Prevalence of Elevated Lipoprotein(a) and Its Association with Cardiovascular Disease in Patients with Overweight and Obesity in Argentina

Abstract

Background

Lipoprotein(a) [Lp(a)] is an established independent risk factor for atherosclerotic cardiovascular disease (ASCVD). Although both obesity and elevated Lp(a) are highly prevalent, their interrelationship and combined impact on ischemic heart disease (IHD) remain incompletely understood, particularly in Latin American populations.

Objective

To evaluate the distribution of elevated Lp(a) across body mass index (BMI) categories, assess its association with ischemic heart disease (IHD), and determine whether this association differs across BMI categories.

Methods

We conducted a multicenter observational study including 2975 adults from specialized centers in Argentina. Patients were stratified into BMI categories: Group 1: <25, Group 2 BMI: 25–29.9, Group 3: 30–34.9 kg/m², with exploratory subgroups ≥35 and ≥40 kg/m². Elevated Lp(a) was defined as >50 mg/dL or >125 nmol/L.

Results

Median Lp(a) levels decreased across higher BMI categories. The prevalence of elevated Lp(a) declined across increasing BMI categories (34.6% in BMI <25 kg/m², 29.8% in BMI 25-29.9, 30.1% in BMI ≥30, 26.4% in BMI ≥35 and 24.6% in BMI ≥40). However, after adjustment for demographic and cardiometabolic factors, overweight and obesity were associated with higher odds of elevated Lp(a) compared with normal weight (BMI 25–29.9: OR 1.97, 95% CI 1.38–2.82; BMI ≥30: OR 1.85, 95% CI 1.22–2.81) suggesting that crude prevalence estimates and adjusted associations were influenced by differences in baseline risk profiles across BMI categories. Elevated Lp(a) was independently associated with IHD across all BMI categories, without significant interaction between BMI and Lp(a).

Conclusions

In this cross-sectional study, elevated Lp(a) was associated with prevalent ischemic heart disease across BMI categories despite lower crude prevalence among individuals with higher BMI. Because of the observational cross-sectional design, causal inferences cannot be made. Nevertheless, these findings support consideration of Lp(a) assessment across the BMI spectrum.

Graphical Abstract

Introduction

Lipoprotein(a) (Lp(a)) is a cholesterol-rich lipoprotein particle composed of an apolipoprotein B-100 molecule covalently bound to apolipoprotein(a). Elevated Lp(a) is independently associated with a higher risk of atherosclerotic cardiovascular disease (ASCVD), including ischemic heart disease (IHD), stroke, and calcific aortic valve stenosis.¹

Lp(a) levels are largely genetically determined and exhibit minimal intraindividual variability. However, it has been suggested that chronic inflammatory states, such as those present in obesity, could modulate plasma Lp(a) levels through inflammatory and hormonal mechanisms.² Obesity is increasingly recognized as a chronic low-grade inflammatory condition characterized by adipose tissue dysfunction, insulin resistance, and altered hepatic metabolism. Several biological pathways have been proposed to explain potential interactions between obesity and Lp(a), including inflammatory cytokine signaling, insulin-mediated regulation of apolipoprotein(a) synthesis, and changes in hepatic lipoprotein metabolism. Several observational studies have reported an inverse association between body mass index (BMI) and circulating Lp(a) concentrations, a phenomenon sometimes referred to as the “Lp(a) paradox” in obesity and metabolic syndrome.³ However, available evidence remains inconsistent, and the biological basis of these observations has not been fully established with studies reporting inverse, positive, or neutral associations between BMI and circulating Lp(a) concentrations.

Lp(a) concentrations exhibit substantial interindividual and ethnic variability, largely determined by genetic factors, particularly variation within the LPA gene. Individuals of African ancestry generally have higher circulating Lp(a) concentrations than European or Asian populations, whereas data from Latin America remain comparatively limited.⁴ Because Latin American populations are characterized by substantial genetic admixture and a high prevalence of cardiometabolic risk factors, extrapolation of findings from other regions may not be entirely appropriate.

