Abstract
Introduction
Medically underserved (US) populations have an increased level of atherosclerotic cardiovascular disease (ASCVD) risk, however, few studies investigated ASCVD risk reduction in US.
Methods
Of 217 subjects with ApoB ≥120 mg/dL and carotid atherosclerosis (≥15% stenosis by ultrasound) enrolled in the Carotid Plaque Composition by MRI (CPC) study between 2005 and 2011, US (n=33) was defined as those without adequate healthcare insurance, while AS (n=184) included those with adequate healthcare coverage. All subjects received atorvastatin-based lipid therapies and lifestyle intervention for 2 years. Metabolic and inflammatory risk factors were compared between AS and US.
Results
At baseline, compared to AS, US displayed higher levels of metabolic and inflammatory risk including systolic blood pressure (140±27 vs. 131±18 mmHg, p=0.04), fasting glucose (125±59 vs. 102±22 mg/dL, p=0.03) and fasting insulin (39±33 vs. 28±20 µU/dL, p=0.03) which resulted in higher insulin resistance (HOMA-IR 2.2±0.4 vs. 1.3±0.1, p=0.03), and hsCRP (5.6±1.5 vs. 2.8±0.2 mg/L, p=0.03). Over 2 years of intervention, US and AS showed similar reductions in LDL-C (-10.7% vs. -16% per year, p=0.2), triglycerides (-16.7% vs. -15.9% per year, p=0.4), and hsCRP (-0.11% vs. -0.04% per year, p=0.1). However, US continued to show significantly higher levels of fasting blood glucose (115±6.0 vs. 101±2.0 mg/dL, p=0.03) and HOMA-IR (1.9±0.2 vs. 1.5±0.1, p=0.047), and hsCRP (3.9±0.7 vs. 1.9±0.2 mg/L, p<0.001) than AS following 2 years of interventions.
Conclusions
US displayed higher ASCVD risk than AS at baseline and over 2 years despite similar reductions following the intervention. These findings highlight the unmet needs for improved intervention strategies and implementation methods for ASCVD risk reduction in US.
Clinical Trial Registration
NCT00715273 at ClinicalTrials.gov
Central Illustration
1. Introduction
Atherosclerotic cardiovascular disease (ASCVD) is remains the leading cause of death in the United States, despite decreases in ASCVD rates over the past 10 years.[1] It is suggested that ASCVD has shifted from a disease of the privileged to one of the disadvantaged.[2] An estimated 7.3 million individuals living with ASCVD in the US are uninsured.[3] In addition to well-established impact of lifestyle factors on ASCVD risk, socioeconomic status and adequate healthcare coverage independently influence ASCVD risk.[4] ASCVD poses a significant health burden as an estimated 45% of individuals with ASCVD report financial hardship due to medical care costs, with the uninsured and low income displaying the highest burden.[5] Furthermore, medically underserved populations are less likely to access treatment and prevention services,[6] because ASCVD risk factors usually show no physical symptoms and are largely undetectable without regularly administered diagnostic tests.
Despite the recognition of impact of the socioeconomic status on elevated ASCVD related morbidity, mortality and financial distress, few studies examine the implementation of ASCVD risk reduction and management strategies in medically underserved populations. In our study, we compared ASCVD-related risk reduction responses to two years of lipid and lifestyle therapies in medically underserved (US) populations compared to adequately served (AS) populations. Through this post-hoc analysis, we seek to inform future intervention methods aimed at reducing ASCVD risk equitably in individuals with varying healthcare coverage.
