Left ventricular hypertrophy (LVH) is defined as an increase in left ventricular mass (LVM) that can be secondary to an increase in ventricular wall thickness or chamber size (1). Electrocardiography (ECG), echocardiography and magnetic resonance imaging are currently the main diagnostic tools for the evaluation of LVH. Although ECG is less accurate than the other two diagnostic methods in diagnosing LVH, it is widely used in epidemiological studies because of its low cost, convenience and easy availability. However, LVH detected by either method is strongly associated with a higher risk of cardiovascular disease (CVD) and CVD-related mortality (2–6). Therefore, the 2018 European Society of Hypertension (ESH)/European Society of Cardiology (ESC) clinical practice guideline added LVH as a high-risk factor to the CVD risk assessment system and recommended screening for LVH in high-risk populations as well as early identification and intervention of controllable risk factors for LVH to prevent premature CVD or CVD-related death (7). Although current evidence suggests that the risk factors for LVH are composed of some non-modifiable and modifiable factors, including uncontrollable factors such as age, gender and genetic susceptibility and controllable risk factors such as hypertension, diabetes, chronic kidney disease, metabolic disorders, obesity, lack of exercise or unhealthy diet, other potential risk factors may still exist (1, 8–10).

Lipoprotein(a) [Lp(a)] is a low-density lipoprotein cholesterol-like particle bound to apolipoprotein(a), which can participate in the occurrence and development of CVD by promoting oxidation, inflammation, calcification, thrombosis and atherosclerosis (11). Because circulating Lp(a) levels are largely regulated by genes, its absolute risk threshold has not yet reached unity, but current evidence suggests that higher Lp(a) is associated with a higher risk of CVD (12). In recent years, studies of Lp(a) and CVD have been widely conducted as Lp(a) has gradually gained attention and improved measurement methods have been developed. As a result, a growing number of observational studies have shown that Lp(a) is closely related to coronary heart disease (CHD) (13), stroke (14), calcific aortic valve disease (CAVD) (15), hypertension (16), atrial fibrillation (17) and venous thromboembolism (18), and evidence from genetic studies has also confirmed the causal association of Lp(a) with CVD and CVAD (19, 20). A recent study revealed the relationship between Lp(a) and LVH assessed by echocardiography only in patients with acute myocardial infarction (21), whereas the correlation between Lp(a) and LVH assessed by ECG in the general population remains unknown.

Therefore, in order to fill the gap in this research field and provide reference for formulating early prevention and treatment strategies of LVH under the background of higher Lp(a), the aim of this study was to explore the relationship between Lp(a) and LVH assessed by ECG in the general population.

In this large population-based cross-sectional observational study, we found that higher Lp(a) was strongly associated with a higher prevalence of LVH assessed by ECG in the general population, and this association was further confirmed to be linear. Additionally, we also demonstrated robustness of the association of Lp(a) with LVH in people aged< 60 or ≥ 60 years, male, with BMI< 30 kg/m², with hypertension or without diabetes.

Current evidence from epidemiological studies suggests that LVH is strongly associated with CVD as well as cardiovascular and all-cause mortality (4, 31–33). Therefore, it is very important to identify the risk factors of LVH and carry out early intervention to prevent the premature occurrence of LVH. Currently, an increasing number of studies have identified independent risk factors for LVH, such as age, hypertension, diabetes and obesity (1, 34, 35). However, to our knowledge, a previous study only showed an association between Lp(a) and LVH assessed by echocardiography in patients with acute myocardial infarction (21), whereas the correlation between Lp(a) and LVH in the general population remains unclear, especially for LVH diagnosed by ECG. Recently, a large observational epidemiological study involving 309,400 participants showed that Lp(a) in the highest tertile was significantly associated with a higher prevalence of LVH diagnosed by echocardiography compared with Lp(a) in the lowest tertile, but not with a higher incidence of LVH during follow-up (36). Although this study demonstrated in the cross-sectional section that participants in the highest tertile of Lp(a) had a 1.3 times higher risk of developing LVH compared with those in the lowest tertile, because LVH in this study was defined by expensive and highly subjective echocardiography, the results may not be applicable to large epidemiological studies based on the general population. Nevertheless, fortunately, our findings were consistent with the conclusion of the above study on the correlation between Lp(a) and LVH, that is, participants in the highest quartile of Lp(a) had a 1.5 times higher risk of suffering from ECG-defined LVH than those in the lowest quartile of Lp(a). Furthermore, in addition to the current evidence confirming the causal association of Lp(a) with atherogenic CVD as well as CAVD (15, 19, 20), several studies have identified other pathogenic phenotypes of Lp(a). For example, Dentali et al. demonstrated in a systematic review and meta-analysis including 14,011 participants and 14 observational studies that higher Lp(a) was significantly associated with higher venous thromboembolism (OR:1.56,95% CI:1.36-1.79) using a random effects model (18). Another observational study showed that although familial hypercholesterolemia did not cause elevated Lp(a), elevated Lp(a) predicted the occurrence of familial hypercholesterolemia (37). However, higher Lp(a) is not always detrimental. For example, Garg et al. found that higher levels of Lp(a) were independently associated with a lower incidence of atrial fibrillation during follow-up in a community-based prospective cohort of 6,593 adults without CVD (17). And Lamina et al. also revealed a negative association of Lp(a) with diabetes (38). However, the above studies only measured the concentration of circulating Lp(a) at baseline, but did not observe the effect of dynamic changes of Lp (a) on CVD. Trinder et al. refined the study design based on the above studies and showed that there was no significant change in the circulating molar concentration of Lp(a) during the median follow-up period of 4.42 years [baseline vs follow-up: 19.50 (7.56-72.50) vs 20.40 (7.70-77.50)], and that Lp(a) at baseline and follow-up were both significantly associated with the occurrence of CVD during follow-up, while changes in Lp(a) had no significant effect on the incidence of CVD (39), indicating that a single measurement of Lp(a) is essential for the primary prevention of CVD in the general population without treatment that significantly changes the level of circulating Lp(a).