Both obesity and elevated Lp(a) increase the risk of cardiovascular events, and despite being frequent conditions among our patients, the relationship between excess body weight and Lp(a) has not yet been clarified; evidence is limited, and findings are inconsistent across population studies.

The presence of elevated Lp(a) in the overweight population could further increase cardiovascular risk, making potential Lp(a) reduction strategies particularly relevant in patients with increased BMI.

A previous study from our research group evaluated the prevalence of elevated Lp(a) and its association with cardiovascular outcomes in a broad population cohort.⁵ However, that analysis did not specifically address the potential influence of BMI categories on the association between elevated Lp(a) and ischemic heart disease. In Argentina, 61.6% of the population has excess body weight.⁶ The current study extends those findings by examining the association between elevated Lp(a), BMI categories, and ischemic heart disease in a large multicenter Argentine cohort.

This approach enables a more refined assessment of cardiovascular risk in patients with excess body weight.

The primary objective of this study was to evaluate whether elevated Lp(a) is associated with IHD across different BMI categories and to explore whether this association differs according to BMI category.

Methods

We conducted a multicenter cross-sectional observational study in specialized centers in Argentina between January 2022 and March 2025. Adult patients with a plasma Lp(a) measurement obtained in a convenience sample were included from local clinical laboratories at each participating center as part of routine patient care. Detailed information regarding assay methodology, calibration procedures, and laboratory harmonization was not systematically collected. Lp(a) concentrations were reported by participating laboratories either in mg/dL or nmol/L. Each participant contributed a single Lp(a) measurement. Elevated Lp(a) was defined according to guideline-recommended thresholds based on the reporting unit available (>50 mg/dL or >125 nmol/L). No unit conversion was performed.

Once 3000 patients had been reached, the first cut-off for sample analysis was made. Data on weight, height, and cardiovascular history were collected. Individuals with acute inflammatory diseases or severe liver diseases were excluded. Exclusion criteria were determined based on information available in medical records and physician documentation at participating centers. No centralized adjudication process was performed. Acute inflammatory diseases and severe liver diseases were identified according to the clinical diagnoses recorded in routine care.

Variables and Statistical Analysis

We analyzed demographic characteristics and the history of diabetes mellitus (DM), hypertension, smoking and documented IHD.

Missing data were handled using a complete-case approach. Patients with missing values for BMI or other key variables required for the primary analyses were excluded from the analytic dataset. No imputation procedures were performed because the proportion of missing data was low (<1% of the study population).

Patients were stratified by BMI: Group 1: BMI <25 kg/m²; Group 2: BMI 25–29.9 kg/m²; Group 3: BMI ≥30 kg/m². As exploratory subgroups, we evaluated BMI ≥35 and BMI ≥40 kg/m².

Elevated Lp(a) was defined as >50 mg/dL or >125 nmol/L, according to European Atherosclerosis Society recommendations.⁷ Continuous variables were compared using non-parametric tests (Kruskal–Wallis); associations were explored using multivariable logistic regression (model 1: Lp(a) prevalence adjusted for age, sex, hypertension, smoking, diabetes mellitus, dyslipidemia).

IHD was defined as history of acute myocardial infarction (diagnosed by clinical criteria, ECG, cardiac biomarkers, or prior hospitalization for MI) as well as history of angioplasty or myocardial surgical revascularization. The study focused on the presence of established ischemic heart disease as a binary variable, and information regarding the timing, extent, or severity of coronary disease was not systematically available.

The primary objective of the study was to evaluate the association between elevated Lp(a) and IHD across body mass index categories. For this purpose, multivariable logistic regression models were constructed with prevalent IHD as the dependent variable and elevated Lp(a) as the main independent variable (model 2: adjusted for age, sex, hypertension, LDL-C, smoking, diabetes, statin status and family history of IHD). An interaction term between elevated Lp(a) and continuous BMI was tested to assess potential effect modification.

To explore whether the association between elevated Lp(a) and IHD differed across BMI categories, analyses were repeated within predefined BMI strata (<25, 25–29.9, and ≥30 kg/m²), and an interaction term between BMI category and elevated Lp(a) was tested.

