Pub Date : 2025-09-19DOI: 10.1016/j.annepidem.2025.09.013
Luisa N. Borrell , Julia Díez , Nerea Lanborena , Sara Yago-Gonzalez , Elena Rodriguez-Alvarez
Purpose
We aimed to examine intersectional BMI inequities across age, sex/gender, immigration status, and education among adults in Spain.
Methods
We analyzed data from 61,844 adults aged ≥ 18 years from the 2014 and 2020 European Health Interview Surveys in Spain and the 2017 Spanish National Health Survey. Using intersectional multilevel analysis of individual heterogeneity and discriminatory accuracy (I-MAIHDA) via linear mixed models, we examined BMI inequities across 180 intersectional strata, defined by age, sex/gender, immigration status, education, and survey year, and quantified the contribution of their intersections to BMI inequities.
Results
The intersectional strata explained 9.6 % of BMI inequities, with 87.1 % of the between-strata variance explained by age, sex/gender, immigration status, education, and survey year. On average, BMI was higher in older adults, immigrants, and those with lower education but lower in women. Interaction effects revealed that immigrant women had a higher predicted BMI than Spanish-born women. BMI inequities between immigrant and Spanish-born women were greatest for those aged 55–64 with middle and high education.
Conclusion
Although most BMI inequities were explained by the social factors considered as inequity axes, interaction effects were present. The latter calls for a universal public health intervention proportionate to the needs of specific groups in the population. I-MAIHDA revealed complex patterns of BMI inequities in Spain, which may inform the interventions needed to address weight-related outcomes.
{"title":"Body mass index inequities among adults in Spain: An intersectional multilevel analysis of individual heterogeneity and discriminatory accuracy of age, sex/gender, immigration status, and education","authors":"Luisa N. Borrell , Julia Díez , Nerea Lanborena , Sara Yago-Gonzalez , Elena Rodriguez-Alvarez","doi":"10.1016/j.annepidem.2025.09.013","DOIUrl":"10.1016/j.annepidem.2025.09.013","url":null,"abstract":"<div><h3>Purpose</h3><div>We aimed to examine intersectional BMI inequities across age, sex/gender, immigration status, and education among adults in Spain.</div></div><div><h3>Methods</h3><div>We analyzed data from 61,844 adults aged ≥ 18 years from the 2014 and 2020 European Health Interview Surveys in Spain and the 2017 Spanish National Health Survey. Using intersectional multilevel analysis of individual heterogeneity and discriminatory accuracy (I-MAIHDA) via linear mixed models, we examined BMI inequities across 180 intersectional strata, defined by age, sex/gender, immigration status, education, and survey year, and quantified the contribution of their intersections to BMI inequities.</div></div><div><h3>Results</h3><div>The intersectional strata explained 9.6 % of BMI inequities, with 87.1 % of the between-strata variance explained by age, sex/gender, immigration status, education, and survey year. On average, BMI was higher in older adults, immigrants, and those with lower education but lower in women. Interaction effects revealed that immigrant women had a higher predicted BMI than Spanish-born women. BMI inequities between immigrant and Spanish-born women were greatest for those aged 55–64 with middle and high education.</div></div><div><h3>Conclusion</h3><div>Although most BMI inequities were explained by the social factors considered as inequity axes, interaction effects were present. The latter calls for a universal public health intervention proportionate to the needs of specific groups in the population. I-MAIHDA revealed complex patterns of BMI inequities in Spain, which may inform the interventions needed to address weight-related outcomes.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"111 ","pages":"Pages 74-81"},"PeriodicalIF":3.0,"publicationDate":"2025-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145114859","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-18DOI: 10.1016/j.annepidem.2025.09.012
Sofia Anania MD , Marian K. Bakker PhD , Jorieke E.H. Bergman MD PhD , Leonie K. Duin MD PhD , Mike Ruettermann MD PhD , Dieuwke C. Broekstra PhD
Objectives
To describe trends in total and live birth (LB) prevalence of orofacial clefts (OFCs) and investigate the effect of prenatal diagnosis on its LB prevalence in the Northern Netherlands (NNL) over a 21-year period.
Study design
Cases with OFCs were selected from Eurocat NNL, a population-based registry of congenital anomalies. Aggregated denominator data (controls) was derived from Statistics Netherlands. We categorized OFCs into cleft lip with or without palate (CL±P) and cleft palate (CP). We analyzed trends in prevalence and temporal patterns in prenatal detection rates and birth type by calculating the annual percentage change (APC) using Joinpoint regression or the X² test for linear trends.
