Global prevalence of dyslipidemias in the general adult population: a systematic review and meta-analysis

Abstract

Introduction

Global regional variations in dyslipidemia prevalence underscore the need for an updated global overview of the magnitude of dyslipidemia over time.

Objective

To estimate the global prevalence of dyslipidemia in adults, considering scientific evidence published in the last 20 years, while analyzing associated sociodemographic and geographic factors.

Methodology

A systematic review and meta-analysis were conducted following PRISMA guidelines. The bibliographic search was performed in PubMed, SCOPUS, Web of Science, and EMBASE, including observational studies published between 2000 and 2025 that reported the prevalence of dyslipidemia components in general adult populations, including hypertriglyceridemia, hypercholesterolemia, hypoalphalipoproteinemia or low high-density lipoprotein cholesterol (HDL-C) levels (< 40 mg/dL for men and < 50 mg/dL for women), and elevated low-density lipoprotein cholesterol (LDL-C). Statistical analyses included meta-analysis with random effects models, stratified analyses by sex and country, and meta-regression by publication year.

Results

A total of 206 studies were included, primarily cross-sectional (98.5%). The global prevalence was 28.8% for hypertriglyceridemia, 24.1% for hypercholesterolemia, 38.4% for low HDL-C, and 18.93% for high LDL-C. Men presented a higher prevalence of hypertriglyceridemia (33.8% vs. 24.5% in women), while women showed a higher frequency of low HDL-C (40.5% vs. 34.1% in men). The highest prevalences of low HDL-C and mixed dyslipidemia patterns were observed in the Middle East and Latin America. Temporal trends revealed a progressive increase in hypertriglyceridemia and low HDL-C, with a decrease in hypercholesterolemia and high LDL-C.

Conclusions

There is an epidemiological transition in the dyslipidemia profile, from classic hypercholesterolemia toward a predominance of hypertriglyceridemia and low HDL-C, with important differences by sex and region. These findings significantly affect cardiovascular prevention and health policies addressing residual risk beyond LDL-C control.

Supplementary Information

The online version contains supplementary material available at 10.1186/s41043-025-01054-3.

Introduction

Dyslipidemia is defined as an alteration in plasma lipid levels that poses cardiovascular risk, specifically including elevated total cholesterol, low-density lipoprotein cholesterol (LDL-C), and triglyceride levels, as well as reduced high-density lipoprotein cholesterol (HDL-C) levels or hypoalphalipoproteinemia. These dyslipidemias represent substantial risk factors for cardiovascular diseases, which constitute the leading cause of death globally [1]. According to a recent report, approximately 1 in 3 adults worldwide presents with hypercholesterolemia, contributing to a considerable proportion of coronary events and cerebrovascular accidents [2]. International guidelines, such as those from the American College of Cardiology/American Heart Association (ACC/AHA), emphasize the importance of early and continuous lipid control to reduce morbidity and mortality associated with dyslipidemia [3].

In recent years, the prevalence of dyslipidemia has shown heterogeneous trends: while some high-income countries have achieved slight reductions in mean population cholesterol levels, various regions of Asia and Latin America report notable increases, attributable in part to drastic changes in dietary habits and increasing sedentary lifestyles [4–6]. In Latin America, the nutritional transition—characterized by greater availability of ultra-processed foods and a Westernized dietary pattern—appears to have potentiated the emergence of unhealthy lipid profiles, particularly in urban areas [7]. These regional variations underscore the need for an updated global overview of the magnitude of dyslipidemia and its possible conditioning factors.

Despite multiple epidemiological studies conducted, knowledge gaps persist regarding the true burden and distribution of dyslipidemia in the adult population. One of the main obstacles lies in the lack of uniformity in diagnostic cut-off points and in the criteria used to classify dyslipidemia subtypes, which hinders direct comparison of results between different investigations [4–7]. Additionally, marked methodological heterogeneity is observed: sample sizes, sampling methods, and geographic coverage vary significantly, with underrepresentation of certain regions and socioeconomic groups. These disparities can lead to divergent prevalence estimates and limit the applicability of conclusions to specific contexts.

For all the aforementioned reasons, the main objective of this systematic review (SR) is to estimate the global prevalence of dyslipidemia in adults, considering the scientific evidence published in the last 20 years. Sociodemographic and geographic factors will be analyzed to identify relevant distribution patterns. Through this comprehensive analysis, we aim to provide robust evidence that clarifies the global magnitude of dyslipidemia and serves as a foundation for developing targeted health policies and clinical interventions to prevent and control dyslipidemia-related cardiovascular diseases.

Methodology

Study design

To meet the proposed objective, an SR with meta-analysis was conducted, following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines [8] adapted for prevalence studies [9, 10].