In secondary exploratory analyses, logistic regression models were also used to examine the distribution of elevated Lp(a) across BMI categories, using BMI <25 kg/m² as the reference group.

Age-stratified analyses were performed using separate multivariable logistic regression models within each age quartile. Complete-case analysis was used, and participants with missing values in any covariate included in the multivariable models were excluded.

Statistical significance was set at a two-sided p-value <0.05. All analyses were performed using Jamovi (version 2.6.24.0) and Epi Info.

Results

Out of 3000 patients, 25 were excluded due to missing data. Because the proportion of missing data was small (0.8%), complete-case analysis was performed without imputation. A total of 2975 patients with complete weight and height data were included. Baseline characteristics stratified by BMI subgroups are shown in Table 1.

Table 1.

Baseline Characteristics of the Population

	Total	Group 1 (BMI <25)	Group 2 (BMI 25–29.9)	Group 3 (BMI ≥30)	BMI ≥35	BMI ≥40	p
N (%)	2975	939 (31.6)	1190 (40)	846 (28.4)	288 (9.7)	106 (3.6)
Age (mean)	57.8 ±14.4	55.1 ± 16.7	59 ± 14.1	58.3 ± 13.1	57.7 ± 13	56.2 ± 13	<0.001
Female (%)	46.1	63.6	36.1	40.4	45.1	51.9	<0.001
Hypertension (%)	52.2	34.8	53.3	70	76.4	78.3	<0.001
Diabetes (%)	13.2	5.4	11.5	24.1	28.5	26.4	<0.001
Smoking (%)	18.6	14.1	18.8	24.5	20.9	22.6	<0.001
IHD (%)	14.4	10.2	16.1	16.7	16.7	18.2	<0.001
PAD (%)	3.2	3.8	2.8	3.2	2.4	2.8	0.2
Stroke (%)	2.2	1.2	2.5	2.7	1.7	0	<0.01
Aortic stenosis (mild/moderate/severe) (%)	6.4	5.1	6.1	8	9.6	10.6	<0.001

Notes: Data are presented as mean ± standard deviation for continuous variables and n (%) for categorical variables. P values refer to comparisons across BMI categories (Group 1, Group 2, and Group 3) using the Kruskal–Wallis test for continuous variables and the χ² test for categorical variables. BMI ≥35 kg/m² and BMI ≥40 kg/m² are exploratory subgroups and are presented descriptively.

Abbreviations: BMI, body mass index; IHD, ischemic heart disease; PAD, peripheral artery disease.

Use of lipid-lowering therapies differed across BMI categories. Statin therapy was progressively more frequent among individuals with higher BMI (52.2%, 56.6%, and 61.2% in BMI <25, BMI 25–29.9, and BMI ≥30 kg/m², respectively; p=0.001). Ezetimibe use showed a similar trend but did not reach statistical significance (p=0.056). PCSK9 inhibitors and inclisiran were infrequently prescribed in this cohort (<1% of participants), with no significant differences across BMI categories (Supplementary Table 1).

A total of 1680 Lp(a) measurements were reported in mg/dL and 1295 in nmol/L. The median Lp(a) values were 23 mg/dL (P25–P75: 9.95–70) and 44 nmol/L (P25–P75: 14–162). Overall, 31.4% of patients had elevated Lp(a). Table 2 and Figure 1 show median Lp(a) levels across BMI categories.

Table 2.

Lp(a) Levels Across BMI Categories Values are Presented as Median [P25–P75]

BMI Category	n (mg/dL)	Lp(a) mg/dL Median [IQR]	p-value	n (nmol/L)	Lp(a) nmol/L Median [IQR]	p-value
Group 1	552	29.4 [10.0–78.0]	0.001*	387	56.0 [20.0–183.0]	0.027*
Group 2	669	20.0 [7.5–68.0]		521	40.0 [14.0–146.0]
Group 3	311	20.0 [7.8–65.0]		247	48.5 [13.8–172.4]
BMI ≥ 35	93	20.0 [9.0–59.4]		89	33.4 [10.0–107.0]
BMI ≥40	55	21.0 [7.9–56.6]		51	34.0 [17.5–119.2]

Notes: *P <0.05 indicates statistical significance. P values were calculated using the Kruskal–Wallis test.

