Health Professionals Follow-up Study Research Articles

Adherence to Healthy and Unhealthy Plant-Based Diets and the Risk of Gout

Plant-based diets are increasing in popularity due, in part, to their health benefits for selected cardiometabolic diseases as well as favorable environmental impact. Little is known about how such a diet is related to gout risk. To examine associations between adherence to a plant-based diet (including healthy and unhealthy versions of this diet), as well as its 18 individual food groups, and incident gout. This prospective cohort study used data from population-based cohorts of US men and women enrolled in the Health Professionals Follow-Up Study (1986-2012) and Nurses' Health Study (1984-2010). Participants were men and women free of gout at baseline. Statistical analyses were performed over March 2020 to August 2023. An overall plant-based diet index (PDI), as well as healthy (hPDI) and unhealthy (uPDI) versions of this index that emphasize healthy and less healthy plant-based foods, respectively. These diet indices were comprised of 18 food groups, assessed using a validated semiquantitative food frequency questionnaire. Incident cases of gout that were confirmed with a supplementary questionnaire to meet the preliminary American College of Rheumatology survey criteria for gout. Cox proportional hazards regression models were used to evaluate multivariable-adjusted associations of all 3 PDIs with incident gout using quintiles (Q) of adherence. Among a total of 122 679 participants (mean [SD] age, 53.8 [9.8] years among 43 703 men; mean [SD] age, 50.9 [7.2] years among 78 976 women) over 2 704 899 person-years of follow-up, 2709 participants experienced incident gout. The overall PDI was not significantly associated with gout in either cohort (Q5 vs Q1 pooled hazard ratio [HR], 1.02 [95% CI, 0.89-1.17]; P for trend = .63). In the pooled analysis, hPDI was significantly inversely associated with risk of gout (Q5 vs Q1 HR, 0.79 [95% CI, 0.69-0.91]; P for trend = .002), while the uPDI was positively associated with risk of gout (Q5 vs Q1 HR, 1.17 [95% CI, 1.03-1.33]; P for trend = .02), particularly in women (Q5 vs Q1 HR, 1.31 [95% CI, 1.05-1.62]; P for trend = .01). In analysis of individual food groups, higher intakes of certain healthy plant foods, such as whole grains (pooled HR per 1 serving/d, 0.93 [95% CI, 0.89-0.97]) and tea and coffee (pooled HR per 1 serving/d, 0.95 [95% CI, 0.92-0.97]), as well as dairy (pooled HR per 1 serving/d, 0.86 [95% CI, 0.82-0.90]), were independently associated with a lower risk of gout, while selected unhealthy plant foods, such as fruit juice (pooled HR per 1 serving/d, 1.06 [95% CI, 1.00-1.13]) and sugar-sweetened beverages (pooled HR per 1 serving/d, 1.16 [95% CI, 1.07-1.26]) were associated with increased risk of gout. The findings of this prospective cohort study of PDIs and gout support current dietary recommendations to increase consumption of healthy plant foods while lowering intake of unhealthy plant foods to mitigate gout risk.

Prediction of 24-Hour Urinary Sodium Excretion Using Machine-Learning Algorithms.

Accurate quantification of sodium intake based on self-reported dietary assessments has been a persistent challenge. We aimed to apply machine-learning (ML) algorithms to predict 24-hour urinary sodium excretion from self-reported questionnaire information. We analyzed 3454 participants from the NHS (Nurses' Health Study), NHS-II (Nurses' Health Study II), and HPFS (Health Professionals Follow-Up Study), with repeated measures of 24-hour urinary sodium excretion over 1 year. We used an ensemble approach to predict averaged 24-hour urinary sodium excretion using 36 characteristics. The TOHP-I (Trial of Hypertension Prevention I) was used for the external validation. The final ML algorithms were applied to 167 920 nonhypertensive adults with 30-year follow-up to estimate confounder-adjusted hazard ratio (HR) of incident hypertension for predicted sodium. Averaged 24-hour urinary sodium excretion was better predicted and calibrated with ML compared with the food frequency questionnaire (Spearman correlation coefficient, 0.51 [95% CI, 0.49-0.54] with ML; 0.19 [95% CI, 0.16-0.23] with the food frequency questionnaire; 0.46 [95% CI, 0.42-0.50] in the TOHP-I). However, the prediction heavily depended on body size, and the prediction of energy-adjusted 24-hour sodium excretion was modestly better using ML. ML-predicted sodium was modestly more strongly associated than food frequency questionnaire-based sodium in the NHS-II (HR comparing Q5 versus Q1, 1.48 [95% CI, 1.40-1.56] with ML; 1.04 [95% CI, 0.99-1.08] with the food frequency questionnaire), but no material differences were observed in the NHS or HPFS. The present ML algorithm improved prediction of participants' absolute 24-hour urinary sodium excretion. The present algorithms may be a generalizable approach for predicting absolute sodium intake but do not substantially reduce the bias stemming from measurement error in disease associations.

Association of ultra-processed food consumption with all cause and cause specific mortality: population based cohort study

ObjectiveTo examine the association of ultra-processed food consumption with all cause mortality and cause specific mortality.DesignPopulation based cohort study.SettingFemale registered nurses from 11 US states in the Nurses’ Health Study...

Appendectomy and Long-term Colorectal Cancer Incidence, Overall and by Tumor Fusobacterium nucleatum Status.