Results
The total prevalence of CL±P decreased non-significantly from 1.65/1000 births in 2001 to 1.17/1000 births in 2021 (APC=-0.82 [95 %CI: −2.18, 0.55]), whereas the LB prevalence decreased significantly from 1.57/1000 births in 2001 to 0.93/1000 births in 2021 (APC=-1.65 [95 %CI −3.03, −0.25]). The total and LB prevalence of CP did not show significant decreasing trends.
The prenatal detection rate of CL±P increased from 15 % to 71 % between 2001 and 2007 (APC=34.38 [95 %CI 24.27, 52.26]), followed by a slower increase, reaching 94 % in 2021 (APC=3.57 [95 %CI 0.59, 6.08]). The percentage of induced abortions increased from 3.2 % to 12.7 % between 2001 and 2021 (X2 for trend P < 0.001), with most induced abortions occurring in the CL±P associated with genetic/syndromic or other congenital anomalies.
Conclusions
The LB prevalence of CL±P decreased, which may be associated with better prenatal detection and increasing pregnancy terminations due to unfavorable prognosis
{"title":"Trends in orofacial cleft prevalence and the effect of prenatal detection on pregnancy outcomes in Northern Netherlands","authors":"Sofia Anania MD , Marian K. Bakker PhD , Jorieke E.H. Bergman MD PhD , Leonie K. Duin MD PhD , Mike Ruettermann MD PhD , Dieuwke C. Broekstra PhD","doi":"10.1016/j.annepidem.2025.09.012","DOIUrl":"10.1016/j.annepidem.2025.09.012","url":null,"abstract":"<div><h3>Objectives</h3><div>To describe trends in total and live birth (LB) prevalence of orofacial clefts (OFCs) and investigate the effect of prenatal diagnosis on its LB prevalence in the Northern Netherlands (NNL) over a 21-year period.</div></div><div><h3>Study design</h3><div>Cases with OFCs were selected from Eurocat NNL, a population-based registry of congenital anomalies. Aggregated denominator data (controls) was derived from Statistics Netherlands. We categorized OFCs into cleft lip with or without palate (CL±P) and cleft palate (CP). We analyzed trends in prevalence and temporal patterns in prenatal detection rates and birth type by calculating the annual percentage change (APC) using Joinpoint regression or the <em>X²</em> test for linear trends.</div></div><div><h3>Results</h3><div>The total prevalence of CL±P decreased non-significantly from 1.65/1000 births in 2001 to 1.17/1000 births in 2021 (APC=-0.82 [95 %CI: −2.18, 0.55]), whereas the LB prevalence decreased significantly from 1.57/1000 births in 2001 to 0.93/1000 births in 2021 (APC=-1.65 [95 %CI −3.03, −0.25]). The total and LB prevalence of CP did not show significant decreasing trends.</div><div>The prenatal detection rate of CL±P increased from 15 % to 71 % between 2001 and 2007 (APC=34.38 [95 %CI 24.27, 52.26]), followed by a slower increase, reaching 94 % in 2021 (APC=3.57 [95 %CI 0.59, 6.08]). The percentage of induced abortions increased from 3.2 % to 12.7 % between 2001 and 2021 (<em>X</em><sup><em>2</em></sup> for trend P < 0.001), with most induced abortions occurring in the CL±P associated with genetic/syndromic or other congenital anomalies.</div></div><div><h3>Conclusions</h3><div>The LB prevalence of CL±P decreased, which may be associated with better prenatal detection and increasing pregnancy terminations due to unfavorable prognosis</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"111 ","pages":"Pages 44-50"},"PeriodicalIF":3.0,"publicationDate":"2025-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145103189","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-16DOI: 10.1016/j.annepidem.2025.09.010
Noah Mancuso , Patrick S. Sullivan
Purpose
To introduce an equity-based method for assessing public transit access to health services and apply it to pre-exposure prophylaxis (PrEP) clinics in metro-Atlanta.
Methods
Census block groups (CBGs) were analyzed with PrEP clinics identified via PrEP Locator. One-way public transit times were estimated using the Google Maps Distance Matrix API. CBGs were classified as public transit deserts if transit options were unavailable or if travel time was > 30 min. T-tests compared sociodemographic characteristics of CBGs with and without public transit. Linear regression assessed the association of a 5 % increase in priority populations with transit times.
Results
Among 2466 CBGs, one-quarter lacked public transit access to PrEP and two-thirds were transit deserts. Median travel time was 32 min. CBGs with transit access had significantly higher proportions of Black, Hispanic/Latinx, young men (aged 25–34), and residents living below the poverty line (P < .001). Increases in the proportion of Hispanic/Latinx residents, young men, and residents living under the poverty line were associated with shorter transit times, with no association for Black residents.
Conclusions
Public transit access to PrEP was low in Atlanta, and overall public transit times were long. Current PrEP locations are aligned with priority populations, but additional work is needed to ensure equity is met for Black and Hispanic/Latinx residents.