Abbreviations: BMI, body mass index; IQR, interquartile range; Lp(a), lipoprotein(a).

Figure 1.

Median Lp(a) value across subgroups. Distribution of Lp(a) levels across BMI categories. Box plots represent median values and interquartile ranges (IQR), with whiskers indicating variability beyond the IQR. Median Lp(a) levels were higher in individuals with lower BMI, with statistically significant differences across groups. Kruskal–Wallis: mg/dL (p = 0.0012); nmol/L (p = 0.027).

Individuals with BMI <25 kg/m² had higher Lp(a) concentrations (median 29.4 mg/dL; IQR 10–78, median 56 nmol/L; IQR 20–183) compared with overweight/obesity (BMI≥25 kg/m²) (median 20 mg/dL; IQR 7.5–68, median 40 nmol/L; IQR 13.4–154). Figure 2 shows the percentage of elevated Lp(a) across the study groups. The prevalence of elevated Lp(a) progressively decreased across increasing BMI categories, from normal weight to severe obesity.

Figure 2.

Percentage of elevated Lp(a) across the study groups (thresholds based on the reporting unit available (>50 mg/dL or >125 nmol/L)). Bars represent the percentage of individuals with elevated Lp(a) within each BMI subgroup. A progressive decline in the prevalence of elevated Lp(a) was observed across higher BMI categories. Differences in the prevalence of elevated Lp(a) across BMI categories were evaluated using Pearson’s chi-square test (p = 0.052).

In multivariable logistic regression analysis, individuals with overweight and obesity showed higher odds of elevated Lp(a) compared with the reference group (BMI <25 kg/m²), with significant associations observed in individuals with overweight and obesity, but associations were not statistically significant in the higher BMI subgroups with a progressive attenuation of the odds ratios observed with increasing BMI, suggesting a weaker adjusted association in the higher BMI subgroups, although these findings should be interpreted cautiously given the smaller sample sizes and wide confidence intervals (Table 3).

Table 3.

Logistic Regression Analysis for Elevated Lp(a)

Group	n	Adjusted OR for Elevated Lp(a)	95% CI	p-value
Group 1 (BMI <25)	939	1.00 (reference)	–	–
Group 2 (BMI 25–29.9)	1190	1.97	1.38–2.82	<0.001
Group 3 (BMI ≥30)	846	1.85	1.22–2.81	0.004
BMI ≥35*	288	1.27	0.59–2.74	0.63
BMI ≥40*	106	0.95	0.27–3.42	0.94

Notes: *Exploratory subgroup. Odds ratios (OR) and 95% confidence intervals (95% CI) were derived from multivariable logistic regression analysis with elevated Lp(a) as the dependent variable. Group 1 (BMI <25 kg/m²) was used as the reference category. Model 1 adjusted for age, sex, hypertension, smoking, diabetes mellitus, and dyslipidemia.

Abbreviations: BMI, body mass index; OR, odds ratio; CI, confidence interval.

In a sensitivity analysis excluding dyslipidemia from the multivariable model, the association between BMI category and elevated Lp(a) changed direction and became consistent with the crude prevalence estimates. Compared with individuals with BMI <25 kg/m², the adjusted odds ratios for elevated Lp(a) were 0.79 (95% CI 0.65–0.95; p=0.014) for BMI 25–29.9 kg/m² and 0.78 (95% CI 0.63–0.97; p=0.025) for BMI ≥30 kg/m².

There were 429 patients with IHD (14.4%). To evaluate the association between BMI categories and IHD, multivariable logistic regression models were constructed with IHD as the dependent variable (model 2). BMI categories were included as the main independent variable, using BMI <25 kg/m² as the reference group. Patients with IHD had a higher elevated Lp(a) frequency across all BMI categories (Table 4).

