PURPOSE:The aim of this study was to conduct a quality improvement study testing 2 novel, web-based, electronic visits (E-visits) enabling men to make a decision regarding prostate cancer screening. MATERIALS AND METHODS:Three thousand two hundred screen-eligible men in Kaiser Permanente, Northern California, equally comprised of 4 self-reported race/ethnicity categories (Asian, Black, Hispanic, and White) were randomly assigned to one of 2 E-visits: one purely informational providing general information on the risks and benefits of screening (informational E-visit) and the second (interactive E-visit), incorporating a personalized estimate for the risk of having an elevated prostate-specific antigen (PSA) level based on the patient's age and race. Completion rates of the E-visits and whether a decision on screening was made were the main outcomes. RESULTS:The overall completion rate was similar by E-visit type (overall 15.6%, informational E-visit 16.2%, interactive E-visit 14.7%, P = .24). White men had the highest completion rate and Black men the lowest (18.5% vs 13.1%, P = .02). Older men aged 60 to 69 years had a higher completion rate than younger men aged 50 to 59 years (18.8% vs 15.4%, P = .045). Among those who completed an E-visit, those completing the interactive E-visit were more likely to make a decision regarding screening (98.7% vs 91.2%, P < .001) and were more likely to request a PSA test (96.2% vs 90.0%; P = .008) compared with those completing the informational E-visit. CONCLUSIONS:Race and age were associated with completion rates of these E-visits. Among men completing the visit, those completing the interactive E-visit were more likely to make a decision regarding screening and order a PSA test than those completing the informational E-visit.

BACKGROUND:The associations between adult lead exposure and late-life cognition are largely unknown. METHODS:In two cohorts, Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE, n = 1638) and Study of Healthy Aging in African Americans (STAR, n = 741), we assessed residential proximity to lead-releasing facilities for association with domain-specific cognition 2 years later. Linear regression models adjusted for age, sex, race/ethnicity, income, education, marital status, smoking status, and alcohol consumption. We meta-analyzed across cohorts. RESULTS:Average age was 76.1 (KHANDLE), 68.8 years (STAR); average residential distance to a lead facility was 8.2 km (KHANDLE), 3.6 km (STAR). In meta-analysis, for every 5 km closer a residence was located to a lead-releasing facility, episodic memory scores 2 years later were -0.05 (95% confidence interval: -0.08, -0.02) standard deviation lower. DISCUSSION:Residential proximity to a lead-releasing facility was associated with poorer cognition 2 years later among adults in two cohorts.

Much remains unknown about the genetics of immune system changes during pregnancy. We used SNP data in a pregnancy cohort to genetically investigate 47 immune biomarkers at two timepoints, along with change between timepoints (delta). We identified 19 biomarkers with significant SNP-based heritability and 34 with genome-wide significant signals, demonstrating genetic regulation. The same biomarkers measured in early- and mid-pregnancy shared about half of significant associations across timepoints, with enrichment for immune pathways. In contrast, delta showed enrichment in transcription factors and developmental processes. About half of suggestive associations overlapped with non-pregnancy associations. However, these data leave a substantial fraction of potentially timepoint-specific and pregnancy-unique findings. Nearby genes were enriched for placental expression, reinforcing the novelty of our results. We additionally explored the relationship between immune genetic associations and prior GWAS of pregnancy complications. Overall, we present the first multi-timepoint genetic study of immune profile in pregnancy.

BACKGROUND:Hypertensive disorders of pregnancy (HDP) cause significant maternal and fetal morbidity, but clinical tools to identify high-risk patients are lacking. Blood pressure (BP) trajectories in early pregnancy have predicted HDP. OBJECTIVES:In this study, we aimed to test these early BP trajectories in a distinct health system. METHODS:Electronic health record data were used to retrospectively identify patients with a singleton delivery from 2012 to 2020 in a large urban academic center. Patients with pre-existing hypertension or serious medical conditions were excluded. Each patient was assigned 1 of 6 previously identified BP trajectory groups from conception through 20 weeks' gestation. The primary outcome of HDP diagnosis, including preeclampsia and gestational hypertension, was identified using International Statistical Classification of Diseases codes. Multivariable logistic regression was used to model the relationship between patient factors and HDP diagnosis, and predictive model discrimination was assessed using C-statistics. RESULTS:Among 25,879 patients, 5,694 (21.9%) were diagnosed with HDP. After adjustment for age, race/ethnicity, BMI, parity, and pregestational diabetes, BP trajectory groups were independently associated with increasing odds of HDP, with adjusted odds of HDP diagnosis progressively increasing from 1.6 (95% CI: 1.4-1.8) to 6 (95% CI: 5.3-6.9) across trajectories. Inclusion of trajectories into predictive models improved discrimination compared to clinical risk factors alone (C-statistic 0.71 [95% CI: 0.70-0.72] vs 0.66 [95% CI: 0.65-0.67]). CONCLUSIONS:BP trajectories, based on early pregnancy BP measurements, improved ability to predict HDP diagnosis in diverse pregnant cohorts. BP trajectories may improve identification of high-risk patients for more intensive BP monitoring during pregnancy.

