Showing papers by "Ross L. Prentice published in 2014"

Pooled Results From 5 Validation Studies of Dietary Self-Report Instruments Using Recovery Biomarkers for Energy and Protein Intake

Abstract: We pooled data from 5 large validation studies of dietary self-report instruments that used recovery biomarkers as references to clarify the measurement properties of food frequency questionnaires (FFQs) and 24-hour recalls. The studies were conducted in widely differing US adult populations from 1999 to 2009. We report on total energy, protein, and protein density intakes. Results were similar across sexes, but there was heterogeneity across studies. Using a FFQ, the average correlation coefficients for reported versus true intakes for energy, protein, and protein density were 0.21, 0.29, and 0.41, respectively. Using a single 24-hour recall, the coefficients were 0.26, 0.40, and 0.36, respectively, for the same nutrients and rose to 0.31, 0.49, and 0.46 when three 24-hour recalls were averaged. The average rate of under-reporting of energy intake was 28% with a FFQ and 15% with a single 24-hour recall, but the percentages were lower for protein. Personal characteristics related to under-reporting were body mass index, educational level, and age. Calibration equations for true intake that included personal characteristics provided improved prediction. This project establishes that FFQs have stronger correlations with truth for protein density than for absolute protein intake, that the use of multiple 24-hour recalls substantially increases the correlations when compared with a single 24-hour recall, and that body mass index strongly predicts under-reporting of energy and protein intakes.

Hormone therapy dose, formulation, route of delivery, and risk of cardiovascular events in women: findings from the Women’s Health Initiative Observational Study

Abstract: Objective Research comparing hormone therapy (HT) doses, regimens, and routes of delivery in relation to cardiovascular disease (CVD) outcomes have been limited. This study directly compared different estrogen doses, routes of delivery, and HT formulations in postmenopausal women in relation to the risk of coronary heart disease (CHD), stroke, CVD mortality, total CVD, and all-cause mortality.

Measurement error corrected sodium and potassium intake estimation using 24-hour urinary excretion

Abstract: Epidemiological studies of the association of sodium and potassium intake with cardiovascular disease risk have almost exclusively relied on self-reported dietary data. Here, 24-hour urinary excretion assessments are used to correct the dietary self-report data for measurement error under the assumption that 24-hour urine recovery provides a biomarker that differs from usual intake according to a classical measurement model. Under this assumption, dietary self-reports underestimate sodium by 0% to 15%, overestimate potassium by 8% to 15%, and underestimate sodium/potassium ratio by ≈20% using food frequency questionnaires, 4-day food records, or three 24-hour dietary recalls in Women's Health Initiative studies. Calibration equations are developed by linear regression of log-transformed 24-hour urine assessments on corresponding log-transformed self-report assessments and several study subject characteristics. For each self-report method, the calibration equations turned out to depend on race and age and strongly on body mass index. After adjustment for temporal variation, calibration equations using food records or recalls explained 45% to 50% of the variation in (log-transformed) 24-hour urine assessments for sodium, 60% to 70% of the variation for potassium, and 55% to 60% of the variation for sodium/potassium ratio. These equations may be suitable for use in epidemiological disease association studies among postmenopausal women. The corresponding signals from food frequency questionnaire data were weak, but calibration equations for the ratios of sodium and potassium/total energy explained ≈35%, 50%, and 45% of log-biomarker variation for sodium, potassium, and their ratio, respectively, after the adjustment for temporal biomarker variation and may be suitable for cautious use in epidemiological studies. Clinical Trial Registration- URL: www.clinicaltrials.gov. Unique identifier: NCT00000611.

Simultaneous Association of Total Energy Consumption and Activity-Related Energy Expenditure With Risks of Cardiovascular Disease, Cancer, and Diabetes Among Postmenopausal Women

