Two Randomized Trials of Low-Dose Calcium Supplementation in Pregnancy.

BACKGROUND: The World Health Organization recommends 1500 to 2000 mg of calcium daily as supplementation, divided into three doses, for pregnant persons in populations with low dietary calcium intake in order to reduce the risk of preeclampsia. The complexity of the dosing scheme, however, has led to implementation barriers. METHODS: We conducted two independent randomized trials of calcium supplementation, in India and Tanzania, to assess the noninferiority of a 500-mg daily dose to a 1500-mg daily dose of calcium supplementation. In each trial, the two primary outcomes were preeclampsia and preterm birth, and the noninferiority margins for the relative risks were 1.54 and 1.16, respectively. RESULTS: A total of 11,000 nulliparous pregnant women were included in each trial. The cumulative incidence of preeclampsia was 3.0% in the 500-mg group and 3.6% in the 1500-mg group in the India trial (relative risk, 0.84; 95% confidence interval [CI], 0.68 to 1.03) and 3.0% and 2.7%, respectively, in the Tanzania trial (relative risk, 1.10; 95% CI, 0.88 to 1.36) - findings consistent with the noninferiority of the lower dose in both trials. The percentage of live births that were preterm was 11.4% in the 500-mg group and 12.8% in the 1500-mg group in the India trial (relative risk, 0.89; 95% CI, 0.80 to 0.98), which was within the noninferiority margin of 1.16; in the Tanzania trial, the respective percentages were 10.4% and 9.7% (relative risk, 1.07; 95% CI, 0.95 to 1.21), which exceeded the noninferiority margin. CONCLUSIONS: In these two trials, low-dose calcium supplementation was noninferior to high-dose calcium supplementation with respect to the risk of preeclampsia. It was noninferior with respect to the risk of preterm live birth in the trial in India but not in the trial in Tanzania. (Funded by the Bill and Melinda Gates Foundation and others; ClinicalTrials.gov number, NCT03350516; Clinical Trials Registry-India number, CTRI/2018/02/012119; and Tanzania Medicines and Medical Devices Authority Trials Registry number, TFDA0018/CTR/0010/5).

24-Hour Urinary Sodium and Potassium Excretion and Cardiovascular Risk.

BACKGROUND: The relation between sodium intake and cardiovascular disease remains controversial, owing in part to inaccurate assessment of sodium intake. Assessing 24-hour urinary excretion over a period of multiple days is considered to be an accurate method. METHODS: We included individual-participant data from six prospective cohorts of generally healthy adults; sodium and potassium excretion was assessed with the use of at least two 24-hour urine samples per participant. The primary outcome was a cardiovascular event (coronary revascularization or fatal or nonfatal myocardial infarction or stroke). We analyzed each cohort using consistent methods and combined the results using a random-effects meta-analysis. RESULTS: Among 10,709 participants, who had a mean (±SD) age of 51.5±12.6 years and of whom 54.2% were women, 571 cardiovascular events were ascertained during a median study follow-up of 8.8 years (incidence rate, 5.9 per 1000 person-years). The median 24-hour urinary sodium excretion was 3270 mg (10th to 90th percentile, 2099 to 4899). Higher sodium excretion, lower potassium excretion, and a higher sodium-to-potassium ratio were all associated with a higher cardiovascular risk in analyses that were controlled for confounding factors (P≤0.005 for all comparisons). In analyses that compared quartile 4 of the urinary biomarker (highest) with quartile 1 (lowest), the hazard ratios were 1.60 (95% confidence interval [CI], 1.19 to 2.14) for sodium excretion, 0.69 (95% CI, 0.51 to 0.91) for potassium excretion, and 1.62 (95% CI, 1.25 to 2.10) for the sodium-to-potassium ratio. Each daily increment of 1000 mg in sodium excretion was associated with an 18% increase in cardiovascular risk (hazard ratio, 1.18; 95% CI, 1.08 to 1.29), and each daily increment of 1000 mg in potassium excretion was associated with an 18% decrease in risk (hazard ratio, 0.82; 95% CI, 0.72 to 0.94). CONCLUSIONS: Higher sodium and lower potassium intakes, as measured in multiple 24-hour urine samples, were associated in a dose-response manner with a higher cardiovascular risk. These findings may support reducing sodium intake and increasing potassium intake from current levels. (Funded by the American Heart Association and the National Institutes of Health.).

