Empirical evidence of correlated biases in dietary assessment instruments and its implications. American Journal of Epidemiology, 153, 394-403

Biometry Research Group, Division of Cancer Prevention, National Cancer Institute, Bethesda, MD 20892-7354, USA.
American Journal of Epidemiology (Impact Factor: 5.23). 03/2001; 153(4):394-403. DOI: 10.1093/aje/153.4.394
Source: PubMed


Multiple-day food records or 24-hour recalls are currently used as "reference" instruments to calibrate food frequency questionnaires (FFQs) and to adjust findings from nutritional epidemiologic studies for measurement error. The common adjustment is based on the critical requirements that errors in the reference instrument be independent of those in the FFQ and of true intake. When data on urinary nitrogen level, a valid reference biomarker for nitrogen intake, are used, evidence suggests that a dietary report reference instrument does not meet these requirements. In this paper, the authors introduce a new model that includes, for both the FFQ and the dietary report reference instrument, group-specific biases related to true intake and correlated person-specific biases. Data were obtained from a dietary assessment validation study carried out among 160 women at the Dunn Clinical Nutrition Center, Cambridge, United Kingdom, in 1988-1990. Using the biomarker measurements and dietary report measurements from this study, the authors compare the new model with alternative measurement error models proposed in the literature and demonstrate that it provides the best fit to the data. The new model suggests that, for these data, measurement error in the FFQ could lead to a 51% greater attenuation of true nutrient effect and the need for a 2.3 times larger study than would be estimated by the standard approach. The implications of the results for the ability of FFQ-based epidemiologic studies to detect important diet-disease associations are discussed.

3 Reads
  • Source
    • ") between urinary N and protein intake (intake ranging from 51 to 212 g/day) was found to be 0.91 (Magee et al., 2004). In a meta-analysis of data by Kipnis et al. (2001) it was found that urinary N is 80% of dietary intake on average. Of the nitrogenous fractions urea is the most predominant, making up between 75% and 90% (Lentner, 1981). "
    [Show abstract] [Hide abstract]
    ABSTRACT: The safe disposal of human excreta is of paramount importance for the health and welfare of populations living in low income countries as well as the prevention of pollution to the surrounding environment. On site sanitation (OSS) systems are the most numerous means of treating excreta in low income countries, these facilities aim at treating human waste at source and can provide a hygienic and affordable method of waste disposal. However, current OSS systems need improvement and require further research and development. Development of OSS facilities that treat excreta at, or close to, its source require knowledge of the waste stream entering the system. Data regarding the generation rate and the chemical and physical composition of fresh faeces and urine was collected from the medical literature as well as the treatability sector. The data were summarised and statistical analysis was used to quantify the major factors that were a significant cause of variability. The impact of this data on biological processes, thermal processes, physical separators and chemical processes was then assessed. Results showed that the median faecal wet mass production was 128 g/cap/day, with a median dry mass of 29 g/cap/day. Faecal output in healthy individuals was 1.20 defecations per 24 hour period and the main factor affecting faecal mass was the fibre intake of the population. Faecal wet mass values were increased by a factor of 2 in low income countries (high fibre intakes) in comparison to values found in high income countries (low fibre intakes). Faeces had a median pH of 6.64 and were composed of 74.6% water. Bacterial biomass is the major component (25-54% of dry solids) of the organic fraction of the faeces. Undigested carbohydrate, fibre, protein and fat comprise the remainder and the amounts depend on diet and diarrhoea prevalence in the population. The inorganic component of the faeces is primarily undigested dietary elements that also depend on dietary supply. Median urine generation rates were 1.42 litres/cap/day with a dry solids content of 59 g/cap/day. Variation in the volume and composition of urine is caused by differences in physical exertion, environmental conditions as well as water, salt and high protein intakes. Urine has a pH 6.2 and contains the largest fractions of nitrogen, phosphorus and potassium released from the body. The urinary excretion of nitrogen was significant (10.98 g/cap/day) with urea the most predominant constituent making up over 50% of total organic solids. The dietary intake of food and fluid is the major cause of variation in both the faecal and urine composition and these variables should always be considered if the generation rate, physical and chemical composition of faeces and urine is to be accurately predicted.
    Critical Reviews in Environmental Science and Technology 02/2015; 45(17):00-00. DOI:10.1080/10643389.2014.1000761 · 3.47 Impact Factor
  • Source
    • "It is, therefore, important to note that the impact of the calibration method in our study was such that the risk estimates from the calibrated data are usually stronger than the non-calibrated results. Nevertheless, measurement error may still have an effect on calibrated RRs to a certain extent because the error structure in the reference method is not entirely independent of that in the FFQ [34,35]. A further methodological strength of the EPIC cohort is the inclusion of individuals from 10 European countries with distinctly diverging meat consumption habits [36]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Recently, some US cohorts have shown a moderate association between red and processed meat consumption and mortality supporting the results of previous studies among vegetarians. The aim of this study was to examine the association of red meat, processed meat, and poultry consumption with the risk of early death in the European Prospective Investigation into Cancer and Nutrition (EPIC). Included in the analysis were 448,568 men and women without prevalent cancer, stroke, or myocardial infarction, and with complete information on diet, smoking, physical activity and body mass index, who were between 35 and 69 years old at baseline. Cox proportional hazards regression was used to examine the association of meat consumption with all-cause and cause-specific mortality. As of June 2009, 26,344 deaths were observed. After multivariate adjustment, a high consumption of red meat was related to higher all-cause mortality (hazard ratio (HR) = 1.14, 95% confidence interval (CI) 1.01 to 1.28, 160+ versus 10 to 19.9 g/day), and the association was stronger for processed meat (HR = 1.44, 95% CI 1.24 to 1.66, 160+ versus 10 to 19.9 g/day). After correction for measurement error, higher all-cause mortality remained significant only for processed meat (HR = 1.18, 95% CI 1.11 to 1.25, per 50 g/d). We estimated that 3.3% (95% CI 1.5% to 5.0%) of deaths could be prevented if all participants had a processed meat consumption of less than 20 g/day. Significant associations with processed meat intake were observed for cardiovascular diseases, cancer, and 'other causes of death'. The consumption of poultry was not related to all-cause mortality. The results of our analysis support a moderate positive association between processed meat consumption and mortality, in particular due to cardiovascular diseases, but also to cancer.
    BMC Medicine 03/2013; 11(1):63. DOI:10.1186/1741-7015-11-63 · 7.25 Impact Factor
  • Source
    • "It has been shown that frequency of intake explains most of the variations in intake. While the FFQ may be an adequate, albeit imperfect, instrument for measuring relative fruit and vegetable intake, the nondifferential misclassification errors observed in the FFQ likely attenuate the estimates toward the null [25]. Consequently, there is another possibility that any true ORs are likely greater than as shown in this analysis. "
    [Show abstract] [Hide abstract]
    ABSTRACT: We investigated the association between dietary intake of citrus fruits and breast cancer risk. The PubMed and EMBASE were searched for relevant articles on diet and breast cancer up to January 2012. All of the epidemiological studies that assessed dietary intake of citrus fruits and presented risk estimates of the association between citrus fruits intake and risk of breast cancer were reviewed. Multivariable-adjusted odds ratios (OR) and associated 95% confidence intervals (CI) for highest versus lowest intake of dietary citrus fruits level were extracted. Overall summary OR was calculated by using a fixed-effect meta-analysis. Six case-control studies out of five articles were eligible. Overall summary OR showed a 10% reduction in risk of breast cancer associated with high intake of citrus fruits (summary OR, 0.90; 95% CI, 0.85-0.96; p<0.001); results were consistent across the studies (I (2)=0). Visual inspection of the results did not suggest a publication bias. Pooled results from observational studies showed an inverse association between citrus fruits intake and the risk of breast cancer.
    Journal of Breast Cancer 03/2013; 16(1):72-76. DOI:10.4048/jbc.2013.16.1.72 · 1.58 Impact Factor
Show more