Nurses' Health Study Participants Research Articles

Association of body mass index and inflammatory dietary pattern with breast cancer pathologic and genomic immunophenotype in the nurses’ health study

BackgroundBreast tumor immune infiltration is clearly associated with improved treatment response and outcomes in breast cancer. However, modifiable patient factors associated with breast cancer immune infiltrates are poorly understood. The Nurses’ Health Study (NHS) offers a unique cohort to study immune gene expression in tumor and adjacent normal breast tissue, immune cell-specific immunohistochemistry (IHC), and patient exposures. We evaluated the association of body mass index (BMI) change since age 18, physical activity, and the empirical dietary inflammatory pattern (EDIP) score, all implicated in systemic inflammation, with immune cell-specific expression scores.MethodsThis population-based, prospective observational study evaluated 882 NHS and NHSII participants diagnosed with invasive breast cancer with detailed exposure and gene expression data. Of these, 262 women (training cohort) had breast tumor IHC for four classic immune cell markers (CD8, CD4, CD20, and CD163). Four immune cell-specific scores were derived via lasso regression using 105 published immune expression signatures’ association with IHC. In the remaining 620 patient evaluation cohort, we evaluated association of each immune cell-specific score as outcomes, with BMI change since age 18, physical activity, and EDIP score as predictors, using multivariable-adjusted linear regression.ResultsAmong women with paired expression/IHC data from breast tumor tissue, we identified robust correlation between novel immune cell-specific expression scores and IHC. BMI change since age 18 was positively associated with CD4+ (β = 0.16; p = 0.009), and CD163 novel immune scores (β = 0.14; p = 0.04) in multivariable analyses. In other words, for each 10 unit (kg/m2) increase in BMI, the percentage of cells positive for CD4 and CD163 increased 1.6% and 1.4%, respectively. Neither physical activity nor EDIP was significantly associated with any immune cell-specific expression score in multivariable analyses.ConclusionsBMI change since age 18 was positively associated with novel CD4+ and CD163+ cell scores in breast cancer, supporting further study of the effect of modifiable factors like weight gain on the immune microenvironment.

External exposome and sleep outcomes in the Nurses’ Health Study

BACKGROUND AND AIM: Multiple environmental exposures, including greenness, light-at-night, and air pollution, have been associated with sleep. However, these associations have not been investigated using large longitudinal cohorts with repeated measures. We examined the association between multiple environmental exposures and sleep duration and quality in the Nurses’ Health Study (NHS). METHODS: Between 2000 and 2014, 81,202 NHS participants self-reported sleep duration on biennial questionnaires, and in 2000, 81,294 NHS participants self-reported sleep quality. We estimated annual average greenness using Landsat Normalized Difference Vegetation Index (NDVI) in 270m buffers, annual average light-at-night (LAN) from the U.S. Defense Meteorological Satellite Program’s Operational Linescan System, and predicted PM2.5 from validated nationwide spatiotemporal models at each residential address. Due to evidence of non-linearity, we modeled exposures in quintiles. Sleep duration was dichotomized (≥7 hours; <7 hours) and modeled with generalized estimating equations. Sleep quality was dichotomized (difficulty sleeping none/a little/some of the time; difficulty sleeping a good bit of/all the time) and modeled using logistic regression. We estimated odds ratios (OR) of adverse sleep outcomes comparing the highest and lowest quintile and 95% confidence intervals (CI) for each exposure. We adjusted for individual-level risk factors and included all exposures simultaneously for mutual adjustment. RESULTS: In mutually-adjusted models, higher annual average residential greenness exposure was associated with suggestive lower odds of short sleep duration (OR: 0.97 95% CI: 0.94, 1.00) with similar associations for poor sleep quality (OR: 0.97; 95% CI: 0.90, 1.00). Higher exposure to LAN was associated higher odds of with short sleep duration (OR: 1.18 95% CI: 1.14, 1.23) and poorer sleep quality (OR: 1.12; 95% CI: 1.03, 1.22). Higher exposure to PM2.5 was not associated with sleep outcomes. CONCLUSION: In the NHS cohort, lower residential greenness and higher LAN exposures were associated with lower self-reported sleep duration and quality.

Prospective Analyses of Sedentary Behavior in Relation to Risk of Ovarian Cancer.