To test hypotheses that appendectomy history might lower long-term colorectal cancer risk and that the risk reduction might be strong for tumors enriched with Fusobacterium nucleatum, bacterial species implicated in colorectal carcinogenesis. The absence of the appendix, an immune system organ and a possible reservoir of certain pathogenic microbes, may affect the intestinal microbiome, thereby altering long-term colorectal cancer risk. Utilizing databases of prospective cohort studies, namely the Nurses' Health Study and the Health Professionals Follow-up Study, we examined the association of appendectomy history with colorectal cancer incidence overall and subclassified by the amount of tumor tissue Fusobacterium nucleatum​​ (Fusobacterium animalis). We used an inverse probability weighted multivariable-adjusted duplication-method Cox proportional hazards regression model. During the follow-up of 139,406 participants (2,894,060 person-years), we documented 2811 incident colorectal cancer cases, of which 1065 cases provided tissue F. nucleatum analysis data. The multivariable-adjusted hazard ratio of appendectomy for overall colorectal cancer incidence was 0.92 (95% CI, 0.84-1.01). Appendectomy was associated with lower F. nucleatum-positive cancer incidence (multivariable-adjusted hazard ratio, 0.53; 95% CI, 0.33-0.85; P=0.0079), but not F. nucleatum-negative cancer incidence (multivariable-adjusted hazard ratio, 0.98; 95% CI, 0.83-1.14), suggesting a differential association by F. nucleatum status (Pheterogeneity=0.015). This differential association appeared to persist in various participant/patient strata including tumor location and microsatellite instability status. Appendectomy likely lowers the future long-term incidence of F. nucleatum-positive (but not F. nucleatum-negative) colorectal cancer. Our findings do not support the existing hypothesis that appendectomy may increase colorectal cancer risk.

Use of melatonin supplements and risk of type 2 diabetes and cardiovascular diseases in the USA: insights from three prospective cohort studies

Use of melatonin supplements has been increasing substantially in both children and adults in the USA; however, their long-term cardiometabolic effects remain unclear. We aimed to assess the associations between regular use of melatonin supplements and the risk of developing type 2 diabetes or cardiovascular disease in adults. In this study, we included individuals from three US cohorts: the Nurses' Health Study (women only), the Health Professionals Follow-up Study (men only), and the Nurses' Health Study II (women only). Women aged 25-55 years and men aged 45-75 years at baseline, who had no diagnosis of cancer at baseline, and who responded to the question about melatonin supplement use (yes or no) were included. We excluded baseline prevalent cardiovascular disease and baseline prevalent type 2 diabetes for the main analyses. The main outcomes were cardiovascular disease and type 2 diabetes incidence. In secondary analyses, we stratified by duration of rotating night shift work in the Nurses' Health Study and Nurses' Health Study II to examine whether the associations with melatonin supplement use differed by rotating night shift work. For the cardiovascular disease analysis, we included 67 202 women from the Nurses' Health Study (follow-up 1998-2019, mean age at baseline: 63·6 years [SD 7·1]), 26 629 men from the Health Professionals Follow-up Study (1998-2020, 62·9 years [8·8], and 65 241 women from the Nurses' Health Study II (2003-19, 48·2 years [4·7]). Follow-up for incident type 2 diabetes was from 1998 to June 30, 2021, for the Nurses' Health Study; 2003 to Jan 31, 2023, for the Nurses' Health Study II; and from 1998 to Jan 31, 2020, for the Health Professionals' Follow-up Study. Melatonin supplement use in the study cohorts doubled over recent decades from less than 2% in 1998-2007 to 4% or higher in 2014-15 (4·0% in men and 5·3% in women). We documented 16 917 incident cardiovascular disease events during 2 609 068 person-years of follow-up and 12 730 incident cases of type 2 diabetes during 2 701 830 person-years of follow-up. In a pooled analysis of the three cohorts, comparing users with non-users of melatonin supplements, the pooled multivariable-adjusted hazard ratios were 0·94 (95% CI 0·83-1·06, p=0·32) for cardiovascular disease and 0·98 (0·86-1·12, p=0·80) for type 2 diabetes. In secondary analyses, melatonin supplement use appeared to attenuate the positive association between long-term shift work (>5 years) and risk of cardiovascular disease (pinteraction=0·013) among the female nurses. With up to 23 years of follow-up of three large prospective cohorts of middle-aged and older men and women, self-reported melatonin supplement use was not associated with risk of type 2 diabetes or cardiovascular disease. Further research is warranted to assess if melatonin supplement use could mitigate the potential risks of type 2 diabetes and cardiovascular disease associated with rotating night shift work. US National Institutes of Health.

