Diane Feskanich’s research while affiliated with Brigham and Women's Hospital and other places

Background: Although a number of studies have examined the association between alcohol intake and hip fractures, few have considered specific alcoholic beverages separately. Objectives: We prospectively assessed total alcohol and specific alcoholic beverage consumption and risk of hip fractures in US men and women. Methods: Health, lifestyle information, and hip fractures were self-reported on biennial questionnaires between 1980 and 2014 in 75,180 postmenopausal women from the Nurses' Health Study, and between 1986 and 2014 in 38,398 men aged ≥50 y from the Health Professionals Follow-Up Study. Diet was assessed approximately every 4 y with a semiquantitative FFQ. RRs were computed for hip fracture using Cox proportional hazards models, adjusting for potential confounders. Results: We ascertained 2360 incident low trauma hip fractures in women and 709 in men. Among women, RRs for low trauma hip fractures compared with nondrinkers were 0.89 (95% CI: 0.80, 0.99) for an average daily consumption of <5.0 g, 0.81 (95% CI: 0.70, 0.94) for 5.0 to <10.0 g, 0.83 (95% CI: 0.71, 0.96) for 10.0 to <20.0 g, and 0.93 (95% CI: 0.78, 1.10) for ≥20.0 g. Among men, risk declined linearly with higher alcohol consumption (P-trend = 0.002). Multivariable RR compared with nondrinkers was 0.77 (95% CI: 0.59, 1.01), 0.69 (0.49, 0.96), and 0.67 (0.48, 0.95) for an average intake of 10 g/d to <20 g/d, 20 g/d to <30 g/d, and 30.0 g/d or more, respectively. In women, the alcoholic beverage most significantly associated with hip fracture risk was red wine (RR per serving = 0.59; 95% CI: 0.45, 0.79). In men, there was no clear association with specific alcoholic beverages. Conclusion: In these 2 US cohorts, low to moderate alcohol consumption, when compared with no consumption, was associated with a lower risk of hip fractures, particularly with red wine consumption among women.

Importance Vitamin supplementation far exceeding recommended doses is popular in segments of the population. However, adverse effects can occur. In a previous secondary analysis of combined data from 2 double-blind randomized clinical trials (RCTs), an unexpected increased risk of hip fracture was found among those treated with high doses of vitamin B6 in combination with vitamin B12. Objectives To study if high intakes of vitamins B6 and B12 from food and supplements were associated with a risk of hip fracture in the Nurses’ Health Study and to investigate whether combined high intakes of both vitamins conferred a particularly increased fracture risk. Design, Setting, and Participants In this prospective cohort study, 75 864 postmenopausal women in the United States were followed up from June 1984 through May 2014. The dates of analysis were July 2016 to June 2018. Information on hip fracture and a wide range of potential confounders was collected at baseline and with biennial follow-up questionnaires. Extensive dietary information was collected approximately every 4 years with a semiquantitative food frequency questionnaire. Relative risks (RRs) were calculated by Cox proportional hazards regression, with cumulative average intakes of vitamins B6 and B12 as main exposures, adjusting for potential confounders. Main Outcome and Measure Hip fracture. Results During follow-up, 2304 of 75 864 women had a hip fracture. Among the women with hip fractures, the median (range) age at hip fracture was 75.8 (46.7-93.0) years and the mean (SD) body mass index (calculated as weight in kilograms divided by height in meters squared) was 24.3 (4.6). Median (interquartile range) cumulative average intakes of total vitamins B6 and B12 were 3.6 (4.8) mg/d and 12.1 (11.7) μg/d, respectively. Both vitamin B6 (RR, 1.29; 95% CI, 1.04-1.59 for an intake of ≥35 vs <2 mg/d; P = .06 for linear trend) and vitamin B12 (RR, 1.25; 95% CI, 0.98-1.58 for an intake of ≥30 vs <5 μg/d; P = .02 for linear trend) were associated with increased fracture risk. Risk was highest in women with a combined high intake of both vitamins (B6 ≥35 mg/d and B12 ≥20 μg/d), exhibiting an almost 50% increased risk of hip fracture (RR, 1.47; 95% CI, 1.15-1.89) compared with women with a low intake of both vitamins (B6 <2 mg/d and B12 <10 μg/d). Conclusions and Relevance In this cohort study, a combined high intake of vitamins B6 and B12 was associated with an increased risk of hip fracture. The intakes were far higher than the recommended dietary allowances. These findings add to previous studies suggesting that vitamin supplements should be used cautiously because adverse effects can occur.