{"title":"Methods for estimating public transit travel times to healthcare services as a measure of equitable healthcare access","authors":"Noah Mancuso , Patrick S. Sullivan","doi":"10.1016/j.annepidem.2025.09.010","DOIUrl":"10.1016/j.annepidem.2025.09.010","url":null,"abstract":"<div><h3>Purpose</h3><div>To introduce an equity-based method for assessing public transit access to health services and apply it to pre-exposure prophylaxis (PrEP) clinics in metro-Atlanta.</div></div><div><h3>Methods</h3><div>Census block groups (CBGs) were analyzed with PrEP clinics identified via <em>PrEP Locator</em>. One-way public transit times were estimated using the <em>Google Maps Distance Matrix</em> API. CBGs were classified as public transit deserts if transit options were unavailable or if travel time was > 30 min. T-tests compared sociodemographic characteristics of CBGs with and without public transit. Linear regression assessed the association of a 5 % increase in priority populations with transit times.</div></div><div><h3>Results</h3><div>Among 2466 CBGs, one-quarter lacked public transit access to PrEP and two-thirds were transit deserts. Median travel time was 32 min. CBGs with transit access had significantly higher proportions of Black, Hispanic/Latinx, young men (aged 25–34), and residents living below the poverty line (P < .001). Increases in the proportion of Hispanic/Latinx residents, young men, and residents living under the poverty line were associated with shorter transit times, with no association for Black residents.</div></div><div><h3>Conclusions</h3><div>Public transit access to PrEP was low in Atlanta, and overall public transit times were long. Current PrEP locations are aligned with priority populations, but additional work is needed to ensure equity is met for Black and Hispanic/Latinx residents.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"111 ","pages":"Pages 24-29"},"PeriodicalIF":3.0,"publicationDate":"2025-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145088041","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-16DOI: 10.1016/j.annepidem.2025.09.009
Kiran Thapa PhD , Ye Shen PhD , José F. Cordero MD, MPH , Emily Anne Vall PhD , Janani Rajbhandari-Thapa PhD
Purpose
We examined whether adverse childhood experiences (ACEs) are associated with obesity in young adulthood, and whether these associations differ by sex.
Methods
We used data from the National Longitudinal Study of Adolescent to Adult Health, a nationally representative cohort of U.S. adolescents followed into adulthood (age 33–43 years) across five waves. Our sample included 5193 participants with measured anthropometrics at wave V (2016–18). Modified Poisson regression estimated risk ratios (RR) for general obesity (body mass index ≥ 30 kg/m²) and abdominal obesity (waist circumference >102 cm for males, >88 cm for females) associated with individual and cumulative ACEs, adjusting for baseline BMI, co-occurring ACEs, and sociodemographic covariates. Sex-stratified models assessed heterogeneity in effects.
Results
Childhood physical abuse was independently associated with higher risk of general obesity, particularly among females (aRR: 1.23; 95 % CI: 1.05–1.45). Exposure to ≥ 4 ACEs was associated with increased risk of both general (aRR: 1.32; 95 % CI: 1.15–1.52) and abdominal obesity (aRR: 1.18; 95 % CI: 1.02–1.37), independent of childhood obesity.
Conclusions
ACEs, especially physical abuse and cumulative exposure, were linked to higher risk of obesity, suggesting that traumatic events may play an important role in young adulthood obesity, especially in females.
{"title":"Associations between adverse childhood experiences and obesity among young US adults","authors":"Kiran Thapa PhD , Ye Shen PhD , José F. Cordero MD, MPH , Emily Anne Vall PhD , Janani Rajbhandari-Thapa PhD","doi":"10.1016/j.annepidem.2025.09.009","DOIUrl":"10.1016/j.annepidem.2025.09.009","url":null,"abstract":"<div><h3>Purpose</h3><div>We examined whether adverse childhood experiences (ACEs) are associated with obesity in young adulthood, and whether these associations differ by sex.</div></div><div><h3>Methods</h3><div>We used data from the National Longitudinal Study of Adolescent to Adult Health, a nationally representative cohort of U.S. adolescents followed into adulthood (age 33–43 years) across five waves. Our sample included 5193 participants with measured anthropometrics at wave V (2016–18). Modified Poisson regression estimated risk ratios (RR) for general obesity (body mass index ≥ 30 kg/m²) and abdominal obesity (waist circumference >102 cm for males, >88 cm for females) associated with individual and cumulative ACEs, adjusting for baseline BMI, co-occurring ACEs, and sociodemographic covariates. Sex-stratified models assessed heterogeneity in effects.</div></div><div><h3>Results</h3><div>Childhood physical abuse was independently associated with higher risk of general obesity, particularly among females (aRR: 1.23; 95 % CI: 1.05–1.45). Exposure to ≥ 4 ACEs was associated with increased risk of both general (aRR: 1.32; 95 % CI: 1.15–1.52) and abdominal obesity (aRR: 1.18; 95 % CI: 1.02–1.37), independent of childhood obesity.</div></div><div><h3>Conclusions</h3><div>ACEs, especially physical abuse and cumulative exposure, were linked to higher risk of obesity, suggesting that traumatic events may play an important role in young adulthood obesity, especially in females.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"111 ","pages":"Pages 51-57"},"PeriodicalIF":3.0,"publicationDate":"2025-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145088071","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-14DOI: 10.1016/j.annepidem.2025.09.003
Akshaya Srikanth Bhagavathula PhD, FACE
Purpose
This study used artificial intelligence (AI) and natural language processing (NLP) to analyze patient reviews of semaglutide, with the goal of better understanding its real-world effectiveness and safety for weight management.
