Using surrogate biomarkers to improve measurement error models in nutritional epidemiology

Reliance on self-reports and estimated food composition data in nutrition research introduces significant bias that can only be addressed with biomarkers

The chemical composition of foods is complex, variable, and dependent on many factors. This has a major impact on nutrition research as it foundationally affects our ability to adequately assess the actual intake of nutrients and other compounds. In spite of this, accurate data on nutrient intake are key for investigating the associations and causal relationships between intake, health, and disease risk at the service of developing evidence-based dietary guidance that enables improvements in population health. Here, we exemplify the importance of this challenge by investigating the impact of food content variability on nutrition research using three bioactives as model: flavan-3-ols, (-)-epicatechin, and nitrate. Our results show that common approaches aimed at addressing the high compositional variability of even the same foods impede the accurate assessment of nutrient intake generally. This suggests that the results of many nutrition studies using food composition data are potentially unreliable and carry greater limitations than commonly appreciated, consequently resulting in dietary recommendations with significant limitations and unreliable impact on public health. Thus, current challenges related to nutrient intake assessments need to be addressed and mitigated by the development of improved dietary assessment methods involving the use of nutritional biomarkers.

Logistic regression with correlated measurement error and misclassification in covariates

Many areas of research, such as nutritional epidemiology, may encounter measurement errors of continuous covariates and misclassification of categorical variables when modeling. It is well known that ignoring measurement errors or misclassification can lead to biased results. But most research has focused on solving these two problems separately. Addressing both measurement error and misclassification simultaneously in a single analysis is less actively studied. In this article, we propose a new correction method for a logistic regression to handle correlated error variables involved in multivariate continuous covariates and misclassification in a categorical variable simultaneously. It is not computationally intensive since a closed-form of the approximate likelihood function conditional on observed covariates is derived. The asymptotic normality of this proposed estimator is established under regularity conditions and its finite-sample performance is empirically examined by simulation studies. We apply this new estimation method to handle measurement error in some nutrients of interest and misclassification of a categorical variable named physical activity in the European Prospective Investigation into Cancer and Nutrition-InterAct Study data. Analyses show that fruit is negatively associated with type 2 diabetes for a group of women doing active physical activity, protein has positive association with type 2 diabetes for the group of less active physical activity, and actual physical activity has a greater effect on reducing the risk of type 2 diabetes than observed physical activity.

Approximate profile likelihood estimation for Cox regression with covariate measurement error

In nutritional epidemiology, measurement error in covariates is a well-known problem since dietary intakes are usually assessed through self-reporting. In this article, we consider an additive error model in which error variables are highly correlated, and propose a new method called approximate profile likelihood estimation (APLE) for covariates measured with error in the Cox regression. Asymptotic normality of this estimator is established under regularity conditions, and simulation studies are conducted to examine the finite sample performance of the proposed estimator empirically. Moreover, the popular correction method called regression calibration is shown to be a special case of APLE. We then apply APLE to deal with measurement error in some nutrients of interest in the EPIC-InterAct Study under a sensitivity analysis framework.

The Association between Plasma Concentration of Phytoestrogens and Hypertension within the Korean Multicenter Cancer Cohort

In order to examine the association between plasma phytoestrogen concentration (genistein, daidzein, equol and enterolactone) and hypertension, we conducted a nested case–control study for 229 hypertension cases including 112 prehypertension and 159 healthy controls derived from the Korean Multi-center Cancer Cohort (KMCC). The concentration of plasma phytoestrogens was measured using time-resolved fluoroimmunoassay. We assessed the association between plasma phytoestrogens and hypertension using logistic regression models using odds ratio (OR) and 95% confidence interval (95%CI). The highest tertile of plasma equol and enterolactone concentration exhibited a significantly decreased risk of hypertension (equol, OR = 0.34, 95%CI 0.20–0.57; enterolactone, OR = 0.32, 95%CI 0.18–0.57), compared with the lowest tertile. Equol and enterolactone showed reduced ORs for prehypertension (the highest tertile relative to the lowest tertile, OR = 0.50, 95%CI 0.26–0.96; OR = 0.38, 95%CI 0.19–0.75, respectively) and hypertension (OR = 0.42, 95%CI 0.22–0.81; OR = 0.28, 95%CI 0.14–0.54, respectively). There was a stronger association in hypertension (the highest tertile relative to the lowest tertile in obesity vs. non-obesity; equol, OR = 0.06 vs. 0.63; enterolactone, OR = 0.07 vs. 0.46; both p-heterogeneity < 0.01). This study suggests that equol and enterolactone may contribute to prevent primarily prehypertension and hypertension, and control cardiovascular disease (CVD) based on the continuum of hypertension and CVD. Further study to assess hypertension risk based on useful biomarkers, including phytoestrogens, may contribute to primary prevention of hypertension.