Identifying modifiable risk factors to prevent aggressive colorectal cancer.

It remains unclear if pre-diagnostic factors influence the developmental pathways of colorectal cancer (CRC) that could enhance tumor aggressiveness. This study used prospective data from 205,489 cancer-free US health professionals to investigate the associations of 31 known or putative risk factors with the risk of aggressive CRC. Tumor aggressiveness was characterized by three endpoints: aggressive CRC (cancer that causes death within 5 years of diagnosis), fatal CRC, and tumor stage at diagnosis. The data augmentation method was used to assess the difference in the associations between risk factors and endpoints. We documented 3201 CRC cases, of which 899 were aggressive. The protective associations of undergoing lower endoscopy (hazard ratios [HR] 0.43, 95% confidence interval (CI) 0.37, 0.49 for aggressive versus HR 0.61, 95% CI 0.56, 0.67 for non-aggressive) and regular use of aspirin (HR 0.70, 95% CI 0.61, 0.81 versus HR 0.84, 95% CI 0.77, 0.92) were stronger for aggressive than non-aggressive CRC (pHeterogeneity <0.05). Lower intake of whole grains or cereal fiber and greater dietary inflammatory potential were associated with a higher risk of aggressive but not non-aggressive CRC. The remaining risk factors showed comparable associations with aggressive CRC and non-aggressive CRC. Aggressive cases were more likely to have KRAS-mutated tumors but less likely to have distal or MSI-high tumors (p < .007). Similar results were observed for fatal CRC and advanced tumor stages at diagnosis. These findings provide initial evidence for the role of pre-diagnostic risk factors in the pathogenesis of aggressive CRC and suggest research priorities for preventive interventions.

All-cause and cause-specific mortality risk and loss in life expectancy associated with incident type 2 diabetes onset age and duration.

BACKGROUND: Evidence on type 2 diabetes onset age and duration on mortality risk has been limited by short follow-up, inadequate control for confounding, missing repeated measurements, and inability to cover the full range of onset age, duration, and major causes of death. Moreover, scarce data dissect how type 2 diabetes onset age and duration shape life expectancy. METHODS: We evaluate prospectively these topics based on 270,075 eligible participants in the Nurses' Health Studies and Health Professionals Follow-up Study, leveraging repeated measurements throughout up to 40 years of follow-up. Cox models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). RESULTS: In fully adjusted analyses, incident early onset type 2 diabetes (diagnosed <40 years of age) was associated with significantly higher mortality from all-causes (HR, 95% CI was 3.16, 2.64-3.79; vs. individuals without type 2 diabetes), cardiovascular disease (6.56, 4.27-10.1), respiratory disease (3.43, 1.38-8.51), neurodegenerative disease (5.13, 2.09-12.6), and kidney disease (8.55, 1.98-36.9). The relative risk elevations declined dramatically with each higher decade of age at diagnosis for deaths from most of these causes, though the absolute risk difference increased continuously. A substantially higher cumulative incidence of mortality and a greater loss in life expectancy were associated with younger age at type 2 diabetes diagnosis. Longer disease duration was associated with generally higher relative and absolute risk of mortality. CONCLUSION: Early onset of type 2 diabetes and longer disease duration are associated with substantially increased risk of all-cause and cause-specific mortality and greater loss in life expectancy.

Causal Selection of Covariates in Regression Calibration for Mismeasured Continuous Exposure.

