) tend to live in the bigger cities. Our research group has shown that second generation Turkish people, who are mainly Islamic, aged 16–30, tend to be (excessive) drinkers less often than native Dutch people (Dotinga, 2005). However, when we controlled for differences in reported religious orientation in the populations over time, significant differences between 1997 and 2005 in binge drinking levels among young people living in highly populated regions remained (analysis by ANCOVA—analysis of covariance).
To summarize, this study shows that the percentage of (past year) drinkers is rather stable. This is in line with the sales figures, with the per capita consumption in the last decade fluctuating around the 8.0 l of pure alcohol a year (World Drink Trends, 2005). This is also in line with other Dutch research. For example, Studies of Statistics Netherlands revealed that in 2001 82% and in 2005 81% of the population of 12 years and older has ‘occasionally’ drunk alcohol (National Drug Monitor, 2006).
Compared to other Western European countries the per capita consumption of alcohol is rather moderate in The Netherlands. Students drink relatively frequently however. (Netherlands National Drug Monitor, 2006; Hibell et al., 2004). The percentages of binge drinkers, however, appear to have decreased sharply for some population subgroups. That finding was also present in a study of Statistics Netherlands, which showed a slightly decreasing, but not very clear, trend whereby the percentage of binge drinkers (defined by them as heavy drinkers, but using the same cut-off points) fluctuated thus: 13% in 2000, 14% in 2001, 12% in 2002, 11% in 2003 and 12% in 2004. However this study of Statistics Netherlands also shows sharper decreases for some population subgroups. For example, the percentage of binge drinking among young men aged 18–24 decreased from 42% in 2002 to 33% in 2005, and for women aged 18–24 the percentage decreased from 18% to 8% (National Drug Monitor, 2006).
From a public health point of view these results are positive, since binge drinking can cause accidents and other alcohol-related problems. Remarkably, however, recently less-positive data were presented too in The Netherlands. Hospital data showed that in 2005 the number of children aged 10–15 treated in a hospital for alcohol intoxication was six times higher than in 2000. An increase was found, too, for people aged 15 plus, but not that high (Valkenberg et al., 2007). It could be that a decrease in binge drinking is taking place in The Netherlands but that at the same time the percentage of people who drink literally to intoxication (much more than six glasses at one occasion) is increasing, especially among very young children, even younger than the age of 15, studied in our research.
Sharp decreases in binge drinking have also been found for younger people living in highly populated regions, compared to people in low populated regions. The sharp decreases found for young people are difficult to explain. No clear explanations have been found yet.
Future research might include qualitative studies to shed light on these findings. Observational studies could yield more valuable information. Few observational studies are aimed exclusively at young people's drinking behaviour. Such studies should take place in natural settings (Goor van de, 1990).
|
References |
|---|
|
TOPABSTRACTIntroductionMethods and MeasurementsResultsDiscussionReferences |
|---|
Abraham MD, Cohen PDA, Til RJ van, et al. Licit And Illicit Drug Use in The Netherlands, 1997 (1999) Amsterdam: CEDRO.
Abraham MD, Kaal HL, Cohen PDA. Licit And Illicit Drug Use in The Netherlands, 2001 (2002) Amsterdam: CEDRO.
Dotinga A. Drinking in a dry culture. Alcohol use among second-generation Turks and Moroccans: measurements and results. In: Thesis (2005) Rotterdam: Erasmus University.
Goor van de LAM, Knibbe RA, Drop MJ. Adolescent drinking behavior: an observational study of the influence of situational factors on adolescents drinking rates. J Stud Alcohol (1990) 51:548–55.[Web of Science][Medline]
Hibell B, Andersson B, Bjarnason T, et al. Alcohol and other drug use among students in 35 European countries. In: The ESPAD report 2003 (2004) Stockholm: CAN.
Laar MV, Cruts A, Verdurmen J, et al. Nationale Drug Monitor Jaarbericht 2006 (2007) Utrecht: Trimbos Institut.
Lahaut VMHCJ. The missing drink. Nonresponse and dataquality in mail surveys on alcohol consumption. In: Thesis (2006) Rotterdam: Erasmus University.
Netherlands National Drug Monitor. (2006) Netherlands Institute for Mental Health and Addiction, Utrecht, 2007.
Rodenburg G, Spijkerman R, Eijnden R, et al. Nationaal Prevalentie Onderzoek Middelengebruik 2005 (2007) Rotterdam: IVO.
Valkenberg H., Lely N Van Der, Brugmans M. Alcohol en jongeren, een ongelukkige combinatie. In: Medisch Contact (2007) 33/34.
World Drink Trends. (2005) World Advertising Research Center, Oxfordshire.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||