Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Feb 6.
Published in final edited form as: Circulation. 2023 Nov 16;149(6):417–426. doi: 10.1161/CIRCULATIONAHA.123.064566

Long-term Mortality in Patients with Severe Hypercholesterolemia Phenotype from a Racial/Ethnically Diverse US Cohort

Jeremy Miles a,#, Andrea Scotti a,#, Francesco Castagna a, Toshiki Kuno a, Pier Pasquale Leone a, Augustin Coisne a, Sebastian Ludwig a, Carl J Lavie b, Parag H Joshi c, Azeem Latib a, Mario J Garcia a, Carlos J Rodriguez a, Michael D Shapiro d, Salim S Virani e, Leandro Slipczuk a
PMCID: PMC10872875  NIHMSID: NIHMS1941443  PMID: 37970713

Abstract

BACKGROUND.

Tools for mortality prediction in patients with the severe hypercholesterolemia phenotype (LDL-C ≥190 mg/dL) are limited and restricted to specific racial/ethnic cohorts. We sought to evaluate the predictors of long-term mortality in a large racial/ethnically diverse US patient cohort with LDL-C ≥190 mg/dL.

METHODS.

We conducted a retrospective analysis of all patients with an LDL-C ≥190 mg/dL seeking care at Montefiore from 2010 through 2020. Patients <18 years of age or with prior malignancy were excluded. The primary endpoint was all-cause mortality. Analyses were stratified by age, sex, and race/ethnicity. Patients were stratified by primary and secondary prevention. Cox-regression analyses were used to adjust for demographic, clinical, and treatment variables.

RESULTS.

A total of 18,740 patients were included (37% non-Hispanic Black, 30% Hispanic, 12% non-Hispanic White and 2% non-Hispanic Asian). The mean age was 53.9 years and median follow-up 5.2 years. Both HDL-C and BMI extremes were associated with higher mortality in univariate analyses. In adjusted models, higher LDL-C and triglyceride levels were associated with an increased 9-year mortality risk (adj-HR 1.08 [1.05–1.11] and 1.04 [1.02–1.06] per 20 mg/dL increase, respectively). Clinical factors associated with higher mortality included male sex (adjHR 1.31 [1.08–1.58]), older age (adjHR 1.19 per 5-year increase [1.15–1.23]), hypertension (adjHR 2.01 [1.57–2.57]), CKD (adjHR 1.68 [1.36–2.09]), diabetes (adjHR 1.79 [1.50–2.15]), heart failure (adjHR 1.51 [1.16–1.95]), myocardial infarction (adjHR 1.41 [1.05–1.90]), and BMI <20 kg/m2 (adjHR 3.36 [2.29–4.93]). A significant survival benefit was conferred by LLT (adj-HR 0.57 [0.42–0.77]). In the primary prevention group, HDL-C <40 mg/dL was independently associated with higher mortality (adj-HR 1.49 [1.06–2.09]). Temporal trend analyses showed a reduction in statin use over time (p<0.001). In the most recent time-period (2019–2020), 56% of patients on primary and 85% of those on secondary prevention were on statin therapy.

CONCLUSIONS.

In a large and diverse cohort of US patients with the severe hypercholesterolemia phenotype, we identified several patient characteristics associated with increased 9-year all-cause mortality and observed a decrease in statin use over time, in particular for primary prevention. Our results support efforts geared towards early recognition and consistent treatment for patients with severe hypercholesterolemia.

Keywords: Severe hypercholesterolemia, lipids, mortality

INTRODUCTION

Low-density lipoprotein cholesterol (LDL-C) has long been recognized as a risk factor for atherosclerotic cardiovascular disease (ASCVD). Severe hypercholesterolemia is commonly defined as an LDL-C ≥190 mg/dL and is strongly linked to higher rates of incident premature atherosclerotic cardiovascular disease (ASCVD). The prevalence of ASCVD has been estimated to be approximately 5% of the United States population1. Multiple guidelines recommend early recognition and initiation of statin therapy for patients with severe hypercholesterolemia but do not recommend further risk stratification within this high-risk group2, mostly due to the lack proven risk stratification strategies in this subgroup of patients. Without appropriate therapy, patients with severe hypercholesterolemia develop ASCVD earlier (10–20 years earlier in men and 20–30 years earlier in women) compared to those with LDL-C <130 mg/dL1. Approximately 2–7% of patients with severe hypercholesterolemia have familial hypercholesterolemia (FH) defining mutations, which portend an even higher risk for earlier and more aggressive ASCVD when compared to those without (11-fold versus 6-fold)3,4, mostly lifelong exposure to high LDL-C levels. Statins represent the foundation of lipid lowering therapies (LLT) and a wealth of data demonstrate that the use of statins decrease ASCVD in patients with severe hypercholesterolemia5. However, emerging evidence from our group and others suggests that ASCVD risk may be heterogeneous as some individuals with LDL-C ≥190 mg/dL develop ASCVD early despite maximal LLT, while others never do, despite marked LDL-C elevation6–9. Indeed, it is plausible that a complex interplay of multiple cardiovascular risk modifiers beyond LDL-C may explain the variability of ASCVD expression in patients with severe hypercholesterolemia10. Understanding the spectrum of risk and independent contributors may allow better treatment strategies.

We therefore sought to evaluate the predictors of long-term mortality in a racial/ethnically diverse US patient cohort with LDL-C ≥190 mg/dL.