Search strategy

The bibliographic search was performed in four electronic databases: PubMed, SCOPUS, Web of Science (including the SciELO catalog), and EMBASE, selected for their extensive coverage of international biomedical literature and based on the Cochrane Collaboration’s methodological recommendations for systematic reviews. The search period was established from January 2000 to March 2025, encompassing a broad time interval to capture contemporary studies and relevant research published in the last 25 years on this topic.

The search strategy was specifically designed to maximize specificity without excessively compromising sensitivity. To this end, the search was limited to terms included in article titles, using keywords related to “dyslipidemia,” “hypercholesterolemia,” “hypertriglyceridemia,” “low HDL,” and “high LDL,” as well as their Spanish equivalents. Additionally, the term “metabolic syndrome” was incorporated as a descriptor, given that many studies report data on dyslipidemias in the context of this condition. The search was restricted to articles that also included terms of prevalence or epidemiology in their titles, thus ensuring that the estimation of population frequency was a primary objective of the research and not a secondary finding. This methodological decision allowed for more precise results and reduced noise in the initial search, prioritizing studies specifically designed to determine the population prevalence of dyslipidemias. Complete details of the search strategies, including specific terms and syntax used in each database, are presented in Supplementary Material 1.

Selection criteria

Regarding inclusion criteria, population-based observational studies with probabilistic sampling (mainly cross-sectional and cohort studies) that reported prevalence data of hypertriglyceridemia, hypercholesterolemia, low HDL-C, or high Low-Density Lipoprotein Cholesterol (LDL-C) in adults (≥ 18 years) were prioritized. This decision is based on the fact that designs with probabilistic sampling provide more precise and representative estimates of prevalence in the general population, reducing the risk of selection bias. Only studies using standardized diagnostic definitions for each type of dyslipidemia according to internationally recognized criteria, such as those of the Adult Treatment Panel III (ATP III) [11], the International Diabetes Federation (IDF) [12], or the Joint ESC/European Atherosclerosis Society (EAS) [13], among others, were accepted.

To enhance transparency regarding the diagnostic criteria employed, specific cut-off values were defined as follows: for hypertriglyceridemia, the most commonly used threshold was ≥ 150 mg/dL (≥ 1.7 mmol/L). For hypercholesterolemia, cut-off points ranged between ≥ 200 mg/dL and ≥ 240 mg/dL for total cholesterol, with 200 mg/dL being the most frequently employed value. Low HDL-C was consistently defined as < 40 mg/dL (< 1.03 mmol/L) for men and < 50 mg/dL (< 1.29 mmol/L) for women across studies. For elevated LDL-C, the most frequent cut-off points were ≥ 130 mg/dL, followed by ≥ 160 mg/dL, depending on cardiovascular risk stratification according to NCEP-ATP III recommendations. These standardized thresholds ensured comparability across studies while maintaining adherence to internationally recognized clinical guidelines. No language or geographical location restrictions were imposed to obtain a global panorama of the prevalence of dyslipidemias and avoid publication bias associated with studies published only in English.

Conversely, case reports, letters to the editor, systematic reviews, bibliometric reviews, and any study without a clear methodology for estimating prevalence were excluded. This exclusion is justified because such formats do not provide primary prevalence data from standardized methods. Likewise, research conducted in very particular populations (for example, only in people with advanced chronic diseases that substantially modified the lipid profile) was discarded, as their inclusion could overestimate the actual prevalence in the general population and distort the combined estimates. Any duplicate publication or those whose data could not be disaggregated to estimate the specific prevalence of each type of dyslipidemia were also left out, thus avoiding double-counting results or imprecise data in the meta-analysis.

Study selection process

The results obtained from the databases were exported to Rayyan software to facilitate reference management and automated detection of duplicates. Once duplicate records were eliminated, two independent reviewers (in blind mode) evaluated titles and abstracts to determine articles’ potential eligibility.

In case of discrepancy, the full text was reviewed or discussed in a joint session to reach a consensus; if disagreement persisted, a third reviewer was consulted. Subsequently, the full texts of potentially eligible studies were examined in detail to verify compliance with all inclusion and exclusion criteria. This process was documented following the phases recommended by PRISMA, without registration on any platform.

Data extraction

Information from the selected studies was independently extracted by two researchers, using an Excel 2023 spreadsheet specifically designed for this review. For each study, the following were documented: bibliographic data (authors, year of publication), design characteristics (type of study, collection period), country or region where the research was conducted, target population and sample size, sampling and diagnostic methods, as well as the main prevalence results of each dyslipidemia evaluated.

