Validity of two new brief instruments to estimate vegetable intake in adults

Abstract

Cost effective population-based monitoring tools are needed for nutritional surveillance and interventions. The aim was to evaluate the relative validity of two new brief instruments (three item: VEG3 and five item: VEG5) for estimating usual total vegetable intake in comparison to a 7-day dietary record (7DDR). Sixty-four Australian adult volunteers aged 30 to 69 years (30 males, mean age ± SD 56.3 ± 9.2 years and 34 female mean age ± SD 55.3 ± 10.0 years). Pearson correlations between 7DDR and VEG3 and VEG5 were modest, at 0.50 and 0.56, respectively. VEG3 significantly (p < 0.001) underestimated mean vegetable intake compared to 7DDR measures (2.9 ± 1.3 vs. 3.6 ± 1.6 serves/day, respectively), whereas mean vegetable intake assessed by VEG5 did not differ from 7DDR measures (3.3 ± 1.5 vs. 3.6 ± 1.6 serves/day). VEG5 was also able to correctly identify 95%, 88% and 75% of those subjects not consuming five, four and three serves/day of vegetables according to their 7DDR classification. VEG5, but not VEG3, can estimate usual total vegetable intake of population groups and had superior performance to VEG3 in identifying those not meeting different levels of vegetable intake. VEG5, a brief instrument, shows measurement characteristics useful for population-based monitoring and intervention targeting.