[Show abstract][Hide abstract] ABSTRACT: We developed, implemented, and evaluated a myocardial infarction (MI) adjudication protocol for cohort research of human immunodeficiency virus. Potential events were identified through the centralized Centers for AIDS Research Network of Integrated Clinical Systems data repository using MI diagnoses and/or cardiac enzyme laboratory results (1995-2012). Sites assembled de-identified packets, including physician notes and results from electrocardiograms, procedures, and laboratory tests. Information pertaining to the specific antiretroviral medications used was redacted for blinded review. Two experts reviewed each packet, and a third review was conducted if discrepancies occurred. Reviewers categorized probable/definite MIs as primary or secondary and identified secondary causes of MIs. The positive predictive value and sensitivity for each identification/ascertainment method were calculated. Of the 1,119 potential events that were adjudicated, 294 (26%) were definite/probable MIs. Almost as many secondary (48%) as primary (52%) MIs occurred, often as the result of sepsis or cocaine use. Of the patients with adjudicated definite/probable MIs, 78% had elevated troponin concentrations (positive predictive value = 57%, 95% confidence interval: 52, 62); however, only 44% had clinical diagnoses of MI (positive predictive value = 45%, 95% confidence interval: 39, 51). We found that central adjudication is crucial and that clinical diagnoses alone are insufficient for ascertainment of MI. Over half of the events ultimately determined to be MIs were not identified by clinical diagnoses. Adjudication protocols used in traditional cardiovascular disease cohorts facilitate cross-cohort comparisons but do not address issues such as identifying secondary MIs that may be common in persons with human immunodeficiency virus.
American journal of epidemiology 03/2014; 179(8). DOI:10.1093/aje/kwu010 · 5.23 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Background: It is unclear how well HIV-specific and conventional CVD risk prediction models stratify risk of events for patients with HIV at high risk for myocardial infarction (MI).
Methods: CNICS is an 8-site HIV cohort with >27,000 patients, a centralized data repository with comprehensive clinical data such as lipids, and centralized MI adjudication facilitating comparisons using a definitive, well-defined outcome. We formed a convenience sample using 731 patients who required MI adjudication based on diagnoses or elevated cardiac enzymes. This group represents a high-risk population that may be challenging for risk scores and thus illuminate potential challenges in prediction of high-risk patients. Of the 731 patients, 200 had an adjudicated MI. Using risk status prior to event dates, we compared 4 scores (estimated by Cox models) using area under the curves (AUCs): Framingham NCEP III (FRS), Framingham refit (using HIV-specific calibration), Framingham + (Framingham refit + HIV factors predictive in CNICS: time on indinavir, tenofovir use, current CD4 count, ex-smoker), and the Data Collection on Adverse events of Anti-HIV Drugs (DAD) score (HIV-specific).
Results: DAD and FRS did not have significantly different AUCs for risk discrimination in this key population. Using re-calibration (Framingham refit) resulted in a higher AUC (0.604 vs. 0.538, p=0.03) compared with the FRS. However adding additional HIV-specific risk factors (Framingham refit + HIV) had no influence on the AUC (also 0.604).
Framingham refit + HIV
Conclusion: We found adding HIV-specific risk factors did not substantially improve a FRS refit model in this difficult to stratify population. This high-risk population results in lower AUC because these individuals are hard to stratify in advance using baseline risk factors. However, performance in these populations does illustrate how robust these scores are in homogenous sub-populations. Future studies will focus on primary MIs rather than all adjudicated MIs and extend our risk score validation in the entire cohort.