To ensure accuracy in information collection, a third researcher compared the records of both reviewers, resolving any inconsistencies by reviewing the original text and joint discussion. This double-review process and final inspection sought to minimize extraction errors and ensure the fidelity of the information included in the analysis.

Risk of bias assessment

Two reviewers independently carried out the risk of bias assessment using the Muun et al. tool [9], which is specifically designed for prevalence studies. This tool evaluates key aspects such as the quality of sampling, clear definition of the study population, validity of diagnostic measurements, response rate, and the possible presence of selection or information biases, assigning an individual score to each domain.

Each study received an overall rating from the sum of scores across different criteria, classifying it as low, moderate, or high risk of bias according to the points obtained. Disagreements between reviewers were resolved through discussion and, if necessary, intervention by a third reviewer. Finally, these assessments were considered when interpreting the findings and the soundness of the review’s conclusions, both in descriptive analyses and the potential meta-analysis.

Statistical analyses

All statistical analyses were performed using R version 4.2.2 and the packages meta, metafor, dplyr, ggplot2, and readxl. First, databases for each dyslipidemia (hypertriglyceridemia, hypercholesterolemia, low HDL-C, and high LDL-C) were imported from Excel files. Subsequently, data cleaning was performed to ensure correct coding of variables; unreported values were recategorized as NA, and all numerical variables were properly converted. For each record, an identification label was created based on the author’s name and year of publication to facilitate the presentation of results in forest plots.

The first analysis consisted of a global meta-analysis for each dyslipidemia, calculating the combined prevalence using a random effects model. The Freeman-Tukey double arcsine transformation was used to stabilize the prevalence variances, the Clopper-Pearson method for confidence intervals, and the DerSimonian-Laird estimation for between-study variance. The Hartung-Knapp adjustment was applied to obtain more conservative confidence intervals. Heterogeneity was evaluated using the I² and tau² statistics, and the results were visualized using forest plots ordered by effect size.

To explore sources of heterogeneity, stratified sensitivity analyses were performed by sex (women and men), continent, and country. For the sex analysis, studies reporting separate data were filtered, and independent meta-analyses were conducted for each group. Additionally, a direct comparison between sexes was performed through a risk ratio meta-analysis (men vs. women), using the Mantel-Haenszel method under a random effects model. Likewise, aggregated analyses at the country level were performed by summing events and sample sizes for each nation, allowing visualization of the geographical variability in the prevalence of dyslipidemias.

Finally, a meta-regression was performed using the year of publication of the studies as a continuous variable to evaluate whether there was a temporal trend in the reported prevalence. The meta-regression was adjusted on the initial estimated random effects model and was graphically represented through bubble diagrams, in which the size of the points was inversely proportional to each estimate’s precision. This analysis allowed for the graphical and statistical exploration of the possible influence of the publication year on the prevalence results.

Additionally, to explore regional heterogeneity in temporal trends, stratified meta-regression analyses were performed by continent/world region (Africa, Asia, Europe, Latin America, the Middle East, and North America) using publication year as a continuous variable. For each dyslipidemia type, separate meta-regression models were fitted for each continent, following the same methodology as the global analysis with Freeman-Tukey double arcsine transformation and random effects models. Regional trends were considered statistically significant at p < 0.05, and the results were visualized through continent-specific bubble plots with regression lines indicating significant temporal trends. This stratified approach allowed identification of region-specific temporal patterns that may be masked in global analyses, providing insights into geographic drivers of worldwide dyslipidemia trends.

Results

Article selection diagram

A total of 23,043 records were identified through database searches (Scopus: 6,104; Embase: 7,097; PubMed: 5,031; Web of Science: 4,811). After removing duplicates, 9,634 records were evaluated, excluding 8,952 for not assessing dyslipidemia prevalence (n = 3,743), including ineligible populations (n = 1,320), inappropriate designs (n = 2,980), lack of cut-off values (n = 545), or inconsistent data (n = 364). Of the 682 full-text articles evaluated, 481 were excluded due to non-probabilistic sampling (n = 284), inadequate definition of dyslipidemia (n = 92), or inappropriate designs (n = 105). Finally, 201 studies were included in the qualitative and quantitative synthesis. One of these studies analyzed two independent population databases, contributing two separate data sets for hypertriglyceridemia, low HDL-C, and hypercholesterolemia. This entire process can be visualized in supplementary material 2.

General characteristics of included studies

The complete summary table is presented in supplementary material 3, while, given the number of articles selected, the reference list of all articles is found in supplementary material 4. In total, 202 observational studies published between 2002 and 2024 were included, the majority being cross-sectional designs (n = 200), and to a lesser extent, population-based longitudinal studies (n = 2). All studies employed probabilistic sampling methods to select participants, including simple random, stratified, cluster, or multistage combinations. Approximately two-thirds used complex multistage or stratified schemes based on national censuses, official registries, household listings, or population databases. In contrast, others conducted direct random sampling on urban or rural communities.

