Objective. To explore whether people who start drinking at an early age are more likely to have been in physical fights after drinking, independent of respondent history of alcohol dependence and frequency of heavy drinking.
Design. In 1992, the US Census Bureau interviewed, in person, 42 862 randomly selected adults age 18 or older, mean age 44, household response rate 91.9%, and in-person response rate 97.4%. The survey included questions regarding the age respondents started drinking, frequency of heavy drinking, and whether respondents were ever or in the past year in a physical fight after drinking alcohol.
Results. Relative to respondents who did not begin drinking until age 21 or older, those who started drinking before age 17 were 2.9 to 4.1 times more likely ever in their lives, and at least 3 times more likely in the past year, to have been in a fight after drinking. These relationships were found even after controlling for personal history of alcohol dependence, years of drinking, frequency of heavy drinking, smoking, drug use, and other personal characteristics associated with the age respondents started drinking.
Conclusions. An early age of drinking onset was associated with alcohol-related violence not only among persons under age 21 but among adults as well. Physicians need to query adolescent patients about the age they began drinking and counsel them about the increased risks associated with early drinking onset, such as an increased risk of being involved in alcohol-related violence.
A recent analysis of the National Longitudinal Alcohol Epidemiologic Survey (NLAES) found that the younger people were when they started drinking, the more likely they were to have experienced alcohol dependence during their life as defined by the Diagnostic and Statistical Manual of Mental Disorders, Fourth Editioncriteria (DSM-IV).1 Among both males and females, and persons with and without a family history of alcoholism, persons who began drinking before age 14 were at least 3 times more likely than those who did not drink until they were over 21 to experience diagnosable alcohol dependence during their life.1 Subsequent analyses of NLAES have revealed that even after controlling for personal history of alcohol dependence, people who started drinking at an earlier age drink heavily with greater frequency during both adolescence and adulthood.2
An earlier age of drinking onset has also been associated with a greater likelihood of experiencing alcohol-related unintentional injuries2 and motor vehicle crash involvement after drinking,3 even after controlling for frequency of heavy drinking, personal history of alcohol dependence, and other characteristics related to the age respondents began drinking. This was found ever in a respondents' life and during the year before the survey, when most respondents were over age 21. These studies revealed that after analytically controlling for respondents' frequency of heavy drinking and personal history of alcohol dependence, people who start drinking at an early age were more likely, on the occasions that they drink, to engage in behavior that increases their chances of being unintentionally injured, such as swimming, using machinery, walking in a dangerous area or in traffic, or driving a motor vehicle after having too much to drink. Among adolescents, some individuals are more likely to engage in a variety of risk taking/deviant behaviors.4Whether people who begin drinking at an early age also have a disproportionate propensity when they drink during adulthood to engage in violent behavior, which could injure themselves or others, has not been explored.
An association between alcohol consumption and violent aggressive behavior has long been recognized.5 Each year, of the 11.1 million victims of violent assaults in the United States, 25% believed the perpetrator had been drinking.6 Roizen7has estimated 37% of assault offenders, 60% of sexual offenders, 57% of men and 27% of women involved in marital violence, and 13% of child abusers had been drinking at the time of the event. Pernanen8 reported that 42% of violent crimes reported to the police involved alcohol. According to the Bureau of Justice Statistics,6 40% of persons convicted of homicide and 25% of victims had been drinking at the time of the event.
Alcohol may encourage aggression by disrupting normal brain mechanisms that normally restrain impulsive behaviors such as aggression.9 A narrowing of attention after drinking may lead to inaccurate assessment of future risks of acting on violent impulses.10 Alcohol may also impair people's ability to process information, and hence, may contribute to misinterpretation of language or behavior thereby increasing conflict.11
Antisocial personality disorder and early onset (type II) alcoholism frequently co-occur.12 Antisocial personality disorder is a psychiatric disorder characterized by a disregard for the rights of others often manifested as a violent criminal lifestyle.
