Congenital heart defects : familial recurrence, screening, and genetic aspects

Background Congenital heart defects are the most common congenital anomalies and comprise a heterogeneous spectrum of structural cardiac malformations. Clinical expression, familial clustering, detection patterns, and genetic architecture are not uniform across lesion types. Contemporary understanding of CHD therefore requires population-based data capable of capturing familial structure, phenotype-specific patterns, real-world screening performance, and molecular heterogeneity within a unified healthcare context. Aims The overall aim was to characterise CHD in a contemporary national population by examining heterogeneity across familial aggregation, phenotype-specific clustering, screening performance, and genetic susceptibility. This included assessment of familial aggregation across kinship structures and familial burden, characterisation of clustering across cardiac phenotypes, evaluation of detection of critical CHD (CCHD) within a nationwide screening system, and exploration of common genetic susceptibility in non-syndromic coarctation of the aorta (CoA), including anatomical and sex-related variation. Methods Studies I-II were nationwide register-based case-control studies including 51,778 individuals with CHD born in Sweden (1987-2017) and 522,543 matched controls. Multigenerational linkage identified parents, siblings, offspring, and half-siblings. Associations were estimated using unconditional logistic regression with robust standard errors clustered on maternal identifier. Lesion-specific analyses used hierarchical classification of mutually exclusive phenotypes complemented by descriptive non-hierarchical mapping of multi-lesion co-occurrence. Study III was a nationwide retrospective cohort including 630 full-term live-born infants with CCHD (2014-2019), using registry linkage and medical record review to classify detection pathways and timing. Study IV was a genome-wide association study including 198 patients with repaired non-syndromic CoA and 4,541 population controls. Logistic regression under an additive genetic model was applied, including predefined anatomical and clinical subgroup and sex-stratified analyses. Results Most individuals with CHD had no recorded affected relatives; however, familial aggregation was evident. Having at least one affected relative was associated with increased odds of CHD (adjusted OR 2.71, 95% CI 2.60-2.83), with a graded pattern by number of affected relatives. Associations were stronger along the maternal line than the paternal line and were not explained by maternal diabetes, hypertension, or obesity. Lesion-specific analyses showed phenotype-specific clustering, with stronger same-lesion than cross-lesion aggregation. Kinship gradients persisted at the lesion level, and maternal predominance was also evident for same-lesion clustering. In the CCHD cohort, live-birth prevalence was 0.91 per 1,000 and remained stable. Overall, 89.2% were diagnosed before discharge or death. Prenatal ultrasound accounted for 41.9% of detections, pulse oximetry screening for 22.5%, and neonatal examination for 13.7%. CoA was the most common phenotype and accounted for most undiagnosed discharges. In Study IV, no genome-wide significant associations were observed when CoA was analysed as a single aggregated phenotype. In contrast, predefined subgroup analyses identified genome-wide significant loci in anatomically and clinically distinct subphenotypes, including CoA with bicuspid aortic valve and left-sided obstructive lesions and early-repair CoA. Sex-stratified analyses and genotype-by-sex interaction testing further demonstrated heterogeneity in association signals between males and females. Conclusions In a contemporary national population, CHD is characterised by structured heterogeneity across familial aggregation, detection performance, and genetic susceptibility. Familial aggregation is measurable and graded by familial burden, with consistent maternal predominance and phenotype-specific clustering. Within a mature nationwide screening system, most infants with CCHD are detected before discharge, yet detection remains uneven across lesion types - most notably for coarctation. Genetic analyses in CoA indicate that phenotypic and sex-specific stratification may reveal distinct association signals within clinically defined subgroups. These findings underscore that CHD represents a heterogeneous spectrum rather than a uniform entity in epidemiological, clinical, and genetic contexts. List of scientific papers I. Congenital heart defects: familial recurrence patterns in Sweden. Kazamia K, Ekberg S, Dietrich CE, Eliasson H, Wahren-Herlenius M, Bergman G. Eur Heart J. 2026. https://doi.org/10.1093/eurheartj/ehaf1048 II. Lesion-specific familial aggregation and co-occurrence patterns of congenital heart defects: a nationwide case-control study. Kazamia K, Ekberg S, Dietrich CE, Eliasson H, Wahren-Herlenius M, Bergman G [Submitted] III. Screening for critical congenital heart defects in Sweden. Lannering K, Kazamia K, Bergman G, Östman-Smith I, Liuba P, Alenius Dahlqvist J, Elfvin A, Mellander M. Pediatrics. 2023. https://doi.org/10.1542/peds.2023-061949 IV. Genome-wide association study of non-syndromic aortic coarctation identifies subphenotype-specific and sex-biased associations. Kazamia K, Björck HM, Weismann C, Lannering K, Fernlund E, Skog Andreasson A, Winberg J, SNAB-CHD, Sonesson SE, Sunnegårdh J, Gunnarsson C, Eriksson P, Kockum I, Ekman-Joelsson BM, Rydberg A, Malm T, Wahren-Herlenius M, Bergman G, Thorlacius GE [Manuscript]