Regression calibration as developed by Rosner, Spiegelman, and Willett is used to adjust the bias in effect estimates due to measurement error in continuous exposures. The method involves two models: a measurement error model relating the mismeasured exposure to the true (or gold-standard) exposure and an outcome model relating the mismeasured exposure to the outcome. However, no comprehensive guidance exists for determining which covariates should be included in each model. In this article, we investigate the selection of the minimal and most efficient covariate adjustment sets under a causal inference framework. We show that to address the measurement error, researchers must adjust for, in both measurement error and outcome models, any common causes (1) of true exposure and the outcome and (2) of measurement error and the outcome. We also show that adjusting for so-called prognostic variables that are independent of true exposure and measurement error in the outcome model, may increase efficiency, while adjusting for any covariates that are associated only with true exposure generally results in efficiency loss in realistic settings. We apply the proposed covariate selection approach to the Health Professional Follow-up Study dataset to study the effect of fiber intake on cardiovascular disease. Finally, we extend the originally proposed estimators to a nonparametric setting where effect modification by covariates is allowed.

21-Gene Recurrence Score and Survival Outcomes in the Phase III Multicenter TAILORx Clinical Trial.

BACKGROUND: Recurrence score (RS) based on a 21-gene genomic assay is frequently used to estimate risk of distant recurrence for choice of adjuvant chemotherapy in breast cancer. It remains unclear whether RS is an independent prognostic factor for breast cancer-specific survival (BCSS) and overall survival (OS) in the TAILORx trial population. METHODS: We evaluated the association of RS with BCSS and OS plus recurrence-free interval (RFI) and invasive disease-free survival (DFS) using multivariable Cox proportional hazards regression analysis, adjusting for clinicopathologic measures, in 8,916 patients with hormone receptor-positive, HER2-negative, node-negative breast cancer. Likelihood ratio (LR) test was used to assess the relative amount of prognostic information provided by RS to BCSS, OS, RFI, and DFS, comparatively. RESULTS: Event rates for BCSS, OS, RFI, and DFS were 1.7%, 5.2%, 5.6%, and 12.6%, respectively, by up to 11.6 years of follow-up. Compared with low-range RS (0-10), patients with midrange (11-25) and high-range (26-100) RS had inferior BCSS (adjusted hazard ratio [aHR], 5.12 [95% CI, 2.09-16.92] and 8.03 [95% CI, 2.91-28.47], respectively) and RFI (aHR, 1.68 [95% CI, 1.23-2.36] and 3.05 [95% CI, 2.02-4.67], respectively), independent of clinicopathologic factors. High-range score was associated with an increased risk of DFS (aHR, 1.56 [95% CI, 1.20-2.04]) but not significantly associated with OS (aHR, 1.44 [95% CI, 0.95-2.18]). Midrange score was associated with neither DFS (aHR, 1.15 [95% CI, 0.96-1.38]) nor OS (HR 1.14 [95% CI, 0.87-1.52]). LR-χ2 values were 83.0 and 65.1 for RFI and BCSS, respectively, and 17.5 and 33.6 for OS and DFS, respectively (P<.0001). CONCLUSIONS: RS is an independent measure for BCSS and recurrence prognoses relative to OS in early-stage breast cancer. It carries more prognostic information for breast cancer-specific outcomes.

Soft classification and regression analysis of audiometric phenotypes of age-related hearing loss.

Age-related hearing loss has a complex etiology. Researchers have made efforts to classify relevant audiometric phenotypes, aiming to enhance medical interventions and improve hearing health. We leveraged existing pattern analyses of age-related hearing loss and implemented the phenotype classification via quadratic discriminant analysis (QDA). We herein propose a method for analyzing the exposure effects on the soft classification probabilities of the phenotypes via estimating equations. Under reasonable assumptions, the estimating equations are unbiased and lead to consistent estimators. The resulting estimator had good finite sample performances in simulation studies. As an illustrative example, we applied our proposed methods to assess the association between a dietary intake pattern, assessed as adherence scores for the dietary approaches to stop hypertension diet calculated using validated food-frequency questionnaires, and audiometric phenotypes (older-normal, metabolic, sensory, and metabolic plus sensory), determined based on data obtained in the Nurses' Health Study II Conservation of Hearing Study, the Audiology Assessment Arm. Our findings suggested that participants with a more healthful dietary pattern were less likely to develop the metabolic plus sensory phenotype of age-related hearing loss.

Weighting estimation in the cause-specific Cox regression with partially missing causes of failure.