2. Methods
2.1. Study Population
To determine ASCVD-related metabolic and inflammatory risk and responses to lipid-lowering therapies in the US individuals comparing to their AS cohort, we analyzed data from the Carotid Plaque Composition by MRI (CPC) study,[7] a longitudinal study that monitored the change in carotid plaque morphology and composition during lipid therapy treatment. Study participants were recruited from both urban and rural regions at three locations: University of Washington in Seattle, WA, the Yakima Heart Center in Yakima, WA, and University of Southern California in Los Angeles, CA. The definition of rural and urban areas was based on 2010 RUCA codes.[8]
A total of 217 subjects with a mean age of 56 years and 60% of male were enrolled in CPC. Among these 217 study participants enrolled between 2005 and 2011, 184 were classified as AS, while 33 classified as US based on self-reported healthcare coverage. US individuals include those who reported having inadequate or a lack of healthcare insurance. AS individuals include those who reported having healthcare insurance with adequate and sufficient coverage. Specific study inclusion criteria included: 1) age <67 years for males and <70 years old for females; 2) family history of cardiovascular disease; 3) medically stable; 4) no contraindications to MRI; 5) angiographically confirmed coronary artery disease (defined as having ≥1 50% stenosis or ≥3 30% coronary lesions) or carotid disease (defined as having a ≥15% stenosis by ultrasound); 6) apolipoprotein (Apo) B ≥120 mg/dL; 7) not receiving lipid therapy >1 year prior to enrollment. All subjects received atorvastatin-based lipid therapy for 2 years and were randomized to 1 of 3 treatment groups: (1) single therapy - atorvastatin (10-80 mg per day) alone, placebos for extended release (ER) niacin and colesevelam or ezetimib; (2) double therapy – atorvastatin plus ER-niacin (2 g/day), and placebo for colesevelam or ezetimib; (3) triple therapy – atorvastatin, ER-niacin plus colesevelam (3.8 g/day) or ezetimibe (10 mg/day). The treatment target for LDL-C was ≤80 mg/dl for the single and double therapy groups and ≤60 mg/dl for the triple therapy group. The mean dose of atorvastatin received was 53 mg and 48 mg daily in the US and AS groups during the study. Additionally, subjects received dietary and lifestyle consultation using American Heart Association recommendations throughout the study period.
All study protocol and procedures received approvals of local Institutional Review Boards.
2.2. Clinical and Laboratory Measures
All subjects were followed every month for the initial 6 months of the study, and then every two months for the duration of the 2-year study period. Clinical data collected at all study visits included blood pressure, heart rate, weight, height, waist circumference, and BMI. Laboratory tests at baseline and during the study included: plasma lipids in mg/dL (total cholesterol, VLDL-C, IDL-C, LDL-C, HDL-C, triglycerides, ApoB, ApoA1, ApoE [mg/dL], and Lp(a) in nmol/L); fasting glucose levels in mg/dL, insulin in µU/dL and calculated HOMA-IR levels using a formula (HOMA-IR = fasting insulin µU/mL * fasting glucose (nmol/L) / 22.5); eGFR (mL/min/1.73m2] was calculated based on creatinine level, age, gender and ethnicity; and white blood cell count (WBC) in k/µL and hsCRP in mg/L.
2.3. Definition of Metabolic and Inflammatory Risk
Metabolic risk was defined as hyperglycemia with fasting glucose levels > or =100 mg/dL, dyslipidemia if high apoB required by the study plus HDL-C<40 mg/dL for men and <50 mg/dL for women, and increased insulin resistance if HOMA-IR >2.9. WBC and hsCRP were used to identify inflammatory risk. In addition, we used 5 common modifiable risk factors (BMI> or =25 Kg/m2, current smoker, type-2 diabetes mellitus, high blood pressure with SBP>140 mmHg or DBP>90 mmHg, and LDL-C>100 mg/dL) to reflect overall ASCVD risk level and room for reduction. We also collected data on treatment status for hypertension, hyperlipidemia and type-2 diabetes.
2.4. Statistical Analysis
Descriptive statistics are presented as mean ± standard error. Normality of outcomes was assessed by Q-Q plots for the mixed models. Demographic, clinical and laboratory variables were compared between US and AS using two sample t-test with a log scale transformation for unequal variances and Chi-square analysis.
Linear mixed models with random intercept and time slope were used to compare rate of changes in LDL-C, triglycerides, and C-reactive protein (hsCRP). The underserved indicator, time and their interaction were used in fixed effects modeling. Logistic mixed model with random intercept and time slope was used to compare rates of changes in the probability of 2+ risk factors. The underserved indicator, time and their interaction were used in fixed effects in modeling. Normal quantile-quantile plots of the random effects and residuals for the mixed model revealed substantial departures from normality for triglycerides and hsCRP outcomes. Therefore, the two outcomes were also analyzed on the log scale which led to less substantial departures from normality (and, thus, more valid confidence intervals and p-values). The model for LDL-C with a random effect for time had a singular fit with a 1.00 correlation between the intercept and the time slope random effects. The random effect for the time was therefore dropped. The model for 2+ risk factors with a random effect for time failed to converge. Therefore, the random effect for the time was dropped.[9]
All analyses were performed in 2019 using R. All p-values <0.05 were considered statistically significant.