Table 4.

% of Elevated Lp(a) and IHD Across Subgroups (Odds Ratio for IHD* and Elevated Lp(a) - Univariate Analysis)

	% of Elevated Lp(a) in Patients with IHD	% of Lp(a) Elevated in Patients Without IHD	OR	95% CI	p
Group 1	40.6	33.8	1.34	0.87–2.05	0.19
Group 2	38	28.2	1.61	1.16–2.23	<0.001
Group 3	46.2	29.2	2.14	1.36–3.35	<0.001

Note: ORs and 95% CIs were calculated using univariate logistic regression.

Abbreviations: IHD, ischemic heart disease; OR, odds ratio; CI, confidence interval.

In patients with overweight, Lp(a) was independently associated with the presence of IHD (BMI <25 OR 1.59 (95% CI 1.1–2.2; p = 0.008) and BMI ≥25 kg/m² OR 1.58 (95% CI 1.03–2.4; p = 0.033)) In a multivariable analysis (model 2) the presence of IHD was independently associated with increased odds of elevated Lp(a) across the subgroups (Figure 3).

Figure 3.

Association between elevated lipoprotein(a) and ischemic heart disease across BMI categories. BMI ≥35 and BMI ≥40: exploratory subgroups; reported descriptively and/or from separate models. Odds ratios (OR) and 95% confidence intervals (95% CI) were derived from multivariable logistic regression adjusted for age, sex, hypertension, LDL-C, smoking, diabetes, statin status and family history of IHD. BMI <25 kg/m² was used as the reference category.

To explore potential age-related differences, analyses were stratified by age quartiles, and adjusted odds ratios (OR) with 95% confidence intervals (CIs) were reported (Table 5).

Table 5.

Multivariable Logistic Regression Analysis of the Association Between BMI Categories and Prevalent Ischemic Heart Disease, Stratified by Age Quartiles

Age Quartile	n	IHD (n)	BMI 25–30 vs <25 OR (95% CI)	p	BMI ≥30 vs <25 OR (95% CI)	p
Q1 (18–48)	658	33	0.90 (0.34–2.38)	0.832	0.77 (0.24–2.44)	0.660
Q2 (49–58)	655	70	1.01 (0.49–2.07)	0.977	0.61 (0.28–1.33)	0.211
Q3 (59–68)	691	100	1.12 (0.60–2.10)	0.718	0.95 (0.48–1.88)	0.884
Q4 (69–101)	654	161	1.28 (0.78–2.11)	0.326	1.38 (0.80–2.38)	0.242

Notes: Models were adjusted for sex, hypertension, smoking, diabetes mellitus, chronic kidney disease, LDL-C <70 mg/dL, and statin use. Sample sizes reflect complete-case multivariable analyses. Participants with missing covariate data were excluded from the corresponding age-stratified regression models.

Abbreviations: BMI, body mass index; IHD, ischemic heart disease; OR, odds ratio; CI, confidence interval.

The prevalence of elevated Lp(a) was similar among patients with and without diabetes mellitus (32.2% vs 31.2%; p = 0.6). Likewise, median Lp(a) levels expressed in mg/dL and nmol/L were similar between patients with and without DM: 25.4 vs 23 mg/dL (p = 0.865); 30 vs 46 nmol/L (p = 0.682), respectively.

We assessed whether BMI modified the association between elevated Lp(a) and IHD by including an interaction term between Lp(a) and BMI in a multivariable model adjusted for age, sex, smoking, diabetes, LDL-C, statin use, and family history of premature cardiovascular disease. The interaction was not significant (OR 1.0017; 95% CI 0.958–1.046; p = 0.94).

Discussion

The principal findings of this study are threefold. First, elevated Lp(a) was independently associated with prevalent ischemic heart disease (IHD) across all BMI categories. Second, median Lp(a) concentrations and the crude prevalence of elevated Lp(a) decreased with increasing BMI. Third, despite these lower crude prevalence estimates, elevated Lp(a) remained associated with IHD irrespective of BMI, and no significant interaction between BMI and Lp(a) was observed. These findings suggest that the cardiovascular risk associated with elevated Lp(a) is largely independent of body weight.