Background:A translation gap exists in how patients with chronic obstructive pulmonary disease (COPD) utilize mitigation strategies to limit exposure to wildfire smoke. This study examines patients’ points of view about barriers and facilitators of strategy uptake. Methods:We performed semistructured, virtual interviews with members of Kaiser Permanente Northern California until thematic saturation. We recruited participants aged ≥65 in the lowest quartile of socioeconomic status because they are disproportionately exposed to air pollution with fewer resources to mitigate exposure. Qualitative analysis was performed using inductive and deductive approaches. Results:Of 90,696 adults, we interviewed 31 in January 2025. Participants were racially and ethnically diverse (19% Black, 10% Hispanic, 3% Native American, 68% non-Hispanic White), from 10 counties. Three major themes were: (1) patients tended to get wildfire and air quality information from internet and smartphone apps, not clinical encounters, but expressed openness to receiving information from clinicians, (2) there appeared to be modifiable barriers to uptake of mitigation strategies, such as education and supplying equipment (e.g., masks, air cleaners), and (3) patients preferred real-time alerts sent to their phones from trusted sources, such as health care entities, to change their behavior during periods of poor air quality. Conclusion:By understanding patients’ perspectives about their relationship with wildfire smoke, we gained practical information to begin designing interventions to protect patients’ health during periods of poor air quality.

BACKGROUND: Organophosphate ester flame retardants and plasticizers (OPEs) have myriad uses in industry and consumer products. Increasing human exposure to OPEs has raised concerns about their potential effects on child neurodevelopment during the pregnancy. Objective: We investigated whether OPE urinary concentrations during pregnancy were associated with child's autism-related outcomes. METHODS: We included 4159 mother-child pairs from 15 cohorts in the NIH Environmental influences on Child Health Outcomes (ECHO) Consortium, with children born from 2006-2020 (median age [interquartile range]: 6 [4,10] years). Nine OPE biomarkers were measured in urine samples collected mid- to late pregnancy. Dilution-adjusted biomarkers were modeled continuously, categorically (high [>median], moderate [≤median], nondetect), or as detect/nondetect depending on their detection frequency. We assessed child autism-related traits via a) parent report on the Social Responsiveness Scale (SRS) and b) clinical autism diagnosis. We examined associations of OPEs with child outcomes, including modification by child sex, using generalized estimating equations to account for clustering by ECHO cohort. RESULTS: Compared with nondetectable concentrations, high exposure to bis-(butoxyethyl) phosphate (BBOEP) was associated with higher autistic trait scores (adj-β 0.97, 95% confidence interval [CI]: 0.42, 1.52) and greater odds of autism diagnosis (adjusted odds ratio [adj-OR]: 1.27, 95% CI: 1.07, 1.50). Bis-(1-chloro-2-propyl) phosphate (BCPP) showed associations with autistic trait scores (BCPP adj-β for high exposure vs nondetect: 0.34, 95% CI: -0.46, 1.13; BCPP adj-β for moderate exposure vs nondetect: 0.72, 95% CI: 0.24, 1.20). High exposure to bis-(2-chloroethyl) phosphate (BCETP) was associated with lower odds of autism diagnosis (adj-OR: 0.76, 95% CI: 0.60, 0.95). Other OPEs showed no associations in adjusted models. Associations between BBOEP and higher autistic trait scores were stronger in males than females. DISCUSSION: Prenatal exposure to OPEs, specifically BCPP and BBOEP, may be associated with a higher risk of autism diagnosis and related traits in childhood.