We examined the association of sedentary behavior with risk of ovarian cancer overall, by tumor subtype, and by participant characteristics in the Nurses' Health Study (NHS) and Nurses' Health Study II (NHS II). A total of 69,558 NHS participants (1992-2016) and 104,130 NHS II participants (1991-2015) who reported on time spent sitting at home, at work, and while watching television were included in the analysis, which included 884 histologically confirmed ovarian cancer cases. Multivariable-adjusted Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for ovarian cancer by sitting time (no mutual adjustment for individual sitting types in primary analyses). We examined potential heterogeneity by tumor histological type (type I or II), body mass index (weight (kg)/height (m)2; < 25 or ≥25), and total physical activity (<15 or ≥15 metabolic equivalent of task-hours/week). We observed an increased risk of ovarian cancer for women who sat at work for 10-19 hours/week (HR=1.25, 95% CI: 1.04, 1.51) and≥20 hours/week (HR=1.40, 95% CI: 1.14, 1.71) versus <5 hours/week. This association did not vary by body mass index, physical activity, or histotype (P for heterogeneity ≥ 0.43). No associations were observed for overall sitting, sitting while watching television, or other sitting at home. Longer sitting time at work was associated with elevated risk of ovarian cancer. Further investigations are required to confirm these findings and elucidate underlying mechanisms.

Splenectomy Results in Venous Thromboembolic Events in Women: A Nurses Health Study

Background:Splenectomy is a therapy for many conditions the most common of which are trauma, Hodgkin lymphoma, thalassemia and hereditary or autoimmune hemolytic anemia, and immune thrombocytopenia (ITP). By 1960 splenectomy had been known to create a risk of over-whelming post splenectomy sepsis (OPSS) with a high mortality rate. More recently population-based studies from Denmark have suggested that there is also an increased risk of thromboembolic (TEE) complications. Thromboembolic events generally include venous thrombosis (deep vein thrombosis (DVT) and pulmonary embolism (PE)) and arterial thrombosis (stroke and myocardial infarction [MI]). ITP and hemolytic anemias have been shown to independently have increased risks of arterial and venous TEE which has complicated assessment of additive effects of splenectomy. While splenectomized patients appear to be at increased risk of arterial and venous thromboses, the incidence and which TEEs predominate remains unclear as does the timing of occurrence of the TEEs.Method:The Nurses' Health Study (NHS) was established in 1976 and enrolled 121,700 female nurses between 30-55 years old. The nurses completed baseline questionnaires and biennial follow up questionnaires. In 2004, NHS participants were asked whether they had undergone splenectomy. The primary outcomes of this study included DVT,PE, and MI, which were identified through self-report on biennial questionnaires. MIs were confirmed through medical record review. End of follow-up for this study went through 2016, date of death, or diagnosis of outcome, whichever came first. Loss to follow up was very low and causes of deaths were carefully tracked.Descriptive statistics compared the splenectomy vs non-splenectomy participants with respect to basic demographic factors and variables that may be related to the outcomes. Cox proportional hazards models were run to evaluate incidence rates for primary and secondary outcomes. We conducted age-adjusted, as well as multivariable adjusted models. Multivariate models were adjusted for age, body mass index, smoking history, diabetes mellitus, high blood pressure, and high cholesterol, thus taking into account known predictors of the specific TEEs. Multivariate models were updated every two years to account for time varying confounders.Results：In 2004, 323 participants reported having had a splenectomy, out of 96,000 completing the questionnaire. There was a strong, significant association between splenectomy and subsequent DVT (n=613) and PE (n=840). The multivariate adjusted HR was 3.73 (95%CI: 1.77-7.86) and 3.80 (95%CI: 2.04-7.10) respectively. When considered together as a composite outcome, the splenectomized participants had a 3.52-fold higher risk (95%CI: 2.12-5.85) of PE/DVT compared with non-splenectomy. There was no association between splenectomy and MI (n=1011; HR = 0.97, 95% CI: 0.31-3.01). Given the very limited number of stroke cases in the splenectomized population, we were underpowered/unable to evaluate the association of splenectomy and subsequent risk of stroke.Conclusion:This study found that splenectomized patients are at a 4-fold increased risk of PE and DVT as compared to not-splenectomized individuals. These findings were not dependent on any of the potentially confounding variable analyzed. We found no association between splenectomy and MI. There are limitations to these results. No information was available on family history and perioperative thrombo-prophylaxis. Also we do not know exactly when or why splenectomy was performed; there were at least 34 women who had splenectomy for ITP but in most cases the underlying disease was unknown. Since the nurses in 1976 were already between 30-55 years old, it is no longer possible to retrospectively obtain sufficient information. Therefore, the splenectomy question has been added to the current survey for NHS#2. We will pursue the relevant information in a follow up letter to respondents who reported having undergone splenectomy while seeing if the findings of NHS#2 confirm those of NHS#1. [Display omitted] DisclosuresBussel: Principia/Sanofi: Consultancy, Membership on an entity's Board of Directors or advisory committees; UCB: Consultancy, Membership on an entity's Board of Directors or advisory committees, Other: DSMB; Momenta/Janssen: Consultancy, Membership on an entity's Board of Directors or advisory committees; Argenx: Consultancy, Membership on an entity's Board of Directors or advisory committees; Dova/Sobi: Consultancy, Membership on an entity's Board of Directors or advisory committees; CSL: Other: DSMB; RallyBio: Consultancy, Membership on an entity's Board of Directors or advisory committees; Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees; Amgen: Consultancy, Membership on an entity's Board of Directors or advisory committees; Rigel: Consultancy, Membership on an entity's Board of Directors or advisory committees; UptoDate: Honoraria.