Previous studies have reported inverse, positive, and neutral associations between adiposity and circulating Lp(a) concentrations. In our cohort, median Lp(a) levels and the prevalence of elevated Lp(a) progressively decreased across increasing BMI categories, consistent with prior reports suggesting an inverse relationship between BMI and Lp(a). Similar findings have been described in large population studies, including the meta-analysis by Willeit et al,⁸ in which higher BMI was associated with slightly lower baseline Lp(a) concentrations.

While the magnitude of the association appeared consistent across BMI strata, formal interaction testing showed no significant effect modification, indicating that BMI category did not meaningfully influence this relationship. The non-significant interaction term suggests that the association between elevated Lp(a) and IHD was broadly similar across BMI categories. However, interaction analyses generally require substantially larger sample sizes than analyses of main effects, and the present study may have had limited power to detect modest effect modification, particularly within the highest BMI categories.

Several biological mechanisms have been proposed, including hyperinsulinemia, insulin resistance, altered hepatic lipoprotein metabolism, and inflammatory regulation of apolipoprotein(a) synthesis.⁹ However, evidence remains inconsistent, and the present study was not designed to evaluate these mechanisms directly.

Clinical studies suggest that diet-induced weight loss may be associated with a transient increase in Lp(a) levels among individuals with obesity, whereas bariatric surgery appears to result in modest reductions.¹⁰ In pediatric populations and in individuals with metabolic syndrome, some studies have reported both positive correlations or inverse associations, indicating that the relationship between adiposity and Lp(a) may vary according to age, sex, and metabolic status.¹¹

Potential mechanisms remain speculative and could involve alterations in hepatic synthesis or clearance pathways,¹² although these hypotheses require further investigation. However, because direct measures of adiposity were not available, the present study cannot provide mechanistic insights into the relationship between body fat distribution and Lp(a).

Previous studies have shown an inverse relationship between Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD) and Lp(a) levels.¹³ However, Kronenberg et al did not observe a significant correlation in European cohorts, suggesting that ethnic and genetic heterogeneity may act as important modulators of this relationship.⁷

An important finding of the present study is the apparent discrepancy between crude prevalence estimates and adjusted associations. Although elevated Lp(a) was less frequent among individuals with higher BMI in unadjusted analyses, multivariable adjustment identified higher odds of elevated Lp(a) in the overweight and obesity categories.

Importantly, these findings should not be interpreted as evidence that excess body weight increases Lp(a) concentrations. Rather, crude prevalence estimates and adjusted odds ratios address different questions. The sensitivity analysis demonstrated that the direction of the association was influenced by model specification, suggesting that residual confounding, referral patterns, indication bias, overadjustment, or collider bias may have contributed to the observed findings. Because participants were recruited from specialized referral centers rather than from a population-based sample, these adjusted associations should be considered exploratory and hypothesis-generating.

We did not observe significant differences in Lp(a) concentrations between individuals with and without diabetes mellitus. This finding contrasts with several previous studies reporting an inverse association between Lp(a) levels and diabetes risk.¹⁴ However, the relatively small number of patients within the lowest Lp(a) strata and the cross-sectional design of the present study limit the interpretation of this observation.

Elevated Lp(a) levels increase the risk of IHD. Our findings are consistent with the overall body of evidence and further show that elevated Lp(a) maintains an independent association with IHD in individuals with overweight and obesity. The association between elevated Lp(a) and IHD persisted after adjustment for cardiometabolic risk factors, suggesting that Lp(a)-related cardiovascular risk is not fully explained by differences in metabolic status. The prevalence estimates observed in the present study should be interpreted within the context of a referral-based centers cohort. Because participants were recruited from specialized centers, both elevated Lp(a) and ischemic heart disease may have been overrepresented relative to the general population.