Methods
A retrospective, cross-sectional analysis was conducted on 772 user-generated reviews of semaglutide posted on Drugs.com (July 2021–March 2025). Sentiment analysis was performed using a transformer-based BERT model on a five-point scale. Topic modeling with Latent Dirichlet Allocation (LDA) and Latent Semantic Analysis (LSA) was used to identify dominant themes. Cluster analysis was applied to segment users based on weight loss outcomes and side effect severity. Reviewers (n = 95) that explicitly reporting both weight loss and treatment duration were analyzed for real-world efficacy and safety.
Results
Users who took semaglutide more than 60 days reported a mean weight loss of 32.2 ± 3.1 lbs (14.6 kg). Frequently mentioned side effects included nausea (46.9 %), headache (18.4 %), vomiting (14.3 %), fatigue (9.2 %), and dizziness (4.8 %). The highest sentiment scores were observed in the ≤ 30-day group (mean: 3.38). Topic modeling identified themes such as appetite suppression, medication cost and access, and long-term experiences. Clusters analysis revealed distinct user profile, including super-responder group with substantial weight loss and another with more side effects.
Conclusions
AI and NLP methods offer valuable tools for analyzing patient-reported outcomes, revealing semaglutide’s real-world efficacy and safety profile for weight management. These findings contribute to ongoing efforts to integrate patient-reported data into post-marketing surveillance and treatment decision-making.
{"title":"Artificial intelligence and natural language processing of patient narratives to evaluate semaglutide for weight loss","authors":"Akshaya Srikanth Bhagavathula PhD, FACE","doi":"10.1016/j.annepidem.2025.09.003","DOIUrl":"10.1016/j.annepidem.2025.09.003","url":null,"abstract":"<div><h3>Purpose</h3><div>This study used artificial intelligence (AI) and natural language processing (NLP) to analyze patient reviews of semaglutide, with the goal of better understanding its real-world effectiveness and safety for weight management.</div></div><div><h3>Methods</h3><div>A retrospective, cross-sectional analysis was conducted on 772 user-generated reviews of semaglutide posted on Drugs.com (July 2021–March 2025). Sentiment analysis was performed using a transformer-based BERT model on a five-point scale. Topic modeling with Latent Dirichlet Allocation (LDA) and Latent Semantic Analysis (LSA) was used to identify dominant themes. Cluster analysis was applied to segment users based on weight loss outcomes and side effect severity. Reviewers (n = 95) that explicitly reporting both weight loss and treatment duration were analyzed for real-world efficacy and safety.</div></div><div><h3>Results</h3><div>Users who took semaglutide more than 60 days reported a mean weight loss of 32.2 ± 3.1 lbs (14.6 kg). Frequently mentioned side effects included nausea (46.9 %), headache (18.4 %), vomiting (14.3 %), fatigue (9.2 %), and dizziness (4.8 %). The highest sentiment scores were observed in the ≤ 30-day group (mean: 3.38). Topic modeling identified themes such as appetite suppression, medication cost and access, and long-term experiences. Clusters analysis revealed distinct user profile, including super-responder group with substantial weight loss and another with more side effects.</div></div><div><h3>Conclusions</h3><div>AI and NLP methods offer valuable tools for analyzing patient-reported outcomes, revealing semaglutide’s real-world efficacy and safety profile for weight management. These findings contribute to ongoing efforts to integrate patient-reported data into post-marketing surveillance and treatment decision-making.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"111 ","pages":"Pages 9-13"},"PeriodicalIF":3.0,"publicationDate":"2025-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145076559","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study aimed to estimate cardiovascular disease (CVD) burden attributable to chronic kidney disease (CKD) and whether this burden varies in participants with different population characteristics.