Validity of the food frequency questionnaire for adults in nutritional epidemiological studies: A systematic review and meta-analysis

As the most widely used tool for assessing dietary intake, the validity of food frequency questionnaires (FFQs) should be evaluated before application. A comprehensive search of the PubMed and Web of Science databases was conducted for publications from January 2000 to April 1, 2020. Pooled estimates were calculated for correlation coefficients and mean differences for energy and 61 nutrients between FFQs and standard methods. The literature search identified 130 articles that included 21,494 participants. Subgroup analyses according to the number of administrations of the reference method, sample size, administration methods, FFQ items, reference periods, quality of the studies, gender, and regions were also performed. We conducted a meta-analysis by summarizing the available evidence to comprehensively assess the validity of FFQs stratified by the reference method type (24-hour recall (24HRs) and food records (FRs). We also performed subgroup analyses to examine the impact on the final summary estimates. After a meta-analysis of the FFQs' validity correlation coefficients of the included studies, this study showed that the range (median) of the validity coefficients of the 24HRs as reference methods was 0.220-0.770 (0.416), and for the FRs, it was 0.173-0.735 (0.373), which indicated that FFQs were suitable to assess the overall dietary intake in nutritional epidemiological studies. The results of the subgroup analysis showed that the number of administrations of the reference method, administration mode, number of items, reference periods, sample size, and gender mainly affected the validity correlation of FFQs.Supplemental data for this article is available online at https://doi.org/10.1080/10408398.2021.1966737 .

Moment estimation method of parameters in additive measurement error model

BACKGROUND

In nutritional epidemiology, covariates in some studies such as the EPIC are prone to measurement error. Estimation of unknown parameters in most measurement error models for food frequency questionnaire (FFQ) and nutrient biomarkers requires replicated measurements. But, the EPIC-InterAct Study did not collect replicated measurements for FFQ or 24-hour dietary recalls (24HR). The method of correcting measurement error in this case is worth studying.

METHODS

A moment method is applied to estimate unknown parameters of the proposed error model with correlated errors between biased measurements of FFQ and 24HR. Then, correction factor and reliability ratio of each error-prone nutrient can be obtained correspondingly. Afterwards, regression calibration (RC) under a Cox model is used to correct measurement error of nutrients of interest in the EPIC-InterAct data.

RESULTS

Compared to the naive estimation, estimation results for dietary intakes could be very different when we take measurement error into consideration. Using RC as the correction method, hazard ratios (HR) of vegetable plus fruit, fat and energy for males become 1.01 (95% CI 0.75-1.35), 1.30 (95% CI 1.12-1.51) and 1.16 (95% CI 1.04-1.28), respectively, and HR of energy for females becomes 0.99 (95% CI 0.91-1.08). These HRs are greatly different from those by naive estimation.

CONCLUSIONS

Although there is no repeated measurement for FFQ and 24HR, we can still estimate all unknown parameters in our proposed error model under four assumptions and then correct measurement error in nutrients of interest in EPIC-InterAct Study by RC for avoiding some misleading results from naive estimation.

The relationship between urinary polyphenol metabolites and dietary polyphenol intakes in young adults

Spot urinary polyphenols have potential as a biomarker of polyphenol-rich food intakes. The aim of this study is to explore the relationship between spot urinary polyphenols and polyphenol intakes from polyphenol-rich food sources. Young adults (18-24 years old) were recruited into a sub-study of an online intervention aimed at improving diet quality. Participants' intake of polyphenols and polyphenol-rich foods was assessed at baseline and 3 months using repeated 24-h recalls. A spot urine sample was collected at each session, with samples analysed for polyphenol metabolites using LC-MS. To assess the strength of the relationship between urinary polyphenols and dietary polyphenols, Spearman correlations were used. Linear mixed models further evaluated the relationship between polyphenol intakes and urinary excretion. Total urinary polyphenols and hippuric acid (HA) demonstrated moderate correlation with total polyphenol intakes (rs = 0·29-0·47). HA and caffeic acid were moderately correlated with polyphenols from tea/coffee (rs = 0·26-0·46). Using linear mixed models, increases in intakes of total polyphenols or polyphenols from tea/coffee or oil resulted in a greater excretion of HA, whereas a negative relationship was observed between soya polyphenols and HA, suggesting that participants with higher intakes of soya polyphenols had a lower excretion of HA. Findings suggest that total urinary polyphenols may be a promising biomarker of total polyphenol intakes foods and drinks and that HA may be a biomarker of total polyphenol intakes and polyphenols from tea/coffee. Caffeic acid warrants further investigation as a potential biomarker of polyphenols from tea/coffee.