Regarding selection criteria, most studies included adults aged 18 years or older who were permanent residents in the evaluated communities. Some studies specified more specific age ranges (for example, 25 to 65 years or 35 to 74 years), depending on local epidemiological surveillance objectives. The most frequent exclusions were pregnant women, hospitalized individuals, people with severe acute illnesses, or those who did not complete essential biochemical measurements (such as lipid profile or glucose). In several studies, subjects with chronic diseases that could alter lipid profiles, such as liver or renal insufficiency or neoplasms, or users of lipid-lowering medications, were also excluded. Most studies also required a minimum fasting period of 8 to 12 h before blood sampling for lipid measurement.

Regarding demographic characterization, the sample size of the studies varied widely, from small community surveys of around 200 participants to large national studies that included tens or hundreds of thousands of adults. The average age of the evaluated populations generally ranged between 36 and 67 years, predominantly reflecting young and middle-aged adult populations. However, in some studies, focusing on cardiovascular risk, the average age exceeded 60 years. Although not all studies reported sex distribution in a standardized manner, in those that did, the percentage of female participants varied between 45% and 55%, generally evidencing a balanced composition between both sexes.

In the included studies, the cut-off points for hypertriglyceridemia were quite homogeneous, situated in most cases at ≥ 150 mg/dL (≥ 1.7 mmol/L), by criteria established by international guidelines such as NCEP-ATP III and IDF. Although these were rare, a few studies used slightly higher thresholds (for example, ≥ 177 mg/dL [≥ 2.0 mmol/L] or ≥ 200 mg/dL [≥ 2.3 mmol/L]). In the case of hypercholesterolemia, the reported cut-off points ranged between ≥ 200 mg/dL (≥ 5.2 mmol/L) and ≥ 240 mg/dL (≥ 6.2 mmol/L) for total cholesterol, with 200 mg/dL (5.2 mmol/L) being the most commonly used value.

Regarding HDL-C, the definition of low HDL-C was very consistent across studies: <40 mg/dL for men and < 50 mg/dL for women, both in conventional units and their equivalents in mmol/L (1.03 mmol/L for men and 1.29 mmol/L for women). This definition remained constant even in studies that employed different reference guidelines. Finally, for elevated LDL-C, when reported, the most frequent cut-off points were ≥ 130 mg/dL, followed to a lesser extent by ≥ 160 mg/dL, depending on stratified cardiovascular risk, according to NCEP-ATP III recommendations. In several cases, the exact value for LDL-C was not specified or was inferred from estimation formulas (for example, the Friedewald formula), without reporting a specific prevalence. Overall, the relative homogeneity in the cut-off points for triglycerides and HDL-C favored comparability between studies. However, there was greater variability in the total cholesterol and LDL-C thresholds.

Finally, regarding methodological quality, all studies achieved scores between 8 and 9 points out of a maximum of 10, corresponding to a low risk of bias. This methodological consistency reinforces confidence in the validity of the extracted data and the robustness of the quantitative syntheses performed in the meta-analysis. The geographic diversity of the studies, with representation from the Americas, Asia, Africa, and Europe, also provides a broad and balanced global vision of the worldwide prevalence of dyslipidemias in adults over the last two decades.

Global meta-analysis and sensitivity analysis of dyslipidemias

Regarding hypertriglyceridemia, the combined global prevalence was 28.77% (95% CI: 27.19–30.41). When stratified by sex, a higher prevalence was observed in men (33.80%, 95% CI: 31.84–35.82) than in women (24.45%, 95% CI: 22.73–26.25). Regarding hypercholesterolemia, the global prevalence was 24.09% (95% CI: 20.25–28.40). The prevalence was similar between men (23.82%, 95% CI: 19.36–28.94) and women (27.16%, 95% CI: 22.42–32.49) (See Table 1; Fig. 1).

Table 1.