Complex diseases are often analyzed using disease subtypes classified by multiple biomarkers to study pathogenic heterogeneity. In such molecular pathological epidemiology research, we consider a weighted Cox proportional hazard model to evaluate the effect of exposures on various disease subtypes under competing-risk settings in the presence of partially or completely missing biomarkers. The asymptotic properties of the inverse and augmented inverse probability-weighted estimating equation methods are studied with a general pattern of missing data. Simulation studies have been conducted to demonstrate the double robustness of the estimators. For illustration, we applied this method to examine the association between pack-years of smoking before the age of 30 and the incidence of colorectal cancer subtypes defined by a combination of four tumor molecular biomarkers (statuses of microsatellite instability, CpG island methylator phenotype, BRAF mutation, and KRAS mutation) in the Nurses' Health Study cohort.

New Quadratic Discriminant Analysis Algorithms for Correlated Audiometric Data.

Paired organs like eyes, ears, and lungs in humans exhibit similarities, and data from these organs often display remarkable correlations. Accounting for these correlations could enhance classification models used in predicting disease phenotypes. To our knowledge, there is limited, if any, literature addressing this topic, and existing methods do not exploit such correlations. For example, the conventional approach treats each ear as an independent observation when predicting audiometric phenotypes and is agnostic about the correlation of data from the two ears of the same person. This approach may lead to information loss and reduce the model performance. In response to this gap, particularly in the context of audiometric phenotype prediction, this paper proposes new quadratic discriminant analysis (QDA) algorithms that appropriately deal with the dependence between ears. We propose two-stage analysis strategies: (1) conducting data transformations to reduce data dimensionality before applying QDA; and (2) developing new QDA algorithms to partially utilize the dependence between phenotypes of two ears. We conducted simulation studies to compare different transformation methods and to assess the performance of different QDA algorithms. The empirical results suggested that the transformation may only be beneficial when the sample size is relatively small. Moreover, our proposed new QDA algorithms performed better than the conventional approach in both person-level and ear-level accuracy. As an illustration, we applied them to audiometric data from the Medical University of South Carolina Longitudinal Cohort Study of Age-related Hearing Loss. In addition, we developed an R package, PairQDA, to implement the proposed algorithms.

Hypothesis Tests for Continuous Audiometric Threshold Data.

OBJECTIVES: Hypothesis tests for hearing threshold data may be challenging due to the special structure of the response variable, which consists of the measurements from the participant's two ears at multiple frequencies. The commonly-used methods may have inflated type I error rates for the global test that examines whether exposure-hearing threshold associations exist in at least one of the frequencies. We propose using both-ear methods, including all frequencies in the same model for hypothesis testing. DESIGN: We compared the both-ear method to commonly used single-ear methods, such as the worse-ear, better-ear, left/right-ear, average-ear methods, and both-ear methods that evaluate individual audiometric frequencies in separate models, through both theoretical consideration and a simulation study. Differences between the methods were illustrated using hypothesis tests for the associations between the Dietary Approaches to Stop Hypertension adherence score and 3-year change in hearing thresholds among participants in the Conservation of Hearing Study. RESULTS: We found that (1) in the absence of ear-level confounders, the better-ear, worse-ear and left/right-ear methods have less power for frequency-specific tests and for the global test; (2) in the presence of ear-level confounders, the better-ear and worse-ear methods are invalid, and the left/right-ear and average-ear methods have less power, with the power loss in the left/right-ear much greater than the average-ear method, for frequency-specific tests and for the global test; and (3) the both-ear method with separate analyses for individual frequencies is invalid for the global test. CONCLUSIONS: For hypothesis testing to evaluate whether there are significant associations between an exposure of interest and audiometric hearing threshold measurements, the both-ear method that includes all frequencies in the same model is the recommended analytic approach.

Risk factors for inadequate and excessive gestational weight gain in 25 low- and middle-income countries: An individual-level participant meta-analysis.