Mass spectrometry-based metabolomics for the discovery of candidate markers of flavonoid and polyphenolic intake in adults

Robust biological markers of dietary exposure are essential in improving the understanding of the link between diet and health outcomes. Polyphenolic compounds, including flavonoids, have been proposed to mitigate the risk of chronic diseases where oxidative stress and inflammation play a central role. Biomarkers can provide objective measurement of the levels of polyphenolic compounds. In this study, we provide methodology to identify potential candidate markers of polyphenol intake in human serum. Seventeen participants from the UK arm of the Global Allergy and Asthma Network of Excellence (GA²LEN) had their dietary intake estimated using a validated food frequency questionnaire, and serum samples were assessed using mass spectrometry to identify potential candidate markers. 144 features were assigned identities, of these we identified four biologically relevant compounds (rhamnazin 3-rutinoside, 2-galloyl-1,4-galactarolactone methyl ester, 2″,32″-di-O-p-coumaroylafzelin and cyclocommunin), which were significantly increased in the serum of participants with high predicted level of fruit and vegetable intake. 2-galloyl-1,4-galactarolactone methyl ester was strongly correlated with total flavonoids (r = 0.62; P = 0.005), flavan-3-ols (r = 0.67; P = 0.002) as well as with other four subclasses. Rhamnazin 3-rutinoside showed strong correlation with pro-anthocyanidins (r = 0.68; P = 0.001), flavones (r = 0.62; P = 0.005). Our results suggest that serum profiling for these compounds might be an effective way of establishing the relative intake of flavonoids and could contribute to improve the accuracy of epidemiological methods to ascertain flavonoid intake.

Biomarker-estimated flavan-3-ol intake is associated with lower blood pressure in cross-sectional analysis in EPIC Norfolk

Flavan-3-ols are a group of bioactive compounds that have been shown to improve vascular function in intervention studies. They are therefore of great interest for the development of dietary recommendation for the prevention of cardio-vascular diseases. However, there are currently no reliable data from observational studies, as the high variability in the flavan-3-ol content of food makes it difficult to estimate actual intake without nutritional biomarkers. In this study, we investigated cross-sectional associations between biomarker-estimated flavan-3-ol intake and blood pressure and other CVD risk markers, as well as longitudinal associations with CVD risk in 25,618 participants of the European Prospective Investigation into Cancer (EPIC) Norfolk cohort. High flavan-3-ol intake, achievable as part of an habitual diet, was associated with a significantly lower systolic blood pressure (- 1.9 (- 2.7; - 1.1) mmHg in men and - 2.5 (- 3.3; - 1.8) mmHg in women; lowest vs highest decile of biomarker), comparable to adherence to a Mediterranean Diet or moderate salt reduction. Subgroup analyses showed that hypertensive participants had stronger inverse association between flavan-3-ol biomarker and systolic blood pressure when compared to normotensive participants. Flavanol intake could therefore have a role in the maintenance of cardiovascular health on a population scale.

Approximate maximum likelihood estimation for logistic regression with covariate measurement error

In nutritional epidemiology, dietary intake assessed with a food frequency questionnaire is prone to measurement error. Ignoring the measurement error in covariates causes estimates to be biased and leads to a loss of power. In this paper, we consider an additive error model according to the characteristics of the European Prospective Investigation into Cancer and Nutrition (EPIC)-InterAct Study data, and derive an approximate maximum likelihood estimation (AMLE) for covariates with measurement error under logistic regression. This method can be regarded as an adjusted version of regression calibration and can provide an approximate consistent estimator. Asymptotic normality of this estimator is established under regularity conditions, and simulation studies are conducted to empirically examine the finite sample performance of the proposed method. We apply AMLE to deal with measurement errors in some interested nutrients of the EPIC-InterAct Study under a sensitivity analysis framework.

