Background
Cardiometabolic disorders, such as type 2 diabetes (T2D), cardiovascular diseases (CVD), obesity, hypercholesterolemia, and hypertension are interrelated conditions, responsible for high mortality and disability rates worldwide. The global burden of these disorders is continuously increasing: the global population with CVD doubled from 271 million in 1990 to 523 million in 2019 [
1], and the case-number of diabetes (20–79 years) is predicted to increase from 537 million in 2021 to 783 million in 2045 [
2]. These cardiometabolic disorders often co-occur within individuals, suggesting a co-pathogenesis of metabolic abnormalities among various cardiometabolic disorders, which is commonly described as metabolic syndrome (MetS), a highly prevalent, multifaceted cluster of metabolic abnormalities [
3‐
9].
A role of genetics is likely: several studies have observed familial aggregation of CVD [
10,
11] and T2D [
12], obesity [
13,
14], and MetS [
15‐
17], while cardiometabolic traits, such as blood pressure, fasting blood glucose, and total cholesterol, were shown to be heritable [
18‐
20]. Some evidence exists that different cardiometabolic disorders co-aggregate within families, i.e., a family history of a specific cardiometabolic disorder associates with elevated risk of another cardiometabolic disorder [
21‐
24]. However, familial aggregation, and especially co-aggregation of a full spectrum of cardiometabolic disorders, has not yet been investigated comprehensively within a single study. Furthermore, the accuracy and generalizability of most family studies are limited by various factors, such as modest sample size ranging from 2302 to 17,954 individuals, specific founder populations, different family relationship included (e.g., only siblings or parent-offspring), or the use of self-reported family history not validated by objective measures.
For above reasons, it remains uncertain to what extent cardiometabolic disorders (co-)aggregate in families in the general population, and to what extent the correlation between cardiometabolic disorders and traits can be explained by genetics. Bridging this knowledge gap may help risk stratification and early detection of cardiometabolic disorders. Furthermore, knowledge of shared genetics between disorders and traits may help advance our understanding of pathophysiology. Therefore, we aimed to: (1) quantify the familial (co-)aggregation of various cardiometabolic disorders; (2) estimate the heritability of a wide array of underlying cardiometabolic traits; and (3) estimate genetic correlations between cardiometabolic traits, by using extensive data from Lifelines, a large multi-generational family study representative of the general Dutch population.
Discussion
In this study, we aimed to estimate the genetic and environmental contribution to the co-occurrence of cardiometabolic disorders and traits within families, using objective measurements in a large multi-generational family study. We quantified the familial (co-)aggregation of six cardiometabolic disorders in first-degree relatives and spouses. Individuals with a first-degree relative affected with one of the cardiometabolic disorders had a higher risk of having the same or related disorders. Similarly, individuals with a spouse affected with cardiometabolic disorders had a higher risk of having the same or related disorders, suggesting an effect of shared environmental factors and/or assortative mating. Also, we estimated the heritability of fifteen cardiometabolic traits. These cardiometabolic traits had moderate heritability, indicating a role for genetics underlying the recurrence of cardiometabolic disorders within a family. Finally, we found moderate genetic correlations between cardiometabolic traits, suggesting genetics as an important but not exclusive underlying mechanism of the interrelation between cardiometabolic disorders.
Familial aggregation is evidence for a role of shared genetics and shared environment within a family in the occurrence of complex disorders. Positive familial aggregation of cardiometabolic disorders has been suggested by previous studies [
10‐
17,
21]. Consistent with the literature, our study also found positive familial aggregation between first-degree relatives although of somewhat lower magnitude. For example, in The Framingham Offspring Study, risks of CVD was approximately 2 times higher in middle-aged adults with at least one parent with CVD [
10] and 1.5 time if they had at least one sibling with CVD [
11]. For T2D, a large Danish study also identified an up to 3.4 times higher risk in first-degree relatives than the general population [
36], even higher than our estimate of 2.48 higher risk in first-degree relatives. Other studies also found higher familial aggregation of MetS compared to our recurrence risk estimate λ
FDR = 1.43 (95% CI 1.39–1.48) in individuals with affected first-degree relatives. A large population-based study in China identified a two to three times higher risk of MetS in younger siblings if their eldest sibling was affected by MetS [
15]. Also, in the Tehran Lipid and Glucose Study, the risk of MetS was higher among offspring with affected parents (OR: 2.29–4.53) [
16]. A possible explanation for the varying aggregation estimates is the heterogeneity between studies due to different family relationship included, age diversity between studies, and differences in ethnicity, lifestyle, and health behaviors between the Netherlands and other countries. Another possible explanation may arise from the diverse definitions of different disease phenotypes. In our study, we utilized both an extended CVD phenotype and a narrow CVD phenotype, the latter encompassing four CVD types: myocardial infarction, heart failure, stroke, and cardiac surgery. We observed a slightly lower familial aggregation in the extended CVD phenotype. This difference could be attributed to varying levels of heterogeneity in the extended CVD phenotype compared to the narrow CVD phenotype, with the latter demonstrating greater homogeneity. Despite these differences, the evidence converges on a major role of shared genetics in determining cardiometabolic risk. In exploratory analysis, we observed a higher recurrence risk in individuals with affected siblings and affected offspring when compared to individuals with affected parents, although possibly this is driven by age differences between offspring and parents. However age-stratified exploratory analysis showed that the recurrence risk estimates were stable across both age and sex.
