|Level of evidence for the association between PCR as protective and as risk factor and types of alcohol behavior|
|Parent–child relationship||Alcohol use||Consistency of findingsc(no of findings in high quality studies)||Level of evidence|
|Protectivea||Initiation of (monthly) use||6 s (0), 16 ns (1)||Inconclusive|
|Frequency and/or amount of use||7 s (2), 21 ns (11)||Weak|
|(Initiation of) heavy drinking||4 s (0), 5 ns (2)||Inconclusive|
|Otherd||0 s (0), 5 ns (2)||Inconclusive|
|Total||17 (2), 47 (16)||Weak|
|Riskb||Initiation of (monthly) use||0 s (0), 2 ns (2)||Inconclusive|
|Frequency and/or amount of use||1 s (0), 6 ns (1)||Inconclusive|
|(Initiation of) heavy drinking||0 s (0), 2 ns (0)||Inconclusive|
|Otherd||0 s (0), 0 s (0)||Inconclusive|
|Total||1 s (0), 10 ns (3)||Inconclusive|
Abbreviation: s=significant, ns=not significant.
a Positive PCR (e.g., warmth, support).
b Negative PCR (e.g., conflict, rejection).
c Some studies showed more than one association and therefore total number of findings is higher than total number of studies.
d Types of alcohol behavior that could not be classified in any of the other categories (e.g., a composite score of frequency of use, frequency of heavy use, and frequency of drunkenness).
Visser et al.
Visser et al. BMC Public Health 2012 12:886 doi:10.1186/1471-2458-12-886