Among individuals with BMI 25–30 and those with BMI ≥30, the odds of IHD were approximately 60% higher when elevated Lp(a) is present, even after adjustment for age, sex, DM, LDL-C, statin use, and family history of premature cardiovascular disease. In a cohort of more than 4600 participants from the MESA study, neither elevated Lp(a) nor a high waist-to-hip ratio (WHR) alone increased cardiovascular disease risk; however, when both factors were present simultaneously, risk increased markedly, more than doubled compared with the group without these abnormalities. In contrast, BMI did not modify this relationship.¹⁵ Our findings differ from those reported in MESA, where the association between Lp(a) and cardiovascular events appeared to vary according to waist-to-hip ratio. One possible explanation is that BMI and waist-to-hip ratio capture different aspects of body composition. Whereas BMI reflects overall body size, waist-to-hip ratio is more closely related to visceral adiposity and fat distribution. Because measures of central adiposity were not available in the present study, we cannot exclude the possibility that effect modification may be present when adiposity is assessed using more specific anthropometric markers. Therefore, the absence of interaction across BMI categories should not be interpreted as evidence that body fat distribution is unrelated to Lp(a)-associated cardiovascular risk. Differences in study design, population characteristics, and statistical power for interaction analyses may also have contributed to the discrepant findings. Other studies have found an additive association between Lp(a) levels, BMI, and the incidence of myocardial infarction.³

These findings support the biological understanding that Lp(a) concentrations are predominantly genetically determined and minimally influenced by other factors. Consequently, the risk associated with elevated Lp(a) appears largely independent of BMI, suggest that the observed association between elevated Lp(a) and IHD was similar across BMI categories, however, causal inference is limited by the observational design. This reinforces the importance of assessing Lp(a) across all BMI categories, including in individuals with overweight or obesity, and emphasizes that elevated Lp(a) should be considered a risk factor regardless of metabolic phenotype.

Limitations

This study has several limitations.

First, the cross-sectional observational design precludes assessment of temporal relationships and does not allow causal inference regarding the associations among BMI, Lp(a), and ischemic heart disease.

Second, participants were recruited from specialized referral centers using a convenience sampling strategy. Consequently, selection and indication biases may have influenced both the likelihood of Lp(a) testing and the prevalence of cardiovascular disease within the cohort, limiting the generalizability of the findings to the broader Argentine population.

Third, Lp(a) measurements were obtained from multiple local laboratories without centralized testing or formal harmonization procedures. Detailed information regarding assay type, calibration, and isoform sensitivity was not available, and inter-laboratory variability may therefore have influenced absolute Lp(a) concentrations.

Fourth, BMI was the only anthropometric measure available. More specific markers of adiposity and fat distribution, including waist circumference, waist-to-hip ratio, waist-to-height ratio, visceral adiposity, and metabolic syndrome components, were not systematically collected. Therefore, we were unable to evaluate whether central adiposity modifies the association between Lp(a) and ischemic heart disease.

Finally, although multivariable adjustment included major cardiovascular risk factors, residual confounding cannot be excluded. The observed discrepancy between crude prevalence estimates and adjusted associations, together with the sensitivity of the results to model specification, suggests that some findings—particularly those regarding the relationship between BMI categories and elevated Lp(a)—should be considered exploratory and hypothesis-generating.

Conclusions

Patients with overweight and obesity exhibited lower average Lp(a) levels than normal-weight individuals. However, elevated Lp(a) was independently associated with prevalent ischemic heart disease in individuals with excess body weight, with a magnitude of association like that observed in normal-weight individuals. These findings suggest that elevated Lp(a) confers cardiovascular risk across all categories of BMI and reinforce the importance of assessing Lp(a) regardless of weight.

Funding Statement

This research did not receive any funding.

Declaration of Generative AI and AI-Assisted Technologies

During the preparation of this work, the author(s) used ChatGPT 5.0 to provide writing assistance (grammar review, technical translation, and journal style adjustment) and Paperpile to manage references. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.

Data Sharing Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request. Due to ethical and privacy restrictions related to patient-level data, the dataset is not publicly available. De-identified data can be shared with qualified researchers for non-commercial purposes, subject to institutional approval and compliance with applicable data protection regulations.