Methods
Our sample included 1988 adults free-of-CVD at baseline who took part in the ATTICA study (2002–2022). Estimated glomerular filtration rate (eGFR) was calculated based on the Chronic Kidney Disease Epidemiology Collaboration equation. CKD was defined in 2002, according to Kidney Disease Improving Global Outcomes (KDIGO) guidelines, as an eGFR ≤ 60 mL/min/1.73 m2. Combined fatal or non-fatal CVD events were assessed in 2006, 2012 and 2022 based on WHO-ICD-10. Population attributable fractions for multiadjusted models were computed based on Miettinen’s formula. Stratified analyses were also performed.
Results
At baseline, CKD prevalence was 4.7 % (n = 94). During the 20-year period, 36.1 % of participants developed CVD. A higher percentage of participants with CKD developed CVD compared to those without (77 % vs. 34 %). Approximately 6 out of a 100 new CVD cases (95 %CI: 1.7 %, 8.1 %) would have been prevented if CKD had been properly managed. Variations in these fractions were observed by sex and presence of comorbidities.
Conclusions
Albeit more research is warranted, our study supports that CKD should become a public health priority, and specifically a CVD priority.
目的：本研究旨在估计慢性肾脏疾病（CKD）引起的心血管疾病（CVD）负担，以及这种负担在不同人群特征的参与者中是否存在差异。方法：我们的样本包括1988名参加ATTICA研究（2002-2022）的无心血管疾病成人。估计肾小球滤过率（eGFR）是根据慢性肾脏病流行病学协作方程计算的。根据肾脏疾病改善全球结局（KDIGO）指南，2002年将CKD定义为eGFR≤60 mL/min/1.73m2。根据WHO-ICD-10在2006年、2012年和2022年评估了合并致死性或非致死性心血管疾病事件。根据Miettinen公式计算多调整模型的人口归因分数。还进行了分层分析。结果：基线时，CKD患病率为4.7% （n=94）。在20年期间，36.1%的参与者患上了心血管疾病。CKD患者发生CVD的比例高于无CKD患者（77% vs. 34%）。如果CKD得到妥善管理，大约100例CVD新病例中有6例（95%CI: 1.7%, 8.1%）本可以预防。这些分数的变化是根据性别和合并症的存在来观察的。结论：尽管需要进行更多的研究，但我们的研究支持CKD应该成为公共卫生的优先事项，特别是心血管疾病。
{"title":"Cardiovascular disease burden attributable to estimated glomerular filtration rate: Insights from the ATTICA study (2002–2022)","authors":"Evangelia Damigou , Matilda Florentin , Christina Chrysohoou , Costas Anastasiou , Fotios Barkas , Evangelos Liberopoulos , Petros P. Sfikakis , Costas Tsioufis , Christos Pitsavos , Demosthenes Panagiotakos","doi":"10.1016/j.annepidem.2025.09.007","DOIUrl":"10.1016/j.annepidem.2025.09.007","url":null,"abstract":"<div><h3>Purpose</h3><div>This study aimed to estimate cardiovascular disease (CVD) burden attributable to chronic kidney disease (CKD) and whether this burden varies in participants with different population characteristics.</div></div><div><h3>Methods</h3><div>Our sample included 1988 adults free-of-CVD at baseline who took part in the ATTICA study (2002–2022). Estimated glomerular filtration rate (eGFR) was calculated based on the Chronic Kidney Disease Epidemiology Collaboration equation. CKD was defined in 2002, according to Kidney Disease Improving Global Outcomes (KDIGO) guidelines, as an eGFR ≤ 60 mL/min/1.73 m<sup>2</sup>. Combined fatal or non-fatal CVD events were assessed in 2006, 2012 and 2022 based on WHO-ICD-10. Population attributable fractions for multiadjusted models were computed based on Miettinen’s formula. Stratified analyses were also performed.</div></div><div><h3>Results</h3><div>At baseline, CKD prevalence was 4.7 % (n = 94). During the 20-year period, 36.1 % of participants developed CVD. A higher percentage of participants with CKD developed CVD compared to those without (77 % vs. 34 %). Approximately 6 out of a 100 new CVD cases (95 %CI: 1.7 %, 8.1 %) would have been prevented if CKD had been properly managed. Variations in these fractions were observed by sex and presence of comorbidities.</div></div><div><h3>Conclusions</h3><div>Albeit more research is warranted, our study supports that CKD should become a public health priority, and specifically a CVD priority.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"111 ","pages":"Pages 1-8"},"PeriodicalIF":3.0,"publicationDate":"2025-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145058725","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-12DOI: 10.1016/j.annepidem.2025.09.008
Glenna Walther , Tilman Brand , Nico Dragano , Claudia Meinke-Franze , Amand Führer , Karin Halina Greiser , Olga Hovardovska , Jamin Kiekert , Lilian Krist , Michael Leitzmann , Wolfgang Lieb , Rafael Mikolajczyk , Ute Mons , Fiona Niedermayer , Nadia Obi , Cara Övermöhle , Marvin Reuter , Börge Schmidt , Ilais Moreno Velásquez , Henry Völzke , Volker Winkler
Aims
Resettlers from the former Soviet Union are one of the largest migrant groups in Germany. Previous studies found lower cardiovascular disease (CVD) mortality among resettlers compared to Germans without migration background (autochthonous). Other studies have shown a higher prevalence of CVD risk factors among resettlers, suggesting a higher CVD mortality. The German National Cohort (NAKO) provides an opportunity to explore these discrepancies.
