Performance of a short percentage energy from fat tool in measuring change in dietary intervention studies

Measurement of percentage energy from fat is important in surveillance of populations and in epidemiologic studies examining relationships between diet and disease as well as for behavioral intervention studies seeking to change dietary behavior. The NCI percentage energy from fat screener (PFat) has adequately predicted percentage of energy from fat compared with 24-h recalls (24HR) in cross-sectional analyses. However, the instrument has not been evaluated for its ability to assess change of percentage energy from fat over time or in response to interventions to change dietary intake of fat. The objective of this analysis is to evaluate the performance of the PFat in assessing change in percentage energy intake from fat in a behavioral intervention setting. Four individual sites participating in the Behavior Change Consortium Nutrition Working Group administered both the PF at and multiple 24HR at baseline and follow-up to 278 participants. A measurement error model was used to assess agreement between the PFat and 24HR at baseline and follow-up. The PFat was consistent with 24HR in finding there was no significant change in percentage energy from fat as a result of the intervention. Both male and female participants in the intervention group demonstrated a significant increase in the correlation between PFat and 24HR from baseline to follow-up. Percentage energy from fat measured by PFat may be useful to provide estimates of change in mean intake of populations over time in longitudinal studies. Further methodologic research is called for in interventions producing significant changes and in diverse populations with adequate sample size.