Abstract Background Lipoprotein(a) Lp(a) is an independent risk factor for atherosclerotic cardiovascular (CV) disease and a one-time measurement is recommended in the guidelines to refine CV risk assessment. However, recent clinical trials have been documenting some changes up to 25% in the placebo groups. Purpose The aim of this study was to evaluate Lp(a) levels variability and to explore possible factors associated with relevant differences in the measurements. Methods This is a retrospective single center analysis of patients with more than one measurement of Lp(a) levels (between 2022 and 2024). The lowest and the highest Lp(a) values, reason for the evaluation, lipid panel, comorbidities and dyslipidemia treatments were collected. Lp(a) variation was calculated as (highest value – lowest value) / lowest value * 100 and it was considered significant if equal to or higher than 25%. The absolute difference on the Lp(a) levels equal or higher than 25 nmol/L was also analyzed. Statistical analysis was done using Qui-square and Mann-Whitney tests and logistic regression. Results A total of 205 patients were analyzed (145 males, 70.7%). Around half of the patients were evaluated in acute context of admission due to CV disease, mainly acute coronary syndrome. Absolute Lp(a) values range from 0.9 nmol/L and 878.7 nmol/L. Lp(a) variability is represented in Graph 1. Median Lp(a) variation was 29.64 56.03. Variation ≥ 25% was verified in 111 patients (54%), with a median Lp(a) variation of 62.50 62.41. Age (z = -2.320, p = 0.020), body mass index (z = -2.048, p = 0.040), history of valvular disease (OR = 0.493, CI 0.257-0.946, p = 0.034) and change in dyslipidemia treatment (OR = 2.857, CI 1.431-5.705, p = 0.003) were different between patients with and without relevant Lp(a) variation. After adjustment, only change in dyslipidemia treatment remain significant (OR = 2.571, CI 1.270 – 5.204, p = 0.009). When considering an absolute change in Lp(a) levels of 25 nmol/L, 65 patients (32%) had a relevant difference that correlated with age (z = -0.387, p = 0.023), overweight (OR = 1.919, CI 1.053-3.495, p = 0.045), on ezetimibe at first evaluation (OR = 2.563, CI 1.210-5.427, p = 0.016) and changes in dyslipidemia treatment (OR = 2.450, CI 1.272-4.720, p = 0.009). Age was the only factor that did not remain significant in the multivariable analysis (age: p = 0.115; overweight: p = 0.016; ezetimibe: p = 0.002; change in treatment: p = 0.007). Conclusion This study elicits that Lp(a) levels variability may be much higher than previously described, up to 54% of the patients in this analysis. Change in dyslipidemia treatment was the strongest predictor of Lp(a) variation. These findings suggest that clinical factors can influence Lp(a), underscoring the need to rethink the concept of a one in a lifetime measurement in CV risk assessment.Graph 1. Lipoprotein(a) variability
Almeida et al. (Sat,) studied this question.