Questions in the NLAES survey permit us to explore whether people who start drinking at an early age are disproportionately likely to engage in violent behavior after drinking both as adolescents and adults, independent of their frequency of heavy drinking, personal history of alcohol dependence, and other behavioral and demographic characteristics associated with early age of drinking onset. Because clinical,13–15 legal,16,,17 and community interventions18,,19 can reduce alcohol consumption and related unintentional and violent injury among adolescents and adults, these questions should be of direct importance to pediatricians and other health care providers.
The NLAES was a national probability survey sponsored by the National Institute on Alcohol Abuse and Alcoholism. In 1992, the US Bureau of the Census administered face-to-face interviews with 42 862 respondents age 18 and older, mean age 44, residing in the noninstitutionalized population of the contiguous states, including the District of Columbia.
The multi-stage sampling approach used for NLAES was described by Massey and colleagues.20 Primary sampling units were stratified according to socioeconomic criteria and were selected with a probability proportional to their population size. Within primary sampling units, geographically defined secondary sampling units, referred to as segments, were selected systematically for the sample. The black population was oversampled at this stage of the sample selection to secure adequate numbers for analytic purposes. Segments were then divided into clusters of 4 to 8 housing units, and all occupied housing units were included in the survey. Within each household, 1 randomly selected person age 18 and older was selected to participate. Young adults age 18 to 29 were oversampled at a ratio of 2.25:1.00 at this stage of sample selection to include a greater representation of this heavier drinking population subgroup. Weighted analyses using SUDAAN21 adjusted for the deliberate oversampling of black individuals and persons age 18 to 29 and accounted for the complex sampling design of NLAES. The household response rate for this representative sample of the US population was 91.9%, and the sample person response rate was 97.4%. The overall response rate was 90%. (A longitudinal follow-up was initially planned but not completed because of financial constraints. A repeat cross-sectional survey is currently being considered).
The primary outcome in this study was having been in a fight after drinking. This was measured by asking respondents, “In your entire life, have you ever been in a physical fight while or right after drinking alcohol? Did that happen in the past 12 months?”
The age of drinking onset was ascertained by asking respondents how old they were when they first started drinking, not counting small tastes or sips of alcohol. Drinking onset data were collected from respondents who were classified as current drinkers (persons who had consumed at least 12 drinks in the past 12 months) and former drinkers (persons who had consumed at least 12 drinks in any 1 year of their lives but not during the year before the interview). Only current and former drinkers were included in these analyses. In an independent study before the full NLAES was conducted, the test-retest reliability of the drinking onset variable was good with a Kappa of 0.72.22
Measures of alcohol use and dependence were derived from the Alcohol Use Disorder and Associated Disabilities Interview Schedule23; a fully structured diagnostic psychiatric interview designed to be administered by trained interviewers who were not clinicians.
The definition of lifetime alcohol dependence was based on the diagnostic criteria of the DSM-IV.24 The Alcohol Use Disorder and Associated Disabilities Interview Schedule included an extensive list of symptom questions that operationalize theDSM-IV criteria for alcohol dependence. Diagnosis of alcohol dependence required that in any 1 year a respondent meet at least 3 of the following 7 criteria for dependence: 1) tolerance, 2) withdrawal or avoidance of withdrawal, 3) persistent desire or unsuccessful attempts to cut down or stop drinking, 4) spending much time drinking, obtaining alcohol, or recovering from its effects, 5) giving up or reducing occupational, social, or recreational activities in favor of drinking, 6) impaired control over drinking and 7) continuing to drink despite a physical or psychological problem caused or exacerbated by drinking. Respondents were classified as showing a lifetime alcohol dependence diagnosis if they experienced an episode of dependence in the past year or at any time before the past year. The independent test retest study determined good reliabilities with Kappas of 0.76, and 0.73 for past year and before past year dependence diagnoses, respectively.23
Current heavy drinking was determined by asking respondents, “during the past 12 months, about how often did you have 5 or more drinks of any type of alcohol in a single day?” Respondents were also asked, “About how often would you say you usually drank enough to feel drunk during the last 12 months? By drunk, I mean times when your speech was slurred, you felt unsteady on your feet, or you had blurred vision.” We also constructed a measure for the frequency with which respondents drank 5 or more drinks during their period of heaviest drinking in their life. Respondents were asked whether there were periods in their life when they drank more than in the past year. Respondents who answered “yes” were then asked their frequency of drinking and the amounts they usually consumed on days they drank during this period. They were also asked what was the greatest amount they drank on any day during that period and how frequently they drank that amount. We used data from these questions, as well as questions about frequency of drinking 5 or more drinks in the past year, to determine their frequency for drinking 5 or more drinks during their lifetime period of heaviest drinking.