Early life physical activity and risk of ovarian cancer in adulthood

Emerging data suggest that exposures in early life may affect ovarian development and contribute to ovarian cancer risk. We evaluated the association between early life physical activity and risk of ovarian cancer in adulthood in two large prospective cohorts, the Nurses' Health Study (NHS) and NHSII. In total, analyses included 28 232 NHS participants (followed from 2004 to 2016) and 56 553 NHSII participants (followed from 1997 to 2017). Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for risk of ovarian cancer overall and by early life body mass index (BMI). Neither physical activity at ages 12-13, 14-17 or 18-22 years nor average physical activity across these three periods was associated with ovarian cancer risk overall (≥78 vs <24 MET-h/wk, HRs=1.34, 1.21, 1.08 and 1.24, respectively), or by categories of early life BMI (Pheterogeneity ≥ .44). No association was observed with the risk of high-grade serous or poorly differentiated tumors or postmenopausal ovarian cancer. Overall, early life physical activity was not clearly related to ovarian cancer risk during adulthood.

Utility of Detecting Rare Clonal Hematopoiesis for Predicting Leukemic Transformation in Healthy Individuals

Clonal hematopoiesis is a recognized feature of the aging hematopoietic compartment (Busque L, et al., Blood 2009). Recent next-generation sequencing (NGS) studies have identified the epigenetic regulators DNMT3A and TET2 as genes mutated frequently in clonal hematopoiesis (Busque L, et al., Nat Genet 2012; Xie M, et al., Nat Med 2014; Jaiswal S, et al., NEJM 2014; Genovese G, et al. NEJM 2014). While intrinsically benign, these hematopoietic clones-termed clonal hematopoiesis of indeterminate potential (CHIP)-are detected in approximately 10% of individuals 70 years of age and associated with an increased risk of hematological malignancy and all-cause mortality (Jaiswal S, et al., NEJM 2014; Genovese G, et al., NEJM 2014; Steensma DP, et al., Blood 2015).However, these studies could only detect common clones-greater than 0.02 variant allele fraction (VAF)-due to the intrinsic error rate of NGS. To characterize the prevalence and mutation profile of rare hematopoietic clones below this threshold, we developed methods for error-corrected sequencing (ECS) targeting 54 AML-associated genes, which improved the limit of detection by two orders of magnitude. Using ECS, we identified clonal mutations ranging in frequency from 0.0003-0.1451 VAF in peripheral blood samples banked in 10-year intervals from 20 healthy 50-70 year-old participants in the Nurses' Health Study (NHS), a longitudinal study of >100,000 US registered nurses. Unexpectedly, we identified hematopoietic clones in the peripheral blood of 95% of these healthy individuals (Young AL, et al. Nat Commun 2016), redefining physiologic clonal hematopoiesis of aging. Interestingly, DNMT3A and TET2 mutations comprised 64% of the clonal mutations identified. These clonal mutations were stable longitudinally and present in both myeloid and lymphoid compartments, suggesting that they arose in long-lived hematopoietic stem and progenitor cells.While clonal hematopoiesis harboring AML-associated somatic mutations is ubiquitous in healthy adults, the clonal characteristics that predict leukemic transformation are still unknown. To address this question, we conducted a blinded case-control study with NHS and Health Professionals Follow-up Study (HPFS) participants. We compared rare clonal variants in peripheral blood samples collected pre-diagnosis from 34 individuals who later developed de novo AML, and from 69 healthy controls. Controls were matched to cases for gender, age, ethnicity and time of blood collection (1 case: 2 controls). Using ECS, we identified 598 total clonal variants in 100/103 individuals at 0.0002-0.35 VAF. On average, we found 7.4 clonal variants per case and 5.0 clonal variants per control. The most frequent mutations in both cases and controls occurred in DNMT3A and TET2 . In conditional logistic regression analysis, only DNMT3A R882H/R882C mutations (detected in 8 cases, 3 controls) were significantly associated with future AML risk (OR 6.0, 95% CI 1.3-28.9, p=0.02), with a median of 12.5 years (range 1-22 years) from blood draw to case diagnosis. Recurrent variants not significantly associated with future AML risk were DNMT3A I780T (1 case, 3 controls), DNMT3A R326C (1 case, 3 controls), ASXL1 E1183K (4 cases, 1 control) and JAK2 V617F (4 cases, 1 control).These findings further support that clonal hematopoiesis is a ubiquitous phenomenon in healthy individuals and merely detecting low frequency clonal hematopoiesis may not be a sensitive, specific, or timely predictor for developing AML. Only, DNMT3A R882H and R882C mutations were significantly associated with long-term risk for developing AML in this study. Future studies must identify additional factors that improve AML risk prediction. DisclosuresNo relevant conflicts of interest to declare.