Introduction and Objective: Food Is Medicine interventions are being promoted to improve glycemic control in type 2 diabetes mellitus (T2D), but limited randomized trial (RCT) data exist. We evaluated the efficacy of a scalable, technology-based grocery prescription program in low-income, high-risk adults with T2D. Methods: We conducted a 1:1 RCT in Kaiser Permanente Northern & Southern California integrated healthcare systems from Oct 2023-Dec 2025 in adults with Medicaid insurance, T2D, and glycosylated hemoglobin (A1c) ≥7.5%. We excluded patients with advanced kidney disease, were institutionalized, pregnant, had no internet access, required an interpreter, or no valid credit/debit card. The intervention used the Instacart online platform and provided a $100/mo stipend for healthy foods (e.g., fruits, vegetables, legumes, nuts/seeds, etc.) with 4 free deliveries/mo for 6 mos. The control arm received 4 free Instacart deliveries/mo for 6 mos with no stipend. The primary outcome was 6-mo A1c using an intent-to-treat approach. Results: Among 1007 subjects (510 intervention, 506 control), both arms were balanced in mean age (53.2 vs. 53.1 years); women (67.9% vs. 66.6%); Asian/Native Hawaiian/Pacific Islander (14.4% vs. 12.3%), Black (26.1% vs. 25.1%), and Hispanic ethnicity (40.9% vs. 41.5%); and mean A1c at entry (9.34% vs. 9.21%). Registration to use Instacart (88.6% vs. 50.8%) and mean monthly orders (1.5 vs. 0.2) were higher in intervention vs. control subjects. Follow-up A1c was available in 93.4% of subjects. Both arms had 6-mo reductions in A1c (Intervention: -0.63%, 95% CI: -0.78%, -0.47%; Control: -0.53%, 95% CI: -0.68%, -0.38%). After adjusting for baseline A1c and stratification factors, the intervention arm did not have a significantly lower 6-mo A1c than the control arm (8.63% vs 8.65%; difference: -0.06%, 95% CI: -0.26%, 0.14%, p=0.55). Conclusion: For low-income, high-risk adults with T2D, provision of monthly funds to purchase healthy foods through an online shopping platform did not improve 6-mo glycemic control vs. usual care. Disclosure C. Nau: Research Support; Current; NewDays.ai. R.W. Grant: None. R.V. Parikh: None. T. Tan: None. J.C. Lo: None. P. Schwartz: None. J.D. Vallejo: None. S. Alexeeff: None. M. Habib: None. H. Zhong: None. C. Chao: None. A. Erkenbeck: None. D. Mozaffarian: Consultant; Current; Google. Consultant; Ended; Amazon Health. Advisory Panel; Current; Instacart Health, Nourish. A. Go: Research Support; Ended; Novartis AG. Research Support; Current; Bristol-Myers Squibb Company. Advisory Panel; Current; Bristol-Myers Squibb Company, Janssen Pharmaceuticals, Inc. Research Support; Ended; Edwards Lifesciences Corporation. Funding The Kaiser Permanente Community HealthFund at the East Bay Community Foundation

Background:Rigorous studies are needed to defend the current National Ambient Air Quality Standards and prompt policymakers to acknowledge a dose-response relationship between air pollution and respiratory outcomes. In the most recent Environmental Protection Agency comprehensive review, scientists determined there was insufficient evidence to claim a causal relationship between chronic particulate matter air pollution (PM2.5) and respiratory outcomes, despite acknowledging one for cardiovascular outcomes (Integrated Science Assessment, Table 5-27). Methods:This study examined the associations between chronic PM2.5 levels and respiratory outcomes in 169,713 patients within Kaiser Permanente Northern California who had chronic obstructive pulmonary disease (COPD) between 2007 and 2016. Results:For every 10 μg/m3 rise in average yearly PM2.5, there was a statistically significant increase in relative risk of death (hazard ratio [HR] = 1.15; 95% confidence interval [CI] = 1.10, 1.20), respiratory-related death (HR = 1.22; 95% CI = 1.15, 1.30), COPD exacerbation (HR = 1.35; 95% CI = 1.29, 1.41), and the combined endpoint of COPD exacerbation or respiratory-related hospitalization with any-cause death as competing risk (HR = 1.30; 95% CI = 1.25, 1.35). We demonstrate a dose-response relationship between 1-year average PM2.5 exposure and respiratory outcomes for key regulatory categories of PM2.5. Conclusions:These findings add rigorous evidence to support the continuation of the current annual PM2.5 standard, which was lowered to 9 μg/m3 in 2024. Our novel contribution is documenting the dose-response relationship between chronic PM2.5 and multiple respiratory outcomes in a longitudinal cohort with patients followed for up to 10 years.