Global prevalence of dyslipidemias (hypertriglyceridemia, hypercholesterolemia, low HDL-C, and high LDL-C) globally and by subgroups

Dyslipidemia	Analysis Category	Group	Studies*	Participants	Prevalence (%)	95% CI
Hypertriglyceridemia	General	Global	190	17,252,139	28.77	27.19–30.41
	By Sex	Men	150	6,352,296	33.80	31.84–35.82
		Women	150	5,496,458	24.45	22.73–26.25
Hypercholesterolemia	General	Global	82	990,164	24.09	20.25–28.40
	By Sex	Men	60	177,571	23.82	19.36–28.94
		Women	61	215,569	27.16	22.42–32.49
Low HDL-C	General	Global	186	16,555,987	38.43	35.04–41.93
	By Sex	Men	151	5,554,571	34.13	30.38–38.08
		Women	150	6,433,878	40.53	35.51–45.76
High LDL-C	General	Global	68	1,013,924	18.93	15.42–23.03
	By Sex	Men	47	270,651	19.57	14.93–25.23
		Women	47	292,828	21.03	16.23–26.78

*The corresponding figures for hypertriglyceridemia, hypercholesterolemia, and low HDL-C are 189, 80, and 185 studies, of which one study contributed data from two independent populations

Fig. 1.

Global prevalence of dyslipidemias (hypertriglyceridemia, hypercholesterolemia, low HDL-C, and high LDL-C) by subgroups: comparison across dyslipidemia types, sex, and publication decades

Regarding the prevalence of low HDL-C, the data showed a global prevalence of 38.43% (95% CI: 35.04–41.93). The prevalence was higher in women (40.53%, 95% CI: 35.51–45.76) than in men (34.13%, 95% CI: 30.38–38.08). The distribution by decades showed relatively stable values. Finally, a global prevalence of 18.93% (95% CI: 15.42–23.03) was obtained for high LDL-C. The prevalence was somewhat higher in women (21.03%, 95% CI: 16.24–26.78) than in men (19.57%, 95% CI: 14.93–25.23) (See Table 1; Fig. 1).

Also, given the observed heterogeneity in LDL-C cut-off points across studies, we performed a stratified analysis to understand better the impact of different diagnostic thresholds on prevalence estimates. Studies using the ≥ 130 mg/dL (≥ 3.4 mmol/L) cut-off point, which represents the more inclusive threshold recommended for moderate cardiovascular risk patients, showed a combined prevalence of 23.35% (95% CI: 18.35–39.24) based on 40 studies. In contrast, studies employing the more stringent ≥ 160 mg/dL (≥ 4.1 mmol/L) threshold, typically used for higher-risk stratification, demonstrated a lower prevalence of 9.41% (95% CI: 6.43–13.58) across 18 studies.

Geographic distribution of dyslipidemia prevalence: analysis by country

Figure 2 presents world heat maps corresponding to the prevalence of the main dyslipidemias evaluated: hypertriglyceridemia, hypercholesterolemia, low HDL-C, and high LDL-C. Regarding hypertriglyceridemia, the highest values were observed in Lithuania (56.71%), Palau (44.37%), and South Africa (43.68%), while the lowest prevalences were recorded in Benin (2.03%), Brazil (7.79%), and China (8.20%). For hypercholesterolemia, the countries with the highest rates were Pakistan (79.81%), Venezuela (70.73%), and Saudi Arabia (69.45%), while the lowest prevalences corresponded to Australia (6.91%), Japan (8.22%), and Lithuania (9.55%).

Fig. 2.

Global geographic distribution of dyslipidemia prevalence: (A) hypertriglyceridemia, (B) hypercholesterolemia, (C) low HDL-C, and (D) high LDL-C

The data for low HDL-C show a pattern similar to hypercholesterolemia, with the highest prevalence in Pakistan (79.81%), Venezuela (70.73%), and Saudi Arabia (69.45%), and the lowest values in Australia (6.91%), Japan (8.22%), and Lithuania (9.55%). Finally, for high LDL-C, the countries with the highest rates were Lithuania (68.35%), Germany (58.70%), and Ecuador (52.88%), while the lowest values corresponded to the Philippines (4.03%), Burkina Faso (2.13%), and Brazil (19.21%). To view the detailed prevalence by sex, please refer to supplementary material 5.

Temporal trends in prevalence: meta-regression analysis

Figure 3 shows the results of meta-regressions by year of publication, both at the global level and stratified by sex, for each of the four types of dyslipidemia analyzed. An ascending and sustained pattern was observed in the prevalence of hypertriglyceridemia between 2000 and 2025, with constant increases in men, women, and globally (Fig. 3A). Although prevalences were consistently higher in men, an increasing trend was also evident in women.

Fig. 3.

Temporal trends in dyslipidemia prevalence by meta-regression analysis: (A) hypertriglyceridemia, (B) hypercholesterolemia, (C) low HDL-C, and (D) high LDL-C

In contrast, for hypercholesterolemia, a progressive decrease in prevalence over time was identified, both globally and in both sexes (Fig. 3B). This decline was more pronounced in women, who initially presented higher values than men but converged toward similar values in the last decade.