Ultra-processed food consumption and mortality among patients with stages I–III colorectal cancer: a prospective cohort study

Ultra-processed foods (UPFs) are emerging as a risk factor for colorectal cancer (CRC), yet how post-diagnostic UPF intake may impact CRC prognosis remains unexplored. Data collected from food frequency questionnaires were used to estimate intakes of total UPFs and UPF subgroups (serving/d) at least 6 months but less than 4 years post-diagnosis among 2498 patients diagnosed with stages I-III CRC within the Nurses' Health Study and Health Professionals Follow-up Study during 1980-2016. Hazard ratios (HR) and 95% confidence intervals (CIs) of all-cause, CRC- and cardiovascular disease (CVD)-specific mortality in association with UPF consumption were estimated using an inverse probability weighted multivariable Cox proportional hazards regression model, adjusted for confounders. The mean (SD) age of patients at diagnosis was 68.5 (9.4) years. A total of 1661 deaths were documented, including 321 from CRC and 335 from CVD. Compared to those in the lowest quintile (median=3.6 servings/d), patients in the highest quintile (median=10 servings/d) of post-diagnostic UPF intake had higher CVD mortality (HR=1.65, 95% CI=1.13-2.40) but not CRC or all-cause mortality. Among UPF subgroups, higher consumption of fats/condiments/sauces was associated with a higher risk of CVD-specific mortality (highest vs. lowest quintile of intake, HR=1.96, 95% CI=1.41-2.73), and higher intake of ice cream/sherbet was associated with an increased risk of CRC-specific mortality (highest vs. lowest quintile, HR=1.86, 95% CI: 1.33-2.61). No statistically significant association was found between UPF subgroups and overall mortality. Higher post-diagnostic intake of total UPFs and fats/condiments/sauces in CRC survivors is associated with higher CVD mortality, and higher ice cream/sherbet intake is linked to higher CRC mortality. US National Institutes of Health and the American Cancer Society.

High-quality fruit and vegetable characterized by cardiometabolic biomarkers and its relation to major chronic disease risk: results from three prospective US cohort studies

BackgroundCurrent guidelines recommend a specified total serving of fruits and vegetables (FV). However, how differences in their nutritional quality influence overall health remains unclear. ObjectiveTo identify high-quality FV using 14 cardiometabolic biomarkers, and assess their consumption, alongside overall FV intake, with chronic disease risk. MethodsWe used data from 3 prospective cohorts, Health Professionals Follow-up Study, Nurses’ Health Study (NHS), and NHS-II. Diet was assessed at baseline and updated every 4 years. Biomarker analysis was conducted on 41,714 participants using generalized linear models. Metabolic quality was ascertained by each FV’s association with biomarkers. Major chronic disease risk analysis involved 207,241 participants followed for 32 years with Cox proportional hazards models. We also analyzed atherosclerotic cardiovascular disease (ASCVD), type 2 diabetes (T2D), cancer, and chronic obstructive pulmonary disease (COPD) as secondary outcomes. ResultsOf 52 FV items, 19 were identified as high-metabolic quality (top five: apples/pears, iceberg/head lettuce, raw spinach, alfalfa sprouts, and eggplant/summer squash). In disease risk analysis, 60,712 major chronic disease events were recorded. A higher proportion of high-metabolic quality FV intake was associated with lower chronic disease risk across total FV quantity levels. In each quantity level stratum (quartiles Q1-Q4), comparing the highest to the lowest quality proportion quartiles, the HR (95% CI) were 0.85 (0.81-0.90), 0.86 (0.82-0.90), 0.84 (0.80-0.89), and 0.89 (0.84-0.94), all P-trend < 0.001. Patterns were similar for ASCVD, T2D, and COPD, but less consistent for cancer. High total FV intake, if consisting mostly of neutral or low-metabolic quality items, was not associated with lower chronic disease risk. For diabetes specifically, these were associated with significantly higher risk (quantity-Q3, HR=1.13 [1.05-1.22]; quantity-Q4, HR=1.17 [1.07-1.28]). ConclusionsOur findings stress the importance of considering both quality and quantity of FV for health, and support dietary guidelines to emphasize high-metabolic quality FV consumption alongside overall intake.

Pre-diagnostic plasma proteomics profile for hepatocellular carcinoma.

Proteomics may discover pathophysiological changes related to hepatocellular carcinoma (HCC), an aggressive and lethal type of cancer with low sensitivity for early-stage diagnosis. We measured 1,305 pre-diagnostic (median 12.7 years) SOMAscan proteins from 54 pairs of healthy individuals who subsequently developed HCC and matched non-HCC controls from the Nurses' Health Study (NHS) and the Health Professionals Follow-up Study (HPFS). Candidate proteins were validated in the independent, prospective UK Biobank Pharma Proteomics Project (UKB-PPP). In NHS/HPFS, we identified 56 elevated proteins in HCC with absolute fold-change >1.2 and Wald test P < .05 in conditional logistic regression analysis. Ingenuity Pathway Analysis identified enrichment of pathways associated with cell viability, adhesion, proteolysis, apoptosis, and inflammatory response. Four proteins, chitinase-3-like protein 1, growth/differentiation factor 15, interleukin-1 receptor antagonist protein, and E-selectin, showed strong positive associations with HCC and were thus validated by ELISA (odds ratio ranged 2.48-14.7, all P < .05) in the NHS/HPFS and by Olink platform (hazard ratio ranged 1.90-3.93, all P < .05) in the UKB-PPP. Adding these four proteins to a logistic regression model of traditional HCC risk factors increased the area under the curve (AUC) from 0.67 to 0.87 in the NHS/HPFS. Consistently, model AUC was 0.88 for HCC risk prediction in the UKB-PPP. However, the limited number of HCC cases in the cohorts necessitates caution in interpreting our findings, emphasizing the need for further validation in high-risk populations.