METHODS

Study population

The electronic medical records of the Montefiore Health System – Albert Einstein College of Medicine (Bronx, NY), a quaternary large health system, were reviewed to identify all lipid panels measured between January 2010 and December 2020 using Clinical Looking GlassTM (CLG), a quality improvement health care surveillance software, a proprietary query tool which integrates all clinical data throughout the Montefiore Health system. Patients were included in this study if they had at least one measurement of LDL-C level ≥190 mg/dL. In case of multiple measurements of LDL-C level ≥190 mg/dL during the study period, the first occurrence was used to define study entry and baseline status for each patient. Patients were included regardless of their statin use or age. Exclusion criteria were 1) age below 18-year-old and 2) history of cancer (926 patients) and 3) history of nephrotic syndrome (475), hypothyroidism (1,003) or biliary obstruction/cholestasis (389). Baseline demographic, clinical, laboratory data, and outcomes were subsequently retrieved from the electronic medical records system. The study cohort was further stratified according to self-identified race/ethnicity, sex, and age categories. Race and ethnicity were self-identified at the time of initial registration for care. We first identified self-reported Hispanics, and then only among non-Hispanics, we specified race. Being Hispanic is an ethnicity, not a race, but self-reported race among Hispanics is notoriously unreliable11. Thus, we categorized race/ethnicity as Hispanic, non-Hispanic Black (NHB), non-Hispanic White (NHW), non-Hispanic Asian (NHA), and others/declined/unknown/not available as has been reported previously12. Summary socioeconomic score (SES), which relate to the wealth, income, education, and occupation of residents in a census-block group (a proxy for neighborhood), was calculated as previously reported13. The modified DLCN score was calculated as follows: patients were analyzed if male <55 years and female <60 years (excluded those older than these cut offs). Possible FH: LDL-C ≤325 mg alone, LDL-C 190–250 mg/dL + CAD or PAD/Stroke; Probable FH: LDL-C>325 mg/dL alone, LDL-C 251–325 mg/dL+ CAD or PAD/Stroke, or LDL-C 190–250 + both CAD and PAD/Stroke); and Definite FH: LDL-C>325 mg/dL + CAD or PAD/Stroke. % LDL-C variation was defined as the lowest LDL-C reduction attained from the maximum.

The study was approved by our Institutional Review Board (Office of Human Research Affairs at Albert Einstein College of Medicine) and was HIPAA compliant. Informed consent was waived due to the retrospective study design. The data that supports the findings of this study is available from the corresponding author upon reasonable request.

Study endpoint

The primary endpoint was all-cause mortality. Patient mortality status was extracted from the electronic medical record through CLG (a system previously known to capture most events with readmissions outside Montefiore occurring rarely (1.9%)14. In addition, in-hospital mortality, deaths occurred elsewhere and notified to our institution were also included through our data mining software.

Statistical analysis

Baseline characteristics are described with mean ± standard deviation (SD) [or medians and 1st to 3rd interquartile ranges (IQRs)] for continuous variables and percentages for categorical variables. Differences among groups were assessed by generating a standardized mean difference, determined to be clinically significant with a value exceeding 0.10. All-cause mortality was estimated using the Kaplan-Meier method and compared using the log-rank test. Hazard ratios (HRs) and 95% confidence intervals (CIs) were determined using the Cox proportional hazards regression. Multivariable analyses were performed with adjustment for the following demographic and clinical variables: age, sex, race/ethnicity, socioeconomical status (SES), smoking, alcohol abuse, BMI, hypertension, coronary artery disease (CAD), myocardial infarction, stroke, heart failure, diabetes, chronic kidney disease (CKD), peripheral artery disease, chronic obstructive pulmonary disease, LDL-C, high-density lipoprotein cholesterol (HDL-C), triglycerides, statin. The relationship between baseline continuous variables and risk of 9-year mortality was further explored with Cox proportional hazards regression models by entering each variable as a restricted cubic spline with 3 knots located at the 10th, 50th, and 90th percentiles. The Cochran-Armitage test was used to assess linear trends in proportions of statin use over time. Interaction terms for were derived from the multivariable Cox Regression model including: age, sex, race, BMI, hypertension, chronic kidney disease, diabetes, COPD, stroke, myocardial infarction, coronary artery disease, heart failure, and lipid lowering therapy. To avoid bias due to complete case analyses, missing data were handled with Multivariate Imputation via Chained Equations using the mice package (v3.13.0; van Buuren & Groothuis-Oudshoorn, 2011). A two-sided p value of <0.05 was considered statistically significant. Statistics were performed using R, version 4.1.3 (The R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

Demographics

As shown in Table 1, a total of 18,740 patients were included (37% non-Hispanic Black (NHB), 30% Hispanic, 12% non-Hispanic White (NHW), and 2% non-Hispanic Asian (NHA)). Median follow-up was 5.2 years (IQR 2.3–8.5 years). Mean age was 53.9 ± 14.7 years and 38.1% were female. Hypertension was present in 47%, diabetes 31%, CKD 20%, smoking 29%, alcohol abuse 4% and CAD 11%, with 4% of patients with prior myocardial infarction. Peripheral artery disease was present in 5.2% and stroke in 4.6%. Mean LDL-C was 212.2 ± 32.6 mg/dL, HDL-C was 54.5 ± 15.5 mg/dL, total cholesterol 298.4 ± 37.5 mg/dL, and triglycerides 158.9 ± 72.5 mg/dL. Median hemoglobin A1c (HbA1c) was 6.0% (IQR 5.6–7.2%) and mean creatinine 1.04 ± 0.70 mg/dL. A total of 75.4% of patients were on statin therapy. Statin prescription was less common in primary (71.8%) than in secondary prevention (93.0%). As seen in Table 2, follow-up LDL-C was measured at 4.9 years [2.3–7.9 years]. When latest LDL-C was analyzed, only 6.9% of patients reached LDL-C <70mg/dL while 17.3% and 20.6% reached LDL-C 70–99 and 100–129 mg/dL, respectively. Among 55.3% of patients, the latest LDL-C remained ≥130 mg/dL. Mean LDL-C reduction was −33.11% ± 24.11% in the total cohort.

Table 1.

Baseline Characteristics According to Race/Ethnicity of Patients with LDL-C ≥190 mg/dL.