STRATOS guidance document on measurement error and misclassification of variables in observational epidemiology: Part 1-Basic theory and simple methods of adjustment

Measurement error and misclassification of variables frequently occur in epidemiology and involve variables important to public health. Their presence can impact strongly on results of statistical analyses involving such variables. However, investigators commonly fail to pay attention to biases resulting from such mismeasurement. We provide, in two parts, an overview of the types of error that occur, their impacts on analytic results, and statistical methods to mitigate the biases that they cause. In this first part, we review different types of measurement error and misclassification, emphasizing the classical, linear, and Berkson models, and on the concepts of nondifferential and differential error. We describe the impacts of these types of error in covariates and in outcome variables on various analyses, including estimation and testing in regression models and estimating distributions. We outline types of ancillary studies required to provide information about such errors and discuss the implications of covariate measurement error for study design. Methods for ascertaining sample size requirements are outlined, both for ancillary studies designed to provide information about measurement error and for main studies where the exposure of interest is measured with error. We describe two of the simpler methods, regression calibration and simulation extrapolation (SIMEX), that adjust for bias in regression coefficients caused by measurement error in continuous covariates, and illustrate their use through examples drawn from the Observing Protein and Energy (OPEN) dietary validation study. Finally, we review software available for implementing these methods. The second part of the article deals with more advanced topics.

Evaluation of (-)-epicatechin metabolites as recovery biomarker of dietary flavan-3-ol intake

Data from dietary intervention studies suggest that intake of (−)-epicatechin mediates beneficial vascular effects in humans. However, population-based investigations are required to evaluate associations between habitual intake and health and these studies rely on accurate estimates of intake, which nutritional biomarkers can provide. Here, we evaluate a series of structurally related (−)-epicatechin metabolites (SREM), particularly (−)-epicatechin-3′-glucuronide, (−)-epicatechin-3′-sulfate and 3′-O-methyl-(−)-epicatechin-5-sulfate (SREM_B), as flavan-3-ol and (−)-epicatechin intake. SREM_B in urine proved to be a specific indicator of (−)-epicatechin intake, showing also a strong correlation with the amount of (−)-epicatechin ingested (R²: 0.86 (95% CI 0.8l; 0.92). The median recovery of (−)-epicatechin as SREM_B in 24 h urine was 10% (IQR 7–13%) and we found SREM_B in the majority of participants of EPIC Norfolk (83% of 24,341) with a mean concentration of 2.4 ± 3.2 µmol/L. Our results show that SREM_B are suitable as biomarker of (−)-epicatechin intake. According to evaluation criteria from IARC and the Institute of Medicine, the results obtained support use of SREM_B as a recovery biomarker to estimate actual intake of (−)-epicatechin.

Combining biomarker and self-reported dietary intake data: A review of the state of the art and an exposition of concepts

Classical approaches to assessing dietary intake are associated with measurement error. In an effort to address inherent measurement error in dietary self-reported data there is increased interest in the use of dietary biomarkers as objective measures of intake. Furthermore, there is a growing consensus of the need to combine dietary biomarker data with self-reported data. A review of state of the art techniques employed when combining biomarker and self-reported data is conducted. Two predominant methods, the calibration method and the method of triads, emerge as relevant techniques used when combining biomarker and self-reported data to account for measurement errors in dietary intake assessment. Both methods crucially assume measurement error independence. To expose and understand the performance of these methods in a range of realistic settings, their underpinning statistical concepts are unified and delineated, and thorough simulation studies are conducted. Results show that violation of the methods' assumptions negatively impacts resulting inference but that this impact is mitigated when the variation of the biomarker around the true intake is small. Thus there is much scope for the further development of biomarkers and models in tandem to achieve the ultimate goal of accurately assessing dietary intake.

Comparison of an interactive 24-h recall and weighed food record for measuring energy and nutrient intakes from complementary foods among 9-10-month-old Malawian infants consuming lipid-based nutrient supplements

