Prevalence of Subclinical Atherosclerosis and Risk of Atherosclerotic Cardiovascular Disease in Older Adults Living with HIV

Abstract

Cardiovascular disease (CVD) is one among the leading causes of mortality in people living with HIV on antiretroviral treatment (ART) worldwide. We examined the prevalence of subclinical atherosclerosis, associated factors, and risk of CVD in older adults living with HIV (OALHIV). A cross-sectional study was conducted with patients aged ≥50 years with HIV infection receiving ART at community hospitals in Chiang Mai, Thailand. Age- and sex-matched patients without documented HIV infection who visited the general outpatient department were enrolled for comparison. Cardio-ankle vascular index (CAVI) and ankle-brachial index (ABI) were measured using the vascular screening system, VaSera System™ (Fukuda Denshi Co., Ltd., Japan) to determine subclinical atherosclerosis (defined as CAVI ≥9.0) and peripheral arterial disease (defined as ABI <0.9), respectively. The Ramathibodi-Electric Generating Authority of Thailand (RAMA-EGAT) scores to predict the risk of coronary stenosis and the 10-year risk of ASCVD by pooled cohort equation were calculated. One hundred fifty-five patients were enrolled (107 HIV/48 comparison). The mean age was 59.0 years (SD 6.1); 67 (43%) were male. Participants in the HIV and comparison group were similar with respect to abnormal CAVI (57% vs. 58%, p = .88), abnormal ABI (6% vs. 8%, p = .50), and the risk of coronary stenosis (34% vs. 44%, p = .23). However, the 10-year risk of ASCVD was lower in HIV than in the comparison group (29% vs. 48%, p = .02). In OALHIV, diabetes mellitus was the only factor associated with abnormal CAVI. The cardiovascular risk among OALHIV in this study was similar to non-HIV population. More than a half of them had abnormal CAVI, and approximately one-third was at ≥10% 10-year risk of ASCVD.

Introduction

One among the leading causes of mortality among people living with HIV (PLHIV) in antiretroviral treatment (ART) era is cardiovascular disease (CVD).^1,2 Pooled data from Europe and North America identified the non-AIDS causes especially cardiovascular events as important contributors to mortality in PLHIV receiving ART.³ In fact, there is evidence showing that untreated HIV-infected population would undergo a rapid vascular aging and early onset of atherosclerosis, ultimately leading to myocardial infarction.⁴ In addition to the conventional cardiovascular risk factors, multiple other factors place PLHIV at higher risk of CVD than general population. These include, for example, immune activation, chronic inflammation, HIV infection itself, and ART-related metabolic disorders.⁵

Prevention of morbidity and mortality due to CVD largely depends on early identification of individuals with high risk of CVD and timely initiation of appropriate care. When most patients remain asymptomatic, monitoring tools or measures to predict the risk could be very helpful. Currently, there are several noninvasive methods to measure arterial stiffness. Cardio-ankle vascular index (CAVI), a recently developed marker of arterial stiffness, which reflects atherosclerosis, has demonstrated a strong correlation with the presence of coronary artery disease (CAD). CAVI has the advantage over other measurements of arterial stiffness, including ankle-brachial index (ABI), for its independence on blood pressure.⁶ A previous study on non-HIV-infected Thai population has shown that high CAVI is an independent risk predictor for CAD, and is associated with excess mortality.⁷ ABI is another noninvasive marker for systemic arterial stiffness. Low ABI has been associated with increased risk of cardiovascular-related mortality.⁸

In Thailand, AIDS-related mortality tremendously decreased after the implementation of the national ART program since the year 2002.⁹ This improved survival has led to an increasing pool of HIV-infected individuals living beyond 50 years of age. We perceived that the prevalence of cardiovascular risk could vary from region by region as a result of difference in genetics, diet, lifestyle, and various health and environmental factors.

This study was conducted in Thailand among older adults living with HIV (OALHIV), determined as patients aged ≥50 years, who were stable on ART in hope that they could be proper representatives of similar population in other Asian countries. A comparison group composed of age- and sex-matched hospital clients who lived in the same geographic area but had no diagnosis of HIV. We aimed to document the prevalence of subclinical atherosclerosis in OALHIV and its associated factors. In addition, we assessed cardiovascular risk in this population.