3. Results
3.1. Baseline Characteristics
As described in Table 1, compared to AS, in the US group a significantly greater percentage of subjects were female (61% vs. 37%, p=0.01) and of Hispanic origin (36% vs. 8%, p<0.001). US displayed a significantly higher prevalence of diabetes (29% vs. 11%, p=0.004) and of >=3 modifiable risk factors (39% vs. 19%, p=0.009). Overall, 28% of subjects with diagnosis of hypertension, but, didn't receive any treatment. Untreated hyperlipidemia was seen in 73% and untreated diabetes was 16%. The treatment status for these 3 modifiable risk factors did not differ significantly between the US and AS groups. No statistical differences were observed in self-reported history of vascular disease including myocardial infarction, clinical diagnosed CAD, stroke or TIA.
Table 1.
Comparison in clinical, metabolic and inflammatory characteristics between US and AS groups at baseline.
| Total | US | AS | p-value | |
|---|---|---|---|---|
| Number of subjects | n=217 | n=33 | n=184 | |
| Age (years) | 56.3±0.54 | 56.3±0.61 | 56.7±1.07 | 0.359 |
| Male gender, n (%) | 128 (59%) | 13 (39%) | 115 (63%) | 0.013 |
| Hispanic/Latino, n (%) | 27 (12%) | 12 (36%) | 15 (8%) | <0.001 |
| Current Smoking, n (%) | 42 (19%) | 9 (27%) | 33 (18%) | 0.211 |
| BMI, kg/m2 | 29.8±5.9 | 30.6±4.6 | 29.7±6.0 | 0.139 |
| History of MI, n (%) | 82 (38%) | 15 (44%) | 67 (36%) | 0.324 |
| Established CAD, n (%) | 104 (48%) | 89 (48%) | 15 (45%) | 0.758 |
| History of stroke or TIA, n (%) | 13 (6%) | 3 (9%) | 10 (5%) | 0.415 |
| Hypertension, n (%) | 129 (59%) | 21 (62%) | 108 (59%) | 0.595 |
| Untreated hypertension, n (%) | 36 (28%) | 6 (29%) | 30 (28%) | 0.941 |
| Hyperlipidemia, n (%) | 188 (87%) | 27 (82%) | 161 (87%) | 0.377 |
| Untreated hyperlipidemia, n (%) | 138 (73%) | 16 (59.3%) | 122 (75.8%) | 0.072 |
| Diabetes, n (%) | 31 (14%) | 10 (29%) | 21 (11%) | 0.004 |
| Untreated diabetes, n (%) | 5 (16%) | 1 (10%) | 4 (19%) | 0.522 |
| With ≥3 modifiable risk factors | 48 (22%) | 13 (39%) | 35 (19%) | 0.009 |
| Systolic blood pressure, mmHg | 133±1.4 | 140±4.7 | 131±1.4 | 0.041 |
| Diastolic blood pressure, mmHg | 80±1.0 | 82±2.7 | 80±1.0 | 0.301 |
| Total cholesterol, mg/dL | 219±3.5 | 216±10 | 220±3.7 | 0.380 |
| VLDL-C, mg/dL | 31±2.1 | 41±8.0 | 29±2.0 | 0.092 |
| IDL-C, mg/dL | 19±0.9 | 22±3.3 | 18±0.8 | 0.130 |
| LDL-C, mg/dL | 145±3.1 | 134±7.1 | 147±3.4 | 0.059 |
| HDL-C, mg/dL | 43±0.8 | 42±2.7 | 43±0.9 | 0.332 |
| Triglycerides, mg/dL | 190±9.2 | 216±21.0 | 185±10.2 | 0.094 |
| ApoB, mg/dL | 120±2.1 | 118±5.3 | 121±2.2 | 0.302 |
| ApoA1, mg/dL | 135±1.6 | 134±5.2 | 135±1.7 | 0.411 |
| ApoE, mg/dL | 4.8±0.2 | 5.5±0.8 | 4.7±0.1 | 0.183 |
| Lp(a), nmol/L | 71±5.9 | 67±15 | 72±6.5 | 0.396 |
| Fasting glucose, mg/dL | 106±2.2 | 125±10.9 | 102±1.6 | 0.025 |
| Insulin level, µU/dL | 30±1.5 | 39±5.7 | 28±1.5 | 0.034 |
| HOMA-IR | 1.5±0.1 | 2.2±0.4 | 1.3±0.1 | 0.031 |
| eGFRb, mL/min/1.73m2 | 86±1.6 | 86±1.8 | 85±2.8 | 0.396 |
| HsCRP, mg/L | 3.2±0.3 | 5.6±1.5 | 2.8±0.2 | 0.032 |
| WBC, k/µL | 6.6±0.1 | 6.9±0.3 | 6.5±0.2 | 0.075 |
AdS, adequately served; ApoA1, apolipoprotein A-1; ApoB, apolipoprotein B; ApoE, apolipoprotein E; CAD, coronary artery disease; eGFR, estimated glomerular filtration rate; HDL-C, high density lipoprotein cholesterol; HOMA-IR, homeostatic model assessment of insulin resistance; HsCRP, high sensitivity C-reactive protein; IDL-C, intermediate density lipoprotein cholesterol; LDL-C, low density lipoprotein cholesterol; Lp(a), lipoprotein(a); MI, myocardial infarction; TIA, transient ischemic attack; UnS, underserved; VLDL-C, very low density lipoprotein cholesterol; WBC, white blood count.