Fortifying complementary foods with lipid-based nutrient supplements (LNS) may improve energy and nutrient intakes of infants at risk for undernutrition. We aimed to determine the relative validity of an interactive 24-h recall (i-24-HR) for assessing the impact of an LNS intervention on dietary intakes of energy and nutrients among rural Malawian 9-10-month-old infants (n 132) participating in the International Lipid-Based Nutrient Supplements Dose (iLiNS-DOSE) trial. Dietary data were collected for the same day via i-24-HR and weighed food records. Inter-method agreements were estimated overall and by intervention group, using Bland-Altman plots and paired t tests; measurement error models (differential error); and percentage of food omissions and intrusions were estimated. Overall, inter-method differences in mean intakes of energy and most nutrients were not significant. When stratified by group, recalled energy intakes were under-estimated (-368 kJ; P=0·01) in the control but not in the intervention group (-42 kJ; P=0·6). This differential reporting error was related to an over-estimation of recalled LNS (8·1 v. 4·5 g; P30 % eating occasions) omissions were milk/fish/eggs, starchy roots/vegetables and sweetened snacks. Common intrusions were milk/yogurt. Starchy staples and LNS were recalled when consumed (>85 %) (i.e. matched). These results emphasise the importance of considering differential error when interpreting dietary results in LNS trials.

Evaluation at scale of microbiome-derived metabolites as biomarker of flavan-3-ol intake in epidemiological studies

The accurate assessment of dietary intake is crucial to investigate the effect of diet on health. Currently used methods, relying on self-reporting and food composition data, are known to have limitations and might not be suitable to estimate the intake of many bioactive food components. An alternative are nutritional biomarkers, which can allow an unbiased assessment of intake. They require a careful evaluation of their suitability, including: (a) the availability of a precise, accurate and robust analytical method, (b) their specificity (c) a consistent relationship with actual intake. We have evaluated human metabolites of a microbiome-derived flavan-3-ol catabolite, 5-(3',4'-dihydroxyphenyl)-[gamma]-valerolactone (gVL), as biomarker of flavan-3-ol intake in large epidemiological studies. Flavan-3-ols are widely consumed plant bioactives, which have received considerable interest due to their potential ability to reduce CVD risk. The availability of authentic standards allowed the development of a validated high-throughput method suitable for large-scale studies. In dietary intervention studies, we could show that gVL metabolites are specific for flavan-3-ols present in tea, fruits, wine and cocoa-derived products, with a strong correlation between intake and biomarker (Spearman's r = 0.90). This biomarker will allow for the first time to estimate flavan-3-ol intake and further investigation of associations between intake and disease risk.

Nutrition epidemiology of flavan-3-ols: The known unknowns

Nutritional epidemiology has an important role, as it can provide long-term data from large populations and does not rely on surrogate markers for morbidity/mortality. Meaningful interpretation and applications of outcomes from epidemiological studies depend on the accurate assessment of dietary intake, which is currently mainly based on a combination of self-reporting and food composition data. Flavan-3-ols are a group of bioactives (non-essential dietary components with significant impact on health) that is a possible candidate for the development of dietary recommendations. The breadth of data available on their effect on health also provides the basis for investigating the suitability of the methods currently used in nutritional epidemiology to assess the health effects of bioactives. The outcomes of this assessment demonstrate that the limitations of currently used methods make it virtually impossible to estimate intake accurately from self-reported dietary data. This is due to the limitations of self-reporting, especially from food-frequency questionnaires, and the inability of currently used methods to deal with the high variability of food composition. Indeed, the estimated intake of flavan-3-ols, can only be interpreted as a marker of specific dietary patterns, but not as the actual intake amount. The interpretation of results from such studies are fraught with serious limitations, especially for establishing associations between intake and health and the development of dietary recommendations. Alternative assessment not affected by these limitations, such as biomarkers, are required to overcome these limitations. The development of nutritional biomarkers is therefore crucial to investigate the health effect of bioactives.

Systematic review of statistical approaches to quantify, or correct for, measurement error in a continuous exposure in nutritional epidemiology

BACKGROUND

Several statistical approaches have been proposed to assess and correct for exposure measurement error. We aimed to provide a critical overview of the most common approaches used in nutritional epidemiology.

METHODS

MEDLINE, EMBASE, BIOSIS and CINAHL were searched for reports published in English up to May 2016 in order to ascertain studies that described methods aimed to quantify and/or correct for measurement error for a continuous exposure in nutritional epidemiology using a calibration study.

RESULTS

We identified 126 studies, 43 of which described statistical methods and 83 that applied any of these methods to a real dataset. The statistical approaches in the eligible studies were grouped into: a) approaches to quantify the relationship between different dietary assessment instruments and "true intake", which were mostly based on correlation analysis and the method of triads; b) approaches to adjust point and interval estimates of diet-disease associations for measurement error, mostly based on regression calibration analysis and its extensions. Two approaches (multiple imputation and moment reconstruction) were identified that can deal with differential measurement error.

