Drug abuse is an important risk factor for the human immunodeficiency virus (HIV) among Hispanics living in the northeastern United States, and both drug abuse and HIV are associated with nutritional deficiencies. The selection of a dietary assessment method most appropriate for Hispanic adults with/without HIV infection who may be drug abusers remains unclear.

Participants were recruited into one of the three groups: HIV-infected drug abusers, HIV-noninfected drug abusers, and HIV-infected non–drug abusers. Subjects who completed two of the three dietary methods were included in cross-sectional pairwise comparisons.

The baseline data from a prospective cohort study of the role of drug abuse in HIV/acquired immunodeficiency syndrome-related weight loss.

The 286 enrolled participants completed 282 food frequency questionnaires (FFQs), 142 3-day diet records, and 270 24-hour recalls. Energy-adjusted and deattenuated correlations between the FFQ and 3-day diet records ranged from 0.11 (carbohydrate) to 0.75 (caffeine). Twenty-seven of 33 nutrient intakes estimated by 3-day diet record were significantly lower than by FFQ (P<0.05). Three-day diet records underestimated dietary intake relative to the FFQ and 24-hour recall methods. Fifty percent of records were not completed. Energy estimates from the FFQ and 24-hour recall were similar to each other.

The 24-hour recall gave higher mean dietary intake estimates, but would require multiple contacts with this difficult-to-reach population. The FFQ that was specifically designed for this Hispanic population performed well relative to the other methods.