Early Life Exposure to Tobacco Smoke and Ovarian Cancer Risk in Adulthood

Abstract Background: Ovarian cancer risk in adulthood may be affected by early life exposure to tobacco smoke. We investigated this relationship in two large prospective cohorts, the Nurses' Health Study (NHS) and NHSII. Methods: In total, analyses included 110,305 NHS participants (1976–2016) and 112,859 NHSII participants (1989–2017). Self-reported early life smoking exposures were queried at baseline or follow-up questionnaires. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for risk of ovarian cancer overall and by tumor histotype. Results: Compared with women who never smoked, ovarian cancer risk was similar for women who started to smoke at age &amp;lt;18 (HR = 0.98, 95%CI: 0.86–1.11) or ≥18 (HR = 1.02, 95%CI: 0.93–1.12). Overall, ovarian cancer risk was not different among participants whose mother did versus did not smoke during pregnancy (HR = 1.05, 95%CI: 0.87–1.27); however, an increased risk was observed among women who themselves were never smokers (HR = 1.38, 95%CI: 1.05–1.81) but not ever smokers (HR = 0.86, 95%CI: 0.66–1.14; Pheterogeneity = 0.02). These associations did not differ by histotype (Pheterogeneity≥0.35). Parental smoking in the home during childhood/adolescence was related to a 15% increased risk of ovarian cancer in adulthood (HR = 1.15, 95%CI: 1.04–1.27) and this association was notably stronger among women with non-serous/ low- grade serous tumors (HR = 1.28, 95%CI: 1.02–1.61) versus high-grade serous/ poorly differentiated tumors (HR = 1.09, 95%CI: 0.93–1.28, Pheterogeneity = 0.25). Conclusions: Exposure to parental tobacco smoke, but not early initiation of smoking, was associated with a modest elevated risk of ovarian cancer. Further investigations are required to confirm these findings and elucidate underlying mechanisms.

Metabolomic predictors for late-life depression

<h3>Introduction</h3> Depression is a major public health problem among older adults and among the leading causes of global burden of disease and disability. Yet, current understanding of pathophysiology and optimal prevention and/or treatment of late-life depression (LLD) is limited. Advanced approaches involving methods such as metabolomics present an innovative path to identify novel biomarkers of LLD outcomes. Several known biological pathways involving tryptophan and lipid-inflammatory molecules (eicosanoids) are relevant to depression; however, there have been few systematic investigations of metabolomic predictors of long-term trajectories of LLD. <h3>Methods</h3> This is a convenience sample of ∼7000 Nurses' Health Study (NHS) participants with known metabolites and who completed depression items every four years from 1992-2016. Between 1989 and 1990, blood samples were collected and stored under standard conditions. In NHS, data on known metabolites were generated by merging the data from multiple case-control studies; however, all metabolites were profiled using liquid chromatography mass spectrometry in the same laboratory. Continuous metabolite values were transformed to probit scores to account for batch effects. We used a previously developed harmonized depressive symptom scale that was indexed to the five-item Mental Health Inventory (MHI-5, range=0-100 points). For cross-sectional analysis, we performed multivariable logistic regression models to evaluate the associations of baseline metabolite levels and likelihood of the moderate or severe depressive symptoms (MHI-5 cut-off=60); for longitudinal analysis, we used latent growth-curve analysis to identify group-based trajectories of depressive symptoms and then related baseline metabolite levels to long-term trajectories. <h3>Results</h3> The mean age of female nurses at blood draw was 57.0 (range=42.6-70.2) years. There were ∼6300 NHS participants with available data on tryptophan metabolites and ∼1300 NHS participants with available data on eicosanoids. We observed stronger signals for cross-sectional associations of higher circulating levels of tryptophan, eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) with lower risk of moderate or severe depressive symptoms, adjusted for age at blood draw, fasting status, case-control status, physical activity, body mass index, diet index, alcohol use, and cigarette smoking [multivariable odds ratio (95% confidence interval): for tryptophan=0.86 (0.80-0.93), p<0.01; for EPA=0.84 (0.71-1.01), p=0.06; for DHA=0.85 (0.71-1.02), p=0.07]. From 1992-2016, we identified four depressive symptom trajectory groups: minimal depressive symptoms (63%), worsening depressive symptoms (12%), improving depressive symptoms (20%), and persistent severe depressive symptoms (5%). Additional results from multinomial models evaluating the associations between metabolites and long-term LLD trajectories will be presented. <h3>Conclusions</h3> In this sample, tryptophan and omega-3 fatty acid metabolites were significantly related to LLD outcomes; these metabolites may play a role in the pathophysiology of LLD. <h3>Funding</h3> HMS Livingston Fellowship Award

The association of resistance training with risk of ovarian cancer.