CONCLUSIONS

For regression calibration, the most common approach to correct for measurement error used in nutritional epidemiology, it is crucial to ensure that its assumptions and requirements are fully met. Analyses that investigate the impact of departures from the classical measurement error model on regression calibration estimates can be helpful to researchers in interpreting their findings. With regard to the possible use of alternative methods when regression calibration is not appropriate, the choice of method should depend on the measurement error model assumed, the availability of suitable calibration study data and the potential for bias due to violation of the classical measurement error model assumptions. On the basis of this review, we provide some practical advice for the use of methods to assess and adjust for measurement error in nutritional epidemiology.

Consumption of citrus and cruciferous vegetables with incident type 2 diabetes mellitus based on a meta-analysis of prospective study

BACKGROUND

Observational studies and meta-analyses suggested that increased total fruits and vegetables consumption have a protective role in incidence of type 2 diabetes mellitus (T2DM). However, we still don't know whether the subtypes, such as citrus fruits and cruciferous vegetables (CV), have a preventive role.

METHODS

We systematically searched the MEDLINE and EMBASE databases up to December 31, 2014. Summary relative risks (SRRs) and 95% confidence intervals (CIs) were calculated using random-effects models.

RESULTS

Seven distinct prospective cohort studies (five articles) were identified for this study. A total of 16,544 incident cases of type 2 diabetes were ascertained among 306,723 participants with follow-up periods ranging from 4.6 to 24 years. Based on four prospective cohort studies, we found that overall, consumption of CV had a protective role in the T2DM incidence (highest vs. lowest analysis: SRR=0.84, 95% CI: 0.73 to 0.96), with evidence of significant heterogeneity (P=0.09, I(2)=54.4%). This association was independent of the main risk factors for cardiovascular disease: smoking, alcohol use, BMI, and physical activity etc. Consumption of citrus fruits did not have a protective role in the T2DM development (highest vs. lowest analysis: SRR=1.02, 95% CI: 0.96 to 1.08), with no evidence of significant heterogeneity (P=0.49, I(2)=0).

CONCLUSIONS

Higher consumption of CV, but not citrus fruits, is associated with a significantly decreased risk of type 2 diabetes. Further large prospective studies are needed to elucidate both relationships.

Consumption of whole grains in relation to mortality from all causes, cardiovascular disease, and diabetes: Dose-response meta-analysis of prospective cohort studies

BACKGROUND

To investigate the correlation between consumption of whole grains and the risk of all-cause, cardiovascular disease (CVD), and diabetes-specific mortality according to a dose-response meta-analysis of prospective cohort studies.

METHODS

Observational cohort studies, which reported associations between whole grains and the risk of death outcomes, were identified by searching articles in the MEDLINE, EMBASE, and the reference lists of relevant articles. The search was up to November 30, 2015. Data extraction was performed by 2 independent investigators, and a consensus was reached with involvement of a third.

RESULTS

Ten prospective cohort studies (9 publications) were eligible in this meta-analysis. During follow-up periods ranging from 5.5 to 26 years, there were 92,647 deaths among 782,751 participants. Overall, a diet containing greater amounts of whole grains may be associated with a lower risk of all-cause, CVD-, and coronary heart disease (CHD)-specific mortality. The summary relative risks (RRs) were 0.93 (95% confidence intervals [CIs]: 0.91-0.95; Pheterogeneity < 0.001) for all-cause mortality, 0.95 (95% CIs: 0.92-0.98; Pheterogeneity < 0.001) for CVD-specific mortality, and 0.92 (95% CIs: 0.88-0.97; Pheterogeneity < 0.001) for CHD-specific mortality for an increment of 1 serving (30 g) a day of whole grain intake. The combined estimates were robust across subgroup and sensitivity analyses. Higher consumption of whole grains was not appreciably associated with risk of mortality from stroke and diabetes.

CONCLUSION

Evidence from observational cohort studies indicates inverse associations of intake of whole grains with risk of mortality from all-cause, CVD, and CHD. However, no associations with risk of deaths from stroke and diabetes were observed.

