Transcript Psykologisen mittarin tilastollinen analysointi

ALCOHOL USE AND AGGRESSION IN
ADOLESCENTS - A META-ANALYSIS
Jouko Miettunen, Adjunct Professor
Department of Psychiatry
University of Oulu, Finland
e-mail: [email protected]
Background




Alcohol use and aggressive (violent)
behaviour have been found to
associate with each other in
adolescent samples.
The strength of this association is
largely unknown
Epidemiological association or
intoxication effect?
Alcohol use risk factor also for
vulnerability for physical violence
Background
Adolescence important period




Onset of substance use
Onset of criminal behaviour
Early alcohol use  early alcoholism?
Early behavioural problems  later
criminality?
Data collection

The literature search for studies was done in a
systematic way. The studies were searched
one database at a time, in abstract and title

Originally time limit (1995-), but now also
older studies included (ongoing study!)

Databases searched included PsycINFO,
Science Direct, PubMed and Google Scholar.
The search was done in October and
November 2008

Also manual search was done and authors
were contacted to get more information
Data collection

The search was limited to studies on the
relation of aggression and alcohol use in the
age group of “normal” adolescents between
age 13 and 18

Studies were included if they measured
aggression and alcohol use in a standardized
and most similar way, at the same point in
time, and in an adolescent cross-sectional
sample.

Results are presented with meta-analytic
methods using pooled odds ratios (OR).
Preliminary results

Eight studies compared proportion of
aggressive or violent persons between
alcohol using and non-using
adolescents
 Included

to the current meta-analysis!
Many other studies compared alcohol
using and non-using adolescents using
various other summary statistics
(correlations, means, …)
 These
studies are not presented here!
Included studies

Choquet et al. Alcohol Alcohol 1991;26:381-90.

Orpinas et al. J Adolesc Health 1995;16:216-25.

Fergusson et al. Addiction 1996;91:483-94.

Ellickson et al. Am J Public Health 1997;87:985-91.

Gudlaugsdottir et al. Int J Epidemiol 2004;33:1046-51.

Eklund & af Klinteberg. J Indiv Dif 2005;26:63-73.

Wade & Pevalin. Can J Criminol Crim Justice 2005;47:619-49.