SARS-CoV-2 infection in pregnancy has become common, yet very little is known about the impact of prenatal exposure on child development. Our objective was to examine the impact of infection with SARS-CoV-2 during pregnancy on child neurodevelopment in the first years of life. We conducted a longitudinal prospective cohort study among a diverse population of 69,987 children born January 2020-September 2021 in Northern California to members of an integrated healthcare system. Maternal SARS-CoV-2 infection during pregnancy was confirmed by polymerase chain reaction (PCR) test. All neurodevelopmental disorders (NDD) diagnosed in children by December 2023 were identified, including autism spectrum disorder (ASD), speech/language delay, and motor delay. Cox proportional hazards regression models estimated hazard ratios (HR) and 95% confidence intervals (CIs), with adjustment for maternal sociodemographic and clinical characteristics, SARS-CoV-2 vaccination status during pregnancy, and child sex. A total of 2777 (3.97%) pregnant individuals had PCR-confirmed SARS-CoV-2 infection during pregnancy. Among 69,987 children aged 27-48 months, 12,006 (17.15%) were diagnosed with NDD; 2724 (3.89%) with ASD, 10,047 (14.36%) with speech/language delay, and 2716 (3.88%) with motor delay. Maternal infection with SARS-CoV-2 during pregnancy was not associated with an increased risk of speech/language delay or motor delay but was associated with an elevated risk of ASD among females (aHR = 1.44, 95% CI 1.05-1.97) but not males (aHR = 1.04, 95% CI 0.83-1.31). Prenatal exposure to SARS-CoV-2 infection may increase risk for autism spectrum disorders among females. Future studies are needed to confirm and extend these findings.

We developed EHRsupport, an electronic health record (EHR)-based measure of social support from data documented within Kaiser Permanente Northern California. This study included 4,145 women from the Pathways Study diagnosed January 2006-September 2021 with invasive breast cancer, who had social support data in the EHR and Medical Outcomes Study social support (MOS-SS) survey data. We used a split sample design to generate random training (N=2,059) and validation (N=2,086) datasets. We applied a natural language processing (NLP) algorithm to capture social support variables in prior work and generated EHRsupport by fitting LASSO models to the MOS-SS score. In the validation dataset, we evaluated the correlation of EHRsupport and MOS-SS and compared associations of the scores with breast cancer treatment and survival outcomes. Our final LASSO model selected all candidate social support variables, missing categories for parenthood status and living situation, and eight interactions by age and/or stage. The resulting EHRsupport score was correlated with MOS-SS (r=0.25, p<0.001). Low EHRsupport (lowest tertile representing low social support) was related to surgery delays>30 days (OR=1.59, 95% confidence interval (CI): 1.15-2.18), adjuvant endocrine therapy discontinuation (hazard ratio (HR)=1.35, 95% CI: 1.05-1.75), overall mortality (HR=1.30, 95% confidence interval (CI): 1.01-1.66) and any invasive event (HR=1.30, 95% CI: 1.06-1.60); MOS-SS was associated with treatment delays only. Neither was associated with recurrence or cancer-specific mortality. EHRsupport was correlated with MOS-SS and was related to treatment and clinical outcomes in breast cancer patients supporting its use as a clinical tool and in research.

Maternal immune activation is a hypothesized mechanism by which prenatal exposure to infection adversely impacts fetal brain development. SARS-CoV-2 IgG class antibodies generated to viral proteins in pregnancy are transferred across the placenta to the newborn. This study aimed to investigate the association between neonatal levels of IgG antibodies to SARS-CoV-2 nucleocapsid and spike proteins and risk of neurodevelopmental disorders (ND) among children with gestational SARS-CoV-2 exposure. Using a longitudinal prospective cohort design within an integrated healthcare system with a diverse patient population in Northern California, the study looked at 399 children born to mothers who had SARS-CoV-2 infection during pregnancy and had not received the COVID-19 vaccination. Exposure was IgG and IgM class antibodies to SARS-CoV-2 nucleocapsid, IgG directed against the full-length spike protein, the s1-receptor binding domain (RBD) and s1 N-terminal domain (NTD) of the spike protein, and IgM to the full length and RBD spike proteins. The main outcome and measure were all ND, including autism spectrum disorder (ASD), speech/language delay, and motor delay. Cox proportional hazards regression models estimated hazard ratios (HR) and 95% confidence intervals (CIs), with adjustment for maternal sociodemographic and clinical characteristics, and child sex. Among the 399 study children (208 females,191 males) followed up to 64 months, 95% had gestational age ≥ 37 weeks and birthweight ≥ 2500 g. The mean maternal age was 31 +- 5 years; 12% were Asian, 5% Black, 45% Hispanic, 33% White, and 5% other/unknown race/ethnicity; 80% had more than a High School education; 86% had commercial insurance; 37% were primiparous; and 35% had pre-pregnancy obesity. A total of 73 children were diagnosed with ND in follow-up (22 ASD, 67 speech/language disorder, 11 motor delay). Associations between SARS-CoV-2 nucleocapsid IgG antibody level and risk of ND differed significantly by child sex, with increased risk observed only among males (ND: aHR = 1.25 [1.01-1.56]; ASD: aHR = 1.38 [0.93-2.06]; speech/language delay: aHR = 1.33 [1.06-1.67]). Risk also differed by trimester of maternal SARS-CoV-2 infection, with over 2-fold increased risks observed with exposure in the first trimester (ND: aHR = 2.12 [1.08-4.15]; speech/language delay: aHR = 2.16 [1.06-4.47]) and second trimester (speech/language delay: aHR = 2.22 [1.23-4.00]). Results were similar for the SARS-CoV-2 spike, RBD and NTD proteins. Few of the neonates had detectable levels of IgM antibodies to the SARS-CoV-2 proteins. In conclusion, higher levels of neonatal IgG antibodies to SARS-CoV-2 proteins following infection in pregnancy may indicate increased risk for ND. If replicated, these findings suggest that newborn antibody testing of infants with prenatal SARS-CoV-2 exposure may provide a novel avenue for early identification of infants at risk for ND and opportunities for early intervention and prevention of adverse child outcomes.