Consumption of Olive Oil and Diet Quality and Risk of Dementia-Related Death

Age-standardized dementia mortality rates are on the rise. Whether long-term consumption of olive oil and diet quality are associated with dementia-related death is unknown. To examine the association of olive oil intake with the subsequent risk of dementia-related death and assess the joint association with diet quality and substitution for other fats. This prospective cohort study examined data from the Nurses' Health Study (NHS; 1990-2018) and Health Professionals Follow-Up Study (HPFS; 1990-2018). The population included women from the NHS and men from the HPFS who were free of cardiovascular disease and cancer at baseline. Data were analyzed from May 2022 to July 2023. Olive oil intake was assessed every 4 years using a food frequency questionnaire and categorized as (1) never or less than once per month, (2) greater than 0 to less than or equal to 4.5 g/d, (3) greater than 4.5 g/d to less than or equal to 7 g/d, and (4) greater than 7 g/d. Diet quality was based on the Alternative Healthy Eating Index and Mediterranean Diet score. Dementia death was ascertained from death records. Multivariable Cox proportional hazards regressions were used to estimate hazard ratios (HRs) and 95% CIs adjusted for confounders including genetic, sociodemographic, and lifestyle factors. Of 92 383 participants, 60 582 (65.6%) were women and the mean (SD) age was 56.4 (8.0) years. During 28 years of follow-up (2 183 095 person-years), 4751 dementia-related deaths occurred. Individuals who were homozygous for the apolipoprotein ε4 (APOE ε4) allele were 5 to 9 times more likely to die with dementia. Consuming at least 7 g/d of olive oil was associated with a 28% lower risk of dementia-related death (adjusted pooled HR, 0.72 [95% CI, 0.64-0.81]) compared with never or rarely consuming olive oil (P for trend < .001); results were consistent after further adjustment for APOE ε4. No interaction by diet quality scores was found. In modeled substitution analyses, replacing 5 g/d of margarine and mayonnaise with the equivalent amount of olive oil was associated with an 8% (95% CI, 4%-12%) to 14% (95% CI, 7%-20%) lower risk of dementia mortality. Substitutions for other vegetable oils or butter were not significant. In US adults, higher olive oil intake was associated with a lower risk of dementia-related mortality, irrespective of diet quality. Beyond heart health, the findings extend the current dietary recommendations of choosing olive oil and other vegetable oils for cognitive-related health.

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.

Appendectomy and risk of prostate cancer in the Health Professionals Follow-up Study.

Appendectomy is a common surgical procedure to treat appendicitis. Limited studies examined its association on prostate cancer, with one large cohort study suggesting significant increased risk of overall and advanced prostate cancer, especially among younger men. A total of 49,104 men in the Health Professionals Follow-up Study were followed from 1986 to 2016. Cox proportional hazards models were applied to evaluate the association between self-reported history of appendectomy and risk of overall and subtype specific prostate cancer, adjusted for multiple risk factors. During 30 years of follow-up, we documented 7,253 overall prostate cancer including 579 advanced and 1,092 lethal events. Compared to men without appendectomy, those who reported at baseline having had appendectomy were not at higher risk of overall (hazard ratio (HR)=1.01, 95% CI = 0.95 to 1.07), advanced (HR=0.99, 95% CI = 0.81 to 1.23) or lethal (HR=1.04, 95% CI = 0.89 to 1.20) prostate cancer. The association remained null when stratified by age. We found no evidence of an association between appendectomy and risk of overall and clinically important prostate cancer. We showed that appendectomy was not associated with overall or advanced prostate cancer adjusted for multiple risk factors among a large population of men with 30 years of follow-up.

Impact of Family History and Germline Genetic Risk Single Nucleotide Polymorphisms on Long-Term Outcomes of Favorable-Risk Prostate Cancer.

Family history and germline genetic risk single nucleotide polymorphisms (SNPs) have been separately shown to stratify lifetime risk of prostate cancer. Here, we evaluate the combined prognostic value of family history of prostate and other related cancers and germline risk SNPs among patients with favorable-risk prostate cancer. A total of 1367 participants from the prospective Health Professionals Follow-up Study diagnosed with low- or favorable intermediate-risk prostate cancer from 1986 to 2017 underwent genome-wide SNP genotyping. Multivariable Cox regression was used to estimate the association between family history, specific germline risk variants, and a 269 SNP polygenic risk score with prostate cancer‒specific death. Family history of prostate, breast, and/or pancreatic cancer was observed in 489 (36%) participants. With median follow-up from diagnosis of 14.9 years, participants with favorable-risk prostate cancer with a positive family history had a significantly higher risk of prostate cancer‒specific death (HR 1.95, 95% CI 1.15-3.32, P = .014) compared to those without any family history. The rs2735839 (19q13) risk allele was associated with prostate cancer‒specific death (HR 1.81 per risk allele, 95% CI 1.04-3.17, P = .037), whereas the polygenic risk score was not. Combined family history and rs2735839 risk allele were each associated with an additive risk of prostate cancer‒specific death (HR 1.78 per risk factor, 95% CI 1.25-2.53, P = .001). Family history of prostate, breast, or pancreatic cancer and/or a 19q13 germline risk allele are associated with an elevated risk of prostate cancer‒specific death among favorable-risk patients. These findings have implications for how family history and germline genetic risk SNPs should be factored into clinical decision-making around favorable-risk prostate cancer.

Abstract NG07: Association between somatic microsatellite instability, hypermutation status, and specific T cell subsets in colorectal cancer tumors

Abstract NG07: Association between somatic microsatellite instability, hypermutation status, and specific T cell subsets in colorectal cancer tumors

Alcohol and colorectal cancer risk subclassified by mutational signatures of DNA mismatch repair deficiency.