BACKGROUND: Many women experience suboptimal gestational weight gain (GWG) in low- and middle-income countries (LMICs), but our understanding of risk factors associated with GWG in these settings is limited. We investigated the relationships between demographic, anthropometric, lifestyle, and clinical factors and GWG in prospectively collected data from LMICs. METHODS AND FINDINGS: We conducted an individual participant-level meta-analysis of risk factors for GWG outcomes among 138,286 pregnant women with singleton pregnancies in 55 studies (27 randomized controlled trials and 28 prospective cohorts from 25 LMICs). Data sources were identified through PubMed, Embase, and Web of Science searches for articles published from January 2000 to March 2019. Titles and abstracts of articles identified in all databases were independently screened by 2 team members according to the following eligibility criteria: following inclusion criteria: (1) GWG data collection took place in an LMIC; (2) the study was a prospective cohort or randomized trial; (3) study participants were pregnant; and (4) the study was not conducted exclusively among human immunodeficiency virus (HIV)-infected women or women with other health conditions that could limit the generalizability of the results. The Institute of Medicine (IOM) body mass index (BMI)-specific guidelines were used to determine the adequacy of GWG, which we calculated as the ratio of the total observed weight gain over the mean recommended weight gain. Study outcomes included severely inadequate GWG (percent adequacy of GWG <70), inadequate GWG (percent adequacy of GWG <90, inclusive of severely inadequate), and excessive GWG (percent adequacy of GWG >125). Multivariable estimates from each study were pooled using fixed-effects meta-analysis. Study-specific regression models for each risk factor included all other demographic risk factors measured in a particular study as potential confounders, as well as BMI, maternal height, pre-pregnancy smoking, and chronic hypertension. Risk factors occurring during pregnancy were further adjusted for receipt of study intervention (if any) and 3-month calendar period. The INTERGROWTH-21st standard was used to define high and low GWG among normal weight women in a sensitivity analysis. The prevalence of inadequate GWG was 54%, while the prevalence of excessive weight gain was 22%. In multivariable models, factors that were associated with a higher risk of inadequate GWG included short maternal stature (<145 cm), tobacco smoking, and HIV infection. A mid-upper arm circumference (MUAC) of ≥28.1 cm was associated with the largest increase in risk for excessive GWG (risk ratio (RR) 3.02, 95% confidence interval (CI) [2.86, 3.19]). The estimated pooled difference in absolute risk between those with MUAC of ≥28.1 cm compared to those with a MUAC of 24 to 28.09 cm was 5.8% (95% CI 3.1% to 8.4%). Higher levels of education and age <20 years were also associated with an increased risk of excessive GWG. Results using the INTERGROWTH-21st standard among normal weight women were similar but attenuated compared to the results using the IOM guidelines among normal weight women. Limitations of the study's methodology include differences in the availability of risk factors and potential confounders measured in each individual dataset; not all risk factors or potential confounders of interest were available across datasets and data on potential confounders collected across studies. CONCLUSIONS: Inadequate GWG is a significant public health concern in LMICs. We identified diverse nutritional, behavioral, and clinical risk factors for inadequate GWG, highlighting the need for integrated approaches to optimizing GWG in LMICs. The prevalence of excessive GWG suggests that attention to the emerging burden of excessive GWG in LMICs is also warranted.

Adolescent dietary patterns in relation to later prostate cancer risk and mortality in the NIH-AARP Diet and Health Study.

BACKGROUND: Although adolescent diet has been proposed to contribute to prostate cancer (PCa) development, no studies have investigated the relation between adolescent dietary patterns and PCa risk or mortality. METHODS: Using data from 164,079 men in the NIH-AARP Diet and Health Study, we performed factor analysis to identify dietary patterns at ages 12-13 years and then used Cox proportional hazards regression to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) of total (n = 17,861), non-advanced (n = 15,499), advanced (n = 2362), and fatal PCa (n = 832). RESULTS: Although not entirely consistent across analyses, a higher adolescent plant-based pattern (characterised by vegetables, fruits, and dark bread) score was associated with slightly reduced risks of total (fully adjusted HRQ5vs.Q1 = 0.93, 95% CI: 0.89-0.98, p trend=0.003) and non-advanced PCa (HR = 0.91, 95% CI: 0.87-0.96, p trend<0.001), whereas no associations were observed for advanced or fatal PCa, or for Western modern (characterised by sweets, processed meat, beef, cheese, and pizza) or Western traditional (characterised gravy, eggs, potatoes and white bread) patterns. CONCLUSION: We found evidence to support a modest, protective role for a plant-based dietary pattern during adolescence on PCa risk. If confirmed in future studies, our findings may help to inform the development of new, primary prevention strategies for PCa.