Much remains unknown about the genetics of immune system changes during pregnancy. We used SNP data in a pregnancy cohort to genetically investigate 47 immune biomarkers at two timepoints, along with change between timepoints (Δ). We identified 19 biomarkers with significant SNP-based heritability and 34 with genome-wide significant signals, demonstrating genetic regulation. The same biomarkers measured in early- and mid-pregnancy shared about half of significant associations across timepoints, with enrichment for immune pathways. In contrast, Δ showed enrichment in transcription factors and developmental processes. About half of suggestive associations overlapped with non-pregnancy associations. However, these data leave a substantial fraction of potentially timepoint-specific and pregnancy-unique findings. Nearby genes were enriched for high expression in decidual cells at the maternal-fetal interface, reinforcing the novelty of our results. We additionally explored the relationship between immune genetic associations and prior GWAS of pregnancy complications. Overall, we present the first two-timepoint genetic study of immune profile in pregnancy.

Introduction: Accurately predicting 10-year ASCVD risk is key to guide primary prevention, yet little is known about how the AHA PREVENT ASCVD equation performs in disaggregated Asian American, Native Hawaiian, and Pacific Islander (AANHPI) adults. Hypothesis: We hypothesized the performance of the PREVENT ASCVD equations varies across contemporary AANHPI groups. Goal: To evaluate discrimination and calibration of the PREVENT ASCVD equations in 8 major AANHPI groups (Chinese, Filipino, Native Hawaiian/other Pacific Islander [NH/PI], Japanese, Korean, South Asian, Vietnamese, other Southeast Asian) and Non-Hispanic Whites (NHW). Methods: We identified all members aged 30-79 years of Kaiser Permanente (KP) Northern California and KP Hawaii integrated healthcare delivery systems from 2012-2022 who had no prior CVD and had complete data to estimate the PREVENT ASCVD risk using the base and full models. Incident ASCVD events (non-fatal and fatal MI and stroke) were identified through December 2023 using validated ICD-9/10 diagnosis codes in EHR and vital status data, and observed risk calculated using a cause-specific risk model that accounted for competing risk of non-ASCVD death. We examined model discrimination (C-index) and calibration (calibration plot and slope) by individual AANHPI group. Results: We identified 1,219,255 eligible adults, with mean (SD) age ranging from 44.4 (13.2) years in South Asian to 61.0 (15.4) years in Japanese; 54.6% were women and there were higher proportions of women in Japanese, Korean, and NHW groups. Discrimination of the PREVENT base and full models was good overall and consistent across AANHPI individuals, with C statistics ranging from 0.77 (Japanese, Korean) to 0.82 (Chinese, South Asian) ( Figure 1-3 ). However, calibration for the PREVENT base equation varied across AANHPI individuals, with underestimation of actual 10-year ASCVD risk in NH/PI, South Asian and other Southeast Asian adults, and overestimation in Chinese, Japanese, Korean and Vietnamese adults. The PREVENT full equation lowered predicted ASCVD risk in all groups ( Figure ). Conclusions: In a large, contemporary, primary prevention population of AANHPI and NHW adults in California and Hawaii, the PREVENT ASCVD equations had good discrimination but variable calibration across 8 major AANHPI individual groups. Given primary prevention decisions rely on predicted absolute risks, our findings support further optimizing ASCVD risk equations in targeted AANHPI groups.