At baseline, systolic blood pressure was on average 9 mmHg higher in the US compared to the AS group (140±27 vs. 131±18 mmgHg, p=0.04), with no differences in diastolic blood pressure between groups. Plasma lipids were not different between groups (Table 1). However, the US group showed significantly higher levels of fasting glucose (125±59 vs. 102±22 mg/dL, p=0.03) and fasting insulin (39±33 vs. 28±20 µU/dL, p=0.03), which resulted in a significantly higher HOMA-IR (2.2±0.42 vs. 1.3±0.09, p=0.03) (Fig. 1A) than the AS group. In addition, US had a significantly higher level of hsCRP (5.6±1.5 vs. 2.8±0.2 mg/L, p=0.03) (Fig. 1B) and non-statistically significant higher count of WBC (6.9±0.3 vs. 6.5±0.2 k/µL, p=0.07). The 2 groups showed a similar eGFR (Table 1).
Fig. 1.
A: Comparison of HOMA-IR between US and AS groups at baseline and 2 years of intensive lipid-lowering therapy. US showed significantly higher HOMA-IR than AS at baseline in mean±SE (2.2±0.42 vs. 1.3±0.09, p=0.03) and at 2 years of study therapy (1.9±0.23 vs. 1.5±0.12, p=0.047). **: p<0.05. B: Comparison of hsCRP levels between US and AS groups at baseline and 2 years of intensive lipid-lowering therapy. US displayed significantly higher levels of hsCRP than AS at both baseline in mean±SE (5.6±1.5 vs. 2.8±0.2 mg/L, p=0.03) and 2 years of lipid-lowering therapy (3.9±0.7 vs. 1.9±0.2 mg/L, p<0.001). **: p<0.05.
3.2. Metabolic and Inflammatory Risk Following Lipid-lowering Therapy
After 2 years of lipid therapy, plasma lipids were effectively decreased among all study participants and were not different between the US and AS groups (Table 2). However, US continued to display significantly higher levels of fasting blood glucose (115±6.0 vs. 101±2.0 mg/dL, p=0.03) and HOMA-IR (1.9±0.23 vs. 1.5±0.12, p=0.047) (Fig. 1A) than AS. Also, hsCRP levels continued to be statistically higher in US (3.9±0.7 vs. 1.9±0.2 mg/L, p<0.001) (Fig. 1B). Furthermore, the US group displayed a non-significant lower study completion rate compared to AS over 2 years (76% vs. 82%, p=0.4) and lower percentages of subjects stayed on double or triple combination lipid therapies (Table 2).
Table 2.