IDWeek 2013 Meeting of the Infectious Diseases Society of America; 10/2013
[Show abstract][Hide abstract] ABSTRACT: Multiple studies have demonstrated an age-related attenuation in risk associations of lipoproteins and lipoprotein ratios with cardiovascular disease events. We recently reported a similar age-related attenuation in risk associations of lipoproteins and lipoprotein ratios with coronary artery calcium. We assessed risk associations of lipoproteins and lipoprotein ratios with carotid intima-media thickness (CIMT), which has not been reported previously. We performed multivariable linear regression using data from the Multi-Ethnic Study of Atherosclerosis (MESA). MESA participants were community-dwelling adults 45 to 84 years of age without clinically apparent cardiovascular disease at baseline, and 4,961 met inclusion criteria for these analyses. In fully adjusted models, differences in CIMT were similar across the MESA age spectrum, with differences in internal CIMT per SD increase in low-density lipoprotein of 0.037 mm (95% confidence interval 0.018 to 0.055) for those 45 to 54 years old and 0.087 mm (95% confidence interval 0.027 to 0.146) for those 75 to 84 years old (p for interaction = 0.2). Similarly, the difference in internal CIMT per SD increase in the total/high-density lipoprotein cholesterol ratio was 0.029 mm (95% confidence interval 0.009 to 0.049) for those 45 to 54 years old and 0.101 mm (95% confidence interval 0.033, 0.169) for those 75 to 84 years old (p for interaction = 0.03). In general, risk associations of lipoproteins and lipoprotein ratios were associated with similar differences in CIMT across all age categories. In conclusion, abnormal lipoproteins and lipoprotein ratios in middle-aged and older patients are powerful risk factors for early atherosclerosis as manifested by an increased CIMT.
The American journal of cardiology 12/2011; 109(5):658-64. DOI:10.1016/j.amjcard.2011.10.021 · 3.28 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The rise in LDL with egg feeding in lean insulin-sensitive (LIS) participants is 2- and 3-fold greater than in lean insulin-resistant (LIR) and obese insulin-resistant (OIR) participants, respectively.
We determined whether differences in cholesterol absorption, synthesis, or both could be responsible for these differences by measuring plasma sterols as indexes of cholesterol absorption and endogenous synthesis.
Plasma sterols were measured by gas chromatography-mass spectrometry in a random subset of 34 LIS, 37 LIR, and 37 OIR participants defined by the insulin sensitivity index (S(I)) and by BMI criteria selected from a parent group of 197 participants. Cholestanol and plant sterols provide a measure of cholesterol absorption, and lathosterol provides a measure of cholesterol synthesis.
The mean (±SD) ratio of plasma total absorption biomarker sterols to cholesterol was 4.48 ± 1.74 in LIS, 3.25 ± 1.06 in LIR, and 2.82 ± 1.08 in OIR participants. After adjustment for age and sex, the relations of the absorption sterol-cholesterol ratios were as follows: LIS > OIR (P < 0.001), LIS > LIR (P < 0.001), and LIR > OIR (P = 0.11). Lathosterol-cholesterol ratios were 0.71 ± 0.32 in the LIS participants, 0.95 ± 0.47 in the LIR participants, and 1.29 ± 0.55 in the OIR participants. After adjustment for age and sex, the relations of lathosterol-cholesterol ratios were as follows: LIS < OIR (P < 0.001), LIS < LIR (P = 0.03), and LIR < OIR (P = 0.002). Total sterol concentrations were positively associated with S(I) and negatively associated with obesity, whereas lathosterol correlations were the opposite.
Cholesterol absorption was highest in the LIS participants, whereas cholesterol synthesis was highest in the LIR and OIR participants. Therapeutic diets for hyperlipidemia should emphasize low-cholesterol diets in LIS persons and weight loss to improve S(I) and to decrease cholesterol overproduction in LIR and OIR persons.
American Journal of Clinical Nutrition 09/2011; 94(5):1182-8. DOI:10.3945/ajcn.110.006668 · 6.77 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The purpose of this study was to determine the association of combinations of lipid parameters with subclinical atherosclerosis.
Carotid intima-media thickness (CIMT) and coronary artery calcium (CAC) are significantly associated with incident cardiovascular disease (CVD). The association between common dyslipidemias (combined hyperlipidemia, [simple] hypercholesterolemia, dyslipidemia of metabolic syndrome, isolated low high-density lipoprotein cholesterol, and isolated hypertriglyceridemia) compared with normolipemia, and CIMT and CAC has not been previously examined.
The MESA (Multi-Ethnic Study of Atherosclerosis) participants were White, Chinese, African-American, or Hispanic adults without clinical CVD. Subjects with diabetes mellitus or who were receiving lipid-lowering therapy were excluded. Every participant was classified into only 1 of 6 groups defined by specific low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, or triglyceride cut points. Multivariate linear and relative risk regressions evaluated the cross-sectional associations with CIMT and CAC after adjusting for CVD risk factors. Interactions with race, sex, and high-sensitivity C-reactive protein were evaluated for CIMT and CAC outcomes.