Abstract: Total energy consumption and activity-related energy expenditure (AREE) estimates that have been calibrated using biomarkers to correct for measurement error were simultaneously associated with the risks of cardiovascular disease, cancer, and diabetes among postmenopausal women who were enrolled in the Women's Health Initiative at 40 US clinical centers and followed from 1994 to the present. Calibrated energy consumption was found to be positively related, and AREE inversely related, to the risks of various cardiovascular diseases, cancers, and diabetes. These associations were not evident in most corresponding analyses that did not correct for measurement error. However, an important analytical caveat relates to the role of body mass index (BMI) (weight (kg)/height (m)(2)). In the calibrated variable analyses, BMI was regarded, along with self-reported data, as a source of information on energy consumption and physical activity, and BMI was otherwise excluded from the disease risk models. This approach cannot be fully justified with available data, and the analyses herein imply a need for improved dietary and physical activity assessment methods and for longitudinal self-reported and biomarker data to test and relax modeling assumptions. Estimated hazard ratios for 20% increases in total energy consumption and AREE, respectively, were as follows: 1.49 (95% confidence interval: 1.18, 1.88) and 0.80 (95% confidence interval: 0.69, 0.92) for total cardiovascular disease; 1.43 (95% confidence interval: 1.17, 1.73) and 0.84 (95% confidence interval: 0.73, 0.96) for total invasive cancer; and 4.17 (95% confidence interval: 2.68, 6.49) and 0.60 (95% confidence interval: 0.44, 0.83) for diabetes.

Potassium Intake and Risk of Stroke in Women With Hypertension and Nonhypertension in the Women’s Health Initiative

Abstract: Background and Purpose—Dietary potassium has been associated with lower risk of stroke, but there are little data on dietary potassium effects on different stroke subtypes or in older women with hypertension and nonhypertension. Methods—The study population consisted of 90 137 postmenopausal women aged 50 to 79 at enrollment, free of stroke history at baseline, followed up prospectively for an average of 11 years. Outcome variables were total, ischemic, and hemorrhagic stroke, and all-cause mortality. Incidence was compared across quartiles of dietary potassium intake, and hazard ratios were obtained from Cox proportional hazards models after adjusting for potential confounding variables, and in women with hypertension and nonhypertension separately. Results—Mean dietary potassium intake was 2611 mg/d. Highest quartile of potassium intake was associated with lower incidence of ischemic and hemorrhagic stroke and total mortality. Multivariate analyses comparing highest to lowest quartile of potassium intak...

Cancer Incidence and Mortality during the Intervention and Postintervention Periods of the Women's Health Initiative Dietary Modification Trial

Abstract: Background: The Women's Health Initiative (WHI) low-fat (20% kcal) dietary modification (DM) trial (1993–2005) demonstrated a nonsignificant reduction in breast cancer, a nominally significant reduction in ovarian cancer, and no effect on other cancers (mean 8.3 years intervention). Consent to nonintervention follow-up was 83% ( n = 37,858). This analysis was designed to assess postintervention cancer risk in women randomized to the low-fat diet (40%) versus usual diet comparison (60%). Methods: Randomized, controlled low-fat diet intervention for prevention of breast and colorectal cancers conducted in 48,835 postmenopausal U.S. women, ages 50 to 79 years at 40 U.S. sites. Outcomes included total invasive cancer, breast cancer, and colorectal cancer, and cancer-specific and overall mortality. Results: There were no intervention effects on invasive breast or colorectal cancer, other cancers, or cancer-specific or overall mortality during the postintervention period or the combined intervention and follow-up periods. For invasive breast cancer, the hazard ratios (HR) and 95% confidence interval (CI) were 0.92 (0.84–1.01) during intervention, 1.08 (0.94–1.24) during the postintervention period, and 0.97 (0.89–1.05) during cumulative follow-up. A reduced risk for estrogen receptor positive/progesterone receptor–negative tumors was demonstrated during follow-up. In women with higher baseline fat intake (quartile), point estimates of breast cancer risk were HR, 0.76 (95% CI, 0.62–0.92) during intervention versus HR, 1.11 (95% CI, 0.84–1.4) during postintervention follow-up ( P diff = 0.03). Conclusions: Dietary fat intake increased postintervention in intervention women; no long-term reduction in cancer risk or mortality was shown in the WHI DM trial. Impact: Dietary advisement to reduce fat for cancer prevention after menopause generally was not supported by the WHI DM trial. Cancer Epidemiol Biomarkers Prev; 23(12); 2924–35. ©2014 AACR .

Post-GWAS gene–environment interplay in breast cancer: results from the Breast and Prostate Cancer Cohort Consortium and a meta-analysis on 79 000 women