Ethical Review Statement

The study was approved by the Clinical Research Ethics Committee of the Faculty of Medical Sciences at the National University of Cuyo (IRB number: 34/2024) and by the individual ethics committees at each participating center. It complies with international and national regulations for clinical research, including the Good Clinical Practice Guidelines and the Declaration of Helsinki and its amendments. Informed consent was waived for retrospectively enrolled patients, while prospective participants provided written informed consent. All patient data were anonymized using a pre-established coding system. The study was conducted as a cross-sectional multicenter registry. Depending on local ethics committee requirements and data collection procedures at participating centers, informed consent was waived for patients whose data were extracted from existing clinical records, whereas written informed consent was obtained when required for contemporaneous data collection.

Author Contributions

Ezequiel Hernán Forte: Conceptualization, formal analysis, investigation, methodology, supervision, written original draft and written review and editing.

Arrupe Matias: Conceptualization, formal analysis, methodology, supervision.

Bufi Leonardo: Validation, writing – original draft, and writing – review and editing, supervision.

Gimenez Sergio: Conceptualization, formal analysis, investigation, methodology, supervision.

Lavalle Cobo Augusto: Validation, writing – original draft, and writing – review and editing, supervision.

Matta Maria Gabriela: Formal analysis, investigation, methodology, supervision.

Nogueira Juan Patricio: Formal analysis, methodology, supervision, writing final version.

Renna Nicolas: Formal analysis, methodology, validation, writing original draft and final version.

Salmeri Emiliano: Conceptualization, formal analysis, investigation, methodology, supervision, writing original draft and final version.

Schreier Laura: Conceptualization, formal analysis, investigation, methodology, supervision, written original draft, writing – review and editing.

Corral Pablo: Conceptualization, formal analysis, investigation, methodology, supervision, writing - original draft and writing - review and editing.

All the author cited contributed to the conception, study design, execution, acquisition of data, analysis, interpretation, investigation, methodology, supervision, validation, writing (original draft, review and editing) of the study. All the authors have agreed on the journal the article be submitted and reviewed and agreed on all versions of the article before submission. All the authors have reviewed and agreed on all versions of the article before submission, and agreed to take responsibility and be accountable for the contents of the article.

Disclosure

Forte Ezequiel H reports honoraria for Novo Nordisk, ADIUM, Boehringer and ELEA. Advisory Board digital for Novo Nordisk. Arrupe Matias reports honoraria for Adium, Novartis, Servier, Novo Nordisk and research funding from Novartis, Novo Nordisk and Astra Zeneca. Bufi Leonardo M, declares no conflicts of interest in this work. Gimenez Sergio, declare honoraria for Adium, Novartis, Elea, Novo Nordisk, Baliarda, Eurofarma, Lafage Ultragenyx and meeting/travel support from Siegfried. Advisor for Adium and Teva. Lavalle Cobo Augusto, reports honoraria for Adium, Novartis, Servier, Novo Nordisk, Elea, Casasco, Servier, Gador, meeting/travel support from Adium and Elea Advisor for MSD, Adium, Elea and research funding from Novartis. Matta Maria Gabriela, declares no conflicts of interest in this work. Nogueira Juan Patricio, declares no conflicts of interest in this work. Renna Nicolas, declare personal fees for lectures, advisory board participation, and/or consulting from Astra-Zeneca and Medtronic, Adium, Gador, Novartis, Casasco, Servier, AZ, Novo Nordisk, all outside the submitted work. Salmeri Emiliano reports honoraria from Novo Nordisk, AstraZeneca, Adium, Elea, Gador and meeting/travel support from Novo Nordisk, AstraZeneca, Adium, Gador and Elea. Schreier Laura, declares no conflicts of interest in this work. Corral Pablo, reports receiving honoraria and/or serving on advisory boards for Amgen, Novartis, MSD, AstraZeneca, and Novo Nordisk, and receiving research funding from Novartis. The authors declare no other conflicts of interest.