Among 4,792 participants, only those with combined hyperlipidemia and hypercholesterolemia demonstrated both increased common CIMT (combined hyperlipidemia 0.048 mm thicker, 95% confidence interval [CI]: 0.016 to 0.080 mm; hypercholesterolemia 0.048 mm thicker, 95% CI: 0.029 to 0.067 mm) and internal CIMT (combined hyperlipidemia 0.120 mm thicker, 95% CI: 0.032 to 0.208 mm; and hypercholesterolemia 0.161 mm thicker, 95% CI: 0.098 to 0.223 mm) as well as increased risk for prevalent CAC (combined hyperlipidemia relative risk: 1.22, 95% CI: 1.08 to 1.38; hypercholesterolemia relative risk: 1.22, 95% CI: 1.11 to 1.34) compared with normolipemia. The interactions between lipid parameters and race, sex, or high-sensitivity C-reactive protein were not significant for any outcomes.
Combined hyperlipidemia and simple hypercholesterolemia were associated with increased CIMT and prevalent CAC in a relatively healthy multiethnic population.
Journal of the American College of Cardiology 09/2010; 56(13):1034-41. DOI:10.1016/j.jacc.2010.01.073 · 16.50 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Although abnormal lipoproteins and lipoprotein ratios are powerful risk factors for clinical cardiovascular events, these associations are stronger in younger than in older subjects. Whether age modifies the relation of lipoproteins and lipoprotein ratios to the relative risk of subclinical cardiovascular disease (CVD), as assessed by coronary artery calcium (CAC) scores, has not been examined in a contemporary, multiethnic cohort. We performed multivariate relative risk regression analyses to determine the relative risks for associations of lipoproteins and lipoprotein ratios with prevalent CAC in participants in Multi-Ethnic Study of Atherosclerosis (MESA). The participants were community-dwelling adults aged 45 to 84 years without clinically apparent CVD at baseline. We excluded those taking lipid-lowering therapy (15%) and stratified the results by decades of age. A total of 5,092 participants met the inclusion criteria. In the fully adjusted models, per SD of low-density lipoprotein, the age-stratified, adjusted relative risk for CAC was 1.17 (95% confidence interval [CI] 1.07 to 1.28) for those aged 45 to 84 years but was 1.05 (95% CI 1.01 to 1.10) for those aged 75 to 84 years (p-interaction = 0.12). The relative risk per SD of total/high-density lipoprotein cholesterol ratio was 1.20 (95% CI 1.12 to 1.29) for those aged 45 to 54 years but only 1.04 (95% CI 1.00 to 1.09) for those aged 75 to 84 years (p-interaction <0.001). The lipoproteins levels and lipoprotein ratios were associated with increased relative risks for CAC across all age categories. However, these associations were markedly attenuated by age. In conclusion, abnormal lipoprotein levels in middle age are a powerful risk factor for early atherosclerosis, as manifested by prevalent CAC.
The American journal of cardiology 02/2010; 105(3):352-8. DOI:10.1016/j.amjcard.2009.09.040 · 3.28 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: It is well known that cardiovascular disease is the number one killer of men and women in the United States and in many parts of the developed world. However, early detection of atherosclerosis remains a challenging area of research and development. Stress echo and myocardial perfusion studies were not designed to be screening tests and the majority of literature using these tests is in populations with a high probability of disease. It must be emphasized that negative stress echo and stress MPI tests only imply a lack of flow limiting disease; they do not indicate lack of atherosclerotic disease. It is important to remember that when these tests are "negative," the implication is favorable short-term prognosis rather than any implication regarding lack of disease. In contrast, carotid intima-media thickness (CIMT) scanning protocols can detect atherosclerotic disease in early and asymptomatic stages. For a number of reasons reviewed in this article, CIMT may be a more optimal screening and risk-stratifying technology: CIMT directly visualizes vasculature unlike biomarkers such as LDL cholesterol, hsCRP, or PLA2.
We performed medline searches for original articles and reviews of carotid IMT from 1985 to the present. We particularly emphasized large multi-center epidemiologic studies of the natural history of patients with carotid IMT measurements.