Materials and Methods

Study design and participants

This cross-sectional study was conducted from August to October 2015. Study participants were recruited from HIV clinics in 12 community hospitals in Chiang Mai, Thailand. They were approached and invited to join the study during routine clinical visits on a first-come first-served basis. The inclusion criteria were (1) having a diagnosis of HIV infection, (2) aged ≥50 years, (3) receiving ART at study enrollment, and (4) willing to join the study. For each hospital, the oldest patients were approached first. For the comparison group, we selected the age- and sex-matched hospital clients visiting the general outpatient department of the same hospital on the same day who had lipid tests performed within the past 12 months and had no documented HIV infection. All the recruiting sites were primary care hospitals that had no specialty clinics.

Data collection, measurements, and definitions

Study participants underwent a face-to-face interview to obtain smoking history, comorbidities (hypertension, diabetes mellitus [DM], dyslipidemia), and treatment. HIV-related medical history and laboratory test results (lipid profile) were retrieved from the hospital records. Cardiovascular assessments including blood pressure and anthropometric measurements were performed. Body mass index (BMI) was calculated by dividing the body weight (in kilograms) by height (in meters) squared (BMI = weight/height²). Underweight and overweight were defined as having BMI <18.5 and >25 kg/m², respectively.¹⁰ The cutoff values for waist circumference used in this study were >80, and >90 cm for female and male, respectively.

CAVI was assessed using the vascular screening system, VaSera System™ (Fukuda Denshi Co., Ltd., Japan). CAVI values <8.0, 8.0 to <9.0, and ≥9.0 were defined as normal, borderline, and subclinical atherosclerosis, respectively. For the purpose of this study, we designated CAVI of ≥8.0 as abnormal and used it as the positive outcome. This cutoff value was shown to be optimal for predicting coronary arterial disease in Thais, with 92% sensitivity, 63% specificity, and 70% accuracy.¹¹ The ABI was measured with the same machine. The ABI values <0.9 and >1.4 were defined as peripheral arterial disease and arterial stiffness, respectively. Both were considered as abnormal ABI, and have been associated with risk of all-cause and CVD mortality in the population-based setting.⁸

The RAMA-Electricity Generating Authority of Thailand (RAMA–EGAT) score was calculated based on data from the Electricity Generating Authority of Thailand staff every 5 years for a period of 30 years. The variables determined as risk factors used in calculation were age (−2: 35–39 years, 0: 40–44 years, 2: 45–49 years, 4: 50–54 years, 6: 55–59 years, 8: 60–65 years, 10: ≥ 65 years), sex (0: female, 3: male), cholesterol (0: < 280 mg/dL, 5: > 280 mg/dL or drug therapy), smoking (0: no, 2: yes), diabetes (0: no, 5: yes), hypertension (0: no, 3: yes), and waist circumference (0: below, 3: above the cutoff ≥36 inches in male and ≥32 inches in female). The summary of scores ranges from −2 to maximum 26. It has been shown to be a good predictor of cardiovascular events in Thai population.¹² Patients with score ≥17 were considered as high-risk population, according to its association with 45.7% prevalence of significant coronary stenosis.

Further, the 10-year risks for ASCVD by pooled cohort equation (PCE) were calculated for each study participant. PCE is a novel risk calculator based on data pooled from several large-scale racially and geographically diverse cohort studies developed by the American College of Cardiology/American Heart Association Task Force. The calculator uses variables easily identified in routine care to estimate the risk of CVD outcomes, including age, sex, race, systolic blood pressure, total cholesterol and HDL, hypertension, DM, and current smoking status.¹³ In this study, we defined participants with 10-year ASCVD risk of ≥10% as a high-risk group.

Ethics statement

The study was approved by the institutional review board at Research Institute for Health Sciences, Chiang Mai University (Certificate approval number 52/2014). Written informed consent was obtained from each participant before enrollment.

Statistical analysis

Statistical analysis was performed using the SPSS version 22.0 (IBM Corporation, Armonk, New York). Univariate analysis was conducted to obtain descriptive statistics, including percentage, range, mean [standard deviation (SD)], median [interquartile range (IQR)] of the selected variables. We compared OALHIV and the age- and gender-matched comparison group of non-infected older adults using Mann–Whitney U test, Fisher's exact test, or chi square as appropriate. We performed univariate and multivariable logistic regression to identify factors associated with abnormal CAVI (the surrogate subclinical of atherosclerosis) in the whole study sample and the subgroup of OALHIV. Only variables signified at p ≤ .2 were included in the adjusted models, and statistical significance was defined at p < .05.