ABSTRACTBackgroundIncreasing evidence, including multiple putative inflammatory risk factors (e.g., c‐reactive protein, and adiposity), supports that inflammation plays an important role in ovarian carcinogenesis. Resistance training (RT) is associated with lower levels of circulating inflammatory markers, independent of physical activity.MethodsWe evaluated the relationship between RT and risk of ovarian cancer accounting for other physical activity (e.g., walking) in two large prospective cohorts, the Nurses’ Health Study (NHS) and NHSII.Key ResultsIn total, analyses included 42,005 NHS participants (2000–2016) and 67,289 NHSII participants (2001–2017) with RT assessed every 4 years. Multivariable Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) of RT with ovarian cancer risk overall and by tumor subtype, adjusting for known and putative ovarian cancer risk factors. We identified a total of 609 cases over 1,748,884 person‐years. No association was observed with overall ovarian cancer risk (RT ≥60 vs 0 min/wk, HR = 0.95, 95%CI: 0.74–1.22) or by histotype (comparable HR = 0.86 and 0.98 for type I and II tumors, respectively). Results did not differ by body mass index (Pinteraction = 0.97), or other physical activity (Pinteraction = 0.31).Conclusions & InferencesWe observed no evidence that moderate levels of RT were associated with risk of ovarian cancer. Further investigations are required to confirm these findings.

Early life exposure to tobacco smoke and ovarian cancer risk in adulthood.

Ovarian cancer risk in adulthood may be affected by early life exposure to tobacco smoke. We investigated this relationship in two large prospective cohorts, the Nurses' Health Study (NHS) and NHSII. In total, analyses included 110305 NHS participants (1976-2016) and 112859 NHSII participants (1989-2017). Self-reported early life smoking exposures were queried at baseline or follow-up questionnaires. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for risk of ovarian cancer overall and by tumour histotype. Overall, ovarian cancer risk was not different among participants whose mothers did versus did not smoke during pregnancy (HR = 1.05, 95% CI: 0.87-1.27); however, an increased risk was observed among women who themselves were never smokers (HR = 1.38, 95% CI: 1.05-1.81) but not among ever smokers (HR = 0.86, 95% CI: 0.66-1.14; Pheterogeneity = 0.02). Compared with women who never smoked, ovarian cancer risk was similar for women who started to smoke at age <18 (HR = 0.98, 95% CI: 0.86-1.11) or ≥18 (HR = 1.02, 95% CI: 0.93-1.12). These associations did not differ by histotype (Pheterogeneity ≥0.35). Parental smoking in the home during childhood/adolescence was related to a 15% increased risk of ovarian cancer in adulthood (HR = 1.15, 95% CI: 1.04-1.27) and this association was suggestively stronger among women with non-serous/low-grade serous tumours (HR = 1.28, 95% CI: 1.02-1.61) versus high-grade serous/poorly differentiated tumours (HR = 1.09, 95% CI: 0.93-1.28; Pheterogeneity = 0.25). Exposure to parental tobacco smoke in the home, but not early initiation of smoking, was associated with a modest elevated risk of ovarian cancer. Further investigations are required to confirm these findings and elucidate underlying mechanisms.

High Dietary Intake of Vegetable or Polyunsaturated Fats Is Associated With Reduced Risk of Hepatocellular Carcinoma

We investigated associations of intake of total fats, specific dietary fats, and fats from different food sources with risk of hepatocellular carcinoma (HCC) using data from the Nurses' Health Study (NHS) and the Health Professionals Follow-up Study (HPFS). We analyzed data from a total of 138,483 women and men who participated in the NHS or HPFS. A validated semi-quantitative food frequency questionnaire was sent to NHS participants in 1980, 1984, 1986, and every 4 years thereafter; dietary information was collected from participants in the HPFS in 1986 and every 4 years thereafter. Multivariable hazard ratios (HRs) and 95% CIs were estimated using Cox proportional hazards regression. After an average follow-up time of 26.6 years, 160 incident HCC cases were documented. Although there was a non-significant association between total fat intake and HCC, intake of vegetable fats reduced risk of HCC (HR for the highest vs lowest quartile, 0.61; 95% CI, 0.39-0.96; Ptrend=.02), but not animal or dairy fats. Replacing animal or dairy fats with an equivalent amount of vegetable fats was associated with a lower risk of HCC (HR per 1 standard deviation, 0.79; 95% CI, 0.65-0.97). Among fat subtypes, monounsaturated and polyunsaturated fatty acids, including n-3 (HR, 0.63; 95% CI, 0.41-0.96; Ptrend=.14) and n-6 polyunsaturated fatty acids (HR, 0.54; 95% CI, 0.34-0.86; Ptrend=.02), were inversely associated with risk of HCC. Higher ratios of monounsaturated or polyunsaturated fat to saturated fat were inversely associated with HCC risk (all Ptrend ≤ .02). In addition, when replacing saturated fats with monounsaturated or polyunsaturated fats, the HR per 1 standard deviation was 0.77 (95% CI, 0.64-0.92). In an analysis of data from 2 large cohort studies, we found higher intake of vegetable fats and polyunsaturated fats to be associated with lower risk of HCC. Replacing animal or dairy fats with vegetable fats, or replacing saturated fats with monounsaturated or polyunsaturated fats, was associated with reduced risk of HCC.

