Are the proposed limits of energy intake:basal metabolic rate and dietary nitrogen:urinary nitrogen ratios suitable for validation of food intake?

Abstract

The validity of 7 d weighed records of diet obtained for pre-menopausal Mexican women was assessed by two independent methods: the energy intake:BMR (EI:BMR) and the dietary N:urinary N (DN:UN). For the latter, complete urine collections are required and completeness was assessed from measurements of para-aminobenzoic acid (PABA) excretion. There were forty-six adult female subjects in the study, thirty-four were from Mexico City and twelve were from a rural population in the Central Highlands, Mexico. However, data were rejected from five urban women for whom the PABA excretion data suggested incomplete urine collection on four or more days. BMR was measured with Oxylog portable O 2 consumption meters, and physical activity level was assessed from a self-completed activity diary. An approximate relationship between the EI:BMR ratio and the DN:UN ratio suggested that the rejection limits on the EI:BMR ratio recommended by Goldberg et al. (1991) are wider than the limits on the DN:UN ratio recommended by Bingham & Cummings (1985). Using the recommended cut-off points for EI:BMR but wider limits for DN:UN, twenty-one and twenty-five women respectively had acceptable intake records by the two methods, and sixteen of them by both methods. In conclusion the modification of the DN:UN limits to 0´92 and 1´70 to set acceptable intake values makes the use of measurements of N and energy balance comparable. Urine values with PABA recoveries greater than 100^15 % should be rejected, as should UN values validated by less than 3 d. Validation: Food intake: MethodologyAn ongoing problem in nutritional research is the uncertain validity of methods for assessing habitual food intake. The 7 d weighed diet record method is often the reference method, as it uses prospectively collected data with precise information on quantity and food type (Bingham et al. 1988). To validate it in free-living individuals is difficult, because it relies on information supplied by the subjects. Validation against some external criterion should therefore be built into the protocol of any planned dietary surveys (Bingham & Cummings, 1985). Isaksson (1980) proposed the use of 24 h urinary N excretion (UN) to validate protein intake; cut-off limits for the UN:dietary N (DN) ratio of 0´7±0´9 were proposed by Bingham & Cummings (1985). Completeness of urine collections may be assessed reliably by the para-aminobenzoic acid (PABA) method (Bingham et al. 1983). Doubly-labelled water ( 18 O and 2 H) measurements of energy expenditure (EE) have been used to assess bias in estimates of energy intake (EI) as measured by the 7 d weighed diet record procedure. These studies have highlighted the problem of under-reporting. Goldberg et al. (1991) proposed validating intakes by detecting unlikely values of the EI:BMR ratio, which we call intake physical activity level (IPAL), assuming a common value for physical activity level (PAL) of 1´55 Â BMRX Using this approach, Black et al. (1991) found that 64 % of published studies using dietary intake records fell below ac...

Under- and overreporting of energy is related to obesity, lifestyle factors and food group intakes in Jamaican adults

“…The recommended equations for predicting BMR 24 may overestimate BMR slightly in tropical countries 45,46 . However, these equations yielded estimates similar to measured BMR in Mexican women with comparable energy intakes (7.3 MJ urban, 9.24 MJ rural) and moderate physical activity levels (1.65 £ BMR in urban and 1.90 £ BMR in rural women) 47 . Furthermore, implausible reporting estimated as intakes within^30% of an independent prediction equation for total energy requirements 27 was highly correlated (Spearman's r ¼ 0.89) with estimates using this method.…”

Underreporting of Energy Intake in Developing Nations

“…Sánchez-Castillo et al 52 used urinary nitrogen and the EI/BMR ratio to evaluate the validity of a 7-day food record reported by 34 women living in Mexico City and 12 women living in a rural Mexican community. BMR was determined by indirect calorimetry, and the cutoff for underreporting was established as EI/BMR Ͻ 1.4.…”