Shepherd et al. J Adolesc 2006;29:539-53.
Choquet et al. Interpersonal aggressive behaviour and alcohol
consumption among young urban adolescents in France. Alcohol &
Alcoholism 1991;26:381-90
•An epidemiological survey was carried out among 1601
adolescents (aged 13-16) from a suburban area of Paris.
•A majority of the adolescents had already demonstrated violent
behaviour, 41% showed repetitive delinquent behaviour, and
12% cumulated several types of such behaviour.
•The aggressive teenagers differed from the non-aggressive
subjects firstly in their alcohol, tobacco and illicit drug
consumption, and secondly with respect to other deviant
behaviour, such as stealing, running away from home or violent
victimization.
Orpinas et al. The co-morbidity of violence-related behaviors with
health-risk behaviors in a population of high school students. J
Adolesc Health 1995;16:216-25.
•All ninth and eleventh graders (n = 2075) of a school
district in Texas.
•Overall, 20% of the students were involved in a
physical fight but had not carried a weapon, 10%
carried a weapon but had not been involved in a
physical fight, and 17% had been involved in a
physical fight and had carried a weapon.
•Students who both fought and carried a weapon
were 19 times more likely to drink alcohol six or
more days than students who did not fight nor carried
a weapon.
Fergusson et al. Alcohol misuse and juvenile offending in
adolescence. Addiction 1996;91:483-94.
•The associations between alcohol misuse and juvenile
offending during the period from 15 to 16 years of age were
studied in a birth cohort of New Zealand adolescents.
•This analysis showed that young people who misused
alcohol had significantly (p <0.001) higher rates of both
violent and property offences.
•A substantial component of the association between alcohol
misuse and juvenile offending arose from shared risk factors
that were common to both outcomes.
•However, young people who abused alcohol had odds of
violent offending that were 3.2 times (p< 0.001) the odds of
those offences for young people who did not misuse alcohol.
Ellickson et al. Profiles of violent youth: substance use and other
concurrent problems. Am J Public Health 1997;87:985-91.
•Longitudinal data for more than 4500 high school
seniors and dropouts from California and Oregon
•More than half the sample had engaged in violence
during the last year, and one in four had committed
predatory violence.
•Violent youth were more likely than their peers to
have poor mental health, use drugs, drop out of
school, and be delinquent.
Gudlaugsdottir et al.Violent behaviour among adolescents in
Iceland: a national survey. Int J Epidemiol 2004;33:1046-51.
•In 1997, a cross-sectional survey was conducted among a
random half of all Icelandic schoolchildren aged 15-16
years (N = 3872).
•The majority of the respondents reported having
committed violence within the last year.
•Smokers were more likely than non-smokers to use
violence (OR = 1.7; 95% CI: 1.2, 2.2)
•Adolescents who had used alcohol >20x in their lifetime
were more than twice as likely to commit violence
compared with those who had never used alcohol (OR =
2.5; 95% CI: 1.8, 3.4).
Eklund & af Klinteberg. Personality Characteristics as Risk
Indications of Alcohol Use and Violent Behavior in Male and
Female Adolescents. J Indiv Dif 2005;26:63–73.
•The present study focused on personality characteristics in
adolescent boys (n = 414) and girls (n = 552) in 8th grade
with self-reported violent behavior and risky alcohol use.
•Adolescents with indications of violent behavior and/or
risky alcohol use, compared to others, were generally more
impulsive, had a stronger need for change and action, were
less adjusted and socially conforming, as well as more
aggressive.
•The findings pointed toward a clustering of problem
behaviors.
• Similar personality profiles between alcohol using and
violent adolescents.
Wade & Pevalin. Adolescent delinquency and health. Can J
Criminol Crim Justice 2005;47:619-49.
•Data from the first two waves of the publicly
available US National Longitudinal Study of
Adolescent Health (n=4,834).
•Results indicate that delinquent behaviours such as
violence, aggression and property damage are
similarly predicted by the same risk factors as are
depression and perceived health, and tobacco,
alcohol, marijuana and hard drug use.
•Most outcomes were associated with one another,
suggesting they may be comorbid manifestations of
risk exposure.
Shepherd et al. Relations between alcohol, violence and
victimization in adolescence. J Adolesc 2006;29:539-53.
•Cross-sectional study of 4187 adolescents aged 11-16 in a
stratified sample of 13 English schools.
•Relationships between fighting, hitting others and
vulnerability to being hit and frequency of drinking and
drunkenness were all highly significant (p<0.0001), and
were evident at all ages.
•The outcome most strongly related to frequency of
drunkenness was hitting others (odds ratio (OR) 6.62),
followed by being hit (OR 4.01) and fighting (OR 2.10).
Shepherd et al. Relations between alcohol, violence and
victimization in adolescence. J Adolesc 2006;29:539-53.
Forest plot with odds ratios
Study
Odds ratio (95% CI)
USA Ellickson et al. 1997
2.1 (1.7 - 2.6)
USA Wade and Pevalin 2005
2.3 (2.0 - 2.6)
Iceland Gudlaugsdottir et al. 2004
2.8 (2.5 - 3.3)
USA Orpinas et al. 1995
4.2 (3.3 - 5.3)
France Choquet et al. 1991
New Zealand Fergusson et al. 1996
5.9 (3.4 - 10.2)
UK Shepherd et al. 2006
7.1 (5.2 - 9.7)
Sweden Eklund and af Klinteberg 2005
7.4 (4.9 - 11.3)
Overall (95% CI)
4.07 (2.97 - 5.58)
6.5 (3.7 - 11.7)
1
1.5
2
3
4
5
Odds ratio
6 7 8
Summary of results

Alcohol using adolescents
had four times more
commonly aggressive
behaviour

Results were consistent
Odds
ratios between 2 and 7
Methodological issues

Self-reported data

Differences in assessment methods
 Alcohol
and behaviour/violence

Attrition in original studies

Differences by age and sex

Similar risk factors (covariates?)

Other substances?

Dose response? Causality?
Conclusions



This study was the first one to pool
studies on association between
alcohol use and aggression in
adolescence.
The results indicate a strong
association between alcohol use and
aggressive behaviour.
Odds ratios (cross-sectional
associations) were quite similar
between studies
Conclusions


Although these results are not a proof for
causal relationship between alcohol use
and aggression, the association is strong
and should be paid attention to both in
research and clinical settings (health and
violence-prevention programs)
Temporal associations between alcohol
(and other substance) use and aggressive
behaviour needs to be further studied
Research Group
Bettina von der Pahlen, Professor
National Institute for Health and Welfare, Helsinki, Finland
Marina Lindholm, Student
Department of Psychology, Abo Akademi University, Turku, Finland
Mauri Marttunen, Professor
National Institute for Health and Welfare, Helsinki, Finland
Jouko Miettunen, Academy Research Fellow
Department of Psychiatry, University of Oulu, Oulu, Finland