Different operationalizations of the 2018 WCRF/AICR cancer prevention recommendations and risk of cancer.

BACKGROUND: The standardized scoring system assessing adherence to the 2018 World Cancer Research Fund (WCRF)/American Institute for Cancer Research (AICR) cancer prevention recommendations assigns equal weight for each recommendation, thereby giving higher weight to dietary factors collectively (5 points) than adiposity (1 point) and physical activity (1 point). An alternative score assigning equal weights to the adiposity, physical activity, alcohol, and other dietary (composite) recommendations may better predict cancer associations. METHODS: We examined associations between standardized and alternative scores with cancer risk in two US prospective cohorts. Multivariable-adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using Cox regression. RESULTS: During 28 years of follow-up, 16,342 incident cancer cases in women and 8729 cases in men occurred. Individuals in the highest versus lowest quintile of the standardized score had a reduced overall cancer risk (women: HR = 0.89, 95% CI: 0.85, 0.94; men: HR = 0.87, 95% CI: 0.81, 0.94). Results were slightly stronger for the alternative score (women: HR = 0.83, 95% CI: 0.79, 0.87; men: HR = 0.81, 95% CI: 0.75, 0.86). Similar patterns were observed for obesity-related, alcohol-related, smoking-related, and digestive system cancers. CONCLUSIONS: Greater adherence to the WCRF/AICR cancer prevention recommendations was associated with lower cancer risk. A score assigning equal weights to the adiposity, physical activity, alcohol, and all remaining diet components yielded stronger associations than the standardized score.

CUL4B enhances the malignant phenotype of esophageal squamous cell carcinoma by suppressing TGFBR3 expression.

Cullin 4B (CUL4B), which acts as a scaffold protein in CUL4B-RING ubiquitin ligase complexes (CRL4B), is frequently overexpressed in cancer and represses tumor suppressors through epigenetic mechanisms. However, the expression and function of CUL4B in esophageal squamous cell carcinoma (ESCC) have not been well illustrated. In this study, we show that upregulation of CUL4B in ESCC cells enhances proliferation, invasion and cisplatin (CDDP)-resistance, while knockdown of CUL4B significantly represses the malignant activities. Mechanistically, we demonstrate that CUL4B promotes proliferation and migration of ESCC cells through inhibiting expression of transforming growth factor beta receptor III (TGFBR3). CRL4B complex binds to the promoter of TGFBR3, and represses its transcription by catalyzing monoubiquitination at H2AK119 and coordinating with PRC2 and HDAC complexes. Taken together, our findings establish a critical role for the CUL4B/TGFBR3 axis in the regulation of ESCC malignancy.

Marginal structural models for multilevel clustered data.

Marginal structural models (MSMs), which adopt inverse probability treatment weighting in the estimating equations, are powerful tools to estimate the causal effects of time-varying exposures in the presence of time-dependent confounders. Motivated by the Conservation of Hearing Study (CHEARS) Audiology Assessment Arm (AAA) where repeated hearing measurements were clustered by study participants, time, and testing sites, we propose two methods to account for the multilevel correlation structure when fitting the MSMs. The first method directly models the covariance of the repeated outcomes when solving the weighted generalized estimating equations for MSMs, while the second two-stage analysis approach fits cluster-specific MSMs first and then combines the estimated parameters using mixed-effects meta-analysis. Finite sample simulation results suggest that our methods can obtain less biased and more efficient estimates of the parameters by accounting for the multilevel correlation. Moreover, we explore the effects of using fixed- or mixed-effects model to estimate the treatment probability on the parameter estimates of the MSMs in the presence of unmeasured cluster-level confounders. Lastly, we apply our methods to the CHEARS AAA data set, to estimate the causal effects of aspirin use on hearing loss.

Iron status among HIV-infected adults during the first year of antiretroviral therapy in Tanzania.

