Background: In diabetes care, an accurate assessment of dietary intake is essential for clinical management. This study assessed the validity of dietary records from people with NIDDM by comparing reported energy and protein intakes with estimated total energy expenditure (TEE)
and urinary nitrogen excretion. Method: To assess energy and nutrient intake 185 patients kept dietary records for two 3‐day periods. Basal metabolic rate (BMRest) was estimated to calculate reported energy intake/BMRest. Appropriate age and sex‐specific
physical activity levels (PAL=TEE/BMR) from published doubly‐labelled water studies of TEE, were used to calculate a cut‐off PAL for energy intake/BMRest, below which intakes were statistically unlikely to be valid. A subgroup (46 patients) collected a complete 24‐h
urine sample to compare nitrogen excretion with protein intake. Results: The mean value of energy intake/BMRest (1.14±0.3) was markedly below the calculated cut‐off point (1.58±0.06) suggesting widespread underreporting. Protein intake (82.5±21.3)
was also lower than estimated protein losses (92.5±19.4). Obese and female NIDDS were more likely to underreport and their intakes were closer to the diabetes dietary recommendations. Conclusion: This study found that people with NIDDM were likely to underreport their food intake.