Reliability of a Timeline Followback measure of caffeinated coffee consumption

Abstract

Background and aims: The Timeline Followback (TLFB) was initially developed to collect retrospective self- reports of alcohol and drug use. Since its development, several peer-reviewed papers have supported it as a sound psychometric measure for substance use and for several other behaviors. Worldwide, coffee is one of the most widely consumed beverages. Although early epidemiological research suggested that drinking coffee is associated with some health problems, several recent studies have found moderate coffee consumption to have an inverse association with mortality. Because of its widespread use, a psychometrically sound measure of coffee consumption would help better inform research and public health policies.

Design: This study investigated the test-retest reliability of a version of the TLFB modified to assess coffee consumption.

Setting and participants: 98 clinical psychology doctoral students completed a 30-day coffee TLFB on two occasions separated by 14 days.

Measurements: Participants used a computerized Survey Monkey™ version of the coffee TLFB to report their caffeinated coffee consumption for the same 30-day interval. For the first administration participants reported their consumption for the preceding 30 days, and for the second administration they reported on the preceding 45 days which included the earlier reporting interval.

Findings: Reliability coefficients were found to be very high (0.90-0.97).

Conclusions: As with other substance use versions of the TLFB, these results suggest the TLFB is a good method for measuring coffee consumption in research studies and in primary health care settings where such data may be important.

Keywords: Caffeine; Calendar; Coffee; Coffee assessment measure; Reliability; Retrospective reports; Timeline Followback.