OBJECTIVES:Tobacco smoking and alcohol use may negatively influence HIV care, but associations have not been examined across cohorts. DESIGN:Multisite international collaboration of cohort studies. METHODS:People with HIV (PWH) were included from 11 cohorts; 5 North American and 6 Western European. Exposures were harmonized smoking and alcohol measures (2010-2018). Loss to care was defined as not having 2+ HIV care visits (HIV RNA and/or CD4 measurement dates) at least 60 days apart, within 12 months following alcohol measure date; HIV viral nonsuppression was defined as >200 copies/ml. Adjusted prevalence ratios (PRs) were estimated using modified Poisson regression; pooled effect estimates and the heterogeneity measure ( I2 ).were derived from a random-effect meta-analysis. RESULTS:Among 83 102 PWH (87.4% male, 46.1% white); 43.7% currently smoked, 44.5% reported low/moderate drinking, 6.9% heavy drinking, 48.6% did not drink. PWH who currently smoked had higher risk of loss to care than nonsmoking PWH (pooled PR [95% CI] = 1.12 [1.08-1.16], I2  = 18.1%); those with heavy drinking had higher risk than those with low/moderate drinking (1.13 [1.03-1.25], I2  = 57.8%). PWH who currently smoked had higher risk of viral nonsuppression than nonsmoking PWH (1.44 [1.25-1.67], I2  = 90.6%); those reporting heavy drinking had higher risk than those with low/moderate drinking (pooled PR [95% CI] = 1.18 [1.02-1.37], I2  = 68.9%). PWH who reported heavy drinking and current smoking, in comparison to low/moderate alcohol use but no current smoking, had highest risk of viral nonsuppression (pooled PR [95% CI] =1.74 [1.37-2.22]), I2  = 81.8%. CONCLUSIONS:Smoking and unhealthy alcohol use were associated with HIV loss to care and viral nonsuppression, with variability between cohorts.

INTRODUCTION:Breastfeeding is recommended. It is unknown whether preconception or prenatal cannabis use is related to breastfeeding behaviors. METHODS:This population-based retrospective cohort study included 200,207 pregnancies in Northern California (2016-2022) with live births screened in early pregnancy for cannabis use. Exposures included prenatal cannabis use, preconception cannabis use only, or no cannabis use. Additional analyses considered the frequency of prenatal cannabis use. Longitudinal breastfeeding outcomes assessed at each well-child visit during the first year included any breastfeeding and full breastfeeding (breastmilk without formula). Adjusted prevalence ratios were calculated using modified Poisson regression for longitudinal binary outcomes. Analyses were conducted in 2024 and 2025. The risk of stopping breastfeeding among those who started was modeled using Cox proportional hazard regression. RESULTS:Overall, 7.6% of pregnancies had preconception cannabis use only, and 7.2% had prenatal use. Most people (94.6% overall) initiated breastfeeding, with only modest differences by cannabis use (94.9% no cannabis use, 95.7% preconception cannabis use only, and 90.5% prenatal cannabis use). However, over time, prenatal cannabis use was associated with earlier discontinuation of breastfeeding (adjusted hazard ratio=1.12; 95% CI=1.09, 1.15) and lower prevalence of breastfeeding (adjusted prevalence ratio=0.84; 95% CI=0.82, 0.85 at 6 months and adjusted prevalence ratio=0.81; 95% CI=0.78, 0.83 at 12 months). Associations were stronger for higher-frequency use. There were small differences in breastfeeding between those with preconception cannabis use only and those with no use. Full breastfeeding results were similar. CONCLUSIONS:Despite high prevalence of breastfeeding initiation, prenatal cannabis use was associated with earlier breastfeeding discontinuation and lower prevalence at 6 and 12 months.

