Validation of brief instruments to measure adult fruit and vegetable consumption.

Four brief food frequency questionnaires were developed and validated to measure an adult's usual daily intake of fruit or vegetable servings over the past month. Fifty males and 50 females, aged 25-52 years, completed two fruit instruments, two vegetable instruments, and a dietary history (the reference method). Individual agreement and group mean estimation were assessed. The 5-item fruit instrument and the 15-item vegetable instrument performed best. At an individual level, the 5-item fruit instrument had 64% sensitivity and 88% specificity for assessing goal attainment (≥ 2 fruit servings/day), while the 15-item vegetable instrument had 67% sensitivity and 82% specificity for measuring goal attainment (≥ 3 vegetable servings/day). At a group level, the 15-item vegetable instrument closely estimated mean intake (ratio of geometric means=0.94), while the 5-item fruit instrument overestimated mean intake by 32%. Nevertheless, when it was used to rank participants, reference method fruit servings increased across the quartiles of intake. These two instruments can be used to assess goal attainment. This vegetable instrument can also be used to estimate group mean intakes, while the fruit instrument can be used to rank participants. Used appropriately, these instruments can be used for screening, monitoring, and evaluation purposes in New Zealand public health and clinical settings.