Milk provides energy and nutrients considered protective for bone. Meta-analyses of cohort studies have found no clear association between milk drinking and risk of hip fracture, and results of recent studies are contradictory. We studied the association between milk drinking and hip fracture in Norway, which has a population characterised by high fracture incidence and a high Ca intake. Baseline data from two population-based cohorts were used: the third wave of the Norwegian Counties Study (1985–1988) and the Five Counties Study (2000–2002). Diet and lifestyle variables were self-reported through questionnaires. Height and weight were measured. Hip fractures were identified by linkage to hospital data with follow-up through 2013. Of the 35 114 participants in the Norwegian Counties Study, 1865 suffered a hip fracture during 613 018 person-years of follow-up. In multivariable Cox regression, hazard ratios (HR) per daily glass of milk were 0·97 (95 % CI 0·92, 1·03) in men and 1·02 (95 % CI 0·96, 1·07) in women. Of 23 259 participants in the Five Counties Study, 1466 suffered a hip fracture during 252 996 person-years of follow-up. HR for hip fractures per daily glass of milk in multivariable Cox regression was 0·99 (95 % CI 0·92, 1·07) in men and 1·02 (95 % CI 0·97, 1·08) in women. In conclusion, there was no overall association between milk intake and risk of hip fracture in Norwegian men and women.

Background: To better understand health habits in older nurses versus the general population, we sought to determine whether the demographics, health care utilization, and Medicare spending by the Nurses' Health Study (NHS) participants enrolled in Medicare and a matched sample of Medicare beneficiaries meaningfully differed. Materials and methods: Analytic cohorts included a random 20% sample of Medicare beneficiaries continuously enrolled in fee for service (FFS) Medicare that were propensity matched to the NHS participants continuously enrolled in FFS Medicare in a single year (2012). Matching was based upon preselected demographic factors and health status, using a nearest-neighbor matching algorithm to obtain a 1:1 match without replacement. Healthcare utilization and spending were compared between the two groups; we also stratified findings by number of chronic comorbidities. Results: Similar rates of utilization of primary care and most outpatient services. However, NHS participants had slightly higher rates of cancer screening, specialist care, and inpatient surgery were observed. When stratified by comorbidity status, the largest differences in utilization and spending were found in women with no comorbidity. Conclusions: The modest differences in observed healthcare utilization and spending suggest that older healthcare professionals may access care in fairly similar ways to the general population, and that health status may be a more important determinant of utilization and spending than health profession in older age groups.

Background: Although a number of studies showed a lower risk of hip fractures with high-quality diets, few of them were conducted in the United States. Objective: This prospective analysis examined the association between several diet quality indexes and risk of hip fractures in US men and women. Design: This is a prospective cohort study. Participants/setting: The participants were 74,446 postmenopausal women from the Nurses' Health Study and 36,602 men aged 50 years and older from the Health Professionals Follow-Up Study in the United States. Main outcome measure: Hip fractures were self-reported on biennial questionnaires between 1980-2012 in women, and between 1986-2012 in men. Statistical analysis: Diet was assessed every 4 years with a validated food frequency questionnaire. Relative risks were computed for hip fracture by quintiles of the Alternate Mediterranean Diet score (aMed), the Alternate Healthy Eating Index-2010 (AHEI-2010), and the Dietary Approaches to Stop Hypertension score using Cox proportional hazards models, adjusting for potential confounders. Results: Two thousand one hundred forty-three incident hip fractures in women and 603 in men were reported during follow-up. A significant inverse trend was observed with the cumulative AHEI-2010 score in women (relative risk comparing extreme quintiles 0.87, 95% CI 0.75 to 1.00; P for trend=0.02). There was also a suggestion of an inverse association with the Dietary Approaches to Stop Hypertension score (P for trend=0.03). In addition, significant inverse trends were observed between all three diet quality scores and hip fractures in women younger than age 75 years but not older women. There was no clear association between diet quality indexes and hip fracture in men. Conclusions: Higher AHEI-2010 scores were associated with a lower risk of hip fractures in US women. The inverse associations with diet quality may be more apparent among those younger than age 75 years.