Methods
This study used baseline data from NAKO and compared age-adjusted percentages of self-reported CVD and associated risk factors between the two groups. Logistic regression models estimated adjusted odds ratios (OR) for associations between resettler status and outcomes.
Results
Among 204,751 participants aged 19–75, 3580 were resettlers and 169,538 autochthonous Germans. Male resettlers had lower odds of risky alcohol consumption (OR: 0.55; 95 %CI: 0.49–0.63) but higher odds of ever smoking (OR: 1.26; 95 %CI: 1.13–1.41) compared to autochthonous German men. Female resettlers showed higher prevalence of hypertension, diabetes mellitus, obesity, and elevated cholesterol/triglycerides, but lower prevalence of risky alcohol consumption and smoking.
In men, the odds of peripheral artery disease (PAD) (OR: 0.46; 95 %CI: 0.21–0.97) and any CVD (OR: 0.81; 95 %CI: 0.66–0.98) were lower among resettlers. No other notable differences in clinical CVDs were observed in men.
Conclusion
Resettlers showed differences regarding CVD risk factor distribution compared to autochthonous Germans. These differences appear to balance out, leading to similar overall CVD prevalence, except for a lower prevalence of PAD and total CVD in male resettlers. Future longitudinal data will allow to explore long-term CVD trajectories.
Lay summary
We compared the prevalence of cardiovascular diseases (CVD) in a special group of migrants (resettlers from the former Soviet Union; German: (Spät-)Aussiedler) and autochthonous Germans, using data from the German National Cohort (NAKO) and investigated risk factors frequencies for these diseases in both groups. We found that male resettlers had less of risky alcohol consumption but smoked more than autochthonous German men. Female resettlers showed higher prevalence of hypertension, diabetes mellitus, obesity, and elevated cholesterol and triglyceride levels, but lower prevalence of risky alcohol consumption and smoking. The prevalence of most CVDs was similar in both groups, except of peripheral artery disease and all CVDs combined which we found less frequent in male resettlers.
{"title":"Migration and cardiovascular disease: A comparative study of prevalence and risk factor profiles in resettlers from the German National Cohort (NAKO)","authors":"Glenna Walther , Tilman Brand , Nico Dragano , Claudia Meinke-Franze , Amand Führer , Karin Halina Greiser , Olga Hovardovska , Jamin Kiekert , Lilian Krist , Michael Leitzmann , Wolfgang Lieb , Rafael Mikolajczyk , Ute Mons , Fiona Niedermayer , Nadia Obi , Cara Övermöhle , Marvin Reuter , Börge Schmidt , Ilais Moreno Velásquez , Henry Völzke , Volker Winkler","doi":"10.1016/j.annepidem.2025.09.008","DOIUrl":"10.1016/j.annepidem.2025.09.008","url":null,"abstract":"<div><h3>Aims</h3><div>Resettlers from the former Soviet Union are one of the largest migrant groups in Germany. Previous studies found lower cardiovascular disease (CVD) mortality among resettlers compared to Germans without migration background (autochthonous). Other studies have shown a higher prevalence of CVD risk factors among resettlers, suggesting a higher CVD mortality. The German National Cohort (NAKO) provides an opportunity to explore these discrepancies.</div></div><div><h3>Methods</h3><div>This study used baseline data from NAKO and compared age-adjusted percentages of self-reported CVD and associated risk factors between the two groups. Logistic regression models estimated adjusted odds ratios (OR) for associations between resettler status and outcomes.</div></div><div><h3>Results</h3><div>Among 204,751 participants aged 19–75, 3580 were resettlers and 169,538 autochthonous Germans. Male resettlers had lower odds of risky alcohol consumption (OR: 0.55; 95 %CI: 0.49–0.63) but higher odds of ever smoking (OR: 1.26; 95 %CI: 1.13–1.41) compared to autochthonous German men. Female resettlers showed higher prevalence of hypertension, diabetes mellitus, obesity, and elevated cholesterol/triglycerides, but lower prevalence of risky alcohol consumption and smoking.</div><div>In men, the odds of peripheral artery disease (PAD) (OR: 0.46; 95 %CI: 0.21–0.97) and any CVD (OR: 0.81; 95 %CI: 0.66–0.98) were lower among resettlers. No other notable differences in clinical CVDs were observed in men.</div></div><div><h3>Conclusion</h3><div>Resettlers showed differences regarding CVD risk factor distribution compared to autochthonous Germans. These differences appear to balance out, leading to similar overall CVD prevalence, except for a lower prevalence of PAD and total CVD in male resettlers. Future longitudinal data will allow to explore long-term CVD trajectories.</div></div><div><h3>Lay summary</h3><div>We compared the prevalence of cardiovascular diseases (CVD) in a special group of migrants (resettlers from the former Soviet Union; German: (Spät-)Aussiedler) and autochthonous Germans, using data from the German National Cohort (NAKO) and investigated risk factors frequencies for these diseases in both groups. We found that male resettlers had less of risky alcohol consumption but smoked more than autochthonous German men. Female resettlers showed higher prevalence of hypertension, diabetes mellitus, obesity, and elevated cholesterol and triglyceride levels, but lower prevalence of risky alcohol consumption and smoking. The prevalence of most CVDs was similar in both groups, except of peripheral artery disease and all CVDs combined which we found less frequent in male resettlers.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"111 ","pages":"Pages 14-23"},"PeriodicalIF":3.0,"publicationDate":"2025-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145066589","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-11DOI: 10.1016/j.annepidem.2025.08.025