Results

Characteristics of study participants

A total of 155 eligible patients were included; 107 participants in the HIV-infected and 48 in the comparison group (Table 1). The majority of participants were female (57%), and the mean age was 59.0 years (SD 6.1). The following parameters were significantly lower among participants in the HIV group than in the comparison group: mean BMI (20.9 vs. 24.6 kg/m², p < .01), systolic blood pressure (133 vs. 143 mmHg, p < .01), diastolic blood pressure (82 vs. 88 mmHg, p < .01), pulse pressure (50 vs. 55, p = .03), and waist circumference above cutoff for gender (28% vs. 60%, p < .01).

Table 1.

Characteristics and Cardiovascular Risk Estimations of Older Adults Living with HIV and Comparison Group

Characteristics	Older adults living with HIV (n = 107)	Comparison group (n = 48)	p-value
Demographic characteristics
Male sex	47 (44)	20 (42)	.79
Mean age (years)	58.7 (5.9)	59.7 (6.5)	.32
Mean body mass index (kg/m²)	20.9 (3.5)	24.6 (4.3)	<.01
Overweight (≥25 kg/m²)	14 (13)	22 (46)	<.01
Underweight (<18.5 kg/m²)	27 (25)	4 (8)	.02
Waist circumference (male)	80.4 (11.4)	89.3 (10.5)	<.01
Waist circumference (female)	75.9 (10.6)	84.3 (10.3)	<.01
Waist circumference > cutoff^a	30 (28)	29 (60)	<.01
Systolic blood pressure (mmHg)	133 (18)	143 (16)	<.01
Systolic blood pressure >130 mmHg	59 (55)	40 (83)	<.01
Diastolic blood pressure (mmHg)	82 (9)	88 (10)	<.01
Diastolic blood pressure >85 mmHg	43 (40)	34 (71)	<.01
Pulse pressure (mmHg)	50 (11)	55 (12)	.03
Pulse pressure >60 mmHg	16 (15)	15 (31)	.02
Comorbidities
Hypertension	48 (45)	30 (63)	.04
Diabetic mellitus	12 (11)	13 (27)	.01
Dyslipidemia	46 (43)	30 (63)	.03
Smoking	14 (13)	8 (17)	.56
HIV-related characteristics
Median duration of HIV diagnosis (years)	10 (7–13)	NA
Median duration on treatment (years)	9 (7–12)	NA
Treatment regimen
Nevirapine based	61 (57)	NA
Efavirenz based	26 (24)	NA
Protease inhibitors based	7 (7)	NA
Data missing	13 (12)	NA
Median CD4 lymphocyte count (cells/mm³)	456 (338–606)	NA
<200	9 (8)	NA
200–500	54 (50)	NA
>500	44 (41)	NA
HIV RNA <400 copies/mL	107 (100)	NA

Data reported in numeric (%), or mean (standard deviation) as appropriate; p-value from t-test, chi square, or Fisher's exact test as appropriate.

The arbitrary cutoff values for increased waist circumference are 80 cm in female and 90 cm in male.

NA, not applicable.

In addition, there were fewer participants of the following comorbidities in OALHIV than in the comparison group: hypertension (45% vs. 63%, p = .04), DM (11% vs. 27%, p = .01), and dyslipidemia (43% vs. 63%, p = .03). Less than a quarter of all participants reported smoking (13% in OALHIV and 17% in the comparison group).

The mean age of OALHIV was 58.7 years (SD 5.9). The median duration of HIV diagnosis was 10 years (IQR 7–13), and the median duration on ART was 9 years (IQR 7–12). The median CD4 lymphocyte count was 456 cells/mm³ (IQR 338–606); all had HIV RNA level <400 copies/mL at the last blood test performed within the 12 months before the time of study. HIV-related characteristics are shown in Table 1.

CAVI, ABI, atherosclerotic CVD risk

There was no difference in mean CAVI (8.26 vs. 8.35, respectively; p = .66) or the prevalence of abnormal CAVI between the OALHIV and the comparison groups (57% vs. 58%, respectively; p = .88) (Table 2). Peripheral arterial disease by ABI was detected in 6% of OALHIV versus 8% of participants in the comparison group, (p = .50). None of the participants in both groups met the criteria for arterial stiffness.