Studies have shown that Raman spectroscopic analysis of fingernail clippings can help differentiate between post-menopausal women who have and who have not suffered a fracture. However, all studies to date have been retrospective in nature, comparing the proteins in nails sourced from women, post-fracture. The objective of this study was to investigate the potential of a prospective test for hip fracture based on spectroscopic analysis of nail tissue. Archived toenail samples from post-menopausal women aged 50 to 63 years in the Nurses’ Health Study were obtained and analysed by Raman spectroscopy. Nails were matched case-controls sourced from 161 women; 82 who underwent a hip fracture up to 20 years after nail collection and 81 age-matched controls. A number of clinical risk factors (CRFs) from the Fracture Risk Assessment (FRAX) tool had been assessed at toenail collection. Using 80% of the spectra, models were developed for increasing time periods between nail collection and fracture. Scores were calculated from these models for the other 20% of the sample and the ability of the score to predict hip fracture was tested in model with and without the CRFs by comparing the odds ratios (ORs) per 1 SD increase in standardised predictive values. The Raman score successfully distinguished between hip fracture cases and controls. With only the score as a predictor, a statistically significant OR of 2.2 (95% confidence interval [CI]: 1.5-3.1) was found for hip fracture for up to 20 years after collection. The OR increased to 3.8 (2.6-5.4) when the CRFs were added to the model. For fractures limited to 13 years after collection, the OR was 6.3 (3.0-13.1) for the score alone. The test based on Raman spectroscopy has potential for identifying individuals who may suffer hip fractures several years in advance. Higher powered studies are required to evaluate the predictive capability of this test.

Background: In an effort to overcome quality and cost constraints inherent in population-based research, diverse data sources are increasingly being combined. In this paper, we describe the performance of a Medicare claims-based incident cancer identification algorithm in comparison with observational cohort data from the Nurses' Health Study (NHS). Methods: NHS-Medicare linked participants' claims data were analyzed using 4 versions of a cancer identification algorithm across 3 cancer sites (breast, colorectal, and lung). The algorithms evaluated included an update of the original Setoguchi algorithm, and 3 other versions that differed in the data used for prevalent cancer exclusions. Results: The algorithm that yielded the highest positive predictive value (PPV) (0.52-0.82) and κ statistic (0.62-0.87) in identifying incident cancer cases utilized both Medicare claims and observational cohort data (NHS) to remove prevalent cases. The algorithm that only used NHS data to inform the removal of prevalent cancer cases performed nearly equivalently in statistical performance (PPV, 0.50-0.79; κ, 0.61-0.85), whereas the version that used only claims to inform the removal of prevalent cancer cases performed substantially worse (PPV, 0.42-0.60; κ, 0.54-0.70), in comparison with the dual data source-informed algorithm. Conclusions: Our findings suggest claims-based algorithms identify incident cancer with variable reliability when measured against an observational cohort study reference standard. Self-reported baseline information available in cohort studies is more effective in removing prevalent cancer cases than are claims data algorithms. Use of claims-based algorithms should be tailored to the research question at hand and the nature of available observational cohort data.

Objectives Barrett’s esophagus (BE) is the only known precursor to esophageal adenocarcinoma. Data examining the association of aspirin with the onset of BE, particularly for women, are scant and conflicting. Methods We leveraged data from 121,700 women enrolled in the Nurses’ Health Study, a large prospective cohort study, who biennially provided detailed information regarding endoscopy and use of aspirin. We used unconditional logistic regression to obtain multivariable (MV)-adjusted odds ratios (ORs) and 95% confidence intervals (CI) to estimate the risk of BE in regular aspirin users (≥2 times/week) compared to non-regular users. Results Among 27,881 women who had undergone upper GI endoscopy, we documented 667 BE cases over 18 years of follow-up. Compared to non-regular users, women who regularly used aspirin had a MV-adjusted OR for BE of 0.85 (95%CI: 0.72, 0.99). The corresponding OR was 0.73 (95%CI: 0.56, 0.96) for BE at least 1 cm long. Compared with women who did not use any aspirin, the MV-adjusted OR for any BE was 0.91 (95% CI, 0.69, 1.20) for women taking 0.5-1.5 tablets/week; 0.92 (95%CI 0.76, 1.11) for 2–5 tablets/week; and 0.71 (95%CI 0.55, 0.92) for ≥6 tablets/week (p-trend=0.01). Compared with non-regular users, the MV-adjusted OR for BE risk was 0.90 (95%CI 0.67, 1.20) for women who regularly used aspirin for 1–5 years, 0.84 (95%CI 0.65, 1.09) for 6–10 years, and 0.81 (95%CI 0.67, 0.97) for >10 years (p-trend=0.03). Conclusion Regular aspirin use was associated with a reduction in the risk of Barrett’s esophagus in women. The reduction in risk appeared related to higher dose and longer duration of use.