Hans Moen , Vishnu Raj , Andrius Vabalas , Markus Perola , Samuel Kaski , Andrea Ganna , Pekka Marttinen
Purpose:
Health registers provide valuable insights into individuals’ health trajectories. This study explores the use of deep learning to model and analyze these trajectories using a nationwide longitudinal dataset containing coded features such as clinical codes, procedures, and drug purchases.
Methods:
We introduce Evolve, a transformer-based deep learning model designed to provide continuous multi-label predictions over time. The model predicts disease onsets at each time step conditioned on the health history up to that time step and the time until a given 5-year forecast window. Evolve is evaluated against several baseline models for basic prediction performance. Additionally, we analyze health trajectories by tracking changes in prediction probabilities and in the latent embedding neighborhood to identify important events.
Results:
Evolve performed comparably to baseline models in disease onset prediction while offering unique trajectory modeling capabilities. The model identified early predictive events and demonstrated that changes in embedding space could indicate shifts in health trajectories. Visualization of evolving health trajectories showed how individuals may become most similar to others with similar profiles and outcomes over time.
Conclusions:
The Evolve model seems promising at enabling continuous health monitoring, early disease detection, and retrospective analysis, making it a promising tool for personalized healthcare interventions.
Code available at: https://github.com/hansmoen/evolvehealth.
{"title":"Towards modeling evolving longitudinal health trajectories with a transformer-based deep learning model","authors":"Hans Moen , Vishnu Raj , Andrius Vabalas , Markus Perola , Samuel Kaski , Andrea Ganna , Pekka Marttinen","doi":"10.1016/j.annepidem.2025.08.025","DOIUrl":"10.1016/j.annepidem.2025.08.025","url":null,"abstract":"<div><h3>Purpose:</h3><div>Health registers provide valuable insights into individuals’ health trajectories. This study explores the use of deep learning to model and analyze these trajectories using a nationwide longitudinal dataset containing coded features such as clinical codes, procedures, and drug purchases.</div></div><div><h3>Methods:</h3><div>We introduce <span>Evolve</span>, a transformer-based deep learning model designed to provide continuous multi-label predictions over time. The model predicts disease onsets at each time step conditioned on the health history up to that time step and the time until a given 5-year forecast window. <span>Evolve</span> is evaluated against several baseline models for basic prediction performance. Additionally, we analyze health trajectories by tracking changes in prediction probabilities and in the latent embedding neighborhood to identify important events.</div></div><div><h3>Results:</h3><div><span>Evolve</span> performed comparably to baseline models in disease onset prediction while offering unique trajectory modeling capabilities. The model identified early predictive events and demonstrated that changes in embedding space could indicate shifts in health trajectories. Visualization of evolving health trajectories showed how individuals may become most similar to others with similar profiles and outcomes over time.</div></div><div><h3>Conclusions:</h3><div>The <span>Evolve</span> model seems promising at enabling continuous health monitoring, early disease detection, and retrospective analysis, making it a promising tool for personalized healthcare interventions.</div><div>Code available at: <span><span>https://github.com/hansmoen/evolvehealth</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"111 ","pages":"Pages 30-43"},"PeriodicalIF":3.0,"publicationDate":"2025-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145058755","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-10DOI: 10.1016/j.annepidem.2025.09.006
Sadiporn Phuthomdee , Sirinya Teeraananchai , Rattaphon Triamwichanon , Patchara Benjarattanaporn , David C. Boettiger , Nittaya Phanuphak
Purpose
To assess HIV incidence and associated factors among key populations (KPs) including men who have sex with men (MSM), transgender women (TGW), sex workers (SW), and people who inject drugs (PWID) using data from the Universal Health Coverage program.