Interaction between apolipoprotein E genotype and hypertension on cognitive function in older women in the Nurses' Health Study.

ObjectiveTo examine the interaction between APOE genotypes and both treated and untreated hypertension on cognitive function in an updated analysis of Nurses’ Health Study (NHS) data.DesignAt baseline (1995–2001) and 3 biennial follow-up assessments over ~6 years, cognitive function was assessed.Setting and participants8300 NHS participants aged 70+ years underwent a cognitive battery, which comprised 6 tests including the Telephone Interview for Cognitive Status (TICS) and tests of verbal memory, category fluency, and working memory.MeasuresWe estimated the mean differences in average cognitive scores across up to 4 assessments using multiple linear regression. We also tested for interaction between APOE e4 allele carrier status and hypertension overall, as well as for apparently untreated and treated hypertension.ResultsWe confirmed that, compared with those with APOE e3/3 genotype, APOE e4 allele carriers scored lower by 0.55 units on the average TICS score (95%CI:-0.67,-0.43). We also observed a significantly worse average TICS score among women with untreated hypertension compared with women without hypertension (difference = -0.23, 95%CI:-0.37,-0.09), while no significant difference was observed for women with treated hypertension. Significant interaction was detected between the APOE e4 allele and untreated hypertension (p-int = 0.02 for the TICS; p-int = 0.045 for global score), but not with treated hypertension. Specifically, compared with normotensive women with the APOE e3/3 genotype, APOE e4 allele carriers with treated hypertension scored lower by 0.50 units (95%CI:-0.69,-0.31); however, the APOE e4 allele carriers with untreated hypertension scored lower by 1.02 units on the TICS score (95%CI:-1.29, -0.76). This interaction of APOE e4 and untreated hypertension was also consistently observed for the global score.ConclusionsWomen with hypertension and at least one APOE e4 allele had worse average cognitive function compared with women without hypertension with the e3/3 genotype; this difference was amplified among APOE e4 allele carriers with untreated hypertension.

PC-FACS

PC-FACS

Serious Illness and End-of-Life Treatments for Nurses Compared with the General Population.

As key team members caring for people with advanced illness, nurses teach patients and families about managing their illnesses and help them to understand their options. Our objective was to determine if nurses' personal healthcare experience with serious illness and end-of-life (EOL) care differs from the general population as was shown for physicians. Observational propensity-matched cohort study. Fee-for-service Medicare. Nurses' Health Study (NHS) and a random 20% national sample of Medicare beneficiaries aged 66 years or older with Alzheimer's disease and related dementias (ADRD) or congestive heart failure (CHF) diagnosed in the hospital. Characteristics of care during the first year after diagnosis and the last 6 months of life (EOL). Among 57 660 NHS participants, 7380 had ADRD and 5375 had CHF; 3227 ADRD patients and 2899 CHF patients subsequently died. Care patterns in the first year were similar for NHS participants and the matched national sample: hospitalization rates, emergency visits, and preventable hospitalizations were no different in either disease. Ambulatory visits were slightly higher for NHS participants than the national sample with ADRD (13.1 vs 12.5 visits; P < .01) and with CHF (13.7 vs 12.5; P < .001). Decedents in the NHS and national sample had similar acute care use (hospitalization and emergency visits) in both diseases, but those with ADRD were less likely to use life-prolonging treatments such as mechanical ventilation (10.9% vs 13.5%; P = .001), less likely to die in a hospital with a stay in the intensive care unit (10.4% vs 12.1%; P = .03), and more likely to use hospice (58.9% vs 54.8%; P < .001). CHF at the EOL results were similar. Nurses with newly identified serious illness experience similar care as the general Medicare population. However, at EOL, nurses are more likely to choose less aggressive treatments than the patients for whom they care. J Am Geriatr Soc 67:1582-1589, 2019.

