The association of xanthelasma palpebrum with cardiovascular events: systematic review with meta-analysis 

Xanthelasmas are lipid depositions in the skin that appear to be associated with qualitative and quantitative abnormalities of lipid metabolism that may also favour deposition in the arterial wall.¹ The relationship between xanthelasma palpebrum (XP) and cardiovascular disease (CVD) is not well known and is understudied. Our aim was to synthesize all the available evidence and evaluate whether xanthelasmas are independently associated with cardiovascular events.

Potentially eligible studies were identified through an electronic search of bibliographic databases from inception to July 2020 (MEDLINE, Cochrane Central Register of Controlled Trials, and Web of Science). Search strategy and detailed methodology are available in Supplementary material online, Data 1. Article screening and data extraction were performed by two reviewers who independently selected studies comparing cardiovascular events between patients with XP and control groups without this characteristic. Ischaemic heart disease, stroke, and overall CVD were the outcomes of interest. The risk of bias was evaluated using the Risk Of Bias In Non-randomized Studies—of Exposures tool.² Pooled estimates of odds ratios (ORs) and 95% confidence intervals (CIs) were derived by random effects meta-analysis. Heterogeneity was assessed using the I² test. The study protocol was registered at PROSPERO:CRD42020219957.

The electronic database search yielded 1133 studies from 1947 to 2020 ( Supplementary material online, Data 2). After applying the inclusion and exclusion criteria, six studies were included: three cohort studies^3–5 and three case–control studies.^6–8 Reasons for exclusion available in Supplementary material online, Data 2 and 3.

The overall risk of bias was moderate, mainly due to the level of risk of confounding implied (i.e. lack of adjustment to relevant risk factors), and only one study (Jee et al.’s case–control study)⁷ was classified as having a serious risk of bias due to the methods for pre-selection of patients for the control group that implied deviation from intended interventions (see Supplementary material online, Data 4).

Overall, 13 476 patients were included. The follow-up range was from 2 to 35 years (Table 1). Pooling all the available data, no significant associations were found regarding XP with ischaemic heart disease (OR 1.81, 95% CI 0.84–3.91; I² = 55%; three studies), stroke (OR 1.53, 95% CI 0.37–6.32; one study), and overall CVD (OR 1.33, 95% CI 0.76–2.32; I² = 36%; four studies) in XP patients (Figure 1). Excluding one study at serious risk of bias, the risk for CVD associated with XP was significantly increased (OR 1.65, 95% CI 1.05–2.61; I² = 0%; three studies; Figure 1).

Figure 1

Forest plot for ischaemic heart disease, stroke, and cardiovascular disease.

Table 1

Main study characteristics

Study, year	Design	Location	Data source	No. of XP patients and controls	Follow-up duration (years)	Age (years), male (%)	Diagnosis methods	Matching	Outcome assessment	Outcome adjustment	CVD/CVR in xanthelasma vs. control considered
Christoffersen, 2013	Prospective cohort study	Copenhagen, Denmark	Copenhagen City Heart Study. From drawn randomly from the Copenhagen Central Person Registry	488 XP 10 397 XP-free	35 years (mean follow-up 23 years) with 100% complete follow-up	20–-80 years 53.4% (n = 5828)	ICD-8-CM codes 410 and 410–414; ICD-10-CM codes I21-I22 and I20-I25	__	Myocardial infarction, ischaemic heart disease, ischaemic stroke, ischaemic cerebrovascular disease, and death	Multifactorial adjustment for: cholesterol, triglycerides, body mass index, hypertension, diabetes, smoking in pack-years, alcohol consumption, physical activity, post-menopausal status, hormonal replacement therapy, education, income, and family history of ischaemic vascular disease	With xanthelasmata: 1.30 (1.12–1.50) for IHD (HR)
Dey, 2013	Case–control study	New Delhi, India	Department of Medicine/Preventive Cardiology at Hamdard Institute of Medical Sciences and Research and HAHC Hospital	61 XP 130 XP-free	∼2 years	21–73 years	Each patient underwent detailed history and physical examination. Blood samples for haemogram, diabetes, and lipids	NR	Coronary artery disease	NR	Coronary artery disease: 6.6% (4–61) vs. 0% (0–130)
						55.7% (n = 34)
Özdöl, 2008	Case–control study	Ankara, Turkey	Departments of Dermatology/Cardiology, Faculty of Medicine, Hacettepe University, Ankara, Turkey	100 XP	∼3 years	48 ± 11 years	Physical examination (for XP diagnosis). Standard questionnaire: for the evaluation of systemic risk factors. Blood Specimens: fasting blood glucose, lipid profile, and LP (a)	Age and sex matched	Clinically overt CVD and future cardiovascular risk, rates of atherosclerotic disease, and serum LP (a) levels	NR	Clinically overt CVD:
				100 XP-free		40% (n = 40)					7% (XP)
											8% (XP-free)
Jee, 2003	Case–control study	Seoul, Korea	Department of Dermatology at Asan Hospital in Seoul	37 XP	∼4 years	17–71 years	Clinical records and telephone interviews, physician’s inspection (for XP diagnosis) and histopathologic confirmation (23 p)	For age	CVD	NR	Incidences of CVD:
				36 XP-free		52.4 (mean)					16.2% (XP)
						35.6% (n = 26)					25% (XP-free)
Menotti, 1998	Prospective cohort study	Italy (rural communities of Crevalcore and Montegiorgio)	Seven countries study on CVD	1527	35 years	49.7 (mean)	8th Revision of the WHO-ICD classification physician’s inspection (for XP diagnosis)	__	CHD-restricted and broad criteria; stroke and CVD according to ICD-8	Multivarial models for 21 measurements (socio-demographic information, family history of relevant cardiovascular and allied conditions, behaviour and life-style characteristics, anthropometric, metabolic measurements, cardiorespiratory variables, and clinical signs)	HR for XP
						100% (n = 1527)					CHD (RC): 2.13 (0.85–5.32)
											CHD (BC): 206 (095–4.45)
											Stroke: 1.53 (0.37–6.32)
											CVD: 1.98 (1.04–3.77)
Rouffy, 1982	Retrospective cohort study	Paris, France	consultation of the specialty of hyperlipoproteinaemia	25 XP	∼2 years	10–79 years	Familial hyperlipoproteinaemia: WHO criteria Physician’s inspection (for XP diagnosis) History examination for: arterial risk (non-lipid), extra-vascular lipid deposits; background or current coronary, cerebral, or limb arteriographic manifestations	__	Arteriopathies (coronary, cerebral, or peripherical)	NR	Arteriopathies in XP patient: 40% (vs 28%)
				575 XP-free		47 years (mean)
						65% (n = 390)

BC, broad criteria; CHD, coronary heart disease; CM, medical codes; CVD, cardiovascular disease; CVR, cardiovascular risk; HR, hazard ratio; ICD, international classification of diseases; IHD, ischaemic heart disease; LP, lipoprotein; NR, not reported; RC, restricted criteria; WHO, World Health Organization; XP, xanthelasma palpebrum.

The main findings of this systematic review were: (i) current evidence does not support a positive association between the presence of XP and increased risk of cardiovascular events (Figure 1); and (ii) evaluating solely the subgroup of studies with lower risk of bias, there was a significant 1.6-fold increase of CVD associated with XP (Figure 1).

None of the studies reported the timeline between XP onset and the occurrence of cardiovascular events, so it is not possible to determine the timing that separated them. For instance, in the 35 years of cohort studies,^4,5 the previous detection of XP was compared with cardiovascular events; however, the retrospective studies, due to the possible presence of a recall bias are not robust to determine the time between XP and the occurrence of cardiovascular events.

Xanthelasma palpebrum onset and CVD seem to share similar pathophysiological mechanisms; therefore, having both clinical entities might be plausible.^1,9 Furthermore, there is a substantial overlap in the risk factors for both stroke and coronary artery disease and XP,^1,9 so the absence of a significant association between XP and cardiovascular risk was not expected. Nevertheless, XP is not always associated with abnormal lipid profiles as approximately 50% have a normal lipid profile.¹⁰ On the contrary, their presence being expected (but not diagnostic) for genetic conditions such as familial hypercholesterolaemia which put people at least at high cardiovascular risk.

Some limitations of these studies may explain part of the results obtained. A first limitation is that it was not possible to perform in-depth analyses based on patient characteristics, such as sex and ethnicity, because of limited data available from most of the original studies. The prevalence of different diseases, as well as the association with increased risk of CVD, may differ among different populations and according to individual characteristics.

Secondly, pooling data of studies with different designs, which evaluated different populations in different settings, increase the power and external validity of the data obtained, but also can be considered a potential limitation as well. It is also worth noting that some included studies were designed prior to the 21st century,^3–5 and not all used database codes to select cases and to collect outcome data, lacking important standardization and clarity regarding the information collected.

Jee et al.’s⁷ case–control study reveals results that were surprising and distinctive from the other studies concerning CVD (higher in the control group). However, this study involves serious selection bias, considering the pre-selection of patients for the control group. According to the authors, patients who had their serum lipid levels measured were preferentially selected, referring that more patients with medical diseases were included, and it is possible that the frequency of CVD in the control group was increased.

Despite efforts to derive less biased estimates, it is essential to note that most of the studies did not adjust for different cardiovascular risk factors, which precludes better understanding of the data.

This systematic review postulates a hypothesis regarding whether XP leads to increased risk of cardiovascular events, an issue requiring further clarification. Thus, further comprehensive cohort studies and detailed data about co-variables in CVD and markers of cardiovascular events (conventional and new) are required to clarify the true magnitude of this risk. The current data, although indicative, are insufficient to enable XP to be used as a strong indicator of CVD in daily practice, because clear evidence has not yet been unequivocally established.

Thus, in conclusion, our pooled analysis does not support an increased risk of cardiovascular events in patients with XP. However, in the subgroup of studies with lower risk of bias a 1.6-fold increased risk of CVD was verified (see Supplementary material online, Data 5). The low number of studies included, and the moderate/serious risk of bias observed preclude more robust conclusions.

Authors’ contributions

D.C. contributed to the conception and design of the work. R.L. and M.A. contributed to the acquisition, analysis, or interpretation of data for the work. R.L. and M.A. drafted the manuscript. C.M., D.B., A.G.A., F.J.P., and D.C. critically revised the manuscript. All gave final approval and agree to be accountable for all aspects of work ensuring integrity and accuracy.

Supplementary material

Supplementary material is available at European Journal of Preventive Cardiology online.

Funding

No funding was received.

Consent

Approval and consent statements were not required since this was a systematic review of the available literature.

Data Availability

The data underlying this article will be shared on reasonable request to the corresponding author.

Author notes

Conflict of interest: D.C. has participated in educational meetings and/or attended a conferences or symposia (including travel, accommodation, and/or hospitality) with Bristol-Myers Squibb, Bayer, Boehringer Ingelheim, Daiichi Sankyo, Merck Serono, Ferrer, Pfizer, Novartis, and Roche. M.A. reported participation in conferences with Boehringer Ingelheim, AstraZeneca, Bayer, Bristol-Myers Squibb, Grünenthal, Tecnimede, and Merck Sharp & Dohme. The remaining authors do not have any conflict of interests to report.