Overall (n=18740) Non-Hispanic Black (6952) Non-Hispanic White (2298) Non-Hispanic Asian (394) Hispanic (n=5622) Other/NA (n=3474) P value SMD*
Age (years) 53.85 ±14.71 54.77 ±14.49 55.55 ±14.98 50.04 ±14.72 52.99 ±14.57 52.70 ±14.92 <0.001 0.178
Female 7131 (38.1) 2514 (36.2) 939 (40.9) 171 (43.4) 2035 (36.2) 1472 (42.4) <0.001 0.085
BMI 30.32 ± 6.37 31.49 ± 6.87 29.54 ± 6.29 27.29 ± 5.15 29.92 ± 5.88 29.31± 5.77 <0.001 0.302
Hypertension 8772 (46.8) 3894 (56.0) 816 (35.5) 159 (40.4) 2623 (46.7) 1280 (36.8) <0.001 0.207
Diabetes 5830 (31.1) 2554 (36.7) 452 (19.7) 110 (27.9) 1789 (31.8) 925 (26.6) <0.001 0.177
CKD 3776 (20.1) 1799 (25.9) 365 (15.9) 52 (13.2) 982 (17.5) 578 (16.6) <0.001 0.138
CAD 2132 (11.4) 813 (11.7) 298 (13.0) 37 (9.4) 690 (12.3) 294 (8.5) <0.001 0.077
Prior MI 830 (4.4) 302 (4.3) 112 (4.9) 14 (3.6) 276 (4.9) 126 (3.6) 0.036 0.039
Heart Failure 1265 (6.8) 589 (8.5) 145 (6.3) 18 (4.6) 347 (6.2) 166 (4.8) <0.001 0.077
PAD 977 (5.2) 421 (6.1) 107 (4.7) 15 (3.8) 327 (5.8) 107 (3.1) <0.001 0.076
Stroke 858 (4.6) 391 (5.6) 95 (4.1) 12 (3.0) 258 (4.6) 102 (2.9) <0.001 0.07
COPD 1765 (9.4) 700 (10.1) 211 (9.2) 25 (6.3) 640 (11.4) 189 (5.4) <0.001 0.114
Smoking 5425 (29.0) 2133 (30.8) 752 (32.8) 84 (21.3) 1763 (31.4) 693 (31.4) <0.001 0.164
Alcohol 768 (4.1) 312 (4.5) 92 (4.0) 10 (2.5) 262 (4.7) 92 (2.7) <0.001 0.066
SES −3.00 (2.89) −3.01 (2.72) −1.36 (2.80) −2.49 (2.71) −3.96 (2.75) −2.54 (2.91) <0.001 0.411
Primary Prevention 15569 (83.1) 5671 (81.6) 1893 (82.4) 339 (86.0) 4610 (82.0) 3056 (88.0) <0.001 0.094
Secondary Prevention 3171 (16.9) 1281 (18.4) 405 (17.6) 55 (14.0) 1012 (18.0) 418 (12.0) <0.001 0.094
Liver Disease 398 (2.1) 137 (2.0) 42 (1.8) 5 (1.3) 149 (2.7) 65 (1.9) 0.02 0.042
Measurements 4.85 (4.86) 5.16 (5.05) 4.42 (4.40) 4.70 (4.60) 5.21 (5.11) 3.96 (4.21) <0.001 0.139
Total Cholesterol (mg/dL) 298.39 ±37.49 297.26 ±39.23 299.27 ±32.84 298.04 ±35.44 299.62 ±37.78 298.10 ±36.51 0.008 0.033
LDL-C (mg/dL) 212.20 ± 32.62 212.66±34.62 211.15±27.27 211.89± 29.42 212.21 ± 33.01 212.01 ± 31.42 0.421 0.022
HDL-C (mg/dL) 54.46 ±15.54 56.52 ±16.21 53.67 ±15.66 52.63 ±13.85 52.75 ±14.51 53.82 ±15.39 <0.001 0.115
Triglycerides (mg/dL) 158.87 ±72.48 140.64 ±64.82 172.39 ±77.31 167.69 ±69.27 173.61 ±75.32 161.54 ±71.45 <0.001 0.218
Creatinine (mg/dL) 1.04 ±0.70 1.14±0.84 0.97 ±0.52 0.96 ±0.55 0.98 ±0.62 1.00 ±0.61 <0.001 0.113
HbA1c (%) 6.00 [5.60, 7.20] 6.20 [5.70, 7.60] 5.80 [5.50, 6.40] 6.10 [5.70, 7.05] 6.00 [5.60, 7.20] 6.00 [5.60, 6.90] <0.001 0.101
Systolic BP (mmHg) 131 [120, 147] 135 [122, 150] 130 [119, 143] 128 [115, 144] 130 [119, 145] 131 [120, 145] <0.001 0.153
Diastolic BP (mmHg) 80 [74, 87] 81 [75, 90] 80 [73, 85] 79 [72, 86] 80 [72, 86] 80 [74, 87] <0.001 0.125
Primary prevention
 Statin 11,182 (71.8) 4,246 (74.9) 1,255 (66.3) 232 (68.4) 3,493 (75.8) 1,956 (64.0) <0.001 0.141
 Ezetimibe 915 (5.9) 367 (6.5) 104 (5.5) 18 (5.3) 283 (6.2) 143 (4.7) 0.012 0.039
 PCSK9i 144 (0.9) 53 (0.9) 20 (1.1) 3 (0.9) 41 (0.9) 27 (0.9) 0.974 0.008
Secondary prevention
 Statin 2,949 (93.0) 1,204 (94.0) 358 (88.4) 53 (96.4) 957 (94.6) 377 (90.2) <0.001 0.156
 Ezetimibe 473 (14.9) 203 (15.9) 53 (13.1) 5 (9.3) 165 (16.3) 47 (11.3) 0.058 0.112
 PCSK9i 149 (4.7) 51 (4.0) 13 (3.2) 2 (3.6) 58 (5.7) 25 (6.0) 0.109 0.073
Open in a new tab

Data are mean ± SD, median [interquartile range], or n (%).