Abstract: We studied the interplay between 39 breast cancer (BC) risk SNPs and established BC risk (body mass index, height, age at menarche, parity, age at menopause, smoking, alcohol and family history of BC) and prognostic factors (TNM stage, tumor grade, tumor size, age at diagnosis, estrogen receptor status and progesterone receptor status) as joint determinants of BC risk. We used a nested case-control design within the National Cancer Institute's Breast and Prostate Cancer Cohort Consortium (BPC3), with 16 285 BC cases and 19 376 controls. We performed stratified analyses for both the risk and prognostic factors, testing for heterogeneity for the risk factors, and case-case comparisons for differential associations of polymorphisms by subgroups of the prognostic factors. We analyzed multiplicative interactions between the SNPs and the risk factors. Finally, we also performed a meta-analysis of the interaction ORs from BPC3 and the Breast Cancer Association Consortium. After correction for multiple testing, no significant interaction between the SNPs and the established risk factors in the BPC3 study was found. The meta-analysis showed a suggestive interaction between smoking status and SLC4A7-rs4973768 (Pinteraction = 8.84 × 10(-4)) which, although not significant after considering multiple comparison, has a plausible biological explanation. In conclusion, in this study of up to almost 79 000 women we can conclusively exclude any novel major interactions between genome-wide association studies hits and the epidemiologic risk factors taken into consideration, but we propose a suggestive interaction between smoking status and SLC4A7-rs4973768 that if further replicated could help our understanding in the etiology of BC.

Use of a urinary sugars biomarker to assess measurement error in self-reported sugars intake in the Nutrition and Physical Activity Assessment Study (NPAAS)

Abstract: Background: Measurement error in self-reported sugars intake may be obscuring the association between sugars and cancer risk in nutritional epidemiologic studies. Methods: We used 24-hour urinary sucrose and fructose as a predictive biomarker for total sugars, to assess measurement error in self-reported sugars intake. The Nutrition and Physical Activity Assessment Study (NPAAS) is a biomarker study within the Women's Health Initiative (WHI) Observational Study that includes 450 postmenopausal women ages 60 to 91 years. Food Frequency Questionnaires (FFQ), four-day food records (4DFR), and three 24-hour dietary recalls (24HRs) were collected along with sugars and energy dietary biomarkers. Results: Using the biomarker, we found self-reported sugars to be substantially and roughly equally misreported across the FFQ, 4DFR, and 24HR. All instruments were associated with considerable intake- and person-specific bias. Three 24HRs would provide the least attenuated risk estimate for sugars (attenuation factor, AF = 0.57), followed by FFQ (AF = 0.48) and 4DFR (AF = 0.32), in studies of energy-adjusted sugars and disease risk. In calibration models, self-reports explained little variation in true intake (5%–6% for absolute sugars and 7%–18% for sugars density). Adding participants' characteristics somewhat improved the percentage variation explained (16%–18% for absolute sugars and 29%–40% for sugars density). Conclusions: None of the self-report instruments provided a good estimate of sugars intake, although overall 24HRs seemed to perform the best. Impact: Assuming the calibrated sugars biomarker is unbiased, this analysis suggests that measuring the biomarker in a subsample of the study population for calibration purposes may be necessary for obtaining unbiased risk estimates in cancer association studies. Cancer Epidemiol Biomarkers Prev; 23(12); 2874–83. ©2014 AACR .

Postmenopausal hormone therapy and the risks of coronary heart disease, breast cancer, and stroke.

Abstract: The principal findings are briefly reviewed from the Women's Health Initiative (WHI) trials of the most commonly used postmenopausal hormone regimens in the US, conjugated equine estrogens and these same estrogens plus medroxyprogesterone acetate. A more detailed review is presented for three major clinical outcomes: coronary heart disease, the primary trial outcome for which a major benefit was hypothesized; invasive breast cancer, the primary safety outcome for which some adverse effect was expected; and stroke which surfaced as an important adverse effect with both regimens, and one that is influential in decisions concerning the continued use of postmenopausal estrogens alone. The review for these outcomes includes an update on interactions of treatment effects with study subject characteristics and exposures and with pre-randomization biomarker levels. It also includes a focus on timing issues that are important to the understanding of treatment effects. Specifically, with combined estrogen plus progestin coronary heart disease risk was elevated early with the elevation dissipating after a few years of treatment, whereas breast cancer elevations increased during the treatment period, and climbed to about a 3-fold increase following 5 years of adherence. Importantly, breast cancer risk elevations appear to be higher among women who initiate treatment at the menopause, or soon thereafter, compared to women having a longer gap time. Stroke effects, on the other hand didn't seem to vary appreciably with these timing issues. The adverse effect was evidently localized to ischemic strokes, for which there was an approximate 50% increase with either regimen. The rather limited knowledge concerning the biomarkers and biological pathways that mediate the hormone therapy effects on these diseases is also briefly reviewed.