Accounting for individualized competing mortality risks in estimating postmenopausal breast cancer risk.

Accurate risk assessment is necessary for decision-making around breast cancer prevention. We aimed to develop a breast cancer prediction model for postmenopausal women that would take into account their individualized competing risk of non-breast cancer death. We included 73,066 women who completed the 2004 Nurses' Health Study (NHS) questionnaire (all≥57years) and followed participants until May 2014. We considered 17 breast cancer risk factors (health behaviors, demographics, family history, reproductive factors) and 7 risk factors for non-breast cancer death (comorbidities, functional dependency) and mammography use. We used competing risk regression to identify factors independently associated with breast cancer. We validated the final model by examining calibration (expected-to-observed ratio of breast cancer incidence, E/O) and discrimination (c-statistic) using 74,887 subjects from the Women's Health Initiative Extension Study (WHI-ES; all were≥55years and followed for 5years). Within 5years, 1.8% of NHS participants were diagnosed with breast cancer (vs. 2.0% in WHI-ES, p=0.02), and 6.6% experienced non-breast cancer death (vs. 5.2% in WHI-ES, p<0.001). Using a model selection procedure which incorporated the Akaike Information Criterion, c-statistic, statistical significance, and clinical judgement, our final model included 9 breast cancer risk factors, 5 comorbidities, functional dependency, and mammography use. The model's c-statistic was 0.61 (95% CI [0.60-0.63]) in NHS and 0.57 (0.55-0.58) in WHI-ES. On average, our model under predicted breast cancer in WHI-ES (E/O 0.92 [0.88-0.97]). We developed a novel prediction model that factors in postmenopausal women's individualized competing risks of non-breast cancer death when estimating breast cancer risk.

Antidepressant use and circulating prolactin levels.

To determine whether antidepressants (AD), specifically selective serotonin reuptake inhibitors (SSRIs), are linked to elevated prolactin levels among the general population. Circulating prolactin levels were available for 4593 healthy participants in the Nurses' Health Study (NHS) and NHS2, including 267 AD users. We fit generalized linear models to calculate and compare adjusted mean prolactin levels between AD users and non-users and further among SSRI users. Multivariable logistic regression was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for "elevated" prolactin levels (>11ng/mL) comparing AD users to non-users. We evaluated AD use and change in prolactin levels among 610 NHS participants with two measurements an average of 11years apart. Adjusted geometric mean prolactin levels were similar among SSRI users (13.2ng/mL, 95% CI 12.2-14.4), users of other classes of ADs (12.7ng/mL, 95% CI 11.0-14.6), and non-users (13.1ng/mL, 95% CI 12.8-13.4). Neither AD use (OR 1.17, 95% CI 0.89-1.53) nor SSRI use (OR 0.95, 95% CI 0.66-1.38) was associated with elevated prolactin levels. Change in prolactin levels was similar across women who started, stopped, consistently used, or never used ADs. This study does not support the hypothesis that AD use would influence breast cancer risk via altered prolactin levels. These results provide some evidence that use of ADs to treat depression or other conditions may not substantially increase prolactin levels in the majority of women.

Abstract B29: Pre- and postdiagnosis analgesic use and ovarian cancer survival.