The prevalence of low HDL-C showed a slightly ascending trend in the three groups analyzed (Fig. 3C). Women maintained the highest rates throughout the period, while men, despite having lower values at the beginning, presented a progressive increase that reduced the gap. At the global level, the slope also suggests a moderate increase in the prevalence of low HDL-C over time. Finally, a continuous decrease in the prevalence of high LDL-C was observed between 2000 and 2025 (Fig. 3D). This downward trend was consistent in men, women, and the global analysis, although slight slope differences existed.

Regional Temporal trends in dyslipidemia prevalence

To address regional heterogeneity in temporal trends, we conducted stratified meta-regression analyses by continent/world region for each dyslipidemia type. These analyses revealed important geographic differences in temporal patterns that were masked in the global meta-regression. For hypertriglyceridemia, significant increasing trends were observed in Latin America (+ 56.3% per decade, p = 0.018) and North America (+ 44.8% per decade, p = 0.003), while Africa, Asia, Europe, and the Middle East showed non-significant changes over time. This pattern suggests that the global increasing trend in hypertriglyceridemia is primarily driven by the Americas, potentially reflecting the impact of westernized dietary patterns and lifestyle changes in these regions.

For hypercholesterolemia, the Middle East showed a significant declining trend (-25.5% per decade, p = 0.003), while other regions demonstrated non-significant changes, indicating that the global decline may be geographically concentrated. Low HDL-C patterns showed regional stability across most continents, with only Africa demonstrating a marginally significant decline (-57.8% per decade, p = 0.053). High LDL-C trends were generally non-significant across regions, with the Middle East showing the only significant decline (-35.7% per decade, p = 0.019). These regional differences highlight the importance of continent-specific epidemiological surveillance and suggest that global dyslipidemia trends reflect varying contributions from different geographic regions, likely influenced by differences in healthcare systems, medication access, dietary transitions, and socioeconomic development patterns.

Discussion

Main findings

In this SR with meta-analysis, which included more than 200 population studies from around the world, the global prevalence of the main dyslipidemias in adults was estimated. The combined prevalence was 28.8% for hypertriglyceridemia, 24.1% for hypercholesterolemia, 38.4% for low HDL-C, and 18.93% for high LDL-C. Notable differences were observed according to sex, country, and year of publication. In general, men showed a higher prevalence of hypertriglyceridemia, while women presented a higher frequency of low HDL-C. The highest prevalences of low HDL-C and hypercholesterolemia were reported in regions of the Middle East, Latin America, and Asia. In contrast, the highest prevalences of high LDL-C were observed primarily in Europe (Lithuania and Germany). Temporal trends showed a progressive increase in hypertriglyceridemia and low HDL-C, while the prevalence of hypercholesterolemia and high LDL-C decreased.

Comparability with other studies

When contextualizing the contribution of this systematic review and meta-analysis within existing literature, it is imperative to acknowledge previous research examining dyslipidemia epidemiology. Although substantial descriptive works on dyslipidemias exist, comprehensive systematic reviews with meta-analyses investigating the complete lipid profile globally remain limited. Notable publications such as Pirillo et al. (2021) [14], Liu (2023) [15], and the NCD Risk Factor Collaboration study (2020) [16] have provided valuable insights, yet present methodological constraints and focused scopes that limit their comprehensiveness. In contrast, our investigation employs robust meta-analytical methodologies across a comprehensive dataset encompassing all adult age cohorts and six continents, examining the complete lipid profile and revealing previously undocumented epidemiological transitions.

Beyond the methodological considerations mentioned above, it is important to contrast our results with those of previous studies. Our global hypertriglyceridemia prevalence of 28.8% aligns closely with regional estimates reported by Pirillo et al. (2021) [14], who documented similar ranges across different populations. However, our hypercholesterolemia prevalence of 24.1% is somewhat lower than the estimates presented in the NCD Risk Factor Collaboration (2020) [16] study, which reported higher mean total cholesterol levels in several regions, possibly reflecting differences in diagnostic cut-off points and population selection criteria.

The regional differences found in our study reflect patterns similar to those reported by Xia et al. (2023) [17] in their meta-analysis on dyslipidemia in Chinese adults, where they found that low HDL-C was the most prevalent dyslipidemia (24.5%), followed by hypertriglyceridemia (15.4%), reinforcing our observations about the growing importance of hypertriglyceridemia and low HDL-C in many regions. Our observation of higher low HDL-C prevalence in women (40.5% vs. 34.1% in men) represents a global pattern that differs from Benjamin et al. (2018) [18] findings in US NHANES data, where women showed higher HDL-C levels and lower prevalence of low HDL-C, highlighting important population differences between US and global patterns.