*

Average SMD among groups. BMI: Body mass index; CKD: Chronic kidney disease; CAD: Coronary artery disease; MI: Myocardial infarction; PAD: Peripheral artery disease; BP: blood pressure; COPD: Chronic obstructive pulmonary disease; LDL-C: low-density lipoprotein cholesterol; HDL-C: high-density lipoprotein cholesterol; HbA1c: Hemoglobin A1c; PCSK9i: Proprotein convertase subtilisin/kexin type 9 inhibitors; IQR: interquartile range; NA: not available; SES: Summary Socioeconomic Score; SMD: standardized mean difference.

Table 2.

LDL-C at Follow-Up According to Race/Ethnicity.

Overall (n=14917) Non-Hispanic Black (n=5709) Non-Hispanic White (n=1762) Non-Hispanic Asian (n=318) Hispanic (n=4639) Other/NA (n=2489) p value SMD*
Baseline LDL-C (mg/dL) 212.20 ± 32.62 212.66 ± 34.62 211.15 ± 27.27 211.89 ± 29.42 212.21 ± 33.01 212.01 ± 31.42 0.421 0.022
Latest LDL-C (mg/dL) 140.16 ± 51.96 141.56 ± 52.51 136.43 ± 48.89 135.50 ± 50.33 139.01 ± 52.46 142.35 ± 51.89 <0.001 0.074
LDL-C Variation (%) −33.11 ± 24.11 −32.50 ± 24.25 −34.52 ± 23.25 −35.32 ± 23.94 −33.82 ± 24.20 −31.92 ± 24.15 <0.001 0.074
Target LDL-C 0.002 0.086
 <70 mg/dL 13,888 (93.1) 5,358 (93.9) 1,640 (93.1) 292 (91.8) 4,267 (92.0) 2,331 (93.7) 0.003 0.044
 <100 mg/dL 11,312 (75.8) 4,387 (76.8) 1,327 (75.3) 229 (72.0) 3,460 (74.6) 1,909 (76.7) 0.027 0.054
 <130 mg/dL 8,244 (55.3) 3,192 (55.9) 920 (52.2) 164 (51.6) 2,545 (54.9) 1,423 (57.2) 0.01 0.06
 ≥130 mg/dL 6,673 (44.7) 2,517 (44.1) 842 (47.8) 154 (48.4) 2,094 (45.1) 1,066 (42.8) 0.01 0.06
Primary Prevention
 LDL-C <100 mg/dL 9,772 (79.0) 3,733 (80.2) 1,149 (78.5) 201 (73.6) 2,952 (78.1) 1,737 (79.1) 0.024 0.068
Secondary Prevention
 LDL-C <70 mg/dL 2,186 (85.7) 920 (87.0) 253 (84.9) 38 (84.4) 719 (83.7) 256 (87.4) 0.266 0.057
Open in a new tab

Data are n (%).

*

Average SMD among groups. LDL-C: low-density lipoprotein cholesterol

NA: not available; SMD: standardized mean difference.

Ethnicity/Race differences

Age was significantly different with NHW patients being the oldest (56 years) and NHA patients the youngest (50 years, p<0.001 for both). As shown on Table 1, NHB patients had the highest burden of comorbidities such as hypertension (56.0%), diabetes (36.7%), chronic kidney disease (25.9%), heart failure (8.5%), peripheral artery disease (6.1%), and stroke (5.6%, p<0.001 for all). Statin therapy was more commonly prescribed in NHB (78.4%) and Hispanic (79.2%) compared to NHW (70.2%) and NHA (72.3%) patients (p<0.001). There were no significant differences among racial/ethnic groups in sex or LDL-C levels.

When follow-up LDL-C measurements were analysed, no clinically meaningful differences based on race/ethnicity were found in terms of LDL-C reduction or percentages of patients at several target LDL-C levels (<70 mg/dL, <100 mg/dL, or <130 mg/dL; Table 2, all SMDs <0.10). Statin intensity is shown in Table S1. When the latest statin dose recorded was considered, only 47.5% were on high-intensity statin therapy. The interaction analysis between clinical and biological variables can be seen on Table S2. Stratified sub-analysis by race/ethnicity is shown in Table S3.

Mortality analysis

A total of 842 (8.0%) patients died during the 9-year follow up. Non-Hispanic Black and NHW patients had higher mortality than Hispanic and NHA patients (9.3% vs. 8.5% vs. 6.9% vs. 6.1%, respectively, p=0.0004; Figure 1A). No difference in mortality was observed between male and female patients (8.0% vs. 7.9%, respectively, p=0.82; Figure 1B). As a continuous variable, higher LDL-C levels were found to be associated with increasing risk of 9-year mortality (Figure 2A). On the contrary, a U-shaped curve was observed for HDL-C with low or significantly elevated levels carrying a higher mortality risk and the lowest mortality was identified between 45 and 80 mg/dL (Figure 2B). Similar findings were observed for BMI that correlated with higher mortality with values outside the 27–40 kg/m2 range (Figure 2C). Older age was associated with increasing 9-year mortality risk, with significant differences in the mortality rates of patients <60 years versus ≥60 years (4.5% vs. 14.5%, p<0.0001; Figure S1).

Figure 1. Kaplan-Meier curves for 9-year mortality stratified by race-ethnicity (A) and sex (B).

Figure 1.

Open in a new tab

Figure 2. Restricted spline curves for 9-year mortality according to LDL-C (A), HDL-C (B), and BMI (C).

Figure 2.

Open in a new tab

BMI: body-mass index; HDL-C: high-density lipoprotein cholesterol; LDL-C: low-density lipoprotein cholesterol.