Abstract Background: Ovarian cancer is the 5th deadliest cancer among US women. Although clinical characteristics (e.g., stage at diagnosis and histology) and treatment (e.g., optimal debulking) predict outcomes, little is known about the contributions of modifiable factors to ovarian cancer prognosis. We evaluated whether use of analgesics (aspirin; other non-steroidal anti-inflammatory drugs [NSAIDs], including ibuprofen, naproxen, and coxibs; and acetaminophen) were associated with improved ovarian cancer prognosis. Methods: The Nurses' Health Study (NHS) is a prospective study of 121,700 US nurses who have completed biennial questionnaires on lifestyle, medications, and disease diagnosis since 1976. Among NHS participants diagnosed with ovarian cancer, we evaluated whether pre- and post-diagnosis use of aspirin, other NSAIDs, or acetaminophen was associated with risk of dying due to any cause or due to ovarian cancer. Regular aspirin use (more than 2 times/week over the prior year) was queried on every questionnaire since 1980; regular use of other NSAIDs or acetaminophen has been queried since 1990. We used Cox proportional hazards regression to calculate relative risks (RRs) and 95% confidence intervals (CI), adjusting for tumor and personal characteristics (stage, histology, invasiveness, age at diagnosis, year of diagnosis, adiposity, smoking, menopause status, oral contraceptive use, parity, tubal ligation, and family history). Women with diagnosed with stage IV ovarian cancer were excluded from the analysis. Results: 865 ovarian cancer patients had data available on pre-diagnosis aspirin; 649 had data on pre-diagnosis use of other NSAIDs and acetaminophen. For post-diagnosis exposures, 602 women had information on post-diagnosis use of aspirin and 491 had information on post-diagnosis use of other NSAIDs/acetaminophen available. Women who regularly used aspirin after diagnosis had a 35% lower risk of death (RR: 0.65; 95% CI: 0.45-0.92) and a 44% lower risk of dying due to ovarian cancer (RR:0.56; 95% CI: 0.36-0.86). Post-diagnosis use of other NSAIDs was associated with a suggestive decreased risk of ovarian cancer death (RR: 0.86; 95% CI: 0.59-1.26). There was no association of pre-diagnosis use of any of the analgesic medications with prognosis, nor was there an association of post-diagnosis acetaminophen use with survival. Conclusion: In this preliminary study of analgesic use among ovarian cancer patients, regular use of aspirin after diagnosis appeared to improve survival. However, there were relatively few current users available for analysis, particularly for non-aspirin NSAIDs (131 women were current users; 66 of whom died during follow-up). Overall, our results suggest that reducing inflammation after diagnosis may be beneficial for improving survival. Additional analyses on patterns of analgesic use (e.g., duration of use and frequency) are on-going, as well as additional data from the Nurses' Health Study II. Further analyses stratifying by stage and histology will also be conducted. Citation Format: Elizabeth M. Poole, Megan S. Rice, Shelley S. Tworoger. Pre- and postdiagnosis analgesic use and ovarian cancer survival. [abstract]. In: Proceedings of the AACR Special Conference on Advances in Ovarian Cancer Research: Exploiting Vulnerabilities; Oct 17-20, 2015; Orlando, FL. Philadelphia (PA): AACR; Clin Cancer Res 2016;22(2 Suppl):Abstract nr B29.

Statin Use and Breast Cancer Risk in the Nurses' Health Study.

Preclinical studies support an anticancer effect of statin drugs, yet epidemiologic evidence remains inconsistent regarding their role in breast cancer primary prevention. Here, we report an updated analysis of the association between statin use and breast cancer incidence in the Nurses' Health Study (NHS) cohort. Postmenopausal NHS participants without a cancer history were followed from 2000 until 2012 (n = 79,518). Data on statin use were retrieved from biennial questionnaires. We fit Cox regression models to estimate associations between longitudinal statin use and breast cancer incidence. Over 823,086 person-years of follow-up, 3,055 cases of invasive breast cancer occurred. Compared with never users, both former and current statin users had similar rates of invasive breast cancer incidence [former users: HRadj, 0.96; 95% confidence interval (CI), 0.82-1.1; current users: HRadj, 1.1; 95% CI, 0.92-1.3]. Associations did not differ by estrogen receptor (ER) status or histology (ductal vs. lobular carcinoma). Statin use was not associated with risk of invasive breast cancer, irrespective of histologic subtype and ER status. Statin drugs do not appear to modify processes involved in breast cancer initiation.

Ambient Air Pollution Exposures and Risk of Rheumatoid Arthritis

Environmental factors may play a role in the development of rheumatoid arthritis (RA). We previously observed increased RA risk among women living closer to major roads (a source of air pollution). Herein, we examined whether long-term exposures to specific air pollutants were associated with RA risk among women in the Nurses' Health Study (NHS). The NHS is a large US cohort of female nurses followed up prospectively every 2 years since 1976. We studied 111,425 NHS participants with information on air pollution exposures as well as data concerning other lifestyle and behavioral exposures and disease outcomes. Outdoor levels of different size fractions of particulate matter (PM10 and PM2.5 ) and gaseous pollutants (SO2 and NO2 ) were predicted for all available residential addresses using monitoring data from the US Environmental Protection Agency. We examined the association of time-varying exposures 6 and 10 years before each questionnaire cycle and cumulative average exposure with the risk of RA, seronegative (rheumatoid factor and anti-citrullinated peptide antibody negative) RA, and seropositive RA. Over the 3,019,424 person-years of followup, 858 incident RA cases were validated by medical record review by 2 board-certified rheumatologists. Overall, we found no evidence of increased risk of RA, seronegative RA, or seropositive RA with exposure to the different pollutants and little evidence of effect modification by socioeconomic status or smoking status, geographic region, or calendar period. In this group of socioeconomically advantaged middle-aged and elderly women, adult exposures to air pollution were not associated with an increased RA risk.

Exposure to ultraviolet-B and risk of developing rheumatoid arthritis among women in the Nurses’ Health Study

ObjectiveTo examine the association between ultraviolet-B (UV-B) light exposure and rheumatoid arthritis (RA) risk among women in two large prospective cohort studies, the Nurses’ Health Study (NHS) and the Nurses’...