A modified Timeline Followback assessment to capture alcohol exposure in pregnant women: Application in the Safe Passage Study

Abstract

Prenatal alcohol exposure (PAE) has been linked to poor pregnancy outcomes, yet there is no recognized standard for PAE assessment, and the specific effects of quantity, frequency, and timing remain largely unknown. The Safe Passage Study was designed to investigate the role of PAE in a continuum of poor peri- and postnatal outcomes. The objective of this manuscript is to describe the rationale for, and feasibility of, modifications to the traditional Timeline Followback (TLFB) for collecting PAE information in a large cohort of pregnant women. Participants from the Northern Plains region (in the United States) and Cape Town, South Africa, were followed prospectively using a modified 30-day TLFB interview, administered up to five times, to obtain detailed PAE information. Required modifications for our population included capturing information regarding sharing, type/brand, container size, and duration, in order to accurately record the amount of alcohol consumed. PAE status was defined for 99.9% of the 11,892 enrolled pregnancies at least once during pregnancy and for 92% across all trimesters. Of 53,823 drinks reported, 98% had all items necessary for standard drink computation. Sharing was reported for 74% of drinks in Cape Town, South Africa and for 10% in the Northern Plains. Compared to referent values from the traditional TLFB, 74% and 67% of drinks had different alcohol-by-volume and container size, respectively. Furthermore, a statistically significant difference was found between the number of containers reported and the number of standard drinks computed, using information from the modified TLFB. This is the first study of this size to wholly encompass all of these changes into a single measure in order to more accurately calculate daily consumption and assess patterns over time. The methods used to collect PAE information and create alcohol exposure measures likely increased the accuracy of standard drinks reported and could be generalized to other populations.

Keywords: Prenatal alcohol exposure; Quantity-frequency methods; Self-report; Standard drink; Timeline followback (TLFB).

Fig. 1

Excerpt from the Timeline Followback (TLFB) case report form used in the Safe Passage Study to collect drink type, size of container, number of containers, number sharing, ice or frozen, and duration for each unique drinking day in the 30-day reference period.

Fig. 2

The array of self-reported alcohol consumption measures collected on the modified Timeline Followback (TLFB). *Countries have differing definitions for what is considered a standard drink. While the US guideline (14 g) was used to define a standard drink in this report, the detail elicited from the modified TLFB allows for derivation based on other national standards. BAC: blood alcohol content; LMP: last menstrual period

Fig. 3

A boxplot of spirit ABV (%) by site representing the minimum, first quartile, median, third quartile, and maximum. NHigh and NLow refer to the number of drink reports clipped from the visual plot where ABV >60% or <20%, respectively. ABV: Alcohol by Volume