Methods
HIV testing data were collected from KP individuals aged ≥ 15 years diagnosed HIV-negative at least once between 2015 and 2022. Flexible parametric survival analysis was used to estimate HIV incidence and identify associated factors.
Results
Among 373,300 individuals tested, 65 % were MSM, 26 % SW, 8 % PWID, and 2 % TGW. HIV incidence rates were highest among MSM (2.23 per 100 person-years, PYs), followed by TGW (2.06 per 100 PYs), PWID (0.46 per 100 PYs), and SW (0.41 per 100 PYs). In multivariable analysis, MSM had significantly higher risk compared to PWID (adjusted hazard ratio (aHR 0.13), SW (aHR 0.20), and TGW (aHR 0.77). Youth among KPs had a 7.30-fold higher risk compared to those aged ≥ 50 years. A decline in incidence was observed in more recent years.
Conclusions
HIV incidence is declining all KPs but remains highest among MSM and TGW, particularly youth, highlighting the need for targeted prevention to end HIV in Thailand.
{"title":"HIV incidence and associated risk factors among key populations accessing routine testing through the national health coverage system in Thailand: A nationwide real-world study","authors":"Sadiporn Phuthomdee , Sirinya Teeraananchai , Rattaphon Triamwichanon , Patchara Benjarattanaporn , David C. Boettiger , Nittaya Phanuphak","doi":"10.1016/j.annepidem.2025.09.006","DOIUrl":"10.1016/j.annepidem.2025.09.006","url":null,"abstract":"<div><h3>Purpose</h3><div>To assess HIV incidence and associated factors among key populations (KPs) including men who have sex with men (MSM), transgender women (TGW), sex workers (SW), and people who inject drugs (PWID) using data from the Universal Health Coverage program.</div></div><div><h3>Methods</h3><div>HIV testing data were collected from KP individuals aged ≥ 15 years diagnosed HIV-negative at least once between 2015 and 2022. Flexible parametric survival analysis was used to estimate HIV incidence and identify associated factors.</div></div><div><h3>Results</h3><div>Among 373,300 individuals tested, 65 % were MSM, 26 % SW, 8 % PWID, and 2 % TGW. HIV incidence rates were highest among MSM (2.23 per 100 person-years, PYs), followed by TGW (2.06 per 100 PYs), PWID (0.46 per 100 PYs), and SW (0.41 per 100 PYs). In multivariable analysis, MSM had significantly higher risk compared to PWID (adjusted hazard ratio (aHR 0.13), SW (aHR 0.20), and TGW (aHR 0.77). Youth among KPs had a 7.30-fold higher risk compared to those aged ≥ 50 years. A decline in incidence was observed in more recent years.</div></div><div><h3>Conclusions</h3><div>HIV incidence is declining all KPs but remains highest among MSM and TGW, particularly youth, highlighting the need for targeted prevention to end HIV in Thailand.</div></div>","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"110 ","pages":"Pages 178-186"},"PeriodicalIF":3.0,"publicationDate":"2025-09-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145049032","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-09-06DOI: 10.1016/j.annepidem.2025.09.005
Athena P. Kourtis , Jeffery Wiener , Weiming Zhu , Minttu M. Rönn , Joshua Salomon , Ya-Lin A. Huang , Cynthia Lyles , Rupa R. Patel , Karen W. Hoover , Robyn Neblett Fanfair , Jonathan Mermin
{"title":"Corrigendum to ‘Estimating the population need for preexposure prophylaxis for HIV in the United States’ [Ann Epidemiol 106 (2025) 48–54]","authors":"Athena P. Kourtis , Jeffery Wiener , Weiming Zhu , Minttu M. Rönn , Joshua Salomon , Ya-Lin A. Huang , Cynthia Lyles , Rupa R. Patel , Karen W. Hoover , Robyn Neblett Fanfair , Jonathan Mermin","doi":"10.1016/j.annepidem.2025.09.005","DOIUrl":"10.1016/j.annepidem.2025.09.005","url":null,"abstract":"","PeriodicalId":50767,"journal":{"name":"Annals of Epidemiology","volume":"110 ","pages":"Page 166"},"PeriodicalIF":3.0,"publicationDate":"2025-09-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145003683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号