Interpretation of results

This epidemiological pattern, characterized by the reduction of LDL-C and total cholesterol along with the increase of hypertriglyceridemia and the decrease of HDL-C, reflects a profound transformation in the profile of dyslipidemias at the population level. This transition represents the shift from the classic pattern of primary hypercholesterolemia toward a predominant hypertriglyceridemia and low HDL-C pattern, a phenotype intimately associated with insulin resistance, metabolic syndrome, and T2DM, as documented by Grundy et al. (2019) [3].

This epidemiological evolution results from the convergence of multiple factors. The global increase in obesity, especially visceral obesity, promotes insulin resistance that significantly alters lipoprotein metabolism, increasing triglyceride-rich VLDL synthesis while compromising peripheral clearance, as characterized by Taskinen and Borén (2015) [19]. Additionally, the westernization of dietary patterns, particularly increased consumption of refined carbohydrates and sugar-sweetened beverages, directly stimulates hepatic lipogenesis and promotes hypertriglyceridemia, as documented by Mente et al. (2017) [20] in the PURE study across 18 countries. Growing sedentarism further deteriorates the lipid profile by reducing lipoprotein lipase activity and decreasing HDL-C synthesis.

The widespread use of statins, while effectively reducing LDL-C and total cholesterol, has modest effects on triglycerides and minimal impact on HDL-C levels. This differential therapeutic impact may have amplified the observed discrepancy between lipid profile components in recent population studies, as noted by Ray et al. (2019) [21].

The clinical implications of this transition toward hypertriglyceridemia and low HDL-C are significant. This phenotype represents a highly pro-atherogenic lipid pattern characterized by small, dense LDL-C particles that more easily penetrate the vascular endothelium and triglyceride-rich remnant lipoproteins with independent atherogenic activity. This phenomenon partially explains the “residual risk” observed in patients achieving LDL-C objectives under statin treatment, as demonstrated by the REDUCE-IT study (Bhatt et al., 2019) [22].

From a public health perspective, this epidemiological transition demands reformulation of preventive strategies beyond LDL-C reduction. Interventions should target elevated triglycerides and dysfunctional HDL-C through strategies addressing insulin resistance, visceral obesity, and westernized dietary patterns. This requires combining population interventions—such as policies reducing sugar consumption and promoting physical activity—with personalized strategies based on individual metabolic profiles, as proposed in current international guidelines.

Implications for public and global health

The findings of this SR and meta-analysis have profound implications for global public health. The high prevalence of dyslipidemias identified represents a considerable cardiovascular disease burden worldwide, particularly in the Middle East and Latin America, where the highest prevalences were documented. Considering that approximately 80% of cardiovascular deaths occur in low- and middle-income countries, these results underscore the urgent need for preventive policies adapted to regional contexts and the epidemiological transition toward predominance of hypertriglyceridemia and low HDL-C.

This epidemiological transition demands rethinking traditional strategies focused exclusively on LDL-C reduction. Public health systems must evolve toward comprehensive approaches recognizing “residual risk” from elevated triglycerides and low HDL-C, especially relevant given the growing global epidemic of obesity, T2DM, and metabolic syndrome. Primary prevention programs should incorporate interventions targeting insulin sensitivity and visceral adiposity, while addressing regional and sex disparities through differentiated strategies that consider higher hypertriglyceridemia prevalence in men and low HDL-C in women. Public policies should target social determinants, including dietary patterns and physical activity, implementing regulatory measures such as taxes on sugar-sweetened beverages, front-of-package labeling, and advertising restrictions.

From economic and research perspectives, the high prevalence of dyslipidemias suggests considerable future budgetary impact from cardiovascular complications, emphasizing the cost-effectiveness of early prevention versus acute event management. Our findings highlight the need to strengthen global epidemiological surveillance with standardized measurement protocols and expanded research on genetic-environmental interactions determining dyslipidemia expression across populations. Only through integrated approaches combining solid scientific evidence with locally-adapted public policies can the global burden of dyslipidemias and their cardiovascular consequences be significantly reduced.

Strengths and limitations

The main strength of this SR and meta-analysis lies in its broad geographic and temporal coverage, including more than 200 population studies from different regions published over 25 years, providing a comprehensive global vision of dyslipidemia epidemiology. The exclusive inclusion of studies with probabilistic sampling, systematic risk of bias evaluation using validated tools, and stratified analyses by sex, geographic region, and temporal trends ensures methodological robustness while identifying differential patterns relevant to public health and clinical practice. The internal consistency of our results with high-quality individual studies and concordance with known pathophysiological mechanisms reinforce the validity of our conclusions about the epidemiological transition toward predominance of hypertriglyceridemia and low HDL-C worldwide.

Among the limitations, methodological heterogeneity stands out, particularly in cut-off points for hypercholesterolemia and elevated LDL-C, which could influence prevalence estimates and hinder direct comparability. Additionally, uneven geographic distribution underrepresents regions such as Africa and Oceania, while the predominance of cross-sectional designs prevents establishing causal relationships between contextual factors and lipid profile changes. Most importantly, the heterogeneous approach regarding participants receiving lipid-lowering medications represents a critical limitation, as some studies excluded individuals using statins or other lipid-lowering drugs. In contrast, others included treated patients without a clear specification. This variability could potentially influence prevalence estimates, given that pharmaceutical agents and nutraceuticals significantly affect lipoprotein profiles. Despite this methodological heterogeneity, prevalence estimates showed relatively consistent patterns within geographic regions and demographic groups, suggesting our overall trends remain robust. Nevertheless, this limitation underscores the urgent need for future epidemiological studies to adopt standardized approaches regarding participants on lipid-modifying therapies, which would enhance the precision of global prevalence estimates and facilitate more accurate international comparisons.

Conclusions and recommendations

In this SR with meta-analysis, encompassing more than 200 population studies from various regions, we could precisely quantify the global burden of dyslipidemias in adults. Our findings reveal significant combined prevalences: 28.8% for hypertriglyceridemia, 24.1% for hypercholesterolemia, 38.4% for low HDL-C, and 18.93% for high LDL-C, confirming that dyslipidemias constitute one of the most prevalent cardiovascular risk factors worldwide. The detailed analysis by sex evidenced an important dimorphism, with a higher prevalence of hypertriglyceridemia in men and a predominance of low HDL-C in women. This pattern suggests important sex-specific differences in dyslipidemia presentation. The geographic mapping identified the Middle East, Latin America, and certain regions of Asia as epidemiological “hot spots,” where particularly high burdens of low HDL-C and mixed dyslipidemias converge, configuring a scenario of high cardiovascular risk. Especially relevant is the analysis of temporal trends, which documents a transformation in the global panorama of dyslipidemias: while hypercholesterolemia and high LDL-C show a progressive decrease, probably influenced by the extended use of statins, a sustained increase in the prevalence of hypertriglyceridemia and low HDL-C is simultaneously observed, reflecting the growing influence of the epidemic of obesity, T2DM, and dietary westernization on population lipid profiles.

Given the findings of this SR, we recommend implementing comprehensive strategies for the prevention and management of dyslipidemias that address not only LDL-C cholesterol but also the emerging components of cardiovascular risk, such as hypertriglyceridemia and low HDL-C. At the clinical level, a more complete characterization of the lipid profile is essential, including, when possible, analysis of lipoprotein subfractions, to identify patients with hypertriglyceridemia and low HDL-C who could benefit from additional interventions beyond statins. Specific policies are required in the public health sphere, directed at regions with high prevalence, such as the Middle East and Latin America, prioritizing interventions on structural determinants such as dietary patterns, sedentarism, and visceral obesity. For future research, we recommend establishing standardized definitions of dyslipidemias to improve comparability between studies, expanding geographic coverage to underrepresented regions, and developing longitudinal studies that allow evaluation of the evolution of the atherogenic dyslipidemia phenotype and its impact on the incidence of cardiovascular events at the population level.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Jhosmer Ballena-Caicedo: Conceptualization, Investigation, Methodology, Resources, Writing - Original Draft, Writing - Review & Editing. Fiorella E. Zuzunaga-Montoya: Investigation, Project administration, Writing - Original Draft, Writing - Review & Editing. Joan A. Loayza-Castro: Methodology, Software, Data Curation, Writing - Review & Editing. Luisa Erika Milagros Vásquez-Romero: Investigation, Methodology, Resources, Writing - Original Draft, Writing - Review & Editing. Rafael Tapia-Limonchi: Investigation, Project administration, Writing - Original Draft, Writing - Review & Editing. Carmen Inés Gutierrez De Carrillo: Methodology, Software, Data Curation, Writing - Review & Editing. Víctor Juan Vera-Ponce: Formal analysis, Validation, Visualization, Funding acquisition, Supervision, Writing - Original Draft, Writing - Review & Editing.

Funding

This study was financed by the Vicerectorado de Investigación de la Universidad Nacional Toribio Rodríguez de Mendoza de Amazonas.

Data availability

Data are available upon request to the corresponding author.

Declarations

Ethics approval and consent to participate

Since this manuscript is a secondary data-based study, it was not required.

Informed consent

Since this is a secondary data analysis, informed consent was not required.

Clinical trial number

Not applicable.

Consent for publication

Not applicable.

Conflict of interest

The authors declare no conflict of interest.