Comparison in metabolic risk factors between US and AS groups following two-years of lipid-lowering therapy.
| Total | US | AS | p-value | |
|---|---|---|---|---|
| Completion, n (%) | 175 (81%) | 25 (76%) | 150 (82%) | 0.4 |
| Treated with double therapy, n (%) | 59 (34%) | 6 (24%) | 53 (35%) | 0.267 |
| Treated with triple therapy, n (%) | 58 (33%) | 7 (28%) | 51 (34%) | 0.555 |
| BMI, Kg/m2 | 29.9±0.4 | 30.8±1.2 | 29.8±0.4 | 0.226 |
| Systolic Blood pressure, mmHg | 126±1.32 | 129±4.60 | 126±1.38 | 0.229 |
| Diastolic Blood pressure, mmHg | 78±0.8 | 80±2.9 | 78±0.8 | 0.321 |
| Total cholesterol, mg/dL | 154±3.8 | 155±4.2 | 155±7.3 | 0.456 |
| Triglycerides, mg/dL | 127±10.4 | 145±14.3 | 125±11.6 | 0.292 |
| LDL-C, mg/dL | 86±1.9 | 89±6.9 | 85±1.9 | 0.292 |
| HDL-C, mg/dL | 51±1.3 | 47±3.5 | 52±1.5 | 0.097 |
| VLDL-C, mg/dL | 19±3.3 | 21±2.7 | 19±3.7 | 0.315 |
| IDL-C, mg/dL | 9±0.5 | 11±1.9 | 9±0.5 | 0.084 |
| ApoB, mg/dL | 75±1.6 | 79±5.0 | 75±1.7 | 0.190 |
| Lp(a), nmol/L | 57±7.0 | 39±15.9 | 59±7.7 | 0.110 |
| Fasting glucose, mg/dL | 103±1.9 | 115±6.0 | 101±2.0 | 0.025 |
| Insulin, µU/dL | 32±1.6 | 37±3.8 | 31±1.7 | 0.102 |
| HOMA-IR | 1.5±0.1 | 1.9±0.2 | 1.5±0.1 | 0.047 |
| eGFR, mL/min/1.73m2 | 85±1.3 | 85±1.4 | 82±3.9 | 0.252 |
| HsCRP, mg/L | 2.1±0.2 | 3.9±0.7 | 1.9±0.2 | <0.001 |
| WBC, k/µL | 6.5±0.2 | 6.8±0.4 | 6.4±0.2 | 0.168 |
Table 3 demonstrates the annualized rates of changes in LDL-C, triglycerides, hsCRP and the number of subjects with >=2 modifiable risk factors. On average per year, the US group displayed a 5.3% greater decrease in LDL-C (p=0.2), 0.9% greater increase in triglycerides (p=0.4), and 0.07% greater increase in hsCRP (p=0.1) compared to the AS group. Given the study provided intensive lipid therapy on dyslipidemia, 2 or more modifiable risk factors, instead of 3 or more, was used to compare risk following the intervention between the 2 groups; this showed a non-statistically significant difference in change of this ASCVD risk status between US and AS (Table 3).
Table 3.
Annualized rates of change in lipids and hsCRP between the US and AS patients.
| US(rate per year; 95% CI) | AS(rate per year; 95% CI) | Difference, US-AS (Est. 95% CI) | p-value | |
|---|---|---|---|---|
| LDL-C, mg/dL | -16.0 (-18.9, -13.1) | -10.7 (-18.2, -3.2) | 5.3 (-2.8, 13.3) | 0.201 |
| Triglycerides, mg/dL | -15.9 (-23.5, -8.2) | -16.7 (-36.3, 2.8) | -0.9 (-20.6, 23.4) | 0.935 |
| HsCRP, mg/L | -0.042 (-0.073, -0.011) | -0.114 (-0.192, -0.034) | -0.071 (-0.156, 0.014) | 0.099 |
| >=2 risk factors (Yes vs. No) | HR=0.31 (0.21-0.43) | HR=0.57 (0.23-1.40) | HR=0.57 (0.23-1.40) | 0.199 |
Note: Linear mixed and logistic mixed model.9
Detailed data on response of metabolic and inflammatory risk factors to 2 years of lipid-lowering by medically served status and therapy groups are presented in supplemental Table.
In addition, the US group showed a trend of progression over 2 years in maximum carotid wall thickness (4.4% vs. -1.3%, p=0.2) and percent wall volume (0.08% vs. -0.5%, p=0.057) compared to regression in the AS group.
4. Discussion
To address ASCVD risk reduction responses in populations with inadequate access to medical care, we compared changes in ASCVD-related metabolic and inflammatory risk between US and AS individuals in response to two-years of lipid lowering pharmacological therapy and lifestyle intervention in the CPC study. We found: (1) US individuals displayed higher levels of metabolic and inflammatory risk including systolic blood pressure, fasting glucose and fasting insulin which resulted in higher insulin resistance, and hsCRP at baseline compared to AS (Table 1). (2) US group had similar reductions in LDL-C, triglycerides and hsCRP over 2 years of intensive lipid-lowering therapy and lifestyle intervention compared to AS (Fig. 1 and Table 3). (3) Despite similar responses to the study therapies, US continued to show significantly higher levels of fasting blood glucose, insulin resistance and hsCRP than AS following the two-year intervention (Fig. 1 and Table 2).