The following demographic and behavior variables were examined as potential confounders of the association between age of drinking onset and later behavior: current age, sex, race/ethnicity (white non-Hispanic, black non-Hispanic, Hispanic, other), education, marital status, smoking status (current, former, never), illicit drug use status (current, former, never), alcohol dependency (current, former, never), years of drinking alcohol, frequency of drinking to intoxication in the past year, and frequency of drinking 5+ drinks on an occasion during the past year and during the respondents lifetime period of heavy drinking.
Like other reports of the NLAES, our statistical analyses were conducted using the SUDAAN statistical package21 to account for the complex survey design and oversampling of NLAES in the estimation of both effects and their standard errors. The pattern of results was very similar with and without use of the SUDAAN weighting. Age of drinking onset was categorized as under 14, each year separately from 14 through 20, and 21 or older (the minimum legal drinking age throughout the United States). We focused on respondents who reported drinking ever in their lifetime (N = 27 081). The univariate associations between age of drinking onset being in a fight while or after drinking and demographic and background behavior characteristics was tested using a modified test of independence that adjusts for the sampling design. The test for independence in SUDAAN is based on the Wald statistic comparing observed and expected values and follows the strategy proposed by Koch, Freeman, and Freeman25 for analyzing complex survey data.
Logistic regression explored whether age of drinking onset was associated with ever or in the past year being in a fight after drinking outcomes, controlling for potential confounding from demographic and behavior characteristics, including alcohol dependency and years of drinking alcohol. The overall significance of the relations between different ages of drinking onset and study outcomes in the logistic models were tested through a χ2statistic comparing models with and without the set of indicator variables representing age of onset. We examined whether the potential association between age of drinking onset and being ever or in the past year in a fight after drinking persisted after controlling for alcohol dependence, years of drinking alcohol, history of drug use and smoking, and other demographic characteristics related to age of drinking onset because of the established relationships between alcohol use5–7 and physical violence and between age of drinking onset and alcohol dependency.1
We then repeated the analysis further controlling for the frequency of heavy drinking by respondents. This was done to see if respondents who started drinking at an early age were more likely to be in fights after drinking simply because they drank heavily more frequently. When we looked at whether respondents were ever in their lives in a fight after drinking, we entered into the regression model how often respondents drank after 5+ drinks during their period of heaviest drinking. When we examined the relation between age of drinking onset and being in a fight in the past year after drinking we entered into the regression past year frequency of drinking 5+ and drinking to intoxication.
Sixty-five percent of respondents ever drank alcohol, 49% before age 21 and 3% before age 14. The mean drinking onset age was 19.0 years. Four percent were classified as alcohol dependent in the past year and 13% ever in their life.
Eleven percent of respondents who ever drank (7% of the entire sample) reported being in a fight while or after drinking at some point in their life, and 2% (1% of the entire sample) reported this behavior in the past year. Forty-three percent of those who reported having been in a fight had not in their life met alcohol dependence criteria.