Sex hormone associations with breast cancer risk and the mediation of randomized trial postmenopausal hormone therapy effects

Abstract: Introduction: Paradoxically, a breast cancer risk reduction with conjugated equine estrogens (CEE) and a risk elevation with CEE plus medroxyprogesterone acetate (CEE + MPA) were observed in the Women’s Health Initiative (WHI) randomized controlled trials. The effects of hormone therapy on serum sex hormone levels, and on the association between baseline sex hormones and disease risk, may help explain these divergent breast cancer findings. Methods: Serum sex hormone concentrations were measured for 348 breast cancer cases in the CEE + MPA trial and for 235 cases in the CEE trial along with corresponding pair-matched controls, nested within the WHI trials of healthy postmenopausal women. Association and mediation analyses, to examine the extent to which sex hormone levels and changes can explain the breast cancer findings, were conducted using logistic regression. Results: Following CEE treatment, breast cancer risk was associated with higher concentrations of baseline serum estrogens, and with lower concentrations of sex hormone binding globulin. However, following CEE + MPA, there was no association of breast cancer risk with baseline sex hormone levels. The sex hormone changes from baseline to year 1 provided an explanation for much of the reduced breast cancer risk with CEE. Specifically, the treatment odds ratio (95% confidence interval) increased from 0.71 (0.43, 1.15) to 0.92 (0.41, 2.09) when the year 1 measures were included in the logistic regression analysis. In comparison, the CEE + MPA odds ratio was essentially unchanged when these year 1 measures were included. Conclusions: Breast cancer risk remains low following CEE use among women having favorable baseline sex hormone profiles, but CEE + MPA evidently produces a breast cancer risk for all women similar to that for women having an unfavorable baseline sex hormone profile. These patterns could reflect breast ductal epithelial cell stimulation by CEE + MPA that is substantially avoided with CEE, in conjunction with relatively more favorable effects of either regimen following a sustained period of estrogen deprivation. These findings may have implications for other hormone therapy formulations and routes of delivery.

A penalized EM algorithm incorporating missing data mechanism for Gaussian parameter estimation

Abstract: Missing data rates could depend on the targeted values in many settings, including mass spectrometry-based proteomic profiling studies. Here, we consider mean and covariance estimation under a multivariate Gaussian distribution with non-ignorable missingness, including scenarios in which the dimension (p) of the response vector is equal to or greater than the number (n) of independent observations. A parameter estimation procedure is developed by maximizing a class of penalized likelihood functions that entails explicit modeling of missing data probabilities. The performance of the resulting "penalized EM algorithm incorporating missing data mechanism (PEMM)" estimation procedure is evaluated in simulation studies and in a proteomic data illustration.

Estimating Risk With Time-to-Event Data: An Application to the Women’s Health Initiative

Abstract: Accurate and individualized risk prediction is critical for population control of chronic diseases such as cancer and cardiovascular disease. Large cohort studies provide valuable resources for building risk prediction models, as the risk factors are collected at the baseline and subjects are followed over time until disease occurrence or termination of the study. However, for rare diseases the baseline risk may not be estimated reliably based on cohort data only, due to sparse events. In this paper, we propose to make use of external information to improve efficiency for estimating time-dependent absolute risk. We derive the relationship between external disease incidence rates and the baseline risk, and incorporate the external disease incidence information into estimation of absolute risks, while allowing for potential difference of disease incidence rates between cohort and external sources. The asymptotic properties, namely, uniform consistency and weak convergence, of the proposed estimators are established. Simulation results show that the proposed estimator for absolute risk is more efficient than that based on the Breslow estimator, which does not utilize external disease incidence rates. A large cohort study, the Women's Health Initiative Observational Study, is used to illustrate the proposed method.

Covariate measurement error correction methods in mediation analysis with failure time data

Abstract: Mediation analysis is important for understanding the mechanisms whereby one variable causes changes in another. Measurement error could obscure the ability of the potential mediator to explain such changes. This paper focuses on developing correction methods for measurement error in the mediator with failure time outcomes. We consider a broad definition of measurement error, including technical error and error associated with temporal variation. The underlying model with the ‘true’ mediator is assumed to be of the Cox proportional hazards model form. The induced hazard ratio for the observed mediator no longer has a simple form independent of the baseline hazard function, due to the conditioning event. We propose a mean-variance regression calibration approach and a follow-up time regression calibration approach, to approximate the partial likelihood for the induced hazard function. Both methods demonstrate value in assessing mediation effects in simulation studies. These methods are generalized to multiple biomarkers and to both case-cohort and nested case-control sampling design. We apply these correction methods to the Women's Health Initiative hormone therapy trials to understand the mediation effect of several serum sex hormone measures on the relationship between postmenopausal hormone therapy and breast cancer risk.