Undoubtably, access to adequate healthcare improves reduces cardiovascular disease risk and improves outcomes[6]. Upon Medicaid expansion provided by the Affordable Care Act (ACA), cardiovascular-related mortality in non-elderly patients decreased 2 years after implementation by 4.3 deaths per 100,000 compared to incidences 3 years prior to implementation in expansion states [10]. Reduced access to care, reduced health literacy, and decreased compliance to lifestyle-based interventions increase cardiovascular disease risk in medically underserved populations [11]. Findings from the Framingham Heart Study and the National Health and Nutrition Examination Surveys (NHANES) highlight the importance of health insurance coverage on ASCVD risk and risk reduction. Within NHANES survey participants, uninsured adults displayed higher LDL and lower HDL cholesterol levels and were less likely to control hypercholesterolemia compared to those on public or private insurance [12]. Among adults ages 19 to 64 years enrolled in the Framingham Heart Study, uninsured men also displayed higher levels of LDL cholesterol, despite no statistical differences in 10-year Framingham risk score or metabolic syndrome. Following treatment in this cohort, uninsured men and women were less likely to achieve blood pressure control (odds ratio: 0.19 and 0.31, respectively) and men were less likely to achieve control of hyperlipidemia (odds ratio 0.17) [25]. Similar reports are observed in minority-specific populations, where Latino/Hispanic adults without health insurance display elevated fasting blood glucose and HbA1c levels compared to those with health insurance [13]. These findings are of concern as disease burden is also elevated in uninsured groups. For example, the microvascular complications of type 2 diabetes are higher in those without adequate healthcare coverage [14]. In addition, lack of healthcare coverage is highest among minority groups [15]. In our study, 8% of AS vs. 36% of US were classified as Hispanic/Latino origin, respectively. The difference in ethnic distribution between AS and US groups in our study may have contributed to observed differences in baseline ASCVD risk[16], [17].[ ] Hispanics and African-Americans have been reported to show higher levels of metabolic risk [18]. and display elevated hsCRP levels relative to Whites after adjusting for socioeconomic factors[19]
A likely contributing factor contributing to increased ASCVD risk and reduced ASCVD control in uninsured groups is access to care [12]. A handful of studies in the United States have shown that interventions targeting risk reduction are effective in medically underserved populations. The WISEWOMAN project investigated the effects of an ASCVD risk reduction program in uninsured woman and observed a combined physical activity and nutritional intervention reduced the incidence of death during the monitoring period [20]. In another study examining the effects of a pharmacist-led diabetes management program, medically underserved minorities displayed improvements in HbA1C, triglyceride and BMI [21]. In a more recent study, nurse management, but not a telehealth intervention, reduced Framingham risk scores in medically underserved rural and urban populations [22]. Our study demonstrated that US group has similar responses to intensive lipid-lowering and lifestyle interventions compared to their AS cohorts. Together these findings suggest interventions targeting ASCVD risk reduction are effective in medically underserved populations, but more comprehensive interventions including lifestyle intervention and medical therapies for management of blood pressure, lipids and diabetes and effective implementation strategies to achieve maximum reduction in ASCVD risk are needed.
Moreover, our study also demonstrated that despite similar responses to the study therapies, US continued to show significantly higher levels of fasting blood glucose, insulin resistance and hsCRP than AS following 2 years of lipid-lowering and lifestyle interventions. The higher baseline risk may require more aggressive treatment regimens to equitably reduce ACSVD risk. In a diabetes management study of medically underserved individuals, those with higher baseline glucose displayed more difficulty in determining insulin requirements to manage their blood glucose levels [23]. It is therefore important to establish baseline physiological, socioeconomic and behavioral risk factors to develop targeted interventions. Future studies are needed to examine the targeted interventions and longer duration for the US populations.