Table 2.

Prevalence of Subclinical Atherosclerosis and Atherosclerotic Cardiovascular Disease Risk Assessment of Older Adults Living with HIV and Comparison Group

Measurements	Older adults living with HIV (n = 107)	Comparison group (n = 48)	p-value
Cardio-ankle vascular index (CAVI)
Mean (SD)	8.26 (1.05)	8.35 (1.31)	.66
Abnormal CAVI
Borderline (≥8.0–<9.0)	36 (34)	14 (29)	.72
Subclinical atherosclerosis (≥9.0)	25 (23)	14 (29)
Ankle-brachial index (ABI)
Mean (SD)	1.06 (0.10)	1.03 (0.13)	.11
Abnormal ABI
Peripheral arterial disease (<0.9)	6 (6)	4 (8)	.50
Arterial stiffness (>1.4)	0	0	1.00
Risk assessments
RAMA-EGAT score	14.0 (9.0–18.0)	15.0 (12.0–19.0)	.04
Proportion with high risk of coronary stenosis (score ≥17)	36 (34%)	21 (44%)	.23
10-year risk of ASCVD	5.6 (3.0–12.1)	9.2 (4.5–17.1)	.02
Proportion with 10-year risk ≥10%	31 (29%)	23 (48%)	.02

Data reported in numeric (%), or mean (standard deviation) as appropriate; p-value from t-test, chi square, or Fisher's exact test as appropriate.

ASCVD, atherosclerotic cardiovascular disease; RAMA-EGAT score, Ramathibodi-Electric Generating Authority of Thailand; CAVI, cardio-ankle vascular index; ABI, ankle-brachial index.

The proportion of overall study participants with high risk of coronary stenosis by RAMA EGAT score and with 10-year risk of ASCVD ≥10% were 57/155 (37%) and 54/155 (35%), respectively. The median RAMA EGAT score was lower in OALHIV than in the comparison group (14 vs. 15, p = .04), as well as the median 10-year ASCVD risk (5.6% vs. 9.2%, p = .02) and the proportion with 10-year risk of ASCVD ≥10% (29% vs. 48%, p = .02). There was a positive correlation between RAMA EGAT and 10-year ASCVD risk by PCE among OALHIV (r = 0.64, p < .01).

When adjusted for comorbidities (hypertension, diabetes, dyslipidemia, and smoking), OALHIV had 1.583 times higher 10-year ASCVD risk than comparison group, but it was not statistically significant (95%CI 0.707–3.543, p = .264).

Factors associated with abnormal CAVI in all study participants

Univariate logistic regression analysis of the total study participants identified age (>60 years), increased systolic blood pressure (>130 mmHg), having a diagnosis of hypertension or on antihypertensive agents, and DM as factors associated with abnormal CAVI. In the multivariable model, age (>60 years) and DM remained significant (Table 3). HIV infection was not found to be a factor associated with higher risk of abnormal CAVI in this study (β = 0.95 [95%CI 0.48–1.89], p = .88).

Table 3.

Factors Associated with Atherosclerosis in All Study Participants (n = 155)

Variables	Univariate logistic regression			Multivariate logistic regression ^a
Variables	β	95% CI	p-value	β	95% CI	p-value
Male sex	1.10	0.88–1.36	.41
Age >60 years	2.69	1.35–5.38	<.01	3.00	1.46–6.14	<.01
Underweight (BMI <18.5 kg/m²)	1.73	0.75–3.97	.20
Overweight (BMI ≥25.0 kg/m²)	0.58	0.27–1.28	.18
Waist circumference > cutoff	0.93	0.71–1.20	.56
Increase in systolic blood pressure >130	1.80	0.93–3.50	.08	1.46	0.72–2.97	.30
Pulse pressure >60 mmHg	1.45	0.64–3.28	.37
Hypertension or on antihypertensive treatment	1.74	0.92–3.31	.09		Not included^b
Diabetic mellitus	4.79	1.56–14.72	<.01	1.39	1.10–1.75	<.01
Dyslipidemia or on lipid lowering agent	0.87	0.46–1.64	.66
Smoking	1.16	0.73–1.86	.53
HIV infection	0.95	0.48–1.89	.88

β, unstandardized regression coefficients; ART, antiretroviral therapy; CI, confidence interval; BMI, body mass index.