BACKGROUND: The influence of inflammation on iron status among people living with HIV (PLWHIV) has not been well explored. We evaluated the trajectory of iron status among PLWHIV during the first year of highly active antiretroviral therapy (HAART), compared alternative approaches for inflammation correction, and assessed the associations of iron status with HIV-1 viral load and anthropometric outcomes. METHODS: We conducted a secondary analysis of data from a randomized trial among 400 adults initiating HAART in Tanzania. Ferritin and C-reactive protein (CRP) were measured at baseline, 1, 6 or 12 months. Ferritin was considered in four ways: unadjusted, and adjusted for inflammation using higher cut-off (HC), Thurnham-corrected (TC) and regression-corrected (RC) approaches. For unadjusted, TC and RC ferritin, iron deficiency (ID) was defined using ferritin < 15 µg/L and elevated iron status was defined using ferritin > 150 µg/L among females and > 200 µg/L among males. For HC ferritin, elevated iron status was defined based on serum ferritin > 500 µg/L, while ID was defined using ferritin < 70 µg/L in the presence of inflammation and < 15 µg/L in the absence of inflammation. Regression models evaluated the trajectory of ferritin concentration across categories of baseline characteristics, and assessed the association of iron status with viral and anthropometric outcomes. RESULTS: The prevalence of iron deficiency at HAART initiation was 9% for unadjusted, 17% for HC, 12% for TC and 22% for RC ferritin. The prevalence of elevated iron status was 42% for unadjusted, 18% for HC, 31% for TC, and 15% for RC ferritin. The prevalence of iron deficiency for all three methods increased during the first year of HAART, while the prevalence of elevated iron status decreased. Baseline elevated iron status defined using HC ferritin was associated with a greater risk of HIV-1 viral load > 1000 copies/mL [relative risk (RR) = 4.29, 95% CI: 1.38-13.3] and incidence of being underweight [body mass index (BMI) < 18.5 kg/m2 , hazard ratio (HR) = 3.65, 95% confidence interval (CI): 1.38-9.67]. Neither baseline-elevated iron status defined using TC or RC ferritin nor baseline iron deficiency defined using any of the three methods was associated with HIV-1 viral load or anthropometric outcomes. CONCLUSIONS: Whether and how inflammation correction is done influences findings of studies of iron status among PLWHIV.

Long-Term Consumption of Nuts (Including Peanuts, Peanut Butter, Walnuts, and Other Nuts) in Relation to Risk of Frailty in Older Women: Evidence from a Cohort Study.

BACKGROUND: Adherence to a healthy diet is inversely associated with frailty. However, the relationship between nuts, a key food group of Mediterranean diet, and frailty is unclear. OBJECTIVES: This study aimed to evaluate the association between nut consumption and frailty in an aging female population. METHODS: This population-based observational study included nonfrail women (≥60 y old) in the NHS from 11 states of the United States. Outcome was incident frailty, defined as having ≥3 of the FRAIL components (fatigue, lower strength, reduced aerobic capacity, multiple chronic conditions, and significant weight loss) and assessed every 4 y from 1992 to 2016. From 1990 to 2014, FFQs were used to assess the intakes of peanuts, peanut butter, walnuts (added in 1998), and other nuts at 4-y intervals. Exposure was total nut consumption, calculated as the sum of intakes of peanuts, peanut butter, walnuts, and other nuts and categorized into <1 serving/mo, 1-3 servings/mo, 1 serving/wk, 2-4 servings/wk, and ≥5 servings/wk. The relations of intakes of peanuts, peanut butter, and walnuts with frailty were also investigated separately. Cox proportional hazards models were used to assess the associations between nut consumption and frailty after adjusting for age, smoking, BMI, EI, diet quality, and medication use. RESULTS: Among 71,704 participants, 14,195 incident frailty cases occurred over 1,165,290 person-years. The adjusted HR (95% CI) for consuming ≥5 servings/wk of nuts was 0.80 (0.73, 0.87), as compared with <1 serving/mo. Higher intakes of peanuts and walnuts, but not peanut butter, were also inversely associated with frailty. CONCLUSIONS: This large prospective cohort study showed a strong and consistent inverse association between regular nut consumption and incident frailty. This suggests that nut consumption should be further tested as a convenient public health intervention for the preservation of health and well-being in older adults.