The use of multiple imputation method for the validation of 24-h food recalls by part-time observation of dietary intake in school

External validation of food recall over 24 h in schoolchildren is often restricted to eating events in schools and is based on direct observation as the reference method. The aim of this study was to estimate the dietary intake out of school, and consequently the bias in such research design based on only part-time validated food recall, using multiple imputation (MI) conditioned on the information on child age, sex, BMI, family income, parental education and the school attended. The previous-day, web-based questionnaire WebCAAFE, structured as six meals/snacks and thirty-two foods/beverage, was answered by a sample of 7-11-year-old Brazilian schoolchildren (n 602) from five public schools. Food/beverage intake recalled by children was compared with the records provided by trained observers during school meals. Sensitivity analysis was performed with artificial data emulating those recalled by children on WebCAAFE in order to evaluate the impact of both differential and non-differential bias. Estimated bias was within ±30 % interval for 84·4 % of the thirty-two foods/beverages evaluated in WebCAAFE, and half of the latter reached statistical significance (P<0·05). Rarely (<3 %) consumed dietary items were often under-reported (fish/seafood, vegetable soup, cheese bread, French fries), whereas some of those most frequently reported (meat, bread/biscuits, fruits) showed large overestimation. Compared with the analysis restricted to fully validated data, MI reduced differential bias in sensitivity analysis but the bias still remained large in most cases. MI provided a suitable statistical framework for part-time validation design of dietary intake over six daily eating events.

A statistical model for measurement error that incorporates variation over time in the target measure, with application to nutritional epidemiology

Most statistical methods that adjust analyses for measurement error assume that the target exposure T is a fixed quantity for each individual. However, in many applications, the value of T for an individual varies with time. We develop a model that accounts for such variation, describing the model within the framework of a meta-analysis of validation studies of dietary self-report instruments, where the reference instruments are biomarkers. We demonstrate that in this application, the estimates of the attenuation factor and correlation with true intake, key parameters quantifying the accuracy of the self-report instrument, are sometimes substantially modified under the time-varying exposure model compared with estimates obtained under a traditional fixed-exposure model. We conclude that accounting for the time element in measurement error problems is potentially important.

Application of a New Statistical Model for Measurement Error to the Evaluation of Dietary Self-report Instruments

Most statistical methods that adjust analyses for dietary measurement error treat an individual's usual intake as a fixed quantity. However, usual intake, if defined as average intake over a few months, varies over time. We describe a model that accounts for such variation and for the proximity of biomarker measurements to self-reports within the framework of a meta-analysis, and apply it to the analysis of data on energy, protein, potassium, and sodium from a set of five large validation studies of dietary self-report instruments using recovery biomarkers as reference instruments. We show that this time-varying usual intake model fits the data better than the fixed usual intake assumption. Using this model, we estimated attenuation factors and correlations with true longer-term usual intake for single and multiple 24-hour dietary recalls (24HRs) and food frequency questionnaires (FFQs) and compared them with those obtained under the "fixed" method. Compared with the fixed method, the estimates using the time-varying model showed slightly larger values of the attenuation factor and correlation coefficient for FFQs and smaller values for 24HRs. In some cases, the difference between the fixed method estimate and the new estimate for multiple 24HRs was substantial. With the new method, while four 24HRs had higher estimated correlations with truth than a single FFQ for absolute intakes of protein, potassium, and sodium, for densities the correlations were approximately equal. Accounting for the time element in dietary validation is potentially important, and points toward the need for longer-term validation studies.

Association of oxidative stress biomarkers with adiposity and clinical staging in women with breast cancer

BACKGROUND/OBJECTIVES

Breast cancer is a disease characterised by both oxidative reactions and inflammation. However, few studies have focused on the oxidative and inflammatory biomarkers. The aim of the present study was to evaluate the association between oxidative stress markers and adiposity and clinical staging, as well as the association between the oxidative and the antioxidant biomarkers of women with breast cancer.

SUBJECTS/METHODS

A total of 135 cases of breast cancer occurring in 2011 and 2012 were assessed. After exclusions, 101 pre- and post-menopausal women with clinical staging I to IV were eligible to participate in the study. The anthropometric evaluation was performed by collecting data on waist circumference, body mass index and body composition. The socioeconomic and clinical profiles were determined using a standard questionnaire. For the oxidative biomarkers, thiobarbituric acid reactive substances (TBARS), oxidative DNA damage (8-hydroxy-2-deoxyguanosine (8-OHdG)), low-density lipoprotein(-) (LDL(-)), autoantibody anti-LDL(-) and liposoluble antioxidants (α-tocopherol, retinol and β-carotene) were analysed. The data were analysed using differences in the mean values, correlation tests and multiple linear regression.

RESULTS

The antioxidant levels were higher in postmenopausal women with clinical staging I and II and negative lymph nodes. The TBARS level was associated with clinical staging. Adiposity was associated with levels of retinol and 8-OHdG, whereas LDL(-), 8-OHdG and TBARS were correlated with liposoluble antioxidants after adjusting for the confounders.