After adjustment for abovementioned demographic and clinical variables, an increased risk of 9-year mortality was observed for male individuals (adj-HR 1.31 [1.09–1.58]) and low BMI (adj-HR 3.36 [2.29–4.93]), Figure 3. No significant amounts of missing data were found, with the exception of the BMI (Table S4). The Multivariable Cox Regression analysis for 9-year all-cause death was performed excluding BMI from the predictor variables (Table S5). The observed results were in line with the primary analysis. The Multivariable Cox Regression analysis for 9-year all-cause death after multiple imputations showed similar findings compared to the main analysis, Table S6.

Figure 3. Forest plot of multivariable cox regression analysis for 9-year all-cause death.

Figure 3.

Open in a new tab

BMI= body-mass index; NHW= non-Hispanic White; COPD= chronic obstructive pulmonary disease; LDL-C= low-density lipoprotein cholesterol; HDL-C= high-density lipoprotein cholesterol.

Compared to NHW ethnicity/race, no difference in mortality was found for NHB and NHA individuals, while a lower risk was observed in the Hispanic patients (adj-HR 0.73 [0.55–0.96]). Higher LDL-C levels were independently associated with higher mortality (adj-HR 1.08 [1.05–1.11] per 20 mg/dL increase), with a significant survival benefit conferred by lipid lowering therapy (adj-HR 0.50 [0.37–0.68]).

When the same adjustments were applied for primary prevention patients (Figure S2), similar associations were found with a particular association with increased mortality for low HDL-C (adj-HR 1.49 [1.06–2.09]).

When a modified DLCN criteria was applied, patients meeting possible, probable and definite FH had graded higher mortality for each (Figure 4). When those with at least two LDL-C measures were analyzed, mortality between those with persistent LDL-C≥190 mg/dL versus improved, did not differ (Figure S3, p=0.24).

Figure 4. Time-trend analysis on statin use stratified by primary and secondary prevention.

Figure 4.

Open in a new tab

Temporal trend analyses showed a reduction in statin use over time in both primary and secondary cardiovascular prevention (overall X2 p <0.001, Cochran-Armitage p <0.001), Figure 3. In the most recent time-period (2019–2020), 56% of patients on primary prevention and 85% of those on secondary prevention were on statin therapy. Moreover, most of the ezetimibe and PCSK9i prescriptions occurred in the latest time-period (Figure S4).

DISCUSSION

Our study represents a large cohort studying risk factors for mortality in patients with LDL-C ≥190 mg/dL. In particular, our cohort is racially/ethnically diverse with a high percentage of usually underrepresented racial/ethnic minorities (37% NHB and 30% Hispanic). Our study has several important findings. First, we demonstrated the heterogeneity of mortality risk in patients with LDL-C ≥190 mg/dL. Secondly, we confirmed undertreatment of patients with severe hypercholesterolemia. In our cohort, 25% of patients were not prescribed statin therapy (28% in primary and 7% in secondary prevention and over half of patients (55.3%) latest LDL-C remained ≥130 mg/dL. Thirdly, we demonstrated that higher LDL-C levels, triglycerides, significantly low HDL-C, older age at diagnosis, male sex, hypertension, diabetes, heart failure, myocardial infarction and lower BMI were independently associated with increased mortality risk while LLT use decreased the risk. Lastly, temporal trend analysis showed a reduction in statin use over time in both primary and secondary cardiovascular prevention.

Even though it is well accepted that LDL-C is a key player for lifetime risk for incident ASCVD, that patients with LDL-C ≥190 mg/dL have significantly higher risk and more premature development of ASCVD, and that statin therapy is widely available and inexpensive, a significant number of patients (28% for primary prevention in our study) do not receive a prescription for a statin. While this treatment gap has been well described in the literature, it remains alarming. Jackson et al studied 227 patients with LDL-C ≥190 mg/dL in Texas and found that only 21% were on statin at diagnosis and despite 90% having a follow-up visit, 41% had no change in treatment and only 13% were referred for specialist care15. Virani et al, using data from the American College of Cardiology National Cardiovascular Data Registry-Practice Innovation and Clinical Excellence registry, studied the use of LLT in 49,447 patients with LDL-C ≥190 mg/dL 16. The authors found the proportion of patients receiving a statin, high-intensity statin, LLT associated with ≥50% LDL-C reduction, ezetimibe, or PCSK9 inhibitor were 58.5%, 31.9%, 34.6%, 8.5%, and 1.5%, respectively. Moreover, they found >200% variation in receipt of several of these medications for patients across practices.

Whilst focused on FH, newer machine learning algorithms are being developed in an attempt to improve detection and treatment void17–19. A study in the UK described an electronic health record audit tool that identified patients with diagnosed FH or possible FH, and flagged those with an LDL-C ≥190 ​mg/dL for further assessment19. Another study, using the FIND FH, machine learning model, identified potential FH patients with good precision (AUC 0.89) in a large database of over 170 million patients20. Our study highlights the importance and imperative for new tools to improve statin prescription and LDL-C goal achievement.

For patients recognized as requiring statin therapy, only 7% and 21% reached LDL-C <70 and <100 mg/dL, respectively. With the current availability of new LLT, there are multiple pharmacological tools to reach current LDL-C goals for primary and secondary prevention. Despite this, a large number of patients fail to achieve their LDL-C goal, mostly due to insufficient prescribing of statin and non-statin therapies, poor adherence and medication discontinuation. Underserved populations such as Hispanic and NHB patients have a high prevalence of dyslipidemia, hypertension, and diabetes, placing them at particularly high risk for recurrent ASCVD. The proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors reduce LDL-C approximately 60% on top of baseline statin therapy and improve cardiovascular outcomes in those with established ASCVD21,22. Even though they are widely covered by insurance companies with usually low co-pays, they are still highly underutilized. In a database of 139,036 healthcare claims, Hispanic patients represented only 3% and NHB 7% of claims, suggesting a large gap of care in ethnic minorities.