We examined whether the association between alcohol consumption and CRC incidence was stronger for tumors with higher contributions of defective MMR (dMMR)-related tumor mutational signatures (TMSs). We used data from 227,916 men and women who participated in the Nurses' Health Study (1980-2016), the Nurses' Health Study II (1991-2017), and the Health Professionals Follow-up Study (1986-2016). Dietary data was collected every 4 years through validated food frequency questionnaires. Relative contributions of two dMMR-related TMSs (c-dMMRa/SBS15 and c-dMMRb/SBS26) were quantified using whole-exome sequencing data in a subset of incident CRC cases. Duplication-method Cox proportional hazards regression models were used to assess the association between alcohol consumption and the risk of CRC subtypes according to different contributions of the TMSs. All statistical tests were 2-sided. We documented 825 incident CRC cases with available TMS data over 26-36 years of follow-up. The association between alcohol consumption and CRC incidence was stronger for tumors with higher contributions of c-dMMRb/SBS26 (P-heterogeneitytrend = 0.02) compared to tumors with lower contributions of this TMS. Compared with nondrinkers, drinkers with ≥15 g/d of alcohol had a high risk of c-dMMRb/SBS26-high CRC [multivariable-adjusted HR: 2.43 (95% CI: 1.55-3.82)], but not c-dMMRb/SBS26-low CRC [0.86 (95% CI: 0.57-1.28)] or c-dMMRb/SBS26-moderate CRC [1.14 (95% CI: 0.76-1.71)]. No significant differential associations were observed for c-dMMRa/SBS15 (P-heterogeneitytrend = 0.41). High alcohol consumption was associated with an increased incidence of CRC containing higher contributions of c-dMMRb/SBS26, suggesting that alcohol consumption may be involved in colorectal carcinogenesis through the DNA mismatch repair pathway.

24-Hour Urinary Chemistries and Kidney Stone Risk

Rationale & ObjectiveMost previous studies of the relationship between urinary factors and kidney stone risk have either assumed a linear effect of urinary parameters on kidney stone risk or implemented arbitrary thresholds suggesting biologically implausible “all-or-nothing” effects. In addition, little is known about the hierarchy of effects of urinary factors on kidney stone risk. This study evaluated the independent associations between urine chemistries and kidney stone formation and examined their magnitude and shape. Study DesignProspective cohort study. Setting & ParticipantsWe analyzed 9,045 24-hour urine collections from 6,217 participants of the Health Professionals Follow-Up Study and Nurses’ Health Studies I and II. ExposuresUrine volume and pH, and concentrations of calcium, citrate, oxalate, potassium, magnesium, uric acid, phosphorus, and sodium. OutcomeIncident symptomatic kidney stones. Analytical ApproachMultivariable logistic regression analysis incorporating restricted cubic splines to explore potentially non-linear relationships between urinary factors and the risk of forming a kidney stone. Optimal inflection point analysis was implemented for each factor and dominance analysis was performed to establish the relative importance of each urinary factor. ResultsEach urinary factor was significantly associated with stone formation except for urine pH. Higher urinary levels of calcium, oxalate, phosphorus, and sodium were associated with a higher risk of stone formation, whereas higher urine volume, uric acid, citrate, potassium, and magnesium were associated with a lower risk. The relationships were substantially linear for urine calcium, uric acid, and sodium. In contrast, the magnitudes of the relationships were modestly attenuated at levels above the inflection points for urine oxalate, citrate, volume, phosphorus, potassium, and magnesium. Dominance analysis identified three categories of factors' relative importance: higher (calcium, volume and citrate), intermediate (oxalate, potassium and magnesium) and lower (uric acid, phosphorus and sodium). LimitationsPredominantly white participants, lack of information on stone composition. ConclusionsUrine chemistries have complex relationships and differential relative associations with the risk of kidney stone formation.

Abstract 2255: Social support with physical and psychosocial quality of life in individuals with prostate cancer in the Health Professionals Follow-up Study (HPFS)