Importance Gestational diabetes (GD) is a heterogeneous condition that predisposes both mother and offspring to metabolic disorders. GD subtypes defined by antepartum testing results have been associated with adverse perinatal outcomes, but little is known about their relationship to maternal metabolic outcomes soon after pregnancy.Objective To evaluate early postpartum glucose tolerance reclassification of GD subtypes.Design, Setting, and Participants This cohort study examined women from the Study of Women, Infant Feeding, and Type 2 Diabetes Mellitus After GD Pregnancy (SWIFT), who were recruited within the Kaiser Permanente Northern California integrated health care system between 2008 and 2011. All women were diagnosed with GD using Carpenter and Coustan criteria with complete glucose measurements at all 4 time points of the diagnostic 3-hour 100-gram oral glucose tolerance test (OGTT). Data analyses were conducted from January to July 2025.ExposureThree subtypes of GD based on the diagnostic OGTT: (1) postload glucose intolerance (GD-P), as having elevations only at 2 or more postload time points; (2) fasting hyperglycemia (GD-F), as having elevations at fasting and 1 postload time point; and (3) both (GD-M), as having elevations at fasting and 2 or more post-load time points.Main Outcomes and Measures At 6 to 9 weeks after delivery, glucose tolerance classification was evaluated using 2-hour, 75-g OGTTs. Modified Poisson regression models were used to estimate adjusted prevalence ratios (PRs) of postpartum prediabetes associated with GD subtypes, without and with adjustments for age, race and ethnicity, prepregnancy body mass index, educational level, and gestational weight gain.Results This study included 1005 women with GD (median [IQR] age, 33.2 [29.8-36.7] years; 368 [36.6%] Asian, 78 [7.8%] Black, 308 [30.6%] Hispanic, 16 [1.6%] multiracial, and 235 [23.4%] White). Prevalence of postpartum prediabetes was 34.5% (347 women), with wide variation across GD subtypes; 23.9% (147 of 616), 41.9% (52 of 124), and 55.8% (148 of 265) for GD-P, GD-F, and GD-M, respectively. Compared with women with GD-P, the adjusted PR for GD-F was 1.74 (95% CI, 1.36-2.24), and for GD-M, it was 2.23 (95% CI, 1.85-2.68) (both P < .001). Pairwise comparisons between GD-F and GD-M were also statistically significant (adjusted PR, 1.28; 95% CI, 1.01-1.61; P = .04).Conclusions and Relevance In this cohort study, GD subtypes had distinct postpartum prediabetes risks. Early action and intervention to address dysglycemia may be most beneficial for women with fasting or mixed defects.

Introduction: Atherosclerotic cardiovascular disease (ASCVD) risk varies substantially across racial and ethnic groups, yet few data exist among disaggregated Asian American, Native Hawaiian, and Pacific Islander (AANHPI) subgroups. Methods: We identified 2,653,007 members of Kaiser Permanente Northern California and Kaiser Permanente Hawaii integrated healthcare delivery systems from 2012-2022 who were aged ≥30 years with no evidence of prior cardiovascular disease. ASCVD events (acute myocardial infarction and stroke) were identified through December 2023 using validated discharge codes and death certificates. We calculated age- and sex-adjusted rates of incident ASCVD by racial/ethnic subgroup. We then examined the multivariable association between AANHPI subgroup with incident ASCVD compared to non-Hispanic White, after adjustment for age, sex, diabetes, hypertension, dyslipidemia, chronic kidney disease, body mass index, and tobacco use. Results: Between 2012-2022, we identified 182,776 Chinese, 193,327 Filipino, 64,488 Native Hawaiian/other Pacific Islander, 45,502 Japanese, 21,989 Korean, 92,738 South Asian, 50,141 Vietnamese, 26,602 other Southeast Asian, and 1,975,444 non-Hispanic White eligible adults. Mean (SD) age was 49 (15) years overall, ranging from 41 (12) years in South Asian to 55 (16) years in Japanese, with 53% women and higher proportions of women in AANHPI subgroups compared to non-Hispanic Whites. Age- and sex-adjusted rates (per 1000 person-years) of incident ASCVD vs. non-Hispanic Whites (3.61) from highest to lowest in AANHPI subgroups were: Native Hawaiian/other Pacific Islander (6.97), other Southeast Asian (5.37), South Asian (4.85), Filipino (4.23), Vietnamese (3.28), Japanese (3.20), and Chinese (2.50). After adjustment for traditional mediators of cardiovascular risk, compared to non-Hispanic Whites, incident ASCVD was higher for Native Hawaiian/other Pacific Islanders, South Asians, and other Southeast Asians; lower for Chinese, Japanese and Korean; and not significantly different for Filipino and Vietnamese ( Figure ). Conclusions: In a large, contemporary population in California and Hawaii, notable variation existed in the risk of incident ASCVD across AANHPI subgroups that was only partially explained by differences in demographic characteristics and clinical ASCVD risk factors. Delineating AANHPI-specific factors will help to reduce disparities in ASCVD in these growing populations within the U.S.