The earlier the age respondents began drinking, the greater the proportions who reported being in physical fights while or after drinking, ever in their lives, and during the year before the survey. Persons who began drinking before age 14 were 11 times more likely than those who began drinking after age 21 to report ever being in a fight while or after drinking, 34% versus 3%, and 14 times more likely to report doing so in the past year, 5% versus 0.36% (Fig 1). The magnitude of differences tended to become smaller as the age of drinking onset became closer to 21.
Several subgroups in the study were significantly more likely to have begun drinking at younger ages: male respondents, younger respondents, those with less than a high school education, persons never married, and persons who currently or ever smoked or used illicit drugs, respondents with a current (past year) or lifetime diagnoses of alcohol dependence, and years of drinking alcohol (all relationsP < .001; data available on request). To control for these potentially confounding characteristics, we entered each of those background characteristics into multiple logistic regression analyses that examined whether or not respondents were in a physical fight while or after drinking as a dichotomous outcome variable. This outcome was examined ever during the respondent's life and during the year before the interview.
Respondents who began drinking at an earlier age were significantly more likely to report that they were in a physical fight after drinking, ever in their lives, and during the past year. For each study outcome, the odds ratio and 95% confidence intervals (CI) for respondents who began drinking each year from age <14 through 20 relative to those who began at age 21 or older was examined separately. Results are summarized in Figs 2 and3.
Figure 2 (Model 1) examines whether respondents were ever in their life in a fight while or after drinking. The strongest relationships were observed when comparing respondents who started drinking at age 14 relative to those starting at age 21 or older. Persons who began drinking before age 14 were 4.7 (95% CI: 3.6, 6.1) times more likely than those who began after age 21 to report ever being in a fight after drinking after controlling for ever being alcohol dependent, years of drinking alcohol, age, gender, race/ethnicity current and previous illicit drug use and smoking, and family history of alcoholism. As can be seen in Fig 2, Model 2, even after further controlling for frequency of drinking 5+ during their period of heaviest drinking, those who started drinking before age 14 were 4.1 (95% CI: 3.2, 5.4) times more likely to have ever been in a fight after drinking. In each individual age group examined, those who started drinking before age 21 were significantly more likely than those who started at age 21 or older to have been in a physical fight after drinking.
Figure 3 presents the relationships between age of drinking onset and being in a physical fight after drinking in the past year. Model 1 analytically controls for age, gender, race/ethnicity, education, family history of alcoholism, personal history of alcohol dependence, and years of drinking alcohol. Model 2 further controls for frequency of drinking 5+ drinks on an occasion and frequency of drinking to intoxication in the past year. In both models, persons who began drinking at age 16 and younger were significantly more likely to report being in a fight after drinking than those who began drinking at age 21 or older. Even after controlling for personal history of alcohol dependence, frequency of drinking 5+ or to intoxication in the past year, years of drinking alcohol, and other characteristics related to the age of drinking onset, respondents who started drinking at ages 16 and younger were at least 3 times as likely to report being in a fight in the past year while or right after drinking. Controlling for frequency of drinking 5+ drinks or to intoxication diminished slightly, but did not eliminate the significantly greater likelihood that persons who began drinking at age 16 or younger were more likely to be in a physical fight in the past year after drinking.
Evaluations of the age people start to drink alcohol reveal that early drinking onset is associated with increased risk of alcohol-related health and social problems among adults including: alcohol dependence later in life,1 frequent heavy drinking even among nondependent persons,2 unintentional injuries under the influence,2 and alcohol-related motor vehicle crashes.3 This study found that the earlier respondents began drinking, the greater the likelihood they would be in a physical fight while or right after drinking. Part of the increased risk for fighting after drinking results from early age drinkers being more likely to develop dependence on alcohol and/or patterns of frequent heavy drinking. However, those who start drinking at an early age were more likely to be in physical fights after drinking even after controlling for history of alcohol dependence, frequency of heavy drinking, years of drinking alcohol, drug use and smoking history, and other characteristics related to the age people start drinking. All of these relations were found ever in a respondents' life and during the year before the survey. Thus, it is possible that delaying the age people start drinking may reduce alcohol dependence, unintentional injuries, and alcohol-related violence not only during adolescence, but adult years as well.