We also found for the first time that HDL-C levels had a U-shaped correlation with mortality with the lowest risk between 45 and 80 mg/dL and higher mortality for those with HDL-C <40 or >80 mg/dL, however only low HDL-C remained independently associated with mortality in the multivariable analysis. Even though others have shown similar findings in different populations, it has not previously been described in patients with severe hypercholesterolemia 20,23,24. The relationship between HDL-C and mortality remains complex, in particular at extreme LDL-C levels with possible interactions between LDL-C and HDL-C and other comorbidities such as alcohol abuse creating competing risks. The U-shaped correlation of HDL-C with mortality in patients with LDL-C ≥190 ​mg/dL has not been previously reported.

Moreover, we observed that compared to NHW patients, Hispanic patients had lower mortality. While this phenomenon has not been previously described in patients with LDL-C ≥190 ​mg/dL, it has been observed in other cohorts and labeled as the Hispanic paradox. Moreover, this finding persisted after adjusting for comorbidities and socioeconomical status. Epidemiological studies have demonstrated that Hispanic Americans tend to have better outcomes than their non-Hispanic counterparts with as much as 17.5% lower mortality rate25,26. Although the mechanisms are unknown, several risk and resilience factors may play a role and deserve further investigation.

In addition, we found a U-shaped correlation between BMI and mortality, with increasing risk for lower and higher values. This goes together with prior findings that despite its numerous deleterious effects on health, several studies have documented an obesity paradox in which overweight and obese people have better prognosis when compared to nonoverweight/nonobese counterparts27.

Last, we demonstrated a worrisome decrease in statin utilization over time with worse impact on primary prevention. Despite significant improvements in detection of severe hypercholesterolemia over time in mostly White populations, there are known significant gaps between different ethnicities and socioeconomic groups28,29. The undertreatment of patients with severe hypercholesterolemia is likely multifactorial in origin, including a lack of proper insurance and access to primary care providers (PCPs), low rates of routine screening by PCPs, a lack of consensus on screening recommendations, insufficient emphasis on LDL-C as a quality-measure4,30,31, lack of recognition of severe hypercholesterolemia as high-risk and hesitance to start treatment in asymptomatic individuals. Analysis of real-world data is essential to recognize these gaps and underscores the need for elucidating specific barriers preventing achievement of proper recognition and treatment across different racial and ethnic subgroups in this population32.

Our study has several limitations. First, given its retrospective design, LDL-C testing post statin initiation was not uniform. The lack of consistent testing precluded us from performing a reliable time dependent covariate analysis. Moreover, we had no data on patient compliance, this however represents a common scenario in ‘real-world’ practice. Second, we cannot exclude unmeasured confounders that could contribute to mortality in our patient cohort. Third, despite having a large and diverse patient population, our patients represent a cohort from a single large healthcare network. Large registries studying patients with severe hypercholesterolemia, including different regions and countries are needed. Fourth, there were more than 3,000 patients whose race/ethnicity was categorized as ‘other’ or ‘not available’, this is however a small number of the overall population. Fifth, all the covariates were included in the multivariable models as baseline factors, with no time-dependent effect. The latter was not analyzed due to the limited amount of data on changes in medications/lipid measurements during follow-up. This precluded us from performing time-dependent covariate analyses. Lastly, while we did our best to account for all of the known confounders, there is a possibility of unmeasured or unobserved confounding due to retrospective study design.

CONCLUSION

In a large and diverse cohort of US patients with the severe hypercholesterolemia phenotype, we identified several patient characteristics associated with increased 9-year all-cause mortality: male sex, older age, hypertension, CKD, diabetes, heart failure, myocardial infarction, and BMI <20. Higher LDL-C levels and triglycerides were associated with higher mortality with a significant survival benefit conferred by LLT therapy and Hispanic ethnicity. In the primary prevention group, low HDL-C was independently associated with higher mortality. Lastly, temporal trend analyses showed a reduction in statin use over time in both primary and secondary cardiovascular prevention. Our results support efforts geared towards early recognition and consistent treatment for patients with severe hypercholesterolemia.

Supplementary Material

Supplemental Publication Material

Figure 5. Kaplan-Meier curves for 9-year mortality stratified by modified DLCN score.

Figure 5.

Open in a new tab

FH = familial hypercholesterolemia.

CLINICAL PERSPECTIVE.

What is new?

  • We found heterogeneity of mortality risk in patients with LDL-C ≥190 mg/dL. Patients with severe hypercholesterolemia were undertreated. In our cohort, 25% of patients were not prescribed statin therapy (28% in primary and 7% in secondary prevention) and over half of patients (55.3%) latest LDL-C remained ≥130 mg/dL.

  • We demonstrated that higher LDL-C levels, triglycerides, significantly low HDL-C, older age at diagnosis, male sex, hypertension, diabetes, heart failure, myocardial infarction and lower BMI were independently associated with increased mortality risk while LLT use decreased the risk. Patients with probable and definite FH had higher mortality when compared to those with possible FH.

  • Temporal trend analysis showed a reduction in statin use over time in both primary and secondary cardiovascular prevention.

  • Lipid lowering therapy show significant mortality risk reduction (HR 0.50 [0.37–0.58]).

What are the clinical implications?

  • Our results support efforts geared towards early recognition and consistent treatment for patients with severe hypercholesterolemia.

  • Treatment of comorbidities associated with all-cause mortality in patients with severe hypercholesterolemia should be prioritized.

Funding

Francesco Castagna is supported by a grant from the National Institute for Health (T32HL144456) and by the National Center for Advancing Translational Science (NCATS) Clinical and Translational Science Award at Einstein-Montefiore (UL1TR001073). Leandro Slipczuk is supported by grants from Amgen and Philips.

Abbreviations:

ASCVD

Atherosclerotic cardiovascular disease

HDL-C

High-Density Lipoprotein Cholesterol

LDL-C

Low-Density Lipoprotein Cholesterol

NHA

Non-Hispanic Asian

NHB

Non-Hispanic Black

NHW

Non-Hispanic White

US

United States

CAD

Coronary artery disease

CKD

Chronic kidney disease

Footnotes

Disclosures:

LS has received consulting honoraria from Philips and Amgen for lectures; participated in an advisory board meeting from BMS; and received Grant support from Amgen and Philips (Institutional). He has also participated as site PI for the Victorian-INITIATE trial.