Importance Chronic hypertension and preeclampsia are leading risk enhancers for maternal-neonatal morbidity and mortality. Severe maternal morbidity (SMM) indicators include heart, kidney, and liver disease, but studies have not excluded patients with preexisting diseases that define SMM. Thus, SMM risks for uncomplicated chronic hypertension specific to preeclampsia remain unclear. Objective To determine SMM rates and estimate relative risks associated with hypertensive disorders of pregnancy among patients with and without chronic hypertension unencumbered by preexisting vascular or end organ diseases. Design, Setting, and Participants This retrospective cohort study used longitudinal health data from electronic health records from patients within a community-based, integrated health care system in northern California. The study cohort selected 263 518 pregnant patients without pregestational heart, kidney, or liver disease entering prenatal care at 14 weeks’ gestation or earlier and delivering a singleton stillbirth or live birth in 2009 to 2019. The data were analyzed between February 2022 and March 2024. Exposures Five joint subgroups combining chronic hypertension status and the hypertensive disorders developing during pregnancy, defined as follows: (1) chronic hypertension with superimposed preeclampsia, (2) chronic hypertension and no preeclampsia, (3) no chronic hypertension with preeclampsia, (4) gestational hypertension, and (5) no chronic hypertension and no preeclampsia or gestational hypertension (reference group). Main Outcomes and Measures The main outcome was SMM rates at delivery hospitalization (cases per 10 000 births) using the Centers for Disease Control and Prevention criteria (≥1 of 21 indicators to define SMM) obtained from electronic health records. Modified Poisson regression models estimated crude and adjusted relative risks (aRRs) and 95% CIs of SMM associated with the chronic hypertension and developing hypertensive disorders of pregnancy groups vs the reference group (no chronic hypertension and no preeclampsia or gestational hypertension) adjusted for clinical, sociodemographic, social, and behavioral covariates. Results The analysis included a total of 263 518 pregnant patients (mean [SD] age at delivery, 31.0 [5.3] years), including 249 892 patients without chronic hypertension (4.7% developed preeclampsia) and 13 626 patients with chronic hypertension (31.5% developed superimposed preeclampsia). The highest SMM rates occurred in the no chronic hypertension with preeclampsia (934.3 [95% CI, 882.3-988.3] cases per 10 000 births) and the chronic hypertension with superimposed preeclampsia (898.3 [95% CI, 814.5-987.8] cases per 10,000 births) groups. Lower SMM rates occurred in the chronic hypertension and no preeclampsia (195.1 [95% CI, 168.0-225.2] cases per 10,000 births), gestational hypertension (312.7 [95% CI, 281.6-346.1] cases per 10,000 births), and no chronic hypertension and no preeclampsia or gestational hypertension (165.8 [95% CI, 160.6-171.2] cases per 10,000 births) groups (P < .001). Compared with the no chronic hypertension and no preeclampsia or gestational hypertension group, risks of SMM were significantly higher for the chronic hypertension with superimposed preeclampsia group (aRR, 4.97 [95% CI, 4.46-5.54]), no chronic hypertension with preeclampsia group (aRR, 5.12 [95% CI, 4.79-5.48]), chronic hypertension and no preeclampsia group (aRR, 1.17 [95% CI, 1.003-1.36]; P = .046), and the gestational hypertension group (aRR, 1.78 [95% CI 1.60-1.99]). Conclusions and Relevance This cohort study found that the highest SMM rates at delivery hospitalization occurred for preeclampsia superimposed on chronic hypertension and preeclampsia without chronic hypertension, while gestational hypertension had intermediate rates of SMM. The patients with chronic hypertension who did not develop preeclampsia had SMM rates that were nearly the same as the lowest-risk patients without chronic hypertension who did not develop preeclampsia or gestational hypertension. These findings provide evidence that prevention of preeclampsia among patients with uncomplicated chronic hypertension is paramount to mitigating maternal morbidity.

Introduction:Anthropogenic climate change, according to the WHO, results in approximately 150,000 deaths annually through mechanisms such as heat-related mortality, altered food production, and the spread of infectious diseases. With climate change predicted to cause over half a million climate-related deaths by 2050, healthcare systems, which contribute significantly to greenhouse gas emissions, must adopt roles in environmental stewardship. Methods:This descriptive case study details how Kaiser Permanente, a large nonprofit organization, became the first carbon-neutral healthcare system in the country. Results:Kaiser Permanente has demonstrated environmental stewardship through initiatives such as an on-site solar program, sustainability scorecards for suppliers, and extensive partnerships with organizations to support community health and environmental building efforts. Initiatives included scaling renewable energy usage, constructing Leadership in Energy and Environmental Design-certified facilities, and reducing water use intensity led to Kaiser Permanente being the first carbon-neutral health system in the country. Conclusions:The WHO has declared climate change as the most significant threat to human health. Kaiser Permanente's journey to carbon neutrality highlights the critical role healthcare systems play in environmental stewardship. Continued focus on climate initiatives by the healthcare sector is essential to address the growing health impacts of climate change. Kaiser Permanente's efforts provide a real-life and practical framework for achieving significant positive climate effects.