The strongest relationship between the age of drinking onset and being in a fight after drinking was seen among respondents who began drinking before age 18. It should be noted that many respondents in this study grew up before 1988 when the legal drinking age ultimately became 21 in every state. However, alcohol purchase at ages below 18 was illegal for study respondents in all states.
Several factors should be considered when interpreting the study results. First, social desirability may have influenced responses. Persons more willing to admit drinking at early ages when drinking is illegal may be more willing to acknowledge fighting after drinking.
Second, the study was cross-sectional and adults may have difficulty remembering when they first started drinking. Longitudinal studies are needed that ask adolescents when they first began to drink and then follow them into adulthood. Studies not based solely on self-report, but official records of fights or assaults involving alcohol would be useful.
Third, the study did not ask respondents about physical fights not involving alcohol. Nor was information collected about who initiated the fight, whether it resulted in injury and arrest and whether the other person(s) involved had been drinking.
Fourth, people who drink may often frequent places where fights are more apt to occur (eg, bars). Whether there is a direct causal relation between alcohol use and violence is still a subject of investigation.
Research is needed to examine why even when personal history of alcohol dependence and measures of lifetime and past year heavy drinking are controlled, persons who begin drinking at an early age are more likely to get into fights after drinking. Several explanations are possible. Drinking and risk-taking behaviors covary,26 and those who begin drinking at early ages may be even greater risk takers in general. Studies show those who begin drinking at an earlier age are more likely to drive after drinking3 and place themselves in situations that increase their risk of unintentional injury.2 Another explanation is that those who begin drinking earlier, because of their longer drinking experiences, may develop strong beliefs that alcohol will make them (and/or others who drink) more aggressive, which may foster violent behavior.7 Several studies have demonstrated these heightened alcohol beliefs are associated with increased episodes of alcohol-related aggression.27,,28 Lastly, peer norms regarding alcohol use are known to be important predictors of adolescent alcohol misuse.29,,30 It may be plausible, therefore, that persons who start drinking at an early age may also associate with others who engage in similar deviant behaviors and, thus, ascribe to social norms encouraging interpersonal violence that persist through their adult ages.
Despite these unresolved questions, our findings underscore the need for physicians' and other health care providers to query their patients about the age they began drinking, and to counsel them about the risks associated with early age of drinking onset such as fighting after drinking.
Treatment interventions to reduce drinking have been found to reduce violence and other harms associated with alcohol abuse. A systematic review of randomized control trials to reduce alcohol dependence and abuse13 reported reductions in aggressive behavior,31 assaults,32 domestic violence,33 and criminal and domestic violence34 associated with post treatment reductions in drinking. A more recent randomized trial evaluated a brief motivational intervention to reduce drinking among injured problem drinkers.14 One year later, the intervention group averaged 3 drinks less per day and experienced a 47% reduction in emergency department, trauma center, and hospital injury admissions. The greatest declines involve intentional injuries and were among mild to moderate drinkers. Similar benefits have been observed in a separate experimental evaluation of adolescents positive for alcohol being treated in an emergency department.15 A brief motivational intervention for older adolescents (mean age 18) produced a significantly lower incidence at 6 month follow-up of alcohol-related injuries, alcohol-related problems with dates, friends, police, parents, at school, and a lower incidence of driving while intoxicated than experienced by those who received standard care. Both intervention and comparison groups experienced significant post treatment declines in drinking.
Furthermore, our study results reinforce the need for policies that reduce adolescent drinking, such as the minimum legal drinking age of 21. That law has been found to reduce drinking, alcohol-related traffic deaths, and deaths from unintentional injuries under the age of 21.15,,16 Some studies,35,,36 but not all,37,,38 have also found that raising the minimum legal drinking age is associated with declines in fighting among the age groups targeted by the law. Community-based programs that use compliance check surveys to assess the extent of sales of alcohol to minors and that increase enforcement to prevent sales can reduce underage drinking18 and alcohol-related traffic crashes and assault injuries.19
The results of this study point not only to a need to further explore the nature of the relation between early drinking onset age and involvement in fights after drinking, they add additional reason to expand clinical, educational, legal, and community interventions to delay drinking onset.