Abstract Introduction: Social support is associated with improved cancer survivorship, but literature on its association with prostate cancer specific physical and psychosocial quality of life (QoL) is lacking. We investigated the associations of social support with physical and psychosocial QoL in individuals with prostate cancer in HPFS. Methods: We included 1,692 men (ages 61-95 years, mean: 71 years) diagnosed with non-metastatic prostate cancer between 2008 and 2016. Social support was measured by the Berkman-Syme Social Network Index (SNI) (a composite measure of marital status, sociability, membership in religious groups or other community organizations: socially isolated, moderately isolated, moderately integrated, socially integrated) and marital status (married, not married). Physical (measured by EPIC-CP: bowel function, urinary incontinence, urinary irritation/obstruction, sexual function, vitality/hormonal function) and psychosocial (memory function and wellbeing) QoL were measured through biennial questionnaires. We fit generalized linear and ordinal mixed effect models for the associations of SNI and marital status with physical and psychosocial QoL, adjusting for age, race, employment status, body mass index, comorbidities, smoking history, and clinical factors. We conducted stratified analyses by baseline age (&amp;lt;70, ≥70), baseline comorbidities (yes, no), primary treatment (active surveillance, radical prostatectomy, radiation), and D’Amico Risk Classification at diagnosis (low, intermediate, high risk). Results: Among those with baseline SNI (N=1,586), 740 were socially integrated, 309 were moderately integrated, 437 were moderately isolated, and 100 were isolated. Among those with baseline marital status (N=1,663), 89% were married. Physical QoL did not differ significantly by SNI. Married individuals (compared to not married) experienced less urinary incontinence (β -0.25, 95% confidence interval [CI]: -0.55, 0.05) and urinary irritation/obstruction (β -0.26, 95% CI: -0.52, 0.00). Those who were socially integrated (compared to isolated) reported better memory function (Odds ratio [OR] 1.47, 95% CI 1.06, 2.03) and wellbeing (OR 1.74, 95% CI 1.31, 2.31). The positive association between social support and QoL was stronger among those with baseline aged &amp;lt;70 years (for vitality/hormonal functioning and memory function), those without comorbidities at baseline (for memory function), those with radiation as primary treatment (for urinary incontinence, vitality/hormonal functioning, memory function, and wellbeing), and those with high risk (for memory function and wellbeing). Conclusion: Prostate cancer survivors with more social support experienced better quality of life, including less urinary incontinence and urinary irritation/obstruction, and better memory function and wellbeing. Citation Format: Naiyu Chen, Colleen B. McGrath, Barbra A. Dickerman, Rachel C. Nethery, Lorelei A. Mucci. Social support with physical and psychosocial quality of life in individuals with prostate cancer in the Health Professionals Follow-up Study (HPFS) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2255.

Abstract 2256: Guideline-based physical activity and physical quality of life among individuals with nonmetastatic prostate cancer: A target trial emulation in the Health Professionals Follow-up Study

Abstract Introduction: Randomized trials have shown that physical activity can improve quality of life (QOL) among individuals with cancer. In the absence of a randomized trial of adequate size and follow-up to comprehensively evaluate this relationship for prostate cancer, we aim to emulate a target trial of guideline-based physical activity strategies and 6-year physical QOL. Methods: We will use observational data on individuals in the Health Professionals Follow-up Study diagnosed with nonmetastatic prostate cancer between 2010-2016 and free of conditions that could preclude following current physical activity recommendations at baseline (first post-diagnostic questionnaire). The physical activity strategies of interest are (1) engage in ≥150 minutes of moderate activity, or ≥75 minutes of vigorous activity per week, or an equivalent combination (i.e., ≥7.5 metabolic equivalent of task [MET]-hours/week of vigorous or moderate activity) and (2) no physical activity intervention. The outcomes of interest are 6-year mean physical QOL scores (based on EPIC-CP measures of bowel function, urinary incontinence, urinary irritation/obstruction, sexual function, and vitality/hormonal function). We will adjust for baseline and time-varying risk factors using the parametric g-formula. Conclusion: These findings may help to inform clinical recommendations to improve QOL among individuals with nonmetastatic prostate cancer. Further, the proposed approach will provide a framework for future studies designed to evaluate the effectiveness of any intervention strategies to improve patient-reported outcome measures in cancer survivorship research. Table 1. Specification and emulation of a target trial using observational data from HPFS Protocol component Target trial specification Target trial emulation Eligibility criteria (1) Individuals diagnosed with non-metastatic prostate cancer between 2010 and 2016; (2) No history of other cancers (Except nonmelanoma skin cancer); (3) No history of conditions that could preclude high levels of physical activity: myocardial infarction, stroke, congestive heart failure, amyotrophic lateral sclerosis, or functional impairment (defined as difficulty climbing a flight of stairs or walking eight blocks due to physical impairment) Same as for the target trial, except: We required complete data on physical activity, birth date, and other risk factors on the pre-diagnostic questionnaire and the first post-diagnostic (baseline) questionnaire. Physical activity strategies Each individual is assigned to one of the following strategies at baseline: (1) No intervention (usual physical activity) (2) Engage in ≥1.25 hours/week of vigorous activity or ≥2.5 hours/week of moderate activity or ≥7.5 MET-hours/week of vigorous or moderate activity (an equivalent combination). Strategies are initiated at baseline and continued over follow-up. Individuals are excused from following the physical activity intervention upon the development of myocardial infarction, stroke, congestive heart failure, amyotrophic lateral sclerosis, cancer metastasis, or functional impairment. Individuals must respond to surveys at every follow-up period to assess time-varying variables. Same as for the target trial, except: (1) Follow-up periods correspond to the questionnaire cycles in the cohort study. (2) We assumed that each 2-year physical activity questionnaire accurately reflects the individual's (1) average physical activity over the previous 2-year period and (2) intended physical activity that the individual would have reported at the start of the current period under no intervention. Strategy assignment Individuals are randomly assigned to a strategy at baseline and are aware of their assigned strategy. We assumed randomization conditional on the following baseline covariates: age; employment status; primary treatment; stage; Gleason score, and prostate-specifc antigen level at diagnosis; pre-baseline values of body-mass index and physical activity. Outcomes 6-year mean quality of life scores in the following physical domains: (1) Bowel function, (2) Urinary incontinence, (3) Urinary irritation/obstruction, (4) Sexual function, (5) Vitality/hormonal function. Individuals self-report their physical quality of life on EPIC (The Explanded Prostate Cancer Index Composite) questionnaires at every follow-up period. Same as for the target trial. Follow-up periods correspond to the questionnaire cycles in the cohort study. Follow-up For each eligible individual, follow-up starts at strategy assignment (baseline) and ends upon death, loss to follow-up, or the end of follow up in December 2018, whichever happens first. Same as for the target trial. Loss to follow-up was defined as questionnaire non-response. Causal contrasts Intention-to-treat and per-protocol effect. Observational analogue of per-protocol effect. Statistical analysis Intention-to-treat analysis. Per protocol analysis: Parametric g-formula to compare 6-year physical quality of life under each strategy via mean differences, with adjustment for pre- and post-baseline prognostic factors associated with adherence and loss to follow-up. Same as for the per-protocol analysis of the target trial. Citation Format: Naiyu Chen, Rachel C. Nethery, Lorelei A. Mucci, Barbra A. Dickerman. Guideline-based physical activity and physical quality of life among individuals with nonmetastatic prostate cancer: A target trial emulation in the Health Professionals Follow-up Study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2256.