5. Limitations
The definitions of US and AS in our study were based on self-reported healthcare coverage which is subjective. Despite a sample size of 217 subjects at baseline, the number in the US cohort is small in proportion to AS. Nevertheless, the smaller US cohort displayed a significant different risk profile and similar response to therapy compared to AS. In addition, the 7 AHA cardiovascular health factors, including smoking, nutrition, physical activity, body weight, blood pressure, cholesterol, and blood sugar, were not collected completely [24]. Future large prospective interventional trials are needed for implementing established effective ASCVD risk reduction strategies in the US populations.
6. Conclusions
The medically underserved individuals displayed an increased ASCVD risk and demonstrated similar improvements in lipids and inflammatory risk as response to intensive lipid-lowering therapy. Despite similar improvements, the medically underserved group continued to display elevated ASCVD risk with higher levels of fasting glucose and hsCRP following therapy compared to the adequately served. These findings highlight unmet needs, warranting more comprehensive ASCVD risk reduction and targeted implementation strategies to successfully reduce ASCVD risk among the underserved populations.
Financial disclosure
All authors stated that there are no financial relationships that need to be disclosed.
Authorship statement
Michael Chu and Gina Many wrote the manuscript. Daniel Isquith and Susan McKeeth performed data collection. Jayne Williamson performed life style modifications. Moni B Neradilek performed statistical analysis. Patrick Colletti supported the study activities conducted at USC. Xue-Qiao Zhao was responsible for overall study design, funding and interpretation of the study results. All authors read and approved the final manuscript.
Declaration of Competing interest
Michael P. Chu: None
Gina Many: None
Daniel A Isquith: None
Susan McKeeth: None
Jayne Williamson: None
Moni B Neradilek: None
Patrick Colletti: None
Xue-Qiao Zhao: Research grants from Amgen, AstraZeneca, and Novartis.
Acknowledgments
This research was supported by an NIH grant (R01 HL63895) from the National Heart, Lung, and Blood Institute with additional funding from Pfizer, Abbott, Daiichi Sankyo and Merck. The study medications were provided by Pfizer, Abbott, Daiichi Sankyo and Merck.
Footnotes
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.ajpc.2021.100227.
Appendix. Supplementary materials
References
- 1.Alfaddagh A, Arps K, Blumenthal R, Martin S. The ABCs of Primary Cardiovascular Prevention: 2019 Update - American College of Cardiology. Am College of Cardiol. 2019 https://www.acc.org/latest-in-cardiology/articles/2019/03/21/14/39/abcs-of-primary-cv-prevention-2019-update-gl-prevention Updated March 21Accessed June 1, 2020. [Google Scholar]
- 2.Wing S. Social inequalities in the decline of coronary mortality. Am J Public Health. 1988;78(11):145–1416. doi: 10.2105/AJPH.78.11.1415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Access to Care. www.heart.org. https://www.heart.org/en/get-involved/advocate/federal-priorities/access-to-care. Published 2020. Accessed June 2, 2020.
- 4.Alcalá HE, Albert SL, Roby DH. Access to Care and Cardiovascular Disease Prevention: A Cross-Sectional Study in 2 Latino Communities. Medicine (Baltimore) 2015;94(34):e1441. doi: 10.1097/MD.0000000000001441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Valero-Elizondo J, Khera R, Saxena A. Financial Hardship From Medical Bills Among Nonelderly U.S. Adults With Atherosclerotic Cardiovascular Disease. J Am Coll Cardiol. 2019;73(6):727–732. doi: 10.1016/j.jacc.2018.12.004. [DOI] [PubMed] [Google Scholar]
- 6.Brooks EL, Preis SR, Hwang SJ. Health insurance and cardiovascular disease risk factors. Am J Med. 2010;123(8):741–747. doi: 10.1016/j.amjmed.2010.02.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Zhao XQ, Phan BA, Chu B. Testing the hypothesis of atherosclerotic plaque lipid depletion during lipid therapy by magnetic resonance imaging: study design of Carotid Plaque Composition Study. Am Heart J. 2007;154(2):239–246. doi: 10.1016/j.ahj.2007.04.035. [DOI] [PubMed] [Google Scholar]
- 8.WWAMI Rural Health Research Center. UW RHRC Rural Urban Commuting Area Codes - RUCA. http://depts.washington.edu/uwruca/index.php. Accessed April 5, 2019.
- 9.Neradilek, MB. The Mountain-Whisper-Light Statistics. CPC/underserved population study – statistical analysis. June 6, 2019.