Variables with p-value <.1 in univariate were included in multivariate analysis.

Not included because of its collineation with systolic blood pressure.

Factors associated with abnormal CAVI in OALHIV

Factors associated with abnormal CAVI among OALHIV in the univariate logistic regression were increased systolic blood pressure (>130 mmHg) and DM. Only DM remained statistically significant in the adjusted model (Table 4).

Table 4.

Factors Associated with Atherosclerosis in HIV-Infected Older Adults (n = 107)

Variables	Univariate logistic regression			Multivariate logistic regression ^a
Variables	OR	95% CI	p-value	OR	95% CI	p-value
Male sex	1.01	0.78–1.31	.94
Age >60 years	1.44	0.65–3.20	.38
Underweight (BMI <18.5 kg/m²)	1.72	0.69–4.29	.24
Overweight (BMI ≥25.0 kg/m²)	0.73	0.20–2.70	.64
Waist circumference > cutoff	0.82	0.62–1.09	.18
Increase SBP >130	2.30	1.05–5.04	.04	2.08	0.93–4.66	.07
Pulse pressure >60 mmHg	1.80	0.58–5.61	.31
Duration on ART ≥10 years	1.40	0.65–3.03	.39
Current CD4 lymphocyte count <200 cells/mm³	1.56	0.37–6.61	.54
Hypertension or on antihypertensive treatment	0.57	0.26–1.24	.16
Diabetic mellitus	1.58	1.04–2.40	.03	1.54	1.01–2.35	.04
Dyslipidemia or on lipid lowering agent	0.89	0.41–1.92	.76
Smoking	0.85	0.48–1.49	.57

Variables with p-value <.1 in univariate were included in multivariate analysis.

OR, odds ratio.

Discussion

We found that more than a half of participants in this study had subclinical atherosclerosis (herein reflected by the abnormal CAVI), and one-third were at risk of CVD. The proportion with abnormal ABI, however, was low in this population. Notably, participants in the HIV and comparison groups were similar with respect to abnormal CAVI, abnormal ABI, and the risk of coronary stenosis (based on the RAMA-EGAT score ≥17). Nevertheless, the median 10-year risk of ASCVD and the proportion with 10-year risk ≥10% were significantly lower in OALHIV than in the comparison group. We observed OALHIV in this study had lower prevalence of comorbidities than comparison group. An explanation for the difference might be their health behaviors, which could have been modified after HIV diagnosis was made, following regular hospital visits, and health supervision in specialized clinic.

As known that CAVI is a more sensitive measure than ABI in screening for arterial change. The documented high proportion of CAVI and low proportion of ABI in OALHIV in our study are consistent with results from a previous study in Thai HIV-infected cohort (mean age 42.2 years) in Bangkok, which reported prevalence of abnormal CAVI and ABI of 37.96% and 0.46%, respectively.¹⁴ In the same study, being a current smoker, high systolic blood pressure, and total cholesterol >200 mg/dL were associated with increased odds of abnormal CAVI. In our study, DM was the only factor associated with increased odds of abnormal CAVI in OALHIV. The link between DM and abnormal CAVI has been documented,¹⁵ suggesting the need for increased awareness of atherosclerotic risk in HIV-infected older adults with DM. Noninvasive screening like CAVI can be offered, as a part of the routine primary prevention, and patients with abnormal CAVI should be referred for further investigation.

We found, however, that older age (>60 years) was associated with abnormal CAVI in total study participants but not in the subgroup analysis restricted to OALHIV. This could imply that abnormal CAVI in HIV-infected individuals can occur regardless of age. Thus, the finding supports accelerated atherosclerosis in OALHIV. However, the statistical power might have lost when considered only OALHIV due to the decrease in number of patients. Meanwhile, a similar proportion of participants in both groups had abnormal ABI. This was in contrast with a previous research in Uganda. The Ugandan study included both HIV-infected and noninfected individuals >40 years, and the prevalence of arterial stiffness (measured by ABI) was 33% and 9%, respectively, in HIV-infected male and female, which was more than twice that documented in the noninfected comparison group.¹⁶