Abstract 2207: Fiber intake and risk of lethal prostate cancer among men at increased genetic risk

Abstract Introduction High fiber intake from grains may lower advanced prostate cancer risk by enhancing insulin sensitivity, lowering insulin-like growth factor 1 and systemic inflammation. Whether a high-fiber diet can lower risk of metastatic disease and prostate cancer death in men at high genetic risk is unclear. Methods Using a new 400-SNP polygenic risk score (PRS) for prostate cancer, we quantified the genetic risk of 10,269 men with available genotype data in the Health Professionals Follow-up Study, followed between 1993 and 2019. Men in the top 50-100% PRS category were considered high-risk and included for analysis. Total fiber intake and different sources (cereal, fruits, vegetables) were assessed by validated food frequency questionnaires. Overall and lethal prostate cancer (metastatic disease/prostate cancer death) were analyzed using Cox proportional hazards models estimating hazard ratios (HRs) and 95% confidence intervals (CIs). Results During 26 years of follow-up, we recorded 1,494 overall and 174 lethal prostate cancer events in the high genetic risk group. As expected, no associations were observed fiber intake with risk of overall prostate cancer. For lethal prostate cancer, the results varied according to the source of fiber. Higher cereal fiber intake was associated with a lower risk of lethal prostate cancer (per 3 g/d higher intake: 0.81 [0.68, 0.97], P-trend=0.02). We further observed a higher disease risk among men with higher fruit fiber intake (per 3g/d higher intake: 1.18 [1.02, 1.36], P-trend=0.03). The associations remained similar when mutually adjusted for each fiber type. No associations were observed for total or vegetable fiber intake. Conclusion In this analysis of men at high prostate cancer genetic risk, we observed that higher cereal fiber intake was associated with lower lethal prostate cancer risk, whereas higher fruit fiber intake showed an opposite trend. Further studies in different populations and investigations into the underlying mechanisms are needed. Fiber intake with risk of overall and lethal prostate cancer among men at high genetic risk. Fiber type Overall prostate cancer: No.event/PY Overall prostate cancer: HR (95% CI) Lethal prostate cancer: No.event/PY Lethal prostate cancer: HR (95% CI) All fiber, g/d T1, median=15.8 490/66047 ref. 42/73826 ref. T2, median=21.8 484/62237 0.96 (0.83-1.11) 61/69849 1.16 (0.77-1.75) T3, median=29.9 520/67374 1.02 (0.86-1.21) 71/75242 1.24 (0.75-2.05) Per 3 g/d 0.98 (0.96-1.01) 1.00 (0.92-1.08) P for trend 0.27 0.93 Cereal fiber, g/d T1, median=4.6 491/63911 ref. 49/72367 ref. T2, median=6.7 512/66251 1.04 (0.91-1.20) 76/73384 1.30 (0.90-1.87) T3, median=9.4 491/65496 0.92 (0.80-1.06) 49/73166 0.75 (0.50-1.12) Per 3 g/d 0.96 (0.90-1.01) 0.81 (0.68-0.97) P for trend 0.14 0.02 Fruit fiber, g/d T1, median=2.5 471/66306 ref. 33/74056 ref. T2, median=4.5 514/63268 1.06 (0.92-1.22) 51/70430 1.21 (0.76-1.91) T3, median=7.2 509/66085 1.00 (0.86-1.16) 90/74431 1.84 (1.18-2.87) Per 3 g/d 0.97 (0.91-1.03) 1.18 (1.02-1.36) P for trend 0.34 0.03 Vegetable fiber, g/d T1, median=4.8 481/64139 ref. 51/71594 ref. T2, median=6.9 505/64938 1.04 (0.91-1.20) 66/72291 1.22 (0.84-1.78) T3, median=9.6 508/66581 1.04 (0.90-1.20) 57/75031 0.98 (0.65-1.45) Per 3 g/d 0.99 (0.93-1.05) 0.97 (0.82-1.15) P for trend 0.72 0.74 Citation Format: Yiwen Zhang, Konrad H. Stopsack, Lorelei A. Mucci, Edward Giovannucci, Anna Plym. Fiber intake and risk of lethal prostate cancer among men at increased genetic risk [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2207.