There is substantial evidence that CIMT is a suitable surrogate for the coronary tree. CIMT is also (along with coronary calcium scoring) recognized by the American Heart Association as a surrogate marker for coronary artery disease. A recent commentary by Stein, et al reviewed the comparison of CIMT to coronary calcium scoring, with favorable findings for CIMT especially in the healthy young and middle-aged populations, as well as women and African American individuals where coronary calcification has more limited utility. Recent findings of the Multi-Ethnic Study of Atherosclerosis indicate further that increased CIMT predicted CVD events in individuals without coronary calcification.
[Show abstract][Hide abstract] ABSTRACT: Dyslipidemia in gestational diabetes (GDM) consists of an ∼50 mg/dL increase in triglyceride, an ∼4 mg/dL decrease in high
density lipoprotein (HDL), and generally lower low density lipoprotein (LDL) levels by ∼30mg/dL, more small dense LDL, and
greater susceptibility of LDL to oxidation. Predictors of increased birth weight are the postprandial hyperglycemia of GDM
and elevated triglyceride. Predictors of diminished birth weight are lower HDL, lower apo-A-I, higher apo A-II and increased
oxidative stress, typified in pre-eclampsia. Birth weight in an individual GDM pregnancy is likely a function of these competing
trends. A therapeutic approach to the management of GDM is needed for the hyperglycemia and hypertriglyceridemia, but also
for the heightened oxidative stress which may be propagated by abnormalities of LDL and HDL function.
[Show abstract][Hide abstract] ABSTRACT: BACKGROUND: The combination of niacin and statin has proven value in hyperlipidemia management and heart disease prevention. However, the efficacy of the non-prescription time-release niacin, Slo-Niacin®, is little studied alone and not at all with atorvastatin. We gave Slo-Niacin® and atorvastatin, singly and together to determine efficacy on the combined abnormalities of triglyceride, LDL and HDL. METHODS: 42 men and women with LDL-C>130mg/dL HDL-C <45 (men or 55mg/dL (women) were randomized to 3 months of atorvastatin 10 mg/day or incremental doses of Slo-Niacin® to 1500 mg/day. The alternate drug was added in the next 3-month segment. Lipid profiles and transaminases were measured monthly and other measures at baseline and the end of each treatment sequence. RESULTS: Mean entry lipids (mg/dL) were: TG 187, LDL-C 171, and HDL-C 39. Mean BMI was 32.6 Kg/m(2). Monotherapy with Slo-Niacin® decreased median triglyceride 15%, mean LDL-C 12% and non-HDL-C 15% and increased HDL-C 8%. Atorvastatin decreased median triglyceride 26%, and mean LDL-C 36%, non-HDL-C 36% and increased HDL-C 6%. Combined therapy decreased median triglyceride 33% and mean LDL-C and non-HDL-C each 43%. HDL-C increased 10% (all p<0.001). Median remnant-like lipoprotein-C decreased 55%, mean apo-B 40%, median hsCRP 23% (all p<0.05), TNFa 12% and no change in IL-6. Mean LDL buoyancy increased 15%, apo-A-I 5% and median HDL(2)-C 20% (all p<0.05). ALT declined with Slo-Niacin® treatment alone compared to atorvastatin and also decreased when Slo-Niacin® was added to atorvastatin. Six subjects dropped out, 3 for niacin related symptoms. CONCLUSIONS: Slo-Niacin® 1.5g/day with atorvastatin 10 mg/day improved lipoprotein lipids, apoproteins and inflammation markers without hepatotoxicity. Slo-Niacin® deserves further study as a cost-effective treatment of hyperlipidemia.
[Show abstract][Hide abstract] ABSTRACT: Along with the rising prevalence of obesity, rates of gestational diabetes mellitus (GDM) and associated adverse outcomes also have increased. We conducted a population-based, retrospective cohort study to assess the association of weight gain between pregnancies with cesarean delivery for the subsequent pregnancy among women with a history of GDM.
Using linked birth-certificate data for women with at least two singleton births in Washington State during the period from 1992-2005, we identified 2,753 women with GDM who delivered vaginally at the baseline pregnancy (first pregnancy on record). The interpregnancy weight change (subsequent-baseline prepregnancy weight) for each woman was calculated and assigned to one of three categories: weight loss (more than 10 lb), weight stable (+/-10 lb), or weight gain (more than 10 lb). Multiple logistic regression was used to calculate the risk (odds ratio [OR]) of cesarean delivery at the subsequent pregnancy among the weight-gain and weight-loss groups relative to the weight-stable category.