PHJ reports grant support from Amgen, Novartis, Novo Nordisk, and NASA.

MDS has served on scientific advisory boards for Amgen, Ionis, Novartis, and Precision Biosciences and served as consultant for Ionis, Novartis, Regeneron and Emendo Biotherapeutics.

SV is supported by grants from the Department of Veterans Affairs, the National Institute of Health, the Tahir and Jooma Family, and has received Honoraria from the American College of Cardiology (Associate Editor for Innovations, ACC.org).

REFERENCES

  • 1.Perak AM, Ning H, De Ferranti SD, Gooding HC, Wilkins JT, Lloyd-Jones DM. Long-term risk of atherosclerotic cardiovascular disease in US adults with the familial hypercholesterolemia phenotype. Circulation. 2016;134:9–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Grundy SM, Stone NJ, Bailey AL, Beam C, Birtcher KK, Blumenthal RS, Braun LT, de Ferranti S, Faiella-Tommasino J, Forman DE, Goldberg R, Heidenreich PA, Hlatky MA, Jones DW, Lloyd-Jones D, Lopez-Pajares N, Ndumele CE, Orringer CE, Peralta CA, Saseen JJ, Smith SC, Sperling L, Virani SS, Yeboah J. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA Guideline on the Management of Blood Cholesterol: Executive Summary: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol. 2019;73:3168–3209. [DOI] [PubMed] [Google Scholar]
  • 3.Khera A v., Won HH, Peloso GM, Lawson KS, Bartz TM, Deng X, van Leeuwen EM, Natarajan P, Emdin CA, Bick AG, Morrison AC, Brody JA, Gupta N, Nomura A, Kessler T, Duga S, Bis JC, van Duijn CM, Cupples LA, Psaty B, Rader DJ, Danesh J, Schunkert H, McPherson R, Farrall M, Watkins H, Lander E, Wilson JG, Correa A, Boerwinkle E, Merlini PA, Ardissino D, Saleheen D, Gabriel S, Kathiresan S. Diagnostic Yield and Clinical Utility of Sequencing Familial Hypercholesterolemia Genes in Patients With Severe Hypercholesterolemia. J Am Coll Cardiol. 2016;67:2578–2589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Bucholz EM, Rodday AM, Kolor K, Khoury MJ, De Ferranti SD. Prevalence and predictors of cholesterol screening, awareness, and statin treatment among US adults with familial hypercholesterolemia or other forms of severe dyslipidemia (1999–2014). Circulation. 2018;137:2218–2230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Vallejo-Vaz AJ, Robertson M, Catapano AL, Watts GF, Kastelein JJ, Packard CJ, Ford I, Ray KK. Low-density lipoprotein cholesterol lowering for the primary prevention of cardiovascular disease among men with primary elevations of low-density lipoprotein cholesterol levels of 190 mg/dL or above: Analyses from the WOSCOPS (West of Scotland coronary prevention study) 5-year randomized trial and 20-year observational follow-up. Circulation. 2017;136:1878–1891. [DOI] [PubMed] [Google Scholar]
  • 6.Galema-Boers AM, Lenzen MJ, Engelkes SR, Sijbrands EJ, Roeters van Lennep JE. Cardiovascular risk in patients with familial hypercholesterolemia using optimal lipid-lowering therapy. J Clin Lipidol. 2018;12:409–416. [DOI] [PubMed] [Google Scholar]
  • 7.Santos RD. Phenotype vs. genotype in severe familial hypercholesterolemia: What matters most for the clinician. Curr Opin Lipidol. 2017;28:130–135. [DOI] [PubMed] [Google Scholar]
  • 8.Mszar R, Nasir K, Santos RD. Coronary Artery Calcification in Familial Hypercholesterolemia. Circulation. 2020;142:1405–1407. [DOI] [PubMed] [Google Scholar]
  • 9.Castagna F, Miles J, Arce J, Leiderman E, Neshiwat P, Ippolito P, Friedmann P, Schenone A, Zhang L, Rodriguez CJ, Blaha MJ, Levsky JM, Garcia MJ, Slipczuk L. Visual Coronary and Aortic Calcium Scoring on Chest Computed Tomography Predict Mortality in Patients With Low-Density Lipoprotein-Cholesterol ≥190 mg/dL. Circ Cardiovasc Imaging. 2022;15:e014135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Bianconi V, Banach M, Pirro M. Why patients with familial hypercholesterolemia are at high cardiovascular risk? Beyond LDL-C levels. Trends Cardiovasc Med. 2021;31:205–215. [DOI] [PubMed] [Google Scholar]
  • 11.Rodriguez CJ, Allison M, Daviglus ML, Isasi CR, Keller C, Leira EC, Palaniappan L, Piña IL, Ramirez SM, Rodriguez B, Sims M. Status of cardiovascular disease and stroke in hispanics/latinos in the united states: A science advisory from the american heart association. Circulation. 2014;130:593–625. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Nasir K, Cainzos-Achirica M, Valero-Elizondo J, Ali SS, Havistin R, Lakshman S, Blaha MJ, Blankstein R, Shapiro MD, Arias L, Saxena A, Feldman T, Budoff MJ, Ziffer JA, Fialkow J, Cury RC. Coronary Atherosclerosis in an Asymptomatic U.S. Population. JACC Cardiovasc Imaging. 2022:1606–1618. [DOI] [PubMed] [Google Scholar]
  • 13.Roux AVD, Merkin SS, Arnett D, Chambless L, Massing M, Nieto FJ, Sorlie P, Szklo M, Tyroler HA, Watson RL. Neighborhood of Residence and Incidence of Coronary Heart Disease. New England Journal of Medicine. 2001;345:99–106. [DOI] [PubMed] [Google Scholar]