Abstract 402: Aneuploidy-associated SQLE gain promotes prostate cancer aggressiveness by altering lipid metabolism

Abstract Prostate cancer (PCa) is the second leading cause of cancer-related death in men in the US. Epidemiology studies on primary PCa cohorts in Physicians' Health Study and Health Professionals Follow-up Study (PHS and HPFS) have shown that high levels of whole-genome aneuploidy, featured by imbalanced chromosome numbers, correlate with lethal progression in PCa. However, details of the mechanisms of how aneuploidy drives PCa aggressiveness are still unclear. Here, we used the case of chromosome 8q (chr 8q, the long arm of chr 8) gain to study aneuploidy-associated prostatic malignancies. Chr 8q gains are the most frequent gain events that occur in approximately 23% of PCa cases. By using the PHS and HPFS cohorts, we modeled the increased expression of each gene located on chr 8q, for predicting the risks for lethal progression, and obtained each corresponding gene’s odds ratio (OR). By ranking the ORs, we revealed that a cholesterol biosynthesis gene, squalene monooxygenase (SQLE), is one of the top associators with lethal progression, amongst all chr 8q genes. SQLE plays a pivotal role in cholesterol synthesis. Previous lymphoma studies have shown that loss of SQLE contributes to cholesterol auxotrophy, and squalene build-up protects against oxidative cell death. In our experimental study, we have used normal and cancerous TMPRSS2-ERG-driven organoid models and found that over-expression of SQLE promotes formation of invasive structures and proliferation in cancer organoids. Interestingly, overexpression of SQLE decreased the protein levels of TMPRSS2-ERG, which appeared to be independent of androgen receptor levels. We also utilized the TMPRSS2-ERG positive VCaP cell line, which harbors gains of SQLE gene copies. We found that knocking down SQLE expression significantly upregulated ERG protein levels. Our recent study has shown that ERG (or other ETS) positive prostate cancers have a strong correlation with downregulation of fatty acid metabolism signature. We speculate that gain of SQLE can drive aggressiveness of prostate cancer by modulating lipid metabolism for growth and migration. Inhibition of SQLE could translate to better clinical outcomes regarding prostate cancer lethality. Citation Format: Thomas Walter Janas, Xiaofeng A. Su, Konrad H. Stopsack, Daniel R. Schmidt, Duanduan Ma, Zhe Li, Kathryn L. Penny, Tamara L. Lotan, Lorelei A. Mucci, Matthew G. Vander Heiden, Elise DeArment, Angelika Amon, Paul A. Scheet. Aneuploidy-associated SQLE gain promotes prostate cancer aggressiveness by altering lipid metabolism [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 402.

Abstract 3467: Tumor immune and microbial characteristics of colorectal cancer according to age at diagnosis and detailed anatomical location

Abstract Background: Colorectal cancer is a heterogeneous disease, exhibiting spectra of tumor characteristics that vary across age at diagnosis and anatomical location. We hypothesized that tumor immune and microbial characteristics of colorectal cancer might change with age at diagnosis and detailed anatomical location. Methods: We utilized a pooled dataset of colorectal cancer cases (N= 958 to1413 depending on variables) with available immune cell density or microbial data in the Nurses' Health Study, Health Professionals Follow-up Study, and Ontario Familial Colon Cancer Registry. We examined whether there were statistically significant trends in tumor immune and microbial according to age at diagnosis and detailed primary tumor location (cecum, ascending, hepatic flexure, transverse, splenic flexure, descending, sigmoid, rectosigmoid, rectum) using Spearman’s correlation test. The densities of immune cells were determined by using multiplex immunofluorescence, including T cells (CD3, CD4, CD8, CD45RA, CD45RO, FOXP3) and macrophages (CD68, M1-like, and M2-like) in tumor intraepithelial and stromal regions. We conducted stratified analyses by microsatellite instability (MSI) status. Results: The density of memory (both CD4+ and CD8+) T cells in tumor epithelial areas continuously increased as age at diagnosis increased (P for trend &amp;lt;0.005). Compared to later-onset colorectal cancer, early-onset colorectal cancer had a lower density of memory T-cells. With respect to tumor site, CD3+CD4+ and CD3+CD8+ T densities in tumor epithelial areas were highest in the cecum compared to other subsites, and there was a statistically significant decreasing trend from the cecum to rectum (P for trend &amp;lt;0.005). We also observed significant decreasing trends for CD68+, M1-like, and M2-like macrophage densities in combined tumor (intraepithelial and stromal) regions from the cecum to rectum (P for trend &amp;lt;0.005). The proportion of F. nucleatum-positive colorectal cancers gradually decreased from the cecum to rectum (P for trend &amp;lt;0.005). Similar trends were observed in separate analyses in non-MSI-high cases. Conclusions: This study suggests that certain tumor immune and microbial characteristics in the tumor microenvironment, specifically T cell and macrophage densities and Fusobacterium nucleatum positivity, change gradually according to age at diagnosis and/or detailed anatomical location. Age and anatomical location spectra in tumor immune and microbial characteristics will shed light on the etiology of colorectal cancer (including early-onset colorectal cancer), leading to better-personalized cancer prevention and treatment. Citation Format: Tomotaka Ugai, Yasutoshi Takashima, Yuxue Zhong, Juha P. Väyrynen, Daniel Buchanan, Jeroen Huyghe, Li Hsu, Conghui Qu, Sushma Thomas, Claire Thomas, Steve Gallinger, Robert Grant, Ulrike Peters, Amanda I. Phipps, Jonathan Nowak, Shuji Ogino. Tumor immune and microbial characteristics of colorectal cancer according to age at diagnosis and detailed anatomical location [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3467.