As many as 34% of OALHIV in our study had high risk of cardiovascular events, as measured by RAMA-EGAT scores. This was much higher compared with the proportion reported in two previous studies done in Bangkok, which reported that 2%–11% of HIV-infected patients stable on ART had a high risk of cardiovascular events.^17,18 In both studies, participants were at younger age. Moreover, in the latter study, they used the lower cutoff (≥6). The Cambodian study reported that 24% of HIV-infected patients (mean age 40.9 years) treated with lopinavir-based ART had a 10-year risk of coronary disease ≥10%.¹⁹ Another study in Korea reported that the use of protease inhibitor was related to metabolic complication in HIV-infected participants at median age 46 years and median BMI 22.2 kg/m².²⁰ With comparable age and BMI, many of our study participants who were mostly on NNRTI-based regimen also had high ASCVD risk.

Despite not reaching statistical significance, we found a trend toward lower CAD estimated by RAMA-EGAT and a significantly lower ASCVD risk estimated by PCE in OALHIV compared with the non-HIV group. This could be explained by the difference in several demographic characteristics. In this study, we matched for age and sex, but did not exclude participants with other noncommunicable diseases, use other medications, or have other metabolic risks; that is, being overweight or more sedentary lifestyle. Meanwhile, OALHIV who have been placed on treatment and came to HIV clinic regularly were in good health, compared with other hospital clients. The ASCVD risk assessment by PCE has been used in several previous studies, including prospective cohort. Mainly, the incidence of CVD in HIV-infected patients was classified as low risk by PCE.²¹ As the PCE came from data in western countries where people were classified as Black versus non-Black, using this equation in Asian population might lead to either over or underestimated risk.

The interpretation of this study finding should be examined in the light of limitations. First, we cannot rule out the possibility that the observed differences and similarities between OALHIV and comparison group may be confounded by unmeasured or unknown factors. Even though the two groups of participants were matched by age and sex to enhance comparability, the nature of the source of participants in the comparison group can influence our results in unpredicted ways. Selection of patients who visit the cardiovascular or endocrinology specialty clinics could bias the results from possible underlying diseases or pre-existing cardiovascular risk. Therefore, we recruited clients who visited the general outpatient department only to minimize possible selection bias. In addition, all the study sites were primary care hospitals without specialty clinics. Although those in the comparison group might not best serve as a so-called healthy control, they represented community people who shared similar physical environment, food culture, and genetics. The inclusion criteria that required lipid testing results within the past year could also create bias. Individuals could have been tested for their other underlying conditions, illnesses, comorbidities, or just been tested as a part of checkup program. This is, however, the most feasible population we could enroll for comparison. At least we learned that risk of CVD in OALHIV who have been on and still required lifelong medication did not exceed age- and sex-matched people with other metabolic risks/diseases in the same community. The message might be useful for patient counseling. Second, we did not exclude individuals with other chronic diseases that might possibly cause vascular pathology. Third, the risk estimation relied on equations from different populations (the PCE), of which their characteristics might not match our cohort. Fourth, we did not further investigate or follow whether those with positive diagnostic screening actually had CAD or vascular change. Finally, the study population might be small. Further study on a larger population and extended study with follow-up period is warranted to learn more about cardiovascular risk in OALHIV.

There were no differences in prevalence of CAVI and ABI between OALHIV and comparison group. However, we observed high prevalence of subclinical atherosclerosis among older adults ≥50 years of age regardless of HIV status. One-third of them also had a high cardiovascular risk estimation. Application of this study results would be an implementation of cardiovascular screening program including available measures, as well as cardiovascular risk calculation, to ensure that further investigation in a timely manner for OALHIV regularly followed up in HIV clinic.

Footnotes

Acknowledgments

This study was part of the HIV/AIDS study among people living with HIV/AIDS in Chiang Mai. We thank all community hospital staff and HIV-infected older adults living with HIV/AIDS in rural Chiang Mai Province who contributed to this research.

Author Disclosure Statement

All authors have no conflict of interest to declare.

Funding Information

The study was supported by the Center of Excellence in HIV/AIDS Chiang Mai University Research and the National Research University Project under Thailand's Office of the Higher Education Commission.

References

Law

, Friis-Møller

, Weber

, et al.: Modelling the 3-year risk of myocardial infarction among participants in the data collection on adverse events of anti-HIV drugs (DAD) study. HIV Med, 2003; 4:1–10.