We considered recurrence risk estimates in spouses a negative control to those in family: given that spouses are unlikely to be genetically related, estimates of spousal recurrence reflect the effects of shared environment and/or assortative mating on cardiometabolic risk. A Danish study observed around 1.5 times higher risk of T2D among individuals with a spouse affected with T2D [
36], and in previous work in both Japanese and Dutch, we corroborated higher spousal cardiometabolic risk, for T2D (OR: 1.20 vs 1.59), hypertension (OR: 1.34 vs 1.45), and MetS (OR: 1.77 vs 1.77) in Japanese vs Dutch [
37]. Although using the same cohort study, the estimates for spousal concordance in the Dutch Lifelines Population [
37] are slightly higher than our current familial aggregation estimates, which may be attributed to differences in the statistical approaches used to estimate the risk of disorders among spouses. The spousal concordance was related to concordance in lifestyle factors, such as physical activity, smoking, and alcohol drinking, indicative of potential cohabitation effects, and assortative mating [
37‐
39].
We found that cardiometabolic traits had moderate genetic extent in which the heritability estimates are consistent with those in the literature although some variation can occur due to differences in age, ethnicity, study design [
40], and type of measurement used [
41]. For example, slightly different estimates of heritability were found in a family study in a 1564 Chinese individuals from 494 families [
20], reporting heritability of fasting glucose (h
2: 0.17), waist circumference (h
2: 0.26), SBP (h
2: 0.24), DBP (h
2: 0.17), triglycerides (h
2: 0.41), HDL-C (h
2: 0.49), LDL-C (h
2: 0.47), total cholesterol (h
2: 0.46), CRP (h
2: 0.38), and BMI (h
2 = 0.38). Also, a Dutch twin study presented moderate to high heritability for a range of cardiometabolic traits from 0.47 for insulin level to 0.78 for BMI [
42], with estimates typically being higher than family studies including the present study.
Genetic similarity between first-degree relatives is likely to contribute to familial co-aggregation of related cardiometabolic disorders [
21‐
24]. A previous study found that parental history of one or more CVD (i.e., myocardial infarction, stroke, and angina) at a younger age < 55 year in the father and < 65 year in the mother significantly increased the risk of MetS in women, with ORs from 1.62 to 1.84 [
21]. Another family study identified increased risk of having multiple cardiometabolic disorders in relation to parental history of diabetes (OR: 1.54, 95% CI 1.01–2.33) and parental history of hypertension (OR 1.42, 95% CI 1.06–1.91). The risk was even higher when both parents were affected with hypertension or diabetes, suggesting an additive genetic effect on the risk of cardiometabolic disease co-occurrence [
24]. Similarly, a population-based study in US found a higher risk of co-occurring cardiometabolic disorders when individuals had a family history of diabetes or hypertension, and only a slightly increased risk with family history of obesity [
23]. These studies mostly found evidence for co-aggregation between CVD, T2D, and hypertension. Compared to these studies, our estimates of co-aggregation between obesity, T2D, and MetS were larger, while hypertension, CVD, and hypercholesterolemia showed only modest familial co-aggregation. A possible reason for this difference is that previous studies used mostly self-reported family history without actual validation with objective laboratory measures in family members.
Cardiometabolic disorders likely share pathophysiological mechanisms such as inflammation and insulin resistance [
4‐
6]. The high insulin resistance in obesity and diabetes is thought to induce inflammation, causing vascular damage and endothelial dysfunction. Such vascular damage and dysfunction lead to increased production of vasoconstrictors, and subsequently to an increase in vascular resistance, a major contributor to CVD and hypertension [
5,
6]. Consistent with this, we found significant phenotypic and genetic correlations between blood pressure, obesity traits, inflammatory markers, and fasting glucose, although these correlations were of modest strength. In the present study, we also found modest genetic correlations between HDL cholesterol and triglyceride with glucose markers, blood pressure, and obesity, which are the traits used for the definition of MetS. The accumulation of Advanced Glycation End products has been considered a potential cross-link between diabetes and cardiovascular events, by increasing inflammation and causing endothelial dysfunction [
43,
44]. We found little evidence of this in the present study: we observed only weak correlations between skin autofluorescence and other cardiometabolic traits, although their genetic correlations are slightly higher than the phenotypic correlation. Furthermore, the association of skin autofluorescence, CVD, and T2D in the previous studies in this population were independent of glucose markers [
43,
45].
Strengths of this study are that it is the largest family and comprehensive study of cardiometabolic outcomes to date, investigating six interrelated cardiometabolic disorders and fifteen intermediate cardiometabolic traits. It was conducted in a large-scale population-based, multi-generational cohort, representative of the general Dutch population [
46]. We used a combination of objective laboratory measurements, questionnaire data, and medication data, resulting in precise outcome definitions and thus precise estimates of recurrence risk and heritability. Our study provides insights into the genetic and/or environmental mechanisms that underly of co-existing cardiometabolic disorders within individuals and families. Our findings highlight the role of shared genetics and environmental factors on the risk of cardiometabolic disorders and suggests overlapping genetic structure between disorders. Furthermore, our estimates of recurrence risk may inform clinicians and health services in diagnosis, patient communication, and potential screening efforts based on family history. However, several study limitations need to be addressed as well. Firstly, only data were available on family members participating in Lifelines. Missing data on non-participating family members may have caused underestimation of recurrence risk and heritability. Secondly, although largely representative of the Dutch population Lifelines predominantly consists of Dutch participants; generalization to other ancestries and maybe even to other European populations is therefore uncertain.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.