Although our study demonstrated the association of Lp(a) with LVH, the mechanisms involved are still unknown. Based on published studies, we proposed the following hypotheses. First, Bergmark et al. used immunoprecipitation and ultracentrifugation experiments, in vitro transfer studies and chemiluminescence ELISAs experiments to evaluate the priority of Lp(a) as a carrier of oxidized phospholipids in human plasma. The results showed that most of oxidized phospholipids and Lp(a) co-precipitated in immunoprecipitation experiments, and most of oxidized phospholipids existed in components containing apolipoprotein(a) after subsequent ultracentrifugation experiments, and apolipoprotein(a) was the most important part of Lp(a) structure. Further in vitro transfer studies showed that oxidized phospholipids could be preferentially transferred to Lp(a) by oxidized LDL in a time-and temperature-dependent manner regardless of the nature of the buffer. Based on these data, we could draw a conclusion that apolipoprotein(a) or Lp(a) could indirectly participate in oxidative stress as the priority carrier of human plasma oxidized phospholipids, and then activate the markers related to oxidative stress, which might eventually promote the occurrence and development of LVH (40). Second, Aung et al. first conducted a single-sample Mendelian randomized study on 17,311 European individuals from British Biobank, which showed that there was an exact causal relationship between higher LDL and higher LVM, and then conducted a two-sample Mendelian randomized study on another genetic data, which also confirmed this conclusion (41), and since Lp(a) as an LDL-like particle has a cholesterol component that acts similarly to LDL-C (42), we hypothesized that the cholesterol component in Lp(a) plays a role in promoting LVH. Third, a previous study has shown that Lp(a) has a ring structure and inactive protease region similar to plasmin precursor protein structure, which can inhibit fibrinolysis and promote thrombosis through competitive binding of fibrinolytic proteins (43), and Lip et al. conducted a cross-sectional study of 178 patients from hypertension clinic, they estimated LVM, LVMI and LVH by echocardiography, the results showed that hypertensive patients had higher Lp(a) levels and left ventricular septum and posterior wall thickness. Further analysis revealed the correlation between plasma fibrinogen level with homology to Lp(a) and LVM, LVMI and LVH. Therefore, based on these complicated relationships, the hypothesis that Lp(a) indirectly affects left ventricular structure and LVH through fibrinogen or thrombogenic state may also be widely recognized (44). Additionally, current evidence confirms the association of Lp(a) with aortic stenosis and hypertension, and aortic stenosis or hypertension has been proved to be closely related to the left ventricular afterload and occurrence and development of LVH (15, 16).Consequently, we assumed that Lp(a) could promote LVH by causing aortic stenosis and hypertension in the early stages of the disease. In addition to the above findings, we believe that there are still other potential mechanisms that need to be further explored.

Despite the valuable findings, this study still had several limitations. First, we failed to confirm the causal association of Lp(a) with LVH due to the limitations of cross-sectional observational studies. Second, we only studied LVH from ECG sources and did not compare it with LVH detected by echocardiography or cardiac magnetic resonance, and the study population was limited to adults in the United States, so the robustness of the association between Lp(a) and LVH still needs to be further explored. Third, although we controlled for some confounding factors in our study, there may still be other potential risk factors for LVH, such as unhealthy diet and genetic susceptibility. Finally, there are up-to-date criteria for assessing LVH by ECG or echocardiography (45), whereas in this study we used previous criteria for assessing LVH by ECG, so the results might not be representative and more studies are needed to further validate the stability and outreach of the results.

In this large adult-based observational study, we found that higher Lp(a) levels were significantly associated with a higher risk of developing LVH assessed by ECG in the general population, and we further confirmed the consistency of this association in some specific populations.These findings suggest that early intervention for excessive Lp(a) levels in adults and the development of prevention and treatment measures matched to high-risk populations may help prevent premature onset and excessive prevalence of LVH. Nevertheless, due to several hypothesized causative mechanisms the intervention for excessive Lp(a) levels is due to be further examined in experimental and clinical studies.

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding authors.

The studies involving humans were approved by National Center for Health Statistics of the Center for Disease Control and Prevention Institutional Review Board. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

XJY, JG, ZWW and CJQ: Writing – review & editing. JG: Data curation. XJY: Data curation, Writing – original draft. XJY and ZWW: Conceptualization, Methodology, Software. FFW and CJQ: Conceptualization, Funding acquisition, Project administration, Supervision. All authors read and approved the final manuscript.