Alcohol use among Inuit pregnant women: Validity of alcohol ascertainment measures over time

Background: Frequency and quantity of alcohol consumed by women are two important indicators of the risks associated with drinking during pregnancy. Some studies have compared the validity of maternal alcohol report obtained during and after pregnancy. However, to date none have examined alcohol use in a Native Canadian population, such as the Inuit. Effective measurement methods are necessary to better understand why children from some communities seem at increased risk of alcohol-related neurodevelopmental disorders. Methods: Prospective and retrospective drinking interviews were obtained from a sub-sample of 67 women included in the Nunavik Child Development Study (NCDS), Nunavik, Canada (1995-2010; N = 248). Number of days of alcohol consumption and binge drinking (five drinks or more per episode) across pregnancy as well as ounces of absolute alcohol per day and per drinking day among users were collected using timeline follow-back interviews administered both during pregnancy and again 11 years after delivery. Consistency of alcohol reports over time, as well as significant differences for alcohol quantities described by users between interviews were examined. Sociodemographic characteristics associated with alcohol use reports were also assessed. Results: The proportion of positive reports of alcohol and binge drinking during pregnancy was higher when women were interviewed prospectively during pregnancy than retrospectively. We observed a fair to moderate agreement of alcohol report between interview periods. By contrast, the number of binge drinking days during pregnancy was slightly higher among alcohol users when documented retrospectively. Conclusions: Our findings endorse the conclusion that prospective alcohol measures provide more reliable ascertainment and likely generate more valid information about the proportion of children prenatally exposed to alcohol in the Inuit population.