Among 2,581 eligible women, 10.9% lost more than 10 lb between pregnancies, 54.0% were weight-stable, and 35.1% gained more than 10 lb. Women who gained more than 10 lb had an adjusted OR for subsequent cesarean delivery of 1.70 (95% confidence interval [CI] 1.16-2.49, 9.7% of women who gained weight), whereas the adjusted OR for women who lost weight was 0.55 (95% CI 0.28-1.10, 4.7% of women who lost weight).
Women with a history of GDM who gained more than 10 lb between pregnancies are at increased risk of future cesarean delivery. Appropriate weight management among women with a history of GDM may result in decreased cesarean delivery rates along with decreases in associated excess risks and costs.
Obstetrics and Gynecology 05/2009; 113(4):817-23. DOI:10.1097/AOG.0b013e31819b33ac · 5.18 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: We have asked whether the prevalence of combined hyperlipidemia (CHL) differs by race/ethnicity, obesity, and insulin resistance in a contemporary, multiethnic, US cohort. We determined the prevalence and adjusted odds of CHL in a cohort of 5923 men and women free of clinically recognized cardiovascular disease and diabetes according to race/ethnicity (white, Chinese, African American, and Hispanic), obesity, and insulin resistance. Untreated lipid values were imputed for those on lipid-lowering therapy. Combined hyperlipidemia was defined using age- and sex-specific greater than or equal to 75th percentile cut points for low-density lipoprotein cholesterol and triglycerides obtained from a predominantly white North American population study. Compared with whites, adjusted odds ratios for CHL were 0.48 in African Americans (95% confidence interval [CI], 0.30-0.75), 1.33 in Hispanics (95% CI, 0.93-1.91), and 1.06 in Asians (95% CI, 0.62-1.82). Within the entire population, the adjusted odds of CHL were over 2-fold higher in overweight and obese participants compared with normal-weight participants and more than 4-fold higher in quartiles 2 through 4 of insulin resistance compared with quartile 1. African Americans had lower odds for CHL than whites despite higher body mass index and abdominal adiposity. Hispanics had a nonsignificantly higher trend, and Asians had no significantly different odds than whites. Modest increases in weight and insulin resistance were associated with significantly higher odds of CHL in a multiethnic US population. Further research is needed to determine the most efficacious diet, exercise, and drug management to decrease the risk of CHL and coronary heart disease among racial/ethnic groups in the United States.
[Show abstract][Hide abstract] ABSTRACT: Five lines of evidence justify comprehensive lipoprotein management over aggressive low-density lipoprotein (LDL) lowering alone in most cases of cardiovascular disease (CVD) prevention. First, lipoprotein lipid transport consists of a single, recycling system involving very-low-density lipoprotein, LDL, and high-density lipoprotein (HDL). Single lipid interventions affect all lipoprotein classes to varying degrees. These effects can be expanded by using different drug classes in combination. Second, observational studies support the unitary nature of lipoprotein risk. A family of curves describes increasing CVD risk from increasing LDL as other risk factors are present. Conversely, a family of curves describes increasing CVD risk from decreasing levels of HDL in mirror image to LDL. The LDL and HDL risks are additive. Third, clinical trials that raise HDL and lower triglyceride ameliorate CVD, as does lowering LDL. Lowering LDL prevents heart disease, but by only 22%-36% with 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor therapy. Studies indicate that better CVD prevention is obtained when drugs for triglyceride and HDL reduction are combined with LDL reduction. Fourth, HDL and its apolipoprotein (apo), apo A-I, as well as apo A-I analogues, decrease atherosclerosis. Each modality decreases atherosclerosis in animal models, and apo A-I Milano acutely decreases human coronary luminal stenosis. Apo A-I analogues have similar promise. Fifth, combined hyperlipidemia is the most common lipid disorder, has the strongest risk for CVD, and combines elevated LDL, hypertriglyceridemia, and low HDL. This condition requires the comprehensive treatment approach described above. In conclusion, 5 lines of evidence justify comprehensive diet and drug treatment for combined hyperlipidemia and, at lesser LDL elevations, the atherogenic dyslipidemias of obesity, diabetes mellitus, and the metabolic syndrome.
The American Journal of Cardiology 05/2008; 101(8A):48B-57B. DOI:10.1016/j.amjcard.2008.02.038 · 3.28 Impact Factor