Correcting for bias due to mismeasured exposure history in longitudinal studies with continuous outcomes.

Epidemiologists are often interested in estimating the effect of functions of time-varying exposure histories in relation to continuous outcomes, for example, cognitive function. However, the individual exposure measurements that constitute the history upon which an exposure history function is constructed are usually mismeasured. To obtain unbiased estimates of the effects for mismeasured functions in longitudinal studies, a method incorporating main and validation studies was developed. Simulation studies under several realistic assumptions were conducted to assess its performance compared to standard analysis, and we found that the proposed method has good performance in terms of finite sample bias reduction and nominal confidence interval coverage. We applied it to a study of long-term exposure to PM 2.5 $\text{PM}_{2.5}$ , in relation to cognitive decline in the Nurses' Health Study Previously, it was found that the 2-year decline in the standard measure of cognition was 0.018 (95% CI, -0.034 to -0.001) units worse per 10 µ g/m 3 $\mu \text{g/m}^3$ increase in PM 2.5 $\text{PM}_{2.5}$ exposure. After correction, the estimated impact of PM 2.5 $\text{PM}_{2.5}$ on cognitive decline increased to 0.027 (95% CI, -0.059 to 0.005) units lower per 10 µ g/m 3 $\mu \text{g/m}^3$ increase. To put this into perspective, effects of this magnitude are about 2/3 of those found in our data associated with each additional year of aging: 0.044 (95% CI, -0.047 to -0.040) units per 1 year older after applying our correction method.

A repeated measures approach to pooled and calibrated biomarker data.

Participant-level meta-analysis across multiple studies increases the sample size for pooled analyses, thereby improving precision in effect estimates and enabling subgroup analyses. For analyses involving biomarker measurements as an exposure of interest, investigators must first calibrate the data to address measurement variability arising from usage of different laboratories and/or assays. In practice, the calibration process involves reassaying a random subset of biospecimens from each study at a central laboratory and fitting models that relate the study-specific "local" and central laboratory measurements. Previous work in this area treats the calibration process from the perspective of measurement error techniques and imputes the estimated central laboratory value among individuals with only a local laboratory measurement. In this work, we propose a repeated measures method to calibrate biomarker measurements pooled from multiple studies with study-specific calibration subsets. We account for correlation between measurements made on the same person and between measurements made at the same laboratory. We demonstrate that the repeated measures approach provides valid inference, and compare it to existing calibration approaches grounded in measurement error techniques in an example describing the association between circulating vitamin D and stroke.

Analytical method for detecting outlier evaluators.

BACKGROUND: Epidemiologic and medical studies often rely on evaluators to obtain measurements of exposures or outcomes for study participants, and valid estimates of associations depends on the quality of data. Even though statistical methods have been proposed to adjust for measurement errors, they often rely on unverifiable assumptions and could lead to biased estimates if those assumptions are violated. Therefore, methods for detecting potential 'outlier' evaluators are needed to improve data quality during data collection stage. METHODS: In this paper, we propose a two-stage algorithm to detect 'outlier' evaluators whose evaluation results tend to be higher or lower than their counterparts. In the first stage, evaluators' effects are obtained by fitting a regression model. In the second stage, hypothesis tests are performed to detect 'outlier' evaluators, where we consider both the power of each hypothesis test and the false discovery rate (FDR) among all tests. We conduct an extensive simulation study to evaluate the proposed method, and illustrate the method by detecting potential 'outlier' audiologists in the data collection stage for the Audiology Assessment Arm of the Conservation of Hearing Study, an epidemiologic study for examining risk factors of hearing loss in the Nurses' Health Study II. RESULTS: Our simulation study shows that our method not only can detect true 'outlier' evaluators, but also is less likely to falsely reject true 'normal' evaluators. CONCLUSIONS: Our two-stage 'outlier' detection algorithm is a flexible approach that can effectively detect 'outlier' evaluators, and thus data quality can be improved during data collection stage.