Importance of cohorts in analyzing trends in safe at-home food-handling practices

Abstract

Safe in-home food preparation is the last line of defense for preventing foodborne illness. The Food Safety Survey assessing consumers' food handling behavior has been conducted every 3–5 years (1993, 1998, 2001, 2006, 2010) using a random digit telephone sample of United States adult consumers. Sample sizes ranged from 1620 to 4547. A previous analysis of this data has examined trends in safe food handling (as measured by washing hands and/or cutting boards after touching/cutting raw meat or chicken and by washing hands after cracking eggs). We continue and expand this analysis by modeling the unique effects of age, survey period (year) and birth cohort on safe food handling. We find that age, period, and cohort effects are relevant in measuring changes in food handling behavior; however, the effects are not similar in size or apparent mediating process. The strongest effect is period, followed by age and cohort. Thus it appears contemporaneous changes in information activity can make relatively large short-run improvements, whereas changes in one's maturation and accumulated experience have quadratic effects, and the unique shared experience of cohort leaves its own definite long-lasting imprint. We propose that the birth cohort effects can be explained by the food safety environment during young adulthood. Those who were young adults in two critical time periods – before 1940 when there were widespread foodborne infections and immediately after the 1993 outbreak of Escherichia coli O157:H7 -- have better food handling behaviors.