CONCLUSIONS

The adiposity and clinical staging of patients were associated with oxidative stress. The oxidative and antioxidant biomarkers showed a negative correlation in patients with breast cancer.

Consumption of nuts and legumes and risk of stroke: a meta-analysis of prospective cohort studies

BACKGROUND AND AIM

The relationships between dietary nuts and legume intake and risk of stroke are inconsistent. We summarized the evidence by a meta-analysis of prospective cohort studies.

METHODS AND RESULTS

We systematically searched the MEDLINE and EMBASE databases up to 31 January 2014. Random-effects models were used to calculate summary relative risks (SRRs) and 95% confidence intervals (CIs). Between-study heterogeneity was assessed using the Cochran's Q and I(2) statistics. Eight prospective studies with a total of 468,887 subjects and 10,493 stroke events were included in the meta-analysis. Overall, a diet containing greater amounts of legumes may be not associated with a lower risk of stroke (SRR = 0.95, 95% CI: 0.84-1.08; P(heterogeneity) = 0.091, I(2) = 43.2%); however, a diet containing greater amounts of nuts may be associated with a lower risk of stroke (SRR = 0.90, 95% CI: 0.81-0.99; P(heterogeneity) = 0.527, I(2) = 0). Gender significantly modified the effects of nut consumption on stroke risk, and high nut intake was associated with reduced risk of stroke in women (SRR = 0.85, 95% CI: 0.75-0.97) other than in men (SRR = 0.95, 95% CI: 0.82-1.11).

CONCLUSION

The current meta-analysis provides some evidences for the hypothesis that high intake of dietary nut was inversely associated with stroke risk, whereas dietary legumes intake was not associated with stroke risk.

Use of two-part regression calibration model to correct for measurement error in episodically consumed foods in a single-replicate study design: EPIC case study

In epidemiologic studies, measurement error in dietary variables often attenuates association between dietary intake and disease occurrence. To adjust for the attenuation caused by error in dietary intake, regression calibration is commonly used. To apply regression calibration, unbiased reference measurements are required. Short-term reference measurements for foods that are not consumed daily contain excess zeroes that pose challenges in the calibration model. We adapted two-part regression calibration model, initially developed for multiple replicates of reference measurements per individual to a single-replicate setting. We showed how to handle excess zero reference measurements by two-step modeling approach, how to explore heteroscedasticity in the consumed amount with variance-mean graph, how to explore nonlinearity with the generalized additive modeling (GAM) and the empirical logit approaches, and how to select covariates in the calibration model. The performance of two-part calibration model was compared with the one-part counterpart. We used vegetable intake and mortality data from European Prospective Investigation on Cancer and Nutrition (EPIC) study. In the EPIC, reference measurements were taken with 24-hour recalls. For each of the three vegetable subgroups assessed separately, correcting for error with an appropriately specified two-part calibration model resulted in about three fold increase in the strength of association with all-cause mortality, as measured by the log hazard ratio. Further found is that the standard way of including covariates in the calibration model can lead to over fitting the two-part calibration model. Moreover, the extent of adjusting for error is influenced by the number and forms of covariates in the calibration model. For episodically consumed foods, we advise researchers to pay special attention to response distribution, nonlinearity, and covariate inclusion in specifying the calibration model.

Nutrient intake in Italian infants and toddlers from North and South Italy: the Nutrintake 636 study

We performed a cross-sectional study to compare the intake of energy, macronutrients, fiber, sodium and iron and the anthropometric status of infants and toddlers living in North (Milano) and South Italy (Catania). Nutrient intake was evaluated using a 7-day weighed food record. Out of 400 planned children aged 6 to 36 months, 390 (98%) were recruited, 189 in Milano and 201 in Catania. The mean (standard deviation) age was 17 (9) months in Milano and 17 (10) months in Catania. Anthropometry, energy intake and macronutrient intake were similar in Milano and Catania. However, iron intake was 27% lower and fiber intake 16% higher in Milano than in Catania. Despite normal anthropometry and energy intake, in the pooled sample there was a high intake of proteins, simple carbohydrates, saturated fats and sodium, and a low intake of iron and fiber compared to Italian reference values. This is the first study to report the macro- and micro-nutrient intake of children aged <12 months using the 7-day weighed food record and one of the very few studies that have employed such reference method in children from the general population.