- Received October 30, 2000.
- Accepted February 7, 2001.
Reprint requests to (R.H.) Social and Behavioral Sciences Department, Boston University School of Public Health, 715 Albany St, TW2, Boston, MA 02118. E-mail:
- NLAES =
- National Longitudinal Alcohol Epidemiologic Survey •
- DSM-IV =
- Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition Criteria •
- CI =
- confidence interval
- ↵Hingson R, Heeren T, Levenson S, Jamanka A, Voas R. Age of drinking onset, driving after drinking and involvement in alcohol-related motor vehicle crashes. Accident Analysis and Prevention. In press
- ↵Reiss AJ, Roth JA. Understanding and Preventing Violence. Washington, DC: National Academy Press; 1994
- ↵Greenfield L. An Analysis of National Data on the Prevalence of Alcohol Involvement in Crime. Washington, DC: US Department of Justice Office of Justice Programs; 1998. NCJ 168632
- ↵Roizen J. Epidemiolgic issues in alcohol related violence. In: Glanter M, ed. Recent Developments in Alcoholism. Vol. 13. New York, NY: Plenum Press; 1997:7–40
- ↵Pernanen K. Alcohol in Human Violence. New York, NY: Guilford Press; 1991
- Gustatson R
- ↵Cook PJ, Moore MJ. Economic perspectives on reducing alcohol related violence In: Martin SE, ed. Alcohol and Interpersonal Violence. Rockville, MD: National Institute of Alcohol Abuse and Alcoholism; 1993:193–212. NIH Publ. No. 93-3496
- ↵Lang AR. Alcohol-related violence: psychological perspectives. In: Martin, SE, ed. Alcohol and Interpersonal Violence: Fostering Multidisciplinary Perspectives. Washington, DC: National Institute of Alcohol Abuse and Alcoholism; 1993. NIH Publ. No. 93-3487
- Virkkunen M,
- et al.
- Dinh-Zarr T,
- Diguiseppi C,
- Heitman E,
- Roberts I
- ↵General Accounting Office. Drinking Age Laws: An Evaluation Synthesis of Their Impact on Highway Safety. Washington, DC: General Accounting Office; 1987. GAO PEMD 87-100
- ↵Massey JT, Parsons VL, Tadros W. Design and estimation for the national health interview survey, 1985–1994;In:Vital and Health Statistics Report Series1989 2(10):Hyattsville, MD: National Center for Health Statistics, U. S. Department of Health and Human Services
- ↵Shah BV, Barnwell BG, Bieler GS. SUDAAN User's Manual, Release 7.0. Research Triangle Park, NC: Research Triangle Institute; 1996
- ↵Grant BF, Hasin DS. The Alcohol Use Disorders and Associated Disabilities Interview Schedule. Rockville, MD: National Institute on Alcohol Abuse and Alcoholism; 1992
- ↵American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (DSM IV). Washington, DC: American Psychiatric Association; 1994
- ↵NIAAA. Ninth Special Report to the US Congress on Alcohol and Health. Washington, DC: US Department of Health and Human Services; 1997. NIH Publ. No. 97–4017
- Derman K,
- George W
- Leigh B
- Dielman TE,
- Butchart AT,
- Shope JT,
- Miller M
- Loveland-Cherry CJ,
- Leech S,
- Laetz VS,
- Ielman TE
- Toteva S,
- Mi'anov I
- Davis JE,
- Reynolds NC
- ↵Parker RN. Alcohol and Homicide a Deadly Combination of Two American Traditions. Albany, NY: State University of New York Press; 1995
- Engs RC,
- Hanson DJ
- Copyright © 2001 American Academy of Pediatrics