Etiological differences between molecular subtypes of breast cancer, defined by estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2), have been examined in previous studies but results are inconsistent, most likely due to small numbers of rarer subtypes. In a recent systematic review of these studies, only six of the possible 44 associations had consistent findings, primarily with the most common luminal A subtype (ER+ or PR+/HER2-). To provide larger numbers from prospective studies for these associations, we utilized a harmonized dataset of nine cohort studies (Cancer Prevention Study-II Nutrition Cohort, Melbourne Cancer Cohort Study, the National Cancer Institute-American Association of Retired Persons cohort study, Nurses’ Health Study, Nurses’ Health Study-2, Prostate, Lung, Colorectal, and Ovarian Cancer Screening cohort, Swedish Mammographic Cohort, Swedish Women’s Lifestyle and Health Study, and Women’s Health Initiative) that had case data on ER, PR, and HER2 from medical records or state tumor registry records. Multivariate, joint Cox proportional hazard regression models were used to calculate hazard ratios (HR) and 95% confidence intervals (CI), and to compare associations across subtypes. Among 616,109 study participants, 11,861 invasive breast cancer cases were diagnosed, including 8,777 luminal A, 1,388 luminal B (ER+ or PR+/ HER2+), 532 HER2-enhancing (ER-/PR-/HER2+), and 1,164 triple negative (ER-/PR-/HER2-) subtypes. The mean age at baseline of the study participants was 54.8 (SD=12.2) years for cases and 54.2 (SD=14.4) years for controls, and the majority self-described as white. We found that the associations of race, benign breast disease, parity, number of live births, age at first birth, age at menopause, alcohol consumption, and smoking initiation relative to first birth were statistically-significantly different for at least one of the molecular subtypes compared to the associations for the ER+ or PR+/HER2- subtype (p-value for tumor homogeneity<0.05). For example, women who drank ≥2 drinks/ day were at increased risk of the luminal A-like (HR=1.44, 95% CI 1.28 - 1.63) and luminal B-like subtypes (HR=1.33, 95% CI 0.97 - 1.81), but not HER2-enhancing (HR=0.92, 95% CI 0.52 - 1.62) or triple negative subtypes (HR=0.89, 95% CI 0.61 - 1.30; p-value for homogeneity<0.001). Younger age at menarche, first degree family history of breast cancer, ever use of oral contraceptives, ever use of menopausal hormones, higher education, and postmenopausal body mass index were positively associated with all molecular subtypes. Our results highlight the importance of considering molecular subtype when evaluating breast cancer associations by risk factors, and provide precise estimates of association for the development of molecular subtype-specific risk prediction models. Citation Format: Mia M. Gaudet, Gretchen Gierach, Brian Carter, Juhua Luo, Roger Milne, Elisabete Weiderpass, Graham Giles, Wendy Chen, Rulla Tamimi, Heather Eliassen, Diane Feskanich, Alicja Wolk, Hans-Olov Adami, Karen Margolis, Susan Gapstur, Montserrat Garcia-Closas, Louise Brinton. Risk factors by molecular subtypes of breast cancer: a pooled analysis of nine cohorts [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2278. doi:10.1158/1538-7445.AM2017-2278

Background: No studies have estimated disability-adjusted life-years (DALYs) lost due to hip fractures using real-life follow-up cohort data. We aimed to quantify the burden of disease due to incident hip fracture using DALYs in prospective cohorts in the CHANCES consortium, and to calculate population attributable fractions based on DALYs for specific risk factors. Methods: We used data from six cohorts of participants aged 50 years or older at recruitment to calculate DALYs. We applied disability weights proposed by the National Osteoporosis Foundation and did a series of sensitivity analyses to examine the robustness of DALY estimates. We calculated population attributable fractions for smoking, body-mass index (BMI), physical activity, alcohol intake, type 2 diabetes and parity, use of hormone replacement therapy, and oral contraceptives in women. We calculated summary risk estimates across cohorts with pooled analysis and random-effects meta-analysis methods. Findings: 223 880 men and women were followed up for a mean of 13 years (SD 6). 7724 (3·5%) participants developed an incident hip fracture, of whom 413 (5·3%) died as a result. 5964 DALYs (27 per 1000 individuals) were lost due to hip fractures, 1230 (20·6%) of which were in the group aged 75–79 years. 4150 (69·6%) DALYs were attributed to disability. Current smoking was the risk factor responsible for the greatest hip fracture burden (7·5%, 95% CI 5·2–9·7) followed by physical inactivity (5·5%, 2·1–8·5), history of diabetes (2·8%, 2·1–4·0), and low to average BMI (2·0%, 1·4–2·7), whereas low alcohol consumption (0·01–2·5 g per day) and high BMI had a protective effect. Interpretation: Hip fracture can lead to a substantial loss of healthy life-years in elderly people. National public health policies should be strengthened to reduce hip fracture incidence and mortality. Primary prevention measures should be strengthened to prevent falls, and reduce smoking and a sedentary lifestyle. Funding: European Community's Seventh Framework Programme.