Elevated Lipoprotein(a) in Asian Populations and its Implications for Cardiovascular Disease

Abstract

Elevated lipoprotein (a) (Lp[a]) is becoming increasingly recognized as a cardiovascular disease risk factor. The relationship of Lp(a) to cardiovascular disease is well characterized in European populations; however, data are lacking in the Asian region. In this scoping review, we aimed to characterize the distribution and prevalence of elevated Lp(a) in the Asian region and its relationship to coronary artery disease, cerebrovascular disease, and peripheral vascular disease. Thirty-one Asian studies met the inclusion criteria and were compared with 6 representative European studies. The greatest representation was in Chinese populations, with much less representation of Japanese, Thai, Indian, and other groups. Levels of Lp(a) in healthy Asian populations were generally lower than in Europid (Caucasian) populations; however, these levels remain lower than those reported in European literature. Within Asian populations, elevated Lp(a) was universally associated with increased risk of coronary artery disease and peripheral vascular disease. The relationship with stroke was less clear.

Central Illustration

Highlights

• •Asian cardiovascular disease burden is increasingly recognized, but Asian Lp(a) data remain limited.
• •Elevated Lp(a) consistently correlates with coronary and PVD; routine assessment should be considered.
• •Ethnic variation in Lp(a) levels between Asian populations highlights the need for population-specific values.
• •Future research should establish Asian-specific risk thresholds, standardize measurement, and expand representation in outcome trials.

Cardiovascular disease (CVD) is becoming an increasingly important cause of morbidity and mortality in the Asia-Pacific region with a forecast doubling of disease burden in the region by 2050.¹ Although statins are effective in reducing cardiovascular (CV) risk by reducing low-density lipoprotein cholesterol, events still occur in treated patients. Elevations in other lipid parameters such as triglyceride-rich lipoproteins and lipoprotein (a) [Lp(a)] account for some of the residual risk.¹ Lp(a) is becoming increasingly recognized as an important risk factor for CVD,² having both proatherogenic and prothrombotic properties.2, 3, 4 Specifically, Mendelian randomization studies and genome-wide association studies have demonstrated a causal effect of Lp(a) on CVD independent of low-density lipoprotein cholesterol.⁵

CV risk is proportional to Lp(a) concentration.² High Lp(a) is strongly correlated with coronary artery disease (CAD), peripheral vascular disease (PVD), and calcific aortic stenosis among Europid populations. There is growing evidence affirming the relationship of Lp(a) with cerebrovascular disease including strokes.⁶ Accordingly, recent lipid guidelines emphasize a need to measure Lp(a) concentration for CV risk stratification.⁷ Screening for index cases among high-risk patients may help improve long-term outcomes by facilitating cascade testing of their relatives.⁸

The Copenhagen City Heart Study was seminal in characterizing the prevalence and CVD outcomes of a relatively homogeneous Danish population.⁹ These results were translatable to other Europids and have prompted extensive study of European groups. The distribution and prevalence of elevated Lp(a) in multiethnic and Asian populations is unclear. It has been suggested that South Asian individuals might have higher Lp(a) than other Asian ethnicities; however, there is little to confirm this assertion as adequately powered studies have predominantly been undertaken only in Chinese, Japanese, and Korean populations. Little is understood about the CV implications of raised Lp(a) in Asian populations. Thus far, there has been limited synthesis of the data available in this field.

In this scoping review, we synthesize the available data on the distribution of Lp(a) levels in the population and the relationship of Lp(a) with CVD across a range of representative studies in Asian populations of interest. We attempt to compare these findings with data from landmark and highly cited studies of Europid populations.

Methodology

This scoping review included studies describing a population in Asia and conducted in Asia, wherein Lp(a) levels are described either in the general population or within a population with CVD (ie, CAD, PVD, or cerebrovascular accident). Where multiple studies were available from a given population, the most recent studies (within 5 years) were prioritized. Research not relevant to countries in Asia, those with primarily Europid population; without reports of Lp(a); studies on pediatric populations; studies with abstracts or full texts that were missing or unavailable; those with small sample sizes (contingent on available studies for a given country); those on special populations such as chronic renal disease, familial hypercholesterolaemia and diabetes; and articles that were not peer-reviewed, were all excluded.

The articles were searched from MEDLINE, EMBASE, EMCARE, and PubMed. From the research question and the study's aims, the following key words were identified: “Asia,” “lipoprotein (a),” and “cardiovascular disease.” The respective MeSH terms were combined using the “AND” Boolean operator. Where a satisfactory MeSH term could not be found for a key word, a fitting term was constructed in its place by combining composite MeSH terms with the “OR” Boolean operator. Key words related to each of the Asian populations of interest (ie, Japan, China, Indonesia, Malaysia, Korea, Thailand, India, Pakistan, the Philippines, Singapore, Sri Lanka, and Taiwan) also were used.

The articles obtained were managed using Endnote X9 software. Search results were sorted by country and duplicates were removed. The abstracts of the remaining papers were assessed independently by 2 reviewers (M.D. and R.O.) according to inclusion and exclusion criteria. Differences of opinion were resolved by discussion between the 2 reviewers.

Papers isolated by abstract screening were obtained in full text (if available). Their eligibility for this review was then evaluated. Because this study aimed to describe Lp(a) in multiple populations of the Asian region, it was necessary to prioritize studies examining underrepresented populations (ie, those with less research on the subject matter). Where a population was overrepresented within the literature, less recent publications and those with comparatively small sample sizes were excluded to ensure more equitable representation of less studied populations with less literature available at the discretion of an expert panel. For example, Chinese and Japanese cohorts have been very heavily examined as part of large cohort studies. In these populations, studies examining cohorts of fewer than 500 participants were generally excluded unless their findings were of particular significance in our synthesis of the available literature. By contrast, far fewer large studies were available examining South Asian populations, including Indian samples. Therefore, South Asian studies with samples of fewer than 500 participants were included to better appreciate the Lp(a) distribution and associated outcomes.

Results

Lp(a) distribution in apparently healthy Asia-Pacific populations

Multiple relevant studies were found that characterized Lp(a) within the Asia-Pacific populations of interest (Table 1).10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25 Lp(a) had a positively skewed distribution, with most subjects having low or moderately increased serum Lp(a) (lower or middle tertiles for Lp[a], where percentiles were defined within a study sample), and a small fraction having high (upper tertile where percentiles were defined within a study sample) or markedly elevated levels (Lp[a] ≥ 50 mg/dL; cutoff recommended by the European Atherosclerosis Society).²⁶ There was noticeable ethnic and geographic variation in serum Lp(a) levels among healthy subjects in the studies included. However, regional differences were not as great nor as clearly defined as suggested by the European reference literature. Also, there was both substantial heterogeneity in the structure of studies and discrepancy in the availability of studies examining each Asian population, making interpretation of results challenging. There was greatest representation of Chinese populations, with much less representation of Japanese, Thai, Indian, and other groups.

Table 1.

Baseline Lp(a) in General Populations

Region	First Author, Year of Publication	Relevant Population Extracted (Total Number of Healthy Subjects)	Mean Lp(a) Values (Where Specified)	Median Lp(a) Values (Where Specified)	Lp(a) Quantiles (Where Specified)
China	Cai et al, 201510	6,125 Chinese Han	-	7.8 mg/dL	50 mg/dL ∼ 94th percentile
	Cui et al, 201811	9,238 Chinese Han	-	5.6 mg/dL	16.7 mg/dL ∼ 80th percentile 30.0 mg/dL ∼ 90th percentile
	Fu et al, 202012	1953 healthy Chinese Han age- and sex-matched control subjects for ischemic cerebrovascular accident (CVA) patients	-	8.60 mg/dL	3.70 mg/dL ∼ 25th percentile 21.30 mg/dL ∼ 75th percentile
		392 healthy Chinese Han age- and sex-matched control subjects for hemorrhagic CVA patients	-	13.40 mg/dL	11.30 mg/dL ∼ 25th percentile 16.10 mg/dL ∼ 75th percentile
	Zhang et al, 201913	8,500 Chinese aged ≥40 y	21.0 mg/dL	-	Tertile 1: Mean Lp(a) = 6.60 ± 2.68 mg/dL Tertile 2: Mean Lp(a) = 18.98 ± 4.66 mg/dL Tertile 3: Mean Lp(a) = 38.22 ± 14.67 mg/dL
	Cheng et al, 199714	200 Chinese (Hong Kong)	27.2 mg/dL	18.2 mg/dL	-
	Wong et al, 199915	352 healthy ethnic Chinese living in Singapore	8.63 mg/dL	9.50 mg/dL	1.30 mg/dL ∼ 5th percentile 4.58 mg/dL ∼ 25th percentile 9.50 mg/dL ∼ 50th percentile 18.86 mg/dL ∼ 75th percentile 42.37 mg/dL ∼ 95th percentile
	Xia et al, 202216	25,343 Chinese Han general population	-	11.9 mg/dL	5.9 mg/dL ∼ 25th percentile 11.9 mg/dL ∼ 50th percentile 23.7 mg/dL ∼ 75th percentile 43.7 mg/dL ∼ 90th percentile 60.8 mg/dL ∼ 95th percentile
	Man et al, 202557	2,955,320 Chinese general population −2,788,206 in mass unit group (mg/dL) −167,114 in molar unit group (nmol/L)	-	Mass unit group: 11.23 mg/dL Molar unit group: 20.00 nmol/L	Mass unit group: 5.50 mg/dL ∼ 25th percentile 22.64 mg/dL ∼ 75th percentile 30 mg/dL ∼ 80th percentile 50 mg/dL ∼ 90th percentile 70 mg/dL ∼ 95th percentile Molar unit group: 9.30 nmol/L ∼ 25th percentile 52.10 nmol/L ∼ 75th percentile
	SUMMARY	26,760 Chinese individuals predominantly living in China, with a fraction in Singapore	Range of mean Lp(a): 8.63-27.2 mg/dL	Range of median Lp(a): 7.8-18.2 mg/dL	16-21 mg/dL ∼ 75th percentile
Japan	Ishikawa et al, 201317	10,494 Japanese individuals	13.8 mg/dL	-	10 mg/dL ∼ 33rd percentile 22 mg/dL ∼ 67th percentile
India	Gambhir et al, 200818	160 Indian individuals age <40 years	-	12.7 mg/dL	30 mg/dL ∼ 78th percentile
	Hoogeveen et al, 200119	46 Asian Indian individuals living in India	9.15 mg/dL	-	-
		206 Asian Indian individuals living in the United States	9.27 mg/dL N Indian	-	-
			9.48 mg/dL S Indian
			5.57 mg/dL E Indiana
			9.45 mg/dL W Indian
			8.67 mg/dL Overall
	Yusuf et al, 201420	150 North Indian individuals	-	20 mg/dL	12 mg/dL ∼ 25th percentile 20 mg/dL ∼ 50th percentile 40 mg/dL ∼ 75th percentile
	Wong et al, 199915	259 healthy ethnic Indian individuals living in Singapore	10.08 mg/dL	12.28 mg/dL	1.10 mg/dL ∼ 5th percentile 5.17 mg/dL ∼ 25th percentile 12.28 mg/dL ∼ 50th percentile 22.43 mg/dL ∼ 75th percentile 43.87 mg/dL ∼ 95th percentile
	SUMMARY	821 healthy Indian individuals living in India, the United States, and Singapore	Range of mean Lp(a): 8.67-10.08 mg/dL	Range of median Lp(a): 12.7-20 mg/dL	22-40 mg/dL ∼ 75th percentileb,c
Sri Lanka	Atukorala et al, 200233	90 healthy volunteers from Sri Lanka	10.9 mg/dL	11.5 mg/dL	-
Singapore	Loh et al, 202671	1,927 Singaporean health care workers	-	16.4 nmol/L (approximately 6.6-6.8 mg/dL)	7.9 nmol/L (approximately 3.2-3.7 mg/dL) ∼ 25th percentile 41.8 nmol/L (approximately 16.7-19.4 mg/dL) ∼ 75th percentile
Korea	Cho et al, 201621	2,611 Korean individuals	29.7 mg/dL	-	50 mg/dL ∼ 84.4th percentile
	Sung et al, 201322	14,583 South Korean individuals	-	21.60 mg/dL	11.30 mg/dL ∼25th percentile 21.60 mg/dL ∼ 50th percentile 38.64 mg/dL ∼ 75th percentile 50 mg/dL ∼ 84th percentile
	SUMMARY	17,191 Korean individuals	29.7 mg/dL	21.60 mg/dL	50 mg/dL ∼ 84th percentile
Taiwan	Chien et al, 200823	3,484 healthy ethnic Chinese individuals living in Taiwan	-	-	34.33 mg/dL ∼ 90th percentile 47.08 mg/dL ∼ 95th percentile 69.30 mg/dL ∼ 99th percentile
Thailand	Leowattana et al, 200024	170 healthy Thai subjects	27.6 mg/dL	-	30 mg/dL ∼ 71.2nd percentile
Malaysia	Hakim et al, 200825	50 healthy Malaysian subjects from the Kuala Lumpur local population	29 mg/dL	17 mg/dL	36 mg/dL ∼ 74th percentile
	Wong et al, 199915	192 healthy ethnic Malaysian individuals living in Singapore	8.46 mg/dL	8.85 mg/dL	5th percentile Lp(a) concentration = 1.62 mg/dL 25th percentile Lp(a) concentration = 4.18 mg/dL 50th percentile Lp(a) concentration = 8.85 mg/dL 75th percentile Lp(a) concentration = 17.92 mg/dL
	SUMMARY	242 healthy ethnic Malaysian individuals in Malaysia and Singapore	Range of Mean Lp(a): 8.46-29 mg/dL	Range of Median Lp(a): 8.85-17 mg/dL	18-36 mg/dL ∼ 75th percentile

CVA = cerebrovascular accident; Lp(a), lipoprotein (a).

^aSignificantly lower than other Indian regions.

^bVaries depending on region in India.

^cNorth Indians may have higher Lp(a).

In studies examining general population samples and in healthy control groups from cohort and case-control studies, Chinese and Japanese subjects had comparable Lp(a) distributions (Table 1). Chinese and Japanese populations had median Lp(a) of less than 15 mg/dL in most studies, with some studies suggesting that Lp(a) may be slightly lower among Chinese patients compared with Japanese patients. Two large Chinese studies estimated the 75th percentile for serum Lp(a) to range between 16 and 22 mg/dL,^11,12 whereas 22 mg/dL approximated the 67th percentile in one large Japanese cohort study.¹⁷ There were few other relevant Japanese papers available to corroborate these findings.

Only a few relevant studies reported Lp(a) in Thai and Malay individuals and these studies had few subjects. The 30-mg/dL level approximated the 71st percentile in one Thai study with 170 participants,²⁴ and in a Malay study with 50 participants, 36 mg/dL represented the 74th percentile of the sample group.²⁵ Healthy Korean individuals had similar serum Lp(a) levels to those quoted for Thai and Malay individuals. In one large study examining 14,583 South Korean individuals, the median Lp(a) was 21.60 mg/dL, and 38.64 mg/dL represented the 75th percentile.²² However, another large Korean study suggested that Lp(a) levels were slightly lower. In the Kangbuk Samsung Health Study cohort comprising 275,430 participants, 18.5 mg/dL and 33.9 mg/dL represented the median and 75th percentile, respectively, for serum Lp(a).²⁷

Indian Lp(a) levels were quoted to be substantially higher by some South Asian studies compared with other ethnic groups. In a prospective case-control study of North Indians, the median Lp(a) of the healthy control group (n = 150) was 20 mg/dL (IQR: 12-40 mg/dL). The 75th percentile corresponded to 40 mg/dL.²⁰ There was, however, considerable variation in measurements provided by the Indian papers included in this review. In a second Indian study, the 78th percentile of a healthy control group was equivalent to 30 mg/dL.¹⁸ The distribution of Lp(a) of an Indian control group was especially low in a third study, with a mean of 9.15 ± 7.33 mg/dL.¹⁹ These studies had relatively small sample sizes, meaning that these observations may be the result of sampling error. Hence, the precise distribution of Lp(a) in healthy Indian individuals was not entirely clear based on the papers included in this review. This group may potentially have higher Lp(a) than other healthy Asian groups; however, larger Indian studies are needed to better elucidate the normal level of Lp(a) in this population.

A series of single-country multiethnic studies were referenced to offer additional insight into the Lp(a) distribution of ethnic Indian individuals (Table 2).^15,28, 29, 30, 31, 32, 33 One study from Sri Lanka found that among 90 healthy volunteers, although only 4.4% had Lp(a) ≥50 mg/dL, 20% had Lp(a) ≥30 mg/dL.³³ In addition, a Singaporean and a Canadian study each compared Lp(a) distributions of healthy ethnic Indian and South Asian individuals with healthy ethnic Chinese individuals living in their respective countries. These studies suggested that Indian and South Asian ethnic subgroups had significantly higher Lp(a) values compared with their ethnic Chinese counterparts.^15,28 In the Singaporean study of note, ethnic Indian individuals were measured to have a median Lp(a) of 12.28 mg/dL, which was significantly higher than that of ethnic Chinese individuals who had a median Lp(a) of 9.50 mg/dL (P < 0.05).¹⁵ The Canadian study expressed Lp(a) levels as natural logarithms of measurements taken in g/L and described mean Lp(a) levels of ethnic South Asian and ethnic Chinese individuals to be 0.0293 and 0.0259, respectively. Subsequent pairwise comparison of South Asian and Chinese distributions were undertaken that showed statistically distinct differences in the 2 distributions, with a higher Lp(a) being recorded among ethnic South Asian individuals (P = 0.0006).²⁸ This difference in Lp(a) levels in individuals of South Asian ethnicity compared with other ethnicities was not as pronounced in the other studies.

Table 2.

Lp(a) in People of Indian Ethnicity Residing in Different Countries

First Author, Year of Publication	Type of Study	Number of Participants	Population	Outcome of Interest	Findings
Anand et al, 200028	Cross-sectional study	985 in total from 3 Canadian cities Participants were between 35 and 75 years old and had lived in Canada for at least 5 years	Different ethnic groups in Canada: •South Asian (ancestors originated from India, Pakistan, Sri Lanka, or Bangladesh) (n = 342) •Chinese (ancestors originated from China, Taiwan, or Hong Kong) (n = 317) •European (n = 326)	Clinical CVD (defined by history or ECG findings) and carotid atherosclerosis (measured with B-mode ultrasonography)	Lp(a) in different ethnic groups living in Canada: South Asian: Mean (SD) Lp(a) = 0.0293 (0.018) g/L = 2.93 (1.8) mg/dL Chinese: Mean (SD) Lp(a) = 0.0259 (0.0119) g/L = 2.59 (1.19) mg/dL European: Mean (SD) Lp(a) = 0.0259 (0.0119) g/L = 2.59 (1.19) mg/dL Pairwise comparison of South Asian and European Lp(a) distributions: P = 0.0005 Pairwise comparison of Chinese and European Lp(a) distributions: P = 0.99 Pairwise comparison of South Asian and Chinese distributions: P = 0.0006 Risk of CVD by ethnic group compared to European multivariate logistic regression OR (95% CI) of CVD if South Asian vs European = 4.51 (1.46-13.89), P = 0.02 OR (95% CI) of CVD if Chinese vs European = 1.05 (0.30-3.63), P = 0.91
Hughes et al, 199729	Cross-sectional study of general population in Singapore	Random sample of 961 men and women (Indian, Malay, and Chinese) aged 30 to 69 years	Singaporean general population •Indian n = 342 (170 men, 172 women) •Malay n = 289 (147 men, 142 women) •Chinese n = 330 (161 men, 169 women)	Association between metabolic syndrome (including risk factors) and ethnicity and sex	Lp(a) in men of each ethnicity (mean; 95% CI) •Indian: 18.1 (15.6-20.6) mg/dL •Malay: 14.8 (12.1-17.5) mg/dL •Chinese: 12.5 (9.8-15.2) mg/dL Pairwise comparison in Singaporean men: •Indian vs Malay: P = 0.09 •Indian vs Chinese: P < 0.01 •Malay vs Chinese: P = 0.25 Lp(a) in women of each ethnicity (mean; 95% CI) •Indian: 22.7 (20.2-25.2) mg/dL •Malay: 17.4 (14.7-20.1) mg/dL •Chinese: 18.5 (16.0-21.0) mg/dL Pairwise comparison in Singaporean men: •Indian vs Malay: P < 0.01 •Indian vs Chinese: P = 0.03 •Malay vs Chinese: P = 0.60
Wong et al, 199915	Cross-sectional study	803 healthy individuals recruited from the Multiphasic Health Screening Programme •Indian (n = 259) •Chinese (n = 352) •Malay (n = 192)	Singaporean general population	Serum Lp(a) concentration	Overall population: Mean Lp(a) concentration (SD) = 9.04 (2.85) mg/dL 5th percentile Lp(a) concentration = 1.33 mg/dL 25th percentile Lp(a) concentration = 4.58 mg/dL 50th percentile Lp(a) concentration = 10.17 mg/dL 75th percentile Lp(a) concentration = 19.81 mg/dL 95th percentile Lp(a) concentration = 43.07 mg/dL The Indian subgroup had significantly higher median Lp(a) concentration than the other 2 subgroups (Chinese and Malay), P < 0.05 Indian subgroup: Mean Lp(a) concentration (SD) = 10.08 (2.89) mg/dL 5th percentile Lp(a) concentration = 1.10 mg/dL 25th percentile Lp(a) concentration = 5.17 mg/dL 50th percentile Lp(a) concentration = 12.28 mg/dL 75th percentile Lp(a) concentration = 22.43 mg/dL 95th percentile Lp(a) concentration = 43.87 mg/dL Chinese subgroup: Mean Lp(a) concentration (SD) = 8.63 (2.89) mg/dL 5th percentile Lp(a) concentration = 1.30 mg/dL 25th percentile Lp(a) concentration = 4.58 mg/dL 50th percentile Lp(a) concentration = 9.50 mg/dL 75th percentile Lp(a) concentration = 18.86 mg/dL 95th percentile Lp(a) concentration = 42.37 mg/dL Malay subgroup: Mean Lp(a) concentration (SD) = 8.46 (2.69) mg/dL 5th percentile Lp(a) concentration = 1.62 mg/dL 25th percentile Lp(a) concentration = 4.18 mg/dL 50th percentile Lp(a) concentration = 8.85 mg/dL 75th percentile Lp(a) concentration = 17.92 mg/dL 95th percentile Lp(a) concentration = 40.15 mg/dL The study also quoted values in Caucasian populations, as follows: 50th percentile Lp(a) concentration = 15.0 mg/dL 75th percentile Lp(a) concentration = 29.0 mg/dL 95th percentile Lp(a) concentration = 60.0 mg/dL
Shaukat et al, 199530	Prospective cohort study	375 in total •102 male individuals of Indian origin and 89 of their sons between 15 and 30 years old •102 age-matched male individuals European patients and of their 82 sons	Male individuals of Indian and North European origin living in the United Kingdom with suspected CAD, and their sons (healthy population)	Father-son correlations for risk markers predicting the severity of parental CAD (including Lp[a]); differences in mean levels of these markers between young male individuals of Indian origin and those of North European origin	Sons of parents of Indian origin had significantly higher Lp(a) than sons of parents of North European origin: 19.1 vs 10.5 mmol/L; P = 0.001 In patients with suspected CAD (mean Lp[a] concentration; 95% CI) Indian origin: 27.2 (25.8-28.5) mg/dL Northern European origin: 19.5 (18.3-20.4) mg/dL In sons of patients with suspected CAD (healthy population) (mean Lp[a] concentration; 95% CI) Sons of Indian origin: 19.1 (16.8-21.8) mg/dL Sons of Northern European origin: 10.5 (8.3-12.8) mg/dL
Tavridou et al, 200331	Cross-sectional study	694 total 373 European 103 Indian 154 Pakistani 64 Bangladeshi	South Asian individuals aged 25-74 (Indian, Pakistani, Bangladeshi) and European population from the Newcastle Heart Project	Lp(a) and various other biochemical markers of CVD	There was no significant difference in Lp(a) levels in South Asian men together compared with European men (P = 0.37). South Asian women combined had higher Lp(a) levels than European women (194 vs 154 mg/L; P = 0.01). Lp(a) levels were not statistically different for men in the 3 South Asian subgroups (P = 0.19). In women, Lp(a) concentrations were significantly higher in Pakistani than Indian individuals (232 vs 162 mg/L; P = 0.01), with Bangladeshi individuals having intermediate levels. Lp(a) concentrations were significantly higher in Pakistani women than Pakistani men (232 vs 148 mg/L; P < 0.0001). Pakistani women had the highest percentage of subjects (40%) with Lp(a) concentrations >300 mg/L.
Loh et al, 202232	Cross-sectional study	2025	Consecutive consenting patients undergoing diagnostic coronary angiography at the National University Heart Centre, Singapore, enrolled in SCADGENS (the Singapore Coronary Artery Disease Genetics Study)	Lp(a) levels	Lp(a) levels were significantly different among each ethnicity in the following order Indian > Malay > Chinese patients Indian vs Malay, P < 0.001 Malay vs Chinese, P = 0.001 Indian vs Chinese, P < 0.001). When stratified by the different cardiovascular phenotypes, there was a consistent trend of Lp(a) levels being highest among the Indians, followed by Malay and then Chinese patients.

CAD = coronary artery disease; CVD = cardiovascular disease; ECG = electrocardiogram; other abbreviation as in Table 1.

Several studies identified sex differences in Lp(a) distribution among Asian samples. One prospective cohort study in China demonstrated that women had greater median Lp(a) across all age quantiles, with many of these differences being statistically significant.¹¹ On average, median female Lp(a) level was 1.98 mg/dL higher than the median male Lp(a) level in each quantile, with the overall male and female median Lp(a) concentrations being 5.1 mg/dL (IQR: 2.5-11.6 mg/dL) and 6.7 mg/dL (IQR: 3.2-16.7 mg/dL), respectively. Postmenopausal women had higher Lp(a) than their premenopausal counterparts. These suggest that sex hormones may have Lp(a)-modulating effects. These findings are supported by other Asian studies and are consistent with those from European populations.34, 35, 36

Asian Lp(a) distribution compared with published data in Europid populations

Several studies have confirmed the role of Lp(a) in CVD in Europid populations, alongside other highly cited Europid studies.^5,9,37, 38, 39, 40 Of these, The Copenhagen City Heart Study is well known for having characterized CVD outcomes and their association with serum Lp(a) in a relatively homogeneous Danish population. It has served as a benchmark, prompting further research describing serum Lp(a) distributions in other populations. The data demonstrate that Europid populations maintain a positively skewed Lp(a) distribution with 50 mg/dL corresponding to the 80th percentile.^37,39 Accordingly, we recognize that 50 mg/dL has been recommended as the cutoff for normal serum Lp(a) in European groups by the European Atherosclerosis Society.²⁶ There remains some variation in Lp(a) distributions between different European ethnic groups; however, the results appear less heterogeneous than what has been documented among Asian individuals.

Serum Lp(a) is lowest among non-Hispanic Caucasians, whose median Lp(a) approximates 12 mg/dL (IQR: 5-32 mg/dL).^41,42 Based on the Asian studies included in this review, healthy Chinese and Japanese samples appear to have comparable Lp(a) distributions to non-Hispanic Caucasians. Several studies even suggest that healthy Chinese individuals had lower serum Lp(a) than Caucasians. One large Chinese study¹¹ described 16.7 mg/dL and 30.0 mg/dL to be the 80th and 90th percentiles, respectively, whereas another described 50 mg/dL as the 94th percentile for Lp(a).¹⁰ The former study suggested that 16.7 mg/dL would serve as a better cutoff for normal Lp(a) among Chinese individuals. Large cohort studies suggest that South Korean individuals may also have a similar serum Lp(a) distribution to Europid individuals. Two of these studies have measured 50 mg/dL to approximate the 84th percentiles for Lp(a).^21,22 In aggregate, these Asian groups may have similar or potentially lower Lp(a) than their Europid counterparts (Central Illustration).

Central Illustration.

Lipoprotein (a) Measurements in Asian Populations and Cardiovascular Implications

Elevated Lp(a) is associated with increased risk of CVD, PVD, and ischemic stroke in Asian populations: Lp(a) testing is recommended. There is significant ethnic variation in Lp(a) measurements between Asian populations, which challenges widely accepted European cutoffs for elevated Lp(a) and illustrates the need for more large-scale studies to establish population-specific Lp(a) cutoffs to assist clinicians with management. CAD = coronary artery disease; CVD = cardiovascular disease; Lp(a) = Lipoprotein (a); PVD = peripheral vascular disease.

Last, the large INTERASPIRE (International Action on Secondary Prevention through Intervention to Reduce Events), which enrolled patients from 13 countries worldwide, found that among enrolled adults hospitalized with coronary heart disease (CHD) 6 to 24 months before enrollment, median Lp(a) levels in Europe ranged from 19.5 nmol/L (Poland) to 59 nmol/L (Portugal). On the other hand, the median Lp(a) in the Western Pacific ranged from 14 nmol/L (Philippines) to 39.5 nmol/L (Malaysia).⁴³ It should be noted that the applicability of these trends to healthy individuals is limited by the history of hospitalization for CHD in this cohort.

It has long been suggested by European literature that Asian Indian individuals have higher serum Lp(a) than Caucasian and other Asian groups. As discussed earlier, there is some evidence suggesting that this could be true.²⁰ However, there are also inconsistencies in the findings of different Indian studies and an overall lack of robust data from which to draw sound conclusions.

The relationship of Lp(a) to CVD in the Asian region

Within the Asian samples included, Lp(a) was associated with preclinical CVD, progression of both preclinical and clinical disease, major adverse cardiovascular events, and CV mortality. This relationship was examined in various ways in the studies included by defining or using predefined cutoff values for high Lp(a); assessing CVD outcomes in different Lp(a) quantiles (tertiles, quartiles, or quintiles); and by comparing the Lp(a) distributions of diseased and healthy groups.

Where predefined Lp(a) cutoffs were used, values chosen varied considerably between studies, most often either 50 mg/dL or 30 mg/dL. The 50-mg/dL cutoff recommended by the European Atherosclerosis Society in Europids²⁶ offered a comparison between some diseased and healthy samples with European reference literature in a range of studies.^{10,21,22,44,45} A 30-mg/dL cutoff was chosen in one study³³ with specific reference to Europid literature; 30 mg/dL represented the 67th percentile for Lp(a) in The Copenhagen City Heart Study.^38,46 Otherwise, 30 mg/dL was chosen predominantly because of factors pertaining to Lp(a) assays, rather than insights relating to the Lp(a) distributions of different ethnic groups.^12,47

A cutoff of 40 mg/dL was used by one Japanese study of patients with acute myocardial infarction (AMI) undergoing primary percutaneous coronary intervention.⁴⁸ The authors based this decision on 40 mg/dL representing the 75th percentile of their study population and 40 mg/dL being the normal upper limit used in their assay system; however, other Japanese studies were not consistent with their choice of Lp(a) cutoff. Studies examining CV outcomes in samples stratified by Lp(a) quantile were of greatest utility in this review. The design of these studies facilitated comparison of Lp(a) distributions between different groups, and the trend in CV risk with increasing Lp(a). They also served as a basis for rationalizing potential Asian-specific Lp(a) cutoff values.

Lp(a) and CAD

There was the greatest amount of data affirming the strong positive correlation between Lp(a) and CAD found in Europid studies; however, multiple studies also suggest a positive correlation among Asian individuals (Table 3).^10,16,18,20, 21, 22, 23, 24^,27,32,45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57 In a cross-sectional study of Indian individuals younger than 40 years with and without angiographic evidence of CAD, median plasma Lp(a) levels were 2.5 times higher in patients with established CAD compared with control subjects (30.0 vs 12.7 mg/dL; P < 0.05).¹⁸ Lp(a) was a determinant of preclinical disease and its progression in studies examining coronary artery calcification (CAC) in healthy patients. In a large Chinese study, individuals with Lp(a) >50 mg/dL had significantly higher odds of CAC (19% higher odds) compared with those with Lp(a) ≤30 mg/dL. This association was stronger for patients with atherosclerotic disease affecting multiple vascular beds such as coronary and cerebrovascular.⁵⁷

Table 3.

Asian Lp(a) CAD Table: Statistical Findings

First Author, Year of Publication	Type of Study	Number of Participants	Population	Outcome of Interest	Elevated Lp(a) Cutoff Value Used (mg/dL)	Findings
Xia et al, 202216	Prospective cohort study	25,343	Chinese Han general population	Diagnoses of CAD, stroke, T2DM and HTN within the study period	Not specified Percentiles defined.	Lp(a) 25th percentile = 5.9 mg/dL Population median Lp(a) = 11.9 mg/dL (IQR: 5.9-23.7 mg/dL) Lp(a) 75th percentile = 23.7 mg/dL Lp(a) 90th percentile = 43.7 mg/dL Lp(a) 95th percentile = 60.8 mg/dL Median age 52 years, 54.4% female Association between CAD (n,%) and Lp(a) percentile 0–25th percentile (345, 5.5%): OR: 1.02 (95% CI: 0.86-1.20; P = 0.844) 25-50th percentile (318, 5.0%): OR: 1.00 50-75th percentile (366, 5.3%): OR: 1.11 (95% CI: 0.94-1.31; P = 0.223) 75-90th percentile (202, 5.3%): OR: 1.10 (95% CI: 0.91-1.34; P = 0.329) 90-95th percentile (76, 6.0%): OR 1.42 (95% CI: 1.08-1.88; P = 0.012) 95-100th percentile (87, 6.9%): OR 1.60 (95% CI: 1.23-2.09; P = 0.001) Overall P value for trend: 0.094
Shi et al, 202149	Prospective cohort study	2,433 patients (1,751 with PCAD, 682 with non-CAD according to coronary angiography)	Chinese patients who underwent coronary angiogram following complaints of angina-like chest pain at Fuwai Hospital between March 2016 and February 2019	PCAD and severity of such CAD; Gensini score; Lp(a) level	Not specified	Mean age = 52.33 ± 7.43 Median Lp(a) = 16.20 mg/dL (IQR: 8.35-31.69) Non-CAD (n = 682): Median Lp(a) = 12.68 mg/dL (IQR: 5.94-26.10 mg/dL) PCAD (n = 1751): Median Lp(a) = 17.69 mg/dL (IQR: 9.56-33.68 mg/dL) P value <0.001
Wei et al, 202250	Retrospective cross-sectional study	2,846	Chinese patients with CHD	Primary: Premature CHD Secondary: MVD	Not specified	Premature CHD defined as onset <45 years old in men and <55 years in women. MVD defined as at least to coronary arteries with ≥50% stenosis Premature CHD (n = 2,161): Median Lp(a) = 22.0 mg/dL (IQR: 8.3-61.6 mg/dL) Mature CHD (n = 685): Median Lp(a) = 26.0 mg/dL (IQR: 8.9-76.9 mg/dL) P = 0.061
Cai et al, 201510	Cross-sectional study	3462 CAD cases; 6,125 non-CAD control subjects	Chinese Han	CAD (subdivided into initial IHD, initial MI, and previous CAD)	50	Non-CAD control subjects: Distribution of Lp(a) conc. was positively skewed Prevalence Lp(a) > 50 mg/dL = 6% Median Lp(a) conc. = 7.8 mg/dL (IQR: 14.32) CAD cases: Median Lp(a) = 9.9 mg/dL (significantly greater than non-CAD control subjects [P < 0.0001]) Prevalence Lp(a) > 50 mg/dL in non-CAD Chinese Han population = 6% Prevalence Lp(a) > 50 mg/dL in Caucasian general population = 20% Based on ROC analysis, proposed Lp(a) cutoff for discriminating CAD from non-CAD in Chinese Han population was 78.9 mg/dL (sensitivity = 58.3%; specificity = 52.3%)
Liu et al, 202051	Prospective cohort study	7,562	Chinese patients with angiography-diagnosed CAD who had experienced a first CVE	RCVE including CV death, nonfatal MI, and stroke	Cutoff not used, instead tertiles defined Tertile 1: <8.88 Tertile 2: 8.88-26.44 Tertile 3: ≥26.45	Mean follow-up of 61.45 ± 19.57 months Baseline characteristics Distribution of Lp(a) was positively skewed. Overall: Median Lp(a) conc. = 15.17 (IQR: 6.79-36.95) Tertile 1: Median Lp(a) conc. = 4.96 (IQR: 3.27-6.79) Tertile 2: Median Lp(a) conc. = 15.17 (IQR: 11.79-20.13) Tertile 3: Median Lp(a) conc. = 52.01 (IQR: 36.95-77.91) P < 0.001 RCVEs 680 (9.0%) RCVEs occurred during follow-up Patients with RCVEs had significantly higher median Lp(a) concentrations than those without (20.58 vs 14.95 mg/dL, P < 0.001) HR for patients in Tertile 3 for Lp(a) relative to Tertile 1 fora: Total RCVE = 2.01 (95% CI: 1.44-2.80; P < 0.001) CVD deaths = 1.98 (95% CI: 1.34-2.92; P < 0.05) Nonfatal MI = 1.57 (95% CI: 1.01-2.46; P < 0.05) Kaplan-Meier analysis: Lp(a) tertile 2 (P = 0.001) and tertile 3 (P < 0.001) groups had significantly reduced cumulative event-free survival rates compared with tertile 1 group Cox regression analysis: Lp(a) was independently associated with RCVEs risk (HR: 2.01; 95% CI: 1.44-2.80; P < 0.001)
Liu et al, 202052	Prospective cohort study	4,078	Chinese stable coronary artery disease patients who recently underwent PCI between March 2011 and March 2016	Cardiovascular events including CV death, nonfatal MI, and stroke	Low Lp(a) level: <15 mg/dL (n = 2005; 49.2%) Medium Lp(a) level: 15-30 mg/dL (n = 826; 20.2%) High Lp(a) level: ≥30 mg/dL (n = 1,247; 30.6)	Average of 4.9 years of follow-up Baseline characteristics Low (Lp[a] < 15 mg/dL) (n = 2005) = 49.2% Medium (Lp[a] = 15-30 mg/dL) (n = 826) = 20.2% High (Lp[a] ≥30 mg/dL) (n = 1247) = 30.6% There was a positive relationship between number of lesions vessels and Lp(a) concentration There were 315 CV eventsb (7.7%) during a median follow-up of 4.9 events group had significantly higher Lp(a) levels than nonevents group (P = 0.013). Events group (median [IQR] Lp[a]): 21.95 mg/dL (8.06-44.41 mg/dL) Nonevents group (median [IQR] Lp[a]): 15.13 mg/dL (IQR: 6.75-37.20 mg/dL) Prevalence of CVEs in low Lp(a) group = 6.3% Prevalence of CVEs in medium Lp(a) group = 7.7% Prevalence of CVEs in high Lp(a) group = 10.0% Kaplan-Meier analysis with the log-rank test: Subjects with Lp(a) levels higher than the median value for Lp(a) had a significantly lower cumulative event-free survival time than those with Lp(a) levels below the median value (P = 0.028) High Lp(a) (≥30 mg/dL) group had significantly worse outcomes than the low Lp(a) group (<15 mg/dL) (P = 0.003) No significant difference in the event-free survival rates of medium and low Lp(a) groups (P = 0.196) HRc for total events in high Lp(a) group (relative to low Lp[a] group) = 2.1 (95% CI: 1.5-3.0; P < 0.001) HRc for CVD death in high Lp(a) group (relative to low Lp[a] group) = 1.9 (95% CI: 1.1-3.4; P = 0.027) HRc for nonfatal MI in high Lp(a) group (relative to low Lp[a] group) = 1.4 (95% CI: 0.8-2.6; P > 0.05) HRc for stroke in high Lp(a) group (relative to low Lp[a] group) = 2.0 (95% CI: 1.3-2.9; P = 0.001) Each 1-SD increment of Lp(a) was associated with a 30.0% increased risk of total CVEs and a 29% increased risk of stroke (P = 0.001, respectively), but a nonsignificant increased risk of CV death or nonfatal MI
Man et al, 202557	Cross-sectional study	2,955,320	Chinese general population	Subclinical atherosclerosis	Lp(a) ≤30 mg/dL Lp(a) >30-50 mg/dL Lp(a) >50 mg/dL	Compared with those with low Lp(a) (≤30 mg/dL), individuals with Lp(a) >50 mg/dL had significantly higher odds of CAC (19% higher odds) This association was stronger for those with "multisite" involvement—meaning atherosclerosis present in 2 or 3 different areas of the body simultaneously
Chien et al, 200823	Prospective cohort study	3,484	Ethnic Chinese living in Taiwan; no CVD history at time of recruitment	CVD (stroke and CHD) and all-cause death	Not specified	During a median follow-up time of 13.8 (IQR: 14.5-14.6) years, there were 122 cases of CHD and 781 deaths (any cause). Incidence of all-cause deaths increased appreciably with Lp(a) quartile For all-cause deaths, the RR for individuals the 4th quartile for Lp(a) compared with the 1st quartile was 1.26 (95% CI: 1.03-1.54; P value for trend = 0.007)
Kim et al, 202327	Prospective cohort study	275,430	Korean individuals enrolled in the Kangbuk Samsung Health Study between 2003 and 2016	Mortality-related outcomes in the Korean population including for CVD, CHD, and cerebrovascular events	50 mg/dL	Multivariate Cox regression analysis, compared with the Lp(a) <50 mg/dL group, patients with Lp(a) 50 mg/dL had significantly higher risk of the following: - All-cause mortality (HR: 1.20; 95% CI: 1.03-1.41) - CV mortality (HR: 1.83; 95% CI: 1.26-2.64) - CHD mortality (HR: 3.07; 95% CI: 1.80-5.60) - MI mortality (HR: 3.96; 95% CI: 2.20-7.13)
Cho et al, 201621	Retrospective cohort study	2,611	Korea; healthy adults who underwent routine health check-ups in 2010 and 2014	CAC progression (measured by CACS)	50	Follow-up was 4 years Positive correlation between Lp(a) and baseline CAC (r = 0.044, P = 0.023) After 4 years, after adjusting for confounding factors, OR for CAC progression for baseline Lp(a) >50 mg/dL compared with baseline Lp(a) <50 mg/dL = 1.333 (95% CI: 1.027-1.730)
Sung et al, 201322	Cross-sectional study	14,583	South Korean occupational cohort (healthy/preclinical population)	CACS >0; measure of preclinical coronary atherosclerosis	50	In men: ORd: 1.21; 95% CI: 1.05-1.40; P = 0.008 for having CACS >0 and being in the top Lp(a) quartile (>38.64 mg/dL) In women, the respective ORd was 1.62; 95% CI: 1.03-2.55; P = 0.038
Ikenaga et al, 201148	Prospective cohort	410; 95 high Lp(a) and 315 low Lp(a)	Japanese patients who underwent primary PCI within 24 hours of the onset of AMI	MACE: cardiac death, AMI, and/or revascularization for new lesions	40	410 patients underwent primary PCI within 24 hours of AMI onset. Lp(a) measured 1 week after AMI. Overall mean Lp(a) = 29.7 ± 26.9 mg/dL High Lp(a) (n = 95): Mean Lp(a) = 67.2 ± 1.8 mg/dL; prevalence (Lp(a) >40 mg/dL) = 23.2% (n = 95) Low Lp(a) (n = 315): Mean Lp(a) = 18.5 ± 1.0 mg/dL; prevalence (Lp[a] <40 mg/dL) = 76.8% (n = 315) After 5 years of follow-up: After 5 years of follow-up, incidence of MACE was significantly higher in the high Lp(a) group compared with low Lp(a) group (34.7% vs 16.5%, P < 0.001). Significantly higher incidence of new lesions requiring revascularizatione in the high Lp(a) group compared with low Lp(a) group (31.6% vs 15.2%, P < 0.001) Multivariate analysis: Lp(a) was an independent predictor of both MACE and new cardiac lesions requiring revascularization MACE and high Lp(a): OR: 1.64 (95% CI: 1.31-2.06; P < 0.001) New lesions requiring revascularization: OR: 1.61 (95% CI: 1.32-2.13; P < 0.001)
Mitsuda et al, 201945	Prospective cohort study	668	Japanese patients with STEMI admitted to hospital	Incidence of MACCE	50	Lp(a) measured in 668 consecutive patients with STEMI at admission Lp(a) and incidence of MACCEs were measured at 3 years after the index STEMI cases Baseline characteristics: Median (IQR) Lp(a) = 16.0 (9.0-26.0) mg/dL Low baseline Lp(a) (<50 mg/dL) = 93.1% (n = 622) High baseline Lp(a) (≥50 mg/dL) group = 6.9% (n = 46) Association between baseline Lp(a) and occurrence of MACCEs in 3 years post-STEMI: Kaplan-Meier survival analysis: Patients with high baseline Lp(a) had significantly lower event-free survival time compared with patients with low baseline Lp(a) (P < 0.001) Multivariate Cox regression analysis: baseline Lp(a), previous CAD, and eGFR were found to be independent predictors of MACCE Baseline Lp(a) (comparing high and low group) HR: 1.017; 95% CI: 1.009-1.025; P < 0.001 Previous CAD HR: 1.705 (95% CI: 1.023-2.843; P = 0.041) Baseline eGFR HR: 0.990 (95% CI: 0.981-0.998; P = 0.017)
Momiyama et al, 201246	Cross-sectional study	143	Japanese patients undergoing coronary angiography	Severity of thoracic and abdominal aortic atherosclerosis (plaque scores), and coronary atherosclerosis (number of stenotic coronary vessels)	30	143 patients undergoing coronary angiography; 42 patients (29%) had high Lp(a) level (>30 mg/dL); 104 patients (73%) had CADf; 89 patients (62%) had thoracic aortic plaques; 131 patients (92%) had abdominal aortic plaques Lp(a) and CAD; Median Lp(a) levels significantly higher in patients with CAD than without (21.4 mg/dL vs 15.7 mg/dL, P < 0.05) Median Lp(a) levels increased stepwise with the number of >50% stenotic coronary vessels: 15.7 mg/dL (CAD[−]), 21.2 mg/dL (1-vessel), 21.4 mg/dL (2-vessel), and 22.9 mg/dL (3-vessel) (P < 0.05 for trend). Lp(a) and thoracic and aortic atherosclerosis: For aortic atherosclerosis, 143 patients were divided into quartiles by plaque scores. Median Lp(a) did not differ significantly between quartiles of thoracic plaques: 17.1 mg/dL (Q1), 19.0 mg/dL (Q2), 23.5 mg/dL (Q3), and 21.2 mg/dL (Q4); P > 0.05 Significant stepwise increased in median Lp(a) with increasing quartiles of abdominal plaques: 17.1 mg/dL (Q1), 19.2 mg/dL (Q2), 19.2 mg/dL (Q3), and 24.0 mg/dL (Q4); P < 0.05 Multivariate analysis: Lp(a) was an independent risk factor for CAD and abdominal aortic plaques, but not thoracic aortic plaques For CAD and high Lp(a) (>30 mg/dL), OR: 3.3 (95% CI: 1.2-9.6; P < 0.05) For abdominal aortic plaques and high Lp(a) (>30 mg/dL), OR: 3.7 (95% CI: 1.4-9.8; P < 0.02)
Leowattana et al, 200024	Case-control study	407; 237 CAD cases, 170 healthy control subjects	Thai coronary artery disease patients	CAD	30	Casesg: 24 AMI: Mean Lp(a) = 29 ± 26 mg/dL; prevalence (Lp[a] >30 mg/dL) = 33% 76 UA: Mean Lp(a) = 44.2 ± 49.0 mg/dL; prevalence (Lp[a] >30 mg/dL) = 43.2% 137 SA: Mean Lp(a) = 35.6 ± 31.3 mg/dL; prevalence (Lp(a) >30 mg/dL) = 45.2% Control subjectsg: Mean Lp(a) = 27.6 ± 25.3 mg/dL; prevalence (Lp[a] > 30 mg/dL) = 28.8% Observations: Significant difference between mean Lp(a) levels in UA compared with the control group; P = 0.0006 Significant difference between mean Lp(a) levels in SA compared with the control group; P = 0.0139 UA patients had a significantly higher prevalence of abnormal Lp(a) (>30 mg/dL) compared with the normal healthy control group (43.2% UA vs 28.8% control, OR: 1.90; 95% CI: 1.08-3.32; P = 0.0248) SA patients had a significantly higher prevalence of abnormal Lp(a) (>30 mg/dL) compared with the normal healthy control group (45.2% SA vs 28.8% control, OR: 2.04; 95% CI: 1.27-3.27; P = 0.0028) AMI patients did not have significantly higher prevalence of abnormal Lp(a) >30 mg/dL compared with the healthy control group OR: 1.65; 95% CI: 0.65-4.31; P > 0.05
Sumarjaya et al, 202053	Prospective Cohort study	66	Indonesian hospitalized patients with acute myocardial infarctionh	MACE post-AMI	10.25; determined by ROC curve analysis (area under the curve of 0.787 [80% sensitivity and 63.4% specificity; P ≤ 0.0001])	Prevalence of AMI in Lp(a) <10.25 mg/dL = 53.0% Prevalence of AMI in Lp(a) >10.25 mg/dL = 47.0% MACEs occurred in 38% of patients (n = 25)i Log-rank test: Among MACE cases, significant difference in mean event-free survival time between groups with high Lp(a) and low Lp(a); P = 0.001 High Lp(a) mean event-free survival = 60.6 hours (95% CI: 43.3-77.9 hours) Low Lp(a) mean event-free survival = 104.3 hours (95% CI: 91.4-117.3 hours) HR for MACE post-AMI if high Lp(a) compared with low Lp(a) = 4.63; P = 0.002 With multivariate analysis and Cox proportional hazards regression test, HR: 4.69; 95% CI: 1.67-13.16; P = 0.003. High Lp(a) was an independent risk factor for occurrence of MACEs following hospitalization post-AMI
Loh et al, 202232	Cross-sectional study	2025	Consecutive consenting patients undergoing diagnostic coronary angiography at the National University Heart Centre, Singapore, enrolled in SCADGENS (the Singapore Coronary Artery Disease Genetics Study)	Correlation between Lp(a) and CV phenotypes	120 nmol/L 130 nmol/L 150 nmol/L 160 nmol/L	In univariable analysis, Lp(a) levels and Lp(a) ≥120 nmol/L were positively associated with the presence of CAD when compared with the CAD− group (OR: 1.10 per 10 nmol/L increment; 95% CI: 1.05-1.16; P < 0.001, and OR: 4.88; 95% CI: 1.79-13.3; P < 0.002, respectively). After adjusting for confounding risk factors, Lp(a) levels and Lp(a) ≥120 nmol/L were significant predictors of the presence of CAD when compared with the CAD− group (OR: 1.12 per 10 nmol/L increment; 95% CI: 1.06-1.19; P < 0.001, and OR: 5.41; 95% CI: 1.69-17.34; P = 0.004 respectively) Among patients with CAD, higher Lp(a) level was associated with increased risk of AMI (OR: 1.02 per 10 nmol/L increment; 95% CI: 1.00-1.04; P = 0.024); when comparing CAD + AMI + group vs CAD + AMI− group, Lp(a) ≥120 nmol/L was not associated with increased AMI risk (P = 0.077), but higher Lp(a) cutoffs were Lp(a) ≥130 nmol/L (OR: 1.47; 95% CI: 1.02-2.10; P = 0.038), Lp(a) ≥150 nmol/L (OR: 1.73; 95% CI: 1.12-2.67; P = 0.013), Lp(a) ≥160 nmol/L (OR: 1.96; 95% CI: 1.20-3.19; P = 0.007), and Lp(a) ≥170 nmol/L (OR: 1.94; 95% CI: 1.12-3.35; P = 0.018)
Gambhir et al, 200818	Cross-sectional study	380; 220 CAD cases, 160 non-CAD control subjects	Indians <40 years old	CAD	Not specified	Median plasma Lp(a) levels were 2.5 times higher in patients with established CAD compared with control subjects (30 mg/dL vs 12.7 mg/dL; P < 0.05) 50% of CAD patients and 22% of control subjects had Lp(a) > 30 mg/dL 33% of control subjects and 62% of CAD patients had Lp(a).20 mg/dL
Yusuf et al, 201420	Prospective case-control study	600; 450 CAD patients and 150 healthy control subjects	North Indian	CAD	Not specified Lp(a) quartiles defined	CAD patients had significantly higher median (IQR) serum Lp(a) compared with control subjects (30.30 [14-55.92] vs 20 [12-40] mg/dL, P < 0.001) Multivariable regression analysis: Lp(a) was an independent risk factor for CAD
Saleheen et al, 201754	Mendelian randomization study	17,503 (PROMIS cohort); 184,305 (CARDIoGRAMplusC4D consortium cohort)	Pakistani patients and control subjects (frequency matched for age and sex) from the PROMIS; patients and control subjects (frequency matched for age and sex) from the CARDIoGRAMplusC4D consortium cohort	AMI	Not specified	After adjustment for Lp(a) concentration and conventional lipid concentrations, OR for AMI was 0.93 (95% CI: 0.90-0.97; P < 0.0001) for every 1-SD incremental increase in LPA KIV2 repeat number After adjusting for LPA KIV2 repeats and conventional lipids, OR for AMI was 1.10 (95% CI: 1.05-1.14) for every 1-SD incremental increase in Lp(a) concentration Hence, many KIV2 repeats correlate with lower Lp(a) (larger particles) and therefore lower AMI risk
Bandara et al, 201647	Cross-sectional study	102	Sri Lankan CAD patients awaiting CABG	CAD severity (Gensini score)	30	102 Sri Lankan CAD patients awaiting CABG; 67 male individuals (65.7%) ∼2/3 of patients had Lp(a) >30 mg/dL Overall mean ± SD Lp(a) = 50 ± 38 mg/dL No significant difference between Lp(a) in men and women
Perera et al, 201555	Case-control study	85 (58 male and 27 female patients)	Sri Lankan patients 40-60 years of age confirmed as having CAD on coronary angiography and 85 age- and sex-matched healthy volunteers as control subjects	Apo A-1 LDL, Apo B, Apo A-I, and Lp(a)	Not specified	Serum Lp(a) cutoff value of 17.3 mg/dL may be an important predictor in ruling out major vessel disease and luminal narrowing by atheroma. Patients (mean): 23.8 mg/dL Control subjects: 19.2 mg/dL
Loh et al, 202458	Retrospective cohort study	521	Singaporean patients with ischemic heart disease	Premature ischemic heart disease	Normal Lp(a) defined as <70 nmol/L (approx. 30 mg/dL)	Median Lp(a) was 35.2 nmol/L (approx. 15 mg/dL) Prevalence (Lp[a] < 70 nmol/L∼30 mg/dL) = 70.8% Prevalence (70 nmol/L < Lp[a] < 120 nmol/L) = 13.4% Prevalence (Lp[a] > 120 nmol/L ∼50 mg/dL) = 15.7% 82.2% of patients were men 46.6% had newly diagnosed IHD and 10% had premature IHD Multivariable regression analysis: Lp(a) ≥155 nmol/L (approx. 62 mg/dL) was associated with development of premature IHD (OR: 2.90; 95% CI: 1.26-6.67; P = 0.012)

AMI = acute myocardial infarction; Apo A-1 = apolipoprotein A-1; Apo B = apolipoprotein B; CABG = coronary artery bypass graft; CAC = coronary artery calcification; CACS = coronary artery calcium score; CARDIoGRAMplusC4D = Coronary Artery Disease Genome-Wide Replication and Meta-Analysis Plus the Coronary Artery Disease Genetics; CHD = coronary heart disease; CV = cardiovascular; CVE = cardiovascular event; eGFR = estimated glomerular filtration rate; HTN = hypertension; IHD = ishaemic heart disease; LDL = low-density lipoprotein; LPA KIV2 = copy number variation domain in the LPA gene; MACE = major adverse cardiovascular event; MACCE = major adverse cardiac and cerebrovascular event; MI = myocardial infarction; MVD = multivessel disease; PCAD = premature CAD; PCI = percutaneous coronary intervention; PROMIS = Pakistan Risk of Myocardial Infarction Study; RCVE = recurrent cardiovascular event; ROC = receiver-operator curve; RR = relative risk; SA = stable angina; STEMI = ST elevation myocardial infarction; T2DM = type 2 diabetes mellitus; UA = unstable angina; other abbreviations as in Tables 1 and 2.

^aAfter adjustment for age, sex, smoking, diabetes mellitus, HTN, left ventricular ejection fraction, revascularization (PCI or coronary artery bypass graft), triglycerides, low-density lipoprotein cholesterol, high-sensitivity C-reactive protein, glycated hemoglobin, and baseline statin use (yes or no).

^bCV event: CV death, nonfatal MI, and stroke.

^cHRs adjusted for covariates: age, sex, current smoking, HTN, diabetes mellitus, systolic blood pressure, glycosylated hemoglobin, high-sensitivity C-reactive protein, number of lesion vessels, baseline statin use, triglyceride, and low-density lipoprotein cholesterol.

^dAfter adjusting for age, sex, lipid-lowering therapy, and multiple CV risk factors.

^eRevascularization for new lesions was defined as revascularization for >75% stenosis for nontarget lesion that progressed more than 1 grade according to the classification of the American Heart Association.

^f1-vessel disease: 44 patients; 2-vessel disease: 34 patients; 3-vessel disease: 26 patients.

^gCV risk factors, including Lp(a), measured in 237 patients diagnosed with CAD via coronary angiography (175 men, 62 women; mean age of 61 ± 10 years). Control subjects: 170 healthy volunteers (95 men, 75 women; mean age 58 ± 15 years).

^hFrom June-August 2018, 66 people hospitalized following AMI were examined for serum Lp(a) at the time of admission and the occurrence of MACE during their hospitalization (63.6% STEMI, 36.4% non-STEMI).

ⁱPercentage of patients: cardiogenic shock (10.8%); heart failure (30.3%); CV death (7.6%); malignant arrhythmias (7.6%); postinfarction angina (7.6%).

One Korean cohort study demonstrated an association between the highest quartile of Lp(a) and coronary artery calcium score (CACS) >0 (OR: 1.23; 95% CI: 1.03-1.46; P = 0.020), even after adjustment for multiple potential confounders.²² Another Korean retrospective cohort study examined the changes in CACS in patients with high and low Lp(a) over the course of 4 years. The study sample was divided into 2 groups (those with Lp[a] ≥50 mg/dL or <50 mg/dL) and preclinical CAD disease progression was defined as a CACS change of >0 after 4 years.²¹ The OR for CAC progression for those with baseline Lp(a) >50 mg/dL compared with baseline Lp(a) <50 mg/dL was 1.333 (95% CI: 1.027-1.730). There also was a stepwise increase in adverse coronary events observed universally in Asian groups with established CAD, echoing the dose-response relationship observed in The Copenhagen City Heart Study.³⁸

A study in Singapore also found that plasma Lp(a) concentration is a positive predictor of CAD and AMI, with individuals with no CAD having the lowest Lp(a) levels (median of 15.8 nmol/L), followed by those with CAD with no history of AMI (20.1 nmol/L), whereas those with a history of AMI having the highest levels (26.2 nmol/L).³² A more recent Singapore study also found similar results. In this study, a baseline Lp(a) ≥70 nmol/L was associated with an increased risk of all-cause mortality (HR: 1.97; 95% CI: 1.20-3.22; P = 0.007) and cardiovascular mortality (HR: 2.01; 95% CI: 1.06-3.82; P = 0.033) in patients with CAD.⁵⁶ An additional Singaporean study of patients with CAD identified on multivariable regression analysis that Lp(a) ≥155 nmol/L was associated with premature onset of CAD.⁵⁸

In European populations, the conventional Lp(a) cutoff of 50 mg/dL was defined empirically. Approximately 20% of participants from European studies have Lp(a) >50 mg/dL¹³ and it has historically been the top 20% to 25% of the Lp(a) distribution, which is associated with a 2- to 3-fold increase in CAD risk.^39,59, 60, 61, 62 This held true in an appreciable proportion of the Asian CAD studies included.^10,20,51,52 As mentioned in an earlier section of this review, one Chinese study applied this principle to its own study population and calculated the appropriate cutoff for myocardial infarction (MI) risk to be 16.7 mg/dL.¹¹ Similarly, a study from Sri Lanka found that in individuals with a positive family history of CAD, serum Lp(a) ≥17.3 mg/dL indicated an increased risk for CAD.⁵⁵ In another Chinese cohort study, a positive inflection was described in the rate of new coronary events above the 80th percentile for Lp(a), where the relative risk of MI increased substantially.¹⁰ An initial MI in the fourth quintile for Lp(a) in a population with known CAD had an OR of 1.933 (95% CI: 1.554-2.406). The OR increased substantially in the fifth quintile to 2.821 (95% CI: 2.281-3.489).

Lp(a) was associated with recurrent cardiac events and cardiac death in numerous Asian studies. Cox regression analysis showed that Lp(a) was independently associated with increased risk of recurrent CV events including CVD death and nonfatal MI in a large Chinese prospective cohort study of patients with CAD who had already had 1 CVD event. In a Japanese cohort of patients who underwent primary percutaneous coronary intervention within 24 hours of onset of AMI, Lp(a) was also an independent predictor of major adverse cardiovascular events including cardiac death and MI, and of new nonculprit lesions on coronary angiogram requiring revascularization.⁴⁸ These findings demonstrate a robust association between elevated serum Lp(a) and adverse coronary outcomes in Asian populations.

Lp(a) and PVD

There were far fewer studies exploring the relationship between Lp(a) and PVD in Asian populations compared with CAD and stroke (Table 4).^14,25,63,64 The Asian studies addressing this topic had comparatively small sample sizes. Despite this, the data depict a reasonably clear positive correlation between elevated Lp(a) and PVD in Asian groups, which parallels the relationship between elevated Lp(a) and CAD discussed previously. This is underpinned by the common pathogenesis the 2 disease processes share. In a prospective case-control study in China, Lp(a) was an established risk factor for PVD and was positively correlated with disease severity.¹⁴ There were consistent findings in a Japanese cross-sectional study, where Lp(a) was significantly higher among patients with critical limb ischemia than those with intermittent claudication.⁶³ In a Singaporean study of patients with PVD, Lp(a) ≥120 nmol/L (approximating ≥50 mg/dL) was associated with the highest severity of limb ischemia (P = 0.032).⁶⁵ In a Taiwanese cross-sectional study, there was a statistically significant increase in serum Lp(a) in patients grouped by increasing severity of PVD.⁶⁴ The Asian data available did not demonstrate Lp(a) to be a predictor of preclinical PVD; however, this can be understood by the absence of conventional proven screening tests for PVD. Nonetheless, we presume that as with CAD, elevated Lp(a) would likely have a positive correlation with both the presence and the progression of preclinical PVD.

Table 4.

Asian Lp(a) PVD Table: Statistical Findings

First Author, Year of Publication	Type of Study	Participants	Population	Outcome of Interest	Elevated Lp(a) Cutoff Value Used (mg/dL)	Findings
Cheng et al, 199714	Prospective case-control	200 cases lower limb PVD and 200 age- and sex-matched control subjects	Chinese (Hong Kong)	PVD	24	Lp(a) was established as an independent risk factor for PVD: 53% (n = 106) of PVD cases had high Lp(a) (>24 mg/dL) 36% of control subjects had high Lp(a) (>24 mg/dL) For high Lp(a) (>24 mg/dL) and PVD, OR: 2.00; P < 0.001 Lp(a) was associated positively with disease severity
Hakim et al, 200825	Case-control	150 PVD cases, 50 control subjects	Malaysia (Kuala Lumpur local population)	PVD	36	PVD cases: Median Lp(a) = 48 mg/dL; proportion (Lp(a) > 36 mg/dL) = 58% Controls: Median Lp(a) = 17 mg/dL; proportion (Lp(a) >36 mg/dL) = 26% Multivariate analysis: OR for PVD and high Lp(a) (>36 mg/dL): 7.69 (95% CI: 2.11-28.01; P = 0.002)
Hikita et al, 201563	Cross-sectional	242 patients with PVD: 42 patients with CLI; 200 patients with IC	Japanese	PVD severity	Not specified	CLI group: Mean Lp(a) = 45.9 ± 23.3 mg/dL IC group: Mean Lp(a) = 26.2 ± 27.7 mg/dL Mean Lp(a) IC group vs CLI group (P = 0.0002)
Tseng, 200464	Cross-sectional	45 patients with PVD 512 control subjects	Ethnic Chinese with diabetes in Taiwan	PVD severity	13.3	Mean Lp(a): No PVD = 17.1 ± 14.4 mg/dL Mild PAD = 23.7 ± 20.3 mg/dL Severe PAD = 36.9 ± 22.8 mg/dL P < 0.001 for trend Risk for PVD with Lp(a) > 13.3 mg/dL, OR: 2.685 (95% CI: 1.203-5.994; P < 0.05)
Loh et al, 202565	Retrospective cohort study vs cross-sectional study	106 patients with PVD	Singaporean patients admitted to hospital for PVD-related limb ischemia	Prevalence of mildly raised vs elevated Lp(a) levels	Mildly raised Lp(a) defined as Lp(a) ≥70 nmol/L (∼30 mg/dL); Elevated Lp(a) defined as Lp(a) ≥120 nmol/L (∼50 mg/dL)	Prevalence(Lp(a) ≥120 nmol/L ∼50 mg/dL) = 22.6% Prevalence(Lp(a) ≥70 nmol/L ∼30 mg/dL) = 34.9% A higher proportion of patients with levels ≥120 nmol/L compared with Lp(a) <120 nmol/L had category 6 classification of chronic limb ischemia by Rutherford classification (95.8% vs 70.7%, P value = 0.011). Lp(a) ≥120 nmol/L and Lp(a) as a continuous variable were associated with the highest severity of limb ischemia, P = 0.032 and P = 0.045, respectively.

CLI = critical limb ischemia; IC = intermittent claudication; PAD = peripheral arterial disease; PVD = peripheral vascular disease; other abbreviation as in Table 1.

Lp(a) and stroke

The relationship of Lp(a) with stroke is contentious. As with CAD, Lp(a) was predictive of preclinical cerebrovascular disease in the studies included. In a prospective study of a general Japanese population assessing carotid intima-media thickness, it was established that over 5 years, every 1 log unit increase in baseline Lp(a) predicted a 0.06-mm increase in carotid intima-media thickness.⁶⁶ A large study of 2,955,320 healthy Chinese participants demonstrated a strong relationship between elevated Lp(a) and subclinical atherosclerotic cerebrovascular disease. Compared with those with low Lp(a) (≤30 mg/dL), individuals with Lp(a) >50 mg/dL had significantly higher odds of detectable carotid plaque (32% higher odds), increased carotid intima-media thickness (15% higher odds), and subclinical brain infarcts (10% higher odds).⁵⁷ In spite of these data, the relationship between Lp(a) and clinical cerebrovascular disease is less well defined. In many studies, there was an overall increase in ischemic stroke and total stroke risk with increasing Lp(a), which resembled the dose-response relationship established between Lp(a) and coronary disease (Table 5).^{12,13,16,17,23,66} In a prospective Taiwanese study, incident stroke risk increased appreciably with baseline Lp(a) quartile: the relative risk for quartile 4 compared with quartile 1 was 1.56 (95% CI: 1.07-2.28; P = 0.003 for the trend).²³ Some studies did not distinguish hemorrhagic from ischemic mechanisms of stroke.

Table 5.