  • 14.Southern WN, Bellin EY, Arnsten JH. Longer Lengths of Stay and Higher Risk of Mortality among Inpatients of Physicians with More Years in Practice. Am J Med. 2011;124:868–874. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Jackson CL, Ahmad Z, Das SR, Khera A. The evaluation and management of patients with LDL-C ≥ 190 ​mg/dL in a large health care system. Am J Prev Cardiol. 2020;1:10002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Virani SS, Kennedy KF, Akeroyd JM, Morris PB, Bittner VA, Masoudi FA, Stone NJ, Petersen LA, Ballantyne CM. Variation in Lipid-Lowering Therapy Use in Patients With Low-Density Lipoprotein Cholesterol ≥190 mg/dL. Circ Cardiovasc Qual Outcomes. 2018;11:e004652. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Banda JM, Sarraju A, Abbasi F, Parizo J, Pariani M, Ison H, Briskin E, Wand H, Dubois S, Jung K, Myers SA, Rader DJ, Leader JB, Murray MF, Myers KD, Wilemon K, Shah NH, Knowles JW. Finding missed cases of familial hypercholesterolemia in health systems using machine learning. NPJ Digit Med. 2019;2: s41746–019-0101–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Safarova MS, Liu H, Kullo IJ. Rapid identification of familial hypercholesterolemia from electronic health records: The SEARCH study. J Clin Lipidol. 2016;10:1230–1239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Green P, Neely D, Humphries SE, Saunders T, Gray V, Gordon L, Payne J, Carter S, Neuwirth C, Rees A, Gallagher H. Improving detection of familial hypercholesterolaemia in primary care using electronic audit and nurse-led clinics. J Eval Clin Pract. 2016;22:341–348. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Myers KD, Knowles JW, Staszak D, Shapiro MD, Howard W, Yadava M, Zuzick D, Williamson L, Shah NH, Banda JM, Leader J, Cromwell WC, Trautman E, Murray MF, Baum SJ, Myers S, Gidding SS, Wilemon K, Rader DJ. Precision screening for familial hypercholesterolaemia: a machine learning study applied to electronic health encounter data. Lancet Digit Health. 2019;1:e393–e402. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Sabatine MS, Giugliano RP, Keech AC, Honarpour N, Wiviott SD, Murphy SA, Kuder JF, Wang H, Liu T, Wasserman SM, Sever PS, Pedersen TR. Evolocumab and Clinical Outcomes in Patients with Cardiovascular Disease. New England Journal of Medicine. 2017;376:1713–1722. [DOI] [PubMed] [Google Scholar]
  • 22.Schwartz GG, Steg PG, Szarek M, Bhatt DL, Bittner VA, Diaz R, Edelberg JM, Goodman SG, Hanotin C, Harrington RA, Jukema JW, Lecorps G, Mahaffey KW, Moryusef A, Pordy R, Quintero K, Roe MT, Sasiela WJ, Tamby J-F, Tricoci P, White HD, Zeiher AM. Alirocumab and Cardiovascular Outcomes after Acute Coronary Syndrome. New England Journal of Medicine. 2018;379:2097–2107. [DOI] [PubMed] [Google Scholar]
  • 23.Castelli WP. Cholesterol and lipids in the risk of coronary artery disease. The Framingham Heart Study. Canadian Journal of Cardiology. 1988;4:5A–10A. [PubMed] [Google Scholar]
  • 24.Fernández-Ruiz I Very high HDL-C levels are associated with higher mortality in patients with CAD. Nat Rev Cardiol. 2022;19:504–504. [DOI] [PubMed] [Google Scholar]
  • 25.Ruiz JM, Steffen P, Smith TB. Hispanic mortality paradox: A systematic review and meta-analysis of the longitudinal literature. Am J Public Health. 2013;103:e52–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Chen Y, Freedman ND, Rodriquez EJ, Shiels MS, Napoles AM, Withrow DR, Spillane S, Sigel B, Perez-Stable EJ, Berrington De González A. Trends in Premature Deaths among Adults in the United States and Latin America. JAMA Netw Open. 2020;3:e1921085. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Lavie CJ, Milani R v., Ventura HO. Obesity and Cardiovascular Disease. Risk Factor, Paradox, and Impact of Weight Loss. J Am Coll Cardiol. 2009;53:1925–1932. [DOI] [PubMed] [Google Scholar]
  • 28.Lassale C, Cené CW, Asselin A, Sims M, Jouven X, Gaye B. Sociodemographic determinants of change in cardiovascular health in middle adulthood in a bi-racial cohort. Atherosclerosis. 2022;346:98–108. [DOI] [PubMed] [Google Scholar]
  • 29.Caleyachetty R, Echouffo-Tcheugui JB, Muennig P, Zhu W, Muntner P, Shimbo D. Association between cumulative social risk and ideal cardiovascular health in US adults: NHANES 1999–2006. Int J Cardiol. 2015;191:296–300. [DOI] [PubMed] [Google Scholar]
  • 30.Virani SS, Aspry K, Dixon DL, Ferdinand KC, Heidenreich PA, Jackson EJ, Jacobson TA, McAlister JL, Neff DR, Gulati M, Ballantyne CM. The Importance of Low-Density Lipoprotein Cholesterol Measurement and Control as Performance Measures: A Joint Clinical Perspective from the National Lipid Association and the American Society for Preventive Cardiology. J Clin Lipidol. 2023;17:208–218. [DOI] [PubMed] [Google Scholar]
  • 31.Jacobs DB, Schreiber M, Seshamani M, Tsai D, Fowler E, Fleisher LA. Aligning Quality Measures across CMS — The Universal Foundation. New England Journal of Medicine. 2023:388:776–779. [DOI] [PubMed] [Google Scholar]
  • 32.Tsao CW, Aday AW, Almarzooq ZI, Anderson CAM, Arora P, Avery CