PT Journal
AU Torvik, FA
   Rognmo, K
   Tambs, K
TI Alcohol use and mental distress as predictors of non-response in a general population health survey: the HUNT study
SO Social Psychiatry and Psychiatric Epidemiology
JI Soc Psychiatry Psychiatr Epidemiol
PY 2012
BP 805
EP 816
VL 47
IS 5
DI 10.1007/s00127-011-0387-3
LA English
DE Adult; Age Distribution; Aged; 80 and over; Alcohol-Related Disorders/epidemiology/*psychology; Depressive Disorder/epidemiology/*psychology; Female; Health Status Indicators; Health Surveys; Humans; Male; Middle Aged; Norway/epidemiology; Odds Ratio; Patient Participation/*psychology/statistics & numerical data; Population Surveillance; Questionnaires; Severity of Illness Index; Spouses/psychology
AB PURPOSE: To investigate to what degree alcohol use and mental distress are associated with non-response in a population-based health study. METHODS: From 1995 to 1997, 91,488 persons were invited to take part in a health study at Nord-Trondelag, Norway, and the response rate was 69.2%. Demographics were available for everyone. Survey answers from a previous survey were available for most of the participants and a majority of non-participants. In addition, the survey responses from spouses and children of the invitees were used to predict participation in the aforementioned study. Crude and adjusted ORs for a number of predictors, among these alcohol consumption and mental distress, are reported. RESULTS: Both heavy drinkers (OR = 1.27) and abstainers (OR = 1.64) had a higher probability of dropping out in comparison to people who usually do not drink. High levels of mental distress (OR = 1.84) also predicted attrition. CONCLUSION: Alcohol use and mental distress are moderately associated with non-response, though probably not a major cause, as controlling for other variables weakened the associations. Nevertheless, the moderate but clear underrepresentation at the crude level of people with high alcohol consumption, abstainers and people with poor mental health should be taken into consideration when interpreting results from health surveys.
ER