Trickey

, May

, Vehreschild

, et al.: Antiretroviral Therapy Cohort Collaboration (ART-CC). Cause-specific mortality in HIV-positive patients who survived ten years after starting antiretroviral therapy. PLoS One, 2016; 11:e0160460.

Antiretroviral Therapy Cohort Collaboration: Causes of death in HIV-1-infected patients treated with antiretroviral therapy, 1996–2006: Collaborative analysis of 13 HIV cohort studies. Clin Infect Dis, 2010; 50:1387–1396.

Baker

, Lundgren

: Cardiovascular implications from untreated human immunodeficiency virus infection. Eur Heart J, 2011; 32:945–951.

Boccara

, Lang

, Meuleman

, et al.: HIV and coronary heart disease: Time for a better understanding. J Am Coll Cardiol, 2013; 61:511–523.

Kotani

, Remaley

: Cardio-Ankle Vascular Index (CAVI) and its potential clinical implications for cardiovascular disease. Cardiol Pharmacol, 2013; 2:100108.

Yingchoncharoen

, Limpijankit

, Jongjirasiri

, Laothamatas

, Yamwong

, Sritara

: Arterial stiffness contributes to coronary artery disease risk prediction beyond the traditional risk score (RAMA-EGAT score). Heart Asia, 2012; 4:77–82.

Resnick

, Lindsay

, McDermott

, et al.: Relationship of high and low ankle brachial index to all-cause and cardiovascular disease mortality: The Strong Heart Study. Circulation, 2004; 109:733–739.

Chasombat

, Lertpiriyasuwat

, Thanprasertsuk

, Suebsaeng

, Lo

: The National Access to Antiretroviral Program for PHA (NAPHA) in Thailand. Southeast Asian J Trop Med Public Health, 2006; 37:704–715.

10.

WHO Expert Consultation: Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet, 2004; 363:157–163.

11.

Yingchoncharoen

, Sritara

: Cardio-ankle vascular index in a Thai population. Pulse (Basel), 2017; 4:8–10.

12.

Pattanaprichakul

, Jongjirasiri

, Yamwong

, Laothammatas

, Sritara

: RAMA-EGAT Risk score for predicting coronary artery disease evaluated by 64- slice CT angiography. Asean Heart J, 2007; 15:18–22.

13.

Goff

Jr. , Lloyd-Jones

, Bennett

, Coady

, D'Agostino

, Gibbons

, et al.: 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation, 2014; 129 Suppl 2:S49–S73.

14.

Pongnonthachai

, Sansanayudh

, Tasukon

, et al.: Prevalence of arterial stiffness measured by cardio-ankle vascular index in human immunodeficiency virus-infected patients (PATCH study). Thai Heart J, 2010; 23:1–12.

15.

Arworn

, Kosachunhanun

, Sony

, et al.: Underrecognition and undertreatment for peripheral arterial disease in diabetic patients in Thailand. J Med Assoc Thai, 2017; 100:149–157.

16.

Siedner

, Kim

, Nakku

, et al.: HIV infection and arterial stiffness among older-adults taking antiretroviral therapy in rural Uganda. AIDS, 2016; 30:667–670.

17.

Suppadungsuk

, Yamwong

, Sungkanuparph

: Risk of cardiovascular events predicted by the Rama-EGAT score among HIV-infected patients in Thailand. Southeast Asian J Trop Med Public Health, 2013; 44:818–824.

18.

Suppadungsuk

, Yamwong

, Sungkanuparph

: Risk of acute myocardial infarction predicted by RAMA-EGAT score among HIV-infected Thai patients. Int J Infect Dis, 2012; 16:e158–e316.

19.

Limsreng

, Marcy

, Ly

, et al.: Dyslipidemias and elevated cardiovascular risk on lopinavir-based antiretroviral therapy in Cambodia. PLoS One, 2016; 11:e0160306.

20.

, Ahn

, Kim

, et al.: Korea HIV/AIDS Cohort Study. Metabolic Complications among Korean Patients with HIV Infection: The Korea HIV/AIDS Cohort Study. J Korean Med Sci, 2017; 32:1268–1274.

21.

Thompson-Paul

, Lichtenstein

, Armon

, et al.: Cardiovascular disease risk prediction in the HIV Outpatient Study. Clin Infect Dis, 2016; 63:1508–1516.number