Self-consistent nonparametric maximum likelihood estimator of the bivariate survivor function

Abstract: SUMMARY As usually formulated the nonparametric likelihood for the bivariate survivor function is overparameterized, resulting in uniqueness problems for the corresponding nonparametric maximum likelihood estimator. Here the estimation problem is redefined to include parameters for marginal hazard rates, and for double failure hazard rates only at informative uncensored failure time grid points where there is pertinent empirical information. Double failure hazard rates at other grid points in the risk region are specified rather than estimated. With this approach the nonparametric maximum likelihood estimator is unique, and can be calculated using a two-step procedure. The first step involves setting aside all doubly censored observations that are interior to the risk region. The nonparametric maximum likelihood estimator from the remaining data turns out to be the Dabrowska (1988) estimator. The omitted doubly censored observations are included in the procedure in the second stage using self-consistency, resulting in a noniterative nonparametric maximum likelihood estimator for the bivariate survivor function. Simulation evaluation and asymptotic distributional results are provided. Moderate sample size efficiency for the survivor function nonparametric maximum likelihood estimator is similar to that for the Dabrowska estimator as applied to the entire dataset, while some useful efficiency improvement arises for the corresponding distribution function estimator, presumably due to the avoidance of negative mass assignments.

Incidence of milk alkali syndrome in the Women's Health Initiative clinical trial and cohort study: response to Neupane.

Abstract: Dear Editor, We thank Dr. Neupane for her letter [1] on our report on calcium and vitamin D supplementation in the Women's Health Initiative (WHI) [2]. Though we did not collect information on the incidence of the rather common milk alkali syndrome, women in the WHI calcium plus vitamin D (CaD) randomized trial were queried twice a year, during the average 7-year intervention period, concerning the occurrence of hypercalcemia and concerning the initiation of kidney dialysis. A total of 51 intervention group and 52 placebo group women reported initiating dialysis during trial follow-up. Our regression analyses that stratify on 5-year baseline age, on randomization assignment in the WHI Hormone Therapy (HT) and Dietary Modification (DM) trials, and on baseline history of kidney stones yield a kidney dialysis hazard ratio (95 % confidence interval) of 0.98 (0.66, 1.44), with no evidence (p=0.72) of interaction with personal supplement use. In comparison, incident hypercalcemia was reported by 422 intervention group women compared to 245 placebo group women. The hypercalcemia HR (95 % CI) was 1.73 (1.47, 2.02) from Cox regression analyses that stratified on baseline age, HTand DM randomization group, and baseline history of hypercalcemia. The HR (95 % CI) was 1.83 (1.39, 2.39) among women not taking personal calcium or vitamin D supplements and 1.69 (1.39, 2.06) among personal supplement users. Censoring the follow-up period when a woman first becomes nonadherent to her assigned study pills gives a hypercalcemia HR (95 % CI) of 1.79 (1.45, 2.21). In summary, the 1,000 mg calcium carbonate plus 400 international units of vitamin D3 studied in the WHI clinical trial evidently increases the incidence of hypercalcemia and, as previously reported, kidney stone occurrence, but did not increase the risk of kidney dialysis during trial follow-up.

Measurement Error in Survival Analysis

Abstract: This entry describes various approaches to estimating the regression coefficient in the Cox failure time regression model when some of the regression variables are poorly measured. Regression parameter estimation may be considered if there is a validation subsample within which the “true” regression variable is available, or if there is a reliability subsample within which multiple estimates of the regression variable having independent measurement errors are available. The former context is closely connected to corresponding missing data estimation problems. Simple ad hoc estimators are reviewed, along with effective but inconsistent regression calibration estimators, and more complex nonparametric maximum likelihood and augmented inverse probability weighted estimators. With only a reliability sample available, estimation procedures include regression calibration and an extended risk set regression calibration approach, along with corrected score estimation procedures. Some comments on procedures for measurement error accommodation in an additive hazards model are also provided. This rapidly developing measurement error research area has implications for the analysis of follow-up studies in a number of important application areas. Keywords: corrected score; Cox regression model; regression calibration; reliability sample; measurement error; validation sample

Performance of Food Frequency Questionnaire, 4-d food record and 24-h dietary recall to measure total sugars (TS) against the urinary TS biomarker in postmenopausal women (36.4)

Abstract: Measurement error (ME) in self-reported sugars may be obscuring the true association of sugars with disease risk in epidemiologic studies. We aimed to assess performance of 3 self-report methods to...