- 10.Khatana SAM, Bhatla A, Nathan AS. Association of Medicaid Expansion With Cardiovascular Mortality. JAMA Cardiol. 2019;4(7):671–679. doi: 10.1001/jamacardio.2019.1651. [published correction appears in JAMA Cardiol. 2019 Jul 1;4(7):714] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Alcalá HE, Albert SL, Roby DH. Access to Care and Cardiovascular Disease Prevention: A Cross-Sectional Study in 2 Latino Communities. Medicine (Baltimore) 2015;94(34):e1441. doi: 10.1097/MD.0000000000001441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Egan BM, Li J, Sarasua SM. Cholesterol Control Among Uninsured Adults Did Not Improve From 2001-2004 to 2009-2012 as Disparities With Both Publicly and Privately Insured Adults Doubled. J Am Heart Assoc. 2017;6(11) doi: 10.1161/jaha.117.006105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Glantz NM, Morales JM, Bevier WC. Insurance Status and Biological and Psychosocial Determinants of Cardiometabolic Risk Among Mexican-Origin U.S. Hispanic/Latino Adults with Type 2 Diabetes. Health Equity. 2020;4(1):142–149. doi: 10.1089/heq.2019.0119. Published 2020 May 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Pugh JA, Tuley MR, Hazuda HP, Stern MP. The influence of outpatient insurance coverage on the microvascular complications of non-insulin-dependent diabetes in Mexican Americans. J Diabetes Complications. 1992;6(4):236–241. doi: 10.1016/1056-8727(92)90058-s. [DOI] [PubMed] [Google Scholar]
- 15.Data Finder - Health, United States - Products. https://www.cdc.gov/nchs/hus/contents2018.htm. Published October 30, 2019. Accessed July 30, 2020
- 16.Muncan B. Cardiovascular disease in racial/ethnic minority populations: illness burden and overview of community-based interventions. Public Health Rev. 2018;39(1) doi: 10.1186/s40985-018-0109-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Frierson GM, Howard EN, DeFina LE, Powell-Wiley TM, Willis BL. Effect of race and socioeconomic status on cardiovascular risk factor burden: the Cooper Center Longitudinal Study. Ethn Dis. 2013;23(1):35–42. [PMC free article] [PubMed] [Google Scholar]
- 18.Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA. 2002;287(3):356–359. doi: 10.1001/jama.287.3.356. [DOI] [PubMed] [Google Scholar]
- 19.Nazmi A, Victora CG. Socioeconomic and racial/ethnic differentials of C-reactive protein levels: a systematic review of population-based studies. BMC Public Health. 2007;7:212. doi: 10.1186/1471-2458-7-212. Published 2007 Aug 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Will JC, Massoudi B, Mokdad A. Reducing risk for cardiovascular disease in uninsured women: combined results from two WISEWOMAN projects. J Am Med Womens Assoc (1972) 2001;56(4):161–165. [PubMed] [Google Scholar]
- 21.Rodríguez JE, Campbell KM, Kirksey OW. Improving diabetes care for minority, uninsured and underserved patients. J Immigr Minor Health. 2014;16(4):747–750. doi: 10.1007/s10903-013-9965-7. [DOI] [PubMed] [Google Scholar]
- 22.Bove AA, Santamore WP, Homko C. Reducing cardiovascular disease risk in medically underserved urban and rural communities. Am Heart J. 2011;161(2):351–359. doi: 10.1016/j.ahj.2010.11.008. [DOI] [PubMed] [Google Scholar]
- 23.Langford AT, Wang B, Orzeck-Byrnes NA. Sociodemographic and clinical correlates of key outcomes from a Mobile Insulin Titration Intervention (MITI) for medically underserved patients. Patient Educ Couns. 2019;102(3):520–527. doi: 10.1016/j.pec.2018.09.016. [DOI] [PubMed] [Google Scholar]
- 24.Lifes Simple 7. www.heart.org. https://www.heart.org/en/professional/workplace-health/lifes-simple-7#:∼:text=The%20American%20Heart%20Association%20has,%2C%20cholesterol%2C%20and%20blood%20pressure. Published 2020. Accessed July 11, 2020.
- 25.Brooks EL, Preis SR, Hwang SJ. Health insurance and cardiovascular disease risk factors. Am J Med. 2010;123(8):741–747. doi: 10.1016/j.amjmed.2010.02.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.