Asian Lp(a) Stroke Table: Statistical Findings

First Author, Year of Publication	Type of Study	Number of Participants	Population	Outcome of Interest	Elevated Lp(a) Cutoff Value Used (mg/dL)	Findings
Xia et al, 202216	Prospective cohort study	25,343	Chinese Han general population	Diagnoses of CAD, stroke, T2DM, and HTN within the study period	Not specified Percentiles defined.	Lp(a) 25th percentile = 5.9 mg/dL Population median Lp(a) = 11.9 mg/dL (IQR: 5.9-23.7 mg/dL) Lp(a) 75th percentile = 23.7 mg/dL Lp(a) 90th percentile = 43.7 mg/dL Lp(a) 95th percentile = 60.8 mg/dL Median age 52 years, 54.4% female Association between stroke (n, %) and Lp(a) percentile 0–25th percentile (173, 2.7%): Median Lp(a) = 3.6 mg/dL (IQR: 2.4-4.7 mg/dL) →OR: 0.92; 95% CI: 0.73-1.15; P = 0.447 25–50th percentile (193, 3.0%): Median Lp(a) = 8.6 mg/dL (IQR: 7.2-10.1 mg/dL) →OR (95% CI) = 1.00 50–75th percentile (193, 3.0%): Median Lp(a) = 16.5 mg/dL (IQR: 13.9-19.6 mg/dL) →OR: 1.12; 95% CI: 0.90-1.39; P = 0.309 75–90th percentile (106, 2.8%): Median Lp(a) = 30.9 mg/dL (IQR: 26.9-36.1 mg/dL) →OR: 1.01; 95% CI: 0.78-1.31; P = 0.927 90–95th percentile (44, 3.5%): Median Lp(a) = 50.0 mg/dL (IQR: 46.6-54.9 mg/dL) →OR: 1.39; 95% CI: 0.98-1.93; P = 0.064 95–100th percentile (43, 3.4%): Median Lp(a) = 78.6 mg/dL (IQR: 68.2-95.6 mg/dL) →OR: 1.27; 95% CI: 0.88-1.81; 0.199 Overall P values for trend: 0.107
Fu et al, 202012	Retrospective case-control	4,494 participants; 2,149 stroke cases, 2,345 healthy control subjects	Chinese Han inpatients with ischemic and hemorrhagic stroke	Ischemic and hemorrhagic stroke	30	1,953 ischemic stroke cases, 1,953 healthy, age- (±5 years) and sex-matched control subjects (1:1) Median Lp(a) in ischemic stroke cases = 12.20 mg/dL (IQR: 5.80-24.80) Median Lp(a) in control subjects = 8.60 mg/dL (IQR: 3.70-21.30) Conditional multivariate analysis: Positive relationship between log-transformed Lp(a) concentration and ischemic stroke compared with patients in the lowest quartile (after adjusting for potential covariates: HTN, DM, hyperlipidemia, hsCRP)
Zhang et al, 201913	Prospective cohort study	8,500	Middle-aged and older Chinese patients	Stroke	Not specified	Mean follow-up of 5.1 years Corresponding stroke incidence over follow-up period per Lp(a) tertile: Tertile 1: Mean Lp(a) = 6.60 ± 2.68 mg/dL; Incidence (stroke) = 4.44% Tertile 2: Mean Lp(a) = 18.98 ± 4.66 mg/dL; Incidence (stroke) = 5.14% Tertile 3: Mean Lp(a) = 38.22 ± 14.67 mg/dL; Incidence (stroke) = 6.14% There was a significant relationship between Lp(a) concentration tertile and incident stroke risk (P < 0.05 for trend)
Man et al, 202557	Cross-sectional study	2,955,320	Chinese general population	Subclinical atherosclerosis	Lp(a) ≤30 mg/dL Lp(a) >30-50 mg/dL Lp(a) >50 mg/dL	Compared with those with low Lp(a) (≤30 mg/dL), individuals with Lp(a) >50 mg/dL had significantly higher odds of the following: Carotid plaque (32% higher odds) Coronary artery calcification (19% higher odds) Increased carotid intima-media thickness (15% higher odds) Subclinical brain infarcts (10% higher odds). These associations were stronger for those with "multisite" involvement—meaning atherosclerosis present in 2 or 3 different areas of the body simultaneously There was a nonlinear dose-response relationship between Lp(a) and carotid plaque (P for nonlinearity 0.028)
Ikezaki et al, 201766	Prospective cohort study	2,245	General population from the Fukuoka area of Japan	Carotid intimal medial thickness (CIMT) of the common carotid artery bilaterally (as a marker of ASCVD risk)	Not specified	5 years of follow-up. Controlling for age and gender, Lp(a) was significantly associated with change in CIMT over the 5-year follow-up period. For every 1 log unit increase in Lp(a) level at baseline, there was a 0.06-mm increase in CIMT over 5 years (P = 0.02)
Ishikawa et al, 201317	Prospective cohort study	10,494	Japanese general population	Incidence of stroke and its subtypes (cerebral hemorrhage, cerebral infarction, and SAH)	Lp(a) Tertiles used	Follow-up duration of 10.7 years Multivariate analysis: after adjusting for age, smoking status, drinking status, BMI, SBP, TC, HDL-C and blood glucose, there appeared to be a reduced risk of cerebral hemorrhage at higher Lp(a) levels (see supplementary table for stroke) in male individuals, including increased cerebral hemorrhage risk
Chien et al, 200823	Prospective cohort study	3,484	Ethnic Chinese living in Taiwan; no CVD Hx at time of recruitment	CVD (stroke and CHD) and all-cause death	Not specified	During a median (IQR) follow-up time 13.8 (14.5-14.6) years: 210 cases of stroke ⇨ Incidence of stroke increased appreciably with Lp(a) quartile ⇨ For stroke, the RR (95% CI) for individuals the 4th quartile for Lp(a) compared with the 1st quartile was 1.56 (1.07-2.28), P value for trend = 0.003 The multivariate RR was significant for stroke at the 90th and 95th percentiles and for total death at the 95th and 99th percentiles 90th percentile for Lp(a) = 34.33 mg/dL; RR for stroke: 1.50; 95% CI: 1.01-2.22; P = 0.045 95th percentile for Lp(a) = 47.08 mg/dL; RR for stroke: 1.75; 95% CI: 1.04-2.93; P = 0.035; RR for all-cause death: 1.39; 95% CI: 1.04-1.86; P = 0.025 99th percentile for Lp(a) = 69.30 mg/dL RR for all-cause death: 1.96; 95% CI: 1.10-3.49; P = 0.022

ASCVD = atherosclerotic cardiovascular disease; BMI = body mass index; DM = diabetes mellitus; HDL-C = high-density lipoprotein cholesterol; hsCRP = high-sensitivity C-reactive protein; Hx = history; RR = relative risk; SAH = sub arachnoid haemorrhage; SBP = systolic blood pressure; TC = total cholesterol; other abbreviations as in Tables 1, 2, and 3.

Where ischemic and hemorrhagic stroke were assessed separately, there were conflicting outcomes. In a Chinese retrospective case-control study with 2,149 stroke cases and 2,345 healthy control individuals, there was a positive relationship between hemorrhagic stroke and log-transformed Lp(a) concentration.¹² By contrast, low Lp(a) conferred increased stroke risk in a large prospective cohort study examining a Japanese general population sample numbering 10,494.¹⁷ There were 393 incident strokes within the mean 10.7-year follow-up period. Three vascular events were accounted for: cerebral hemorrhage, cerebral infarction, and subarachnoid hemorrhage. The same general trend was observed in all 3 categories: an overall decrease in stroke incidence and multivariate adjusted HRs with increasing Lp(a) tertile. This was statistically significant when examining cerebral hemorrhage in the total sample and male individuals alone, but not among patients with cerebral infarction or subarachnoid hemorrhage. There are several factors to consider: there were no other Asian studies that distinguished hemorrhagic from ischemic stroke included in this study; these 2 papers compare different Asian populations that seem to have different Lp(a) distributions; and they have substantially different sample sizes. It is difficult to draw meaningful conclusions other than to comment that more studies are required to investigate the effect of Lp(a) on stroke risk, especially hemorrhagic stroke. Recommendations for testing Lp(a) as part of CVD risk management in Asian populations.

Current guidelines recommend measuring Lp(a) at least once during adulthood to detect individuals with genetically elevated concentrations, particularly in those with a personal or family history of premature atherosclerotic CVD (Table 6).67, 68, 69 Many international guidelines use a threshold of 50 mg/dL, a value largely derived from The Copenhagen City Heart Study. However, emerging evidence and recent consensus statements from studies in Asian populations—particularly Chinese cohorts—suggest that a lower cutoff of around 30 mg/dL may better identify individuals at increased cardiovascular risk.⁶⁷

Table 6.

Recommendations for Population Screening of Lp(a) as Part of CVD Risk Management in Asian Populations

First Author, Year of Publication	Population	Society	Recommendations	Lp(a) Threshold
Blumenthal et al, 2025	Europid	American College of Cardiology/American Heart Association (Joint Committee on Clinical Practice Guidelines)	Lp(a) measurement at least once during adulthood, particularly in those with a personal or family history of premature atherosclerotic CVD.	≥50 mg/dL
Mach et al, 202669	Europid	2025 Focused Update of the 2019 ESC/EAS Guidelines for the management of dyslipidaemias	It is reasonable to consider elevated Lp(a) levels >50 mg/dL (>105 nmol/L) to refine CV risk estimation and potentially re-classify risk in individuals at moderate risk or close to treatment decision thresholds. Advocate for Lp(a) testing at least once in each adult's lifetime.	Lp(a) ≥105 nmol/L (approximately 50 mg/dL)
Koh et al, 202172	Asian Pacific	Asian Pacific Society of Cardiology	Lp(a) measurement should be performed at least once in each adult person's lifetime, especially those with family history of premature atherosclerotic CVD.	Not specified
Li et al, 202267	Chinese	Beijing Heart Society	Lp(a) testing is recommended for the following populations: 1) populations at extremely high risk for ASCVD 2) individuals with a family history of early-onset ASCVD (men <55 y of age, women <65 y of age) 3) individuals whose lineal relatives have Lp(a) levels >90 mg/dL (200 nmol/L) 4) FH or other types of hereditary dyslipidemia 5) Patients with calcific aortic valve stenosis	≥30 mg/dL
Puri et al 202068	Indian	Lipid Association of India (LAI) - LAI Expert Consensus Statement on the Management of Dyslipidaemia in Indians 2020	Lp(a) measurement at the time of initial screening at 18 years old. Targeted: premature ASCVD FH. Family history of premature ASCVD and/or high Lp(a). Recurrent ASCVD patients after acute coronary syndrome.	≥50 mg/dL

CV = cardiovascular; ESC/EAS = European Society of Cardiology/European Atherosclerosis Society; FH = familial hypercholesterolemia; other abbreviations as in Tables 1, 2, and 5.

Based on our synthesis of the available literature, it is evident that Lp(a) should be acknowledged as a cardiovascular risk factor. It has been shown that Lp(a) testing in primary prevention can refine CVD risk classification and may reduce costs from both health care and societal perspectives.⁷⁰ More studies in Asian populations are needed to aid in the development of appropriate population-specific risk reduction pathways for testing and initiation of Lp(a)-lowering therapies, which can be applied effectively with implementation science.

Study limitations

Most of the studies are cross-sectional studies, which limits the ability of these studies to determine causality of Lp(a) in the development of atherosclerotic CVD. Furthermore, the studies varied widely in terms of populations, outcomes reported, and data analysis. The studies also used varied assays—Lp(a) has many isoforms, which underscore the importance of isoform-independent assays. All these study variations made between-study comparisons difficult. Despite the growing number of studies on Lp(a), more studies are still needed, such as those that could aid in the incorporation of Lp(a) into cardiovascular risk prediction tools for individual populations.

Conclusions

Lp(a) levels reported in Asian studies are broadly similar to those of Europeans; however, there did appear to be some interethnic variation. Hence, there is a need to develop ethnic-specific Lp(a) levels and accompanying guidelines for the Asian regions when determining and managing CV risk. The Asia-Pacific literature included in this review overwhelmingly describes a dose-response relationship between serum Lp(a) concentration and CV risk. This is consistent with the findings of The Copenhagen City Heart Study and the broader Europid literature. Our review confirms that Lp(a) has pathological effects in both Asian and Europid individuals and contributes to the body of evidence purporting that an excess of Lp(a) might be causative of CVD (Central Illustration). These findings are also logical considering the proatherogenic, prothrombotic, and proinflammatory properties of Lp(a).

Our review affirms that, in contrast to the Caucasian population, there is comparable scarcity of data on the distribution of Lp(a) in Asian individuals. More studies in general, including increased enrollment in cardiovascular outcome trials, are encouraged to strengthen the knowledge based on this underrepresented population. There is also a need for the region to build capacity in Lp(a) measurement, harmonization of units, and data linkage across nations.

Funding Support and Author Disclosures

The authors have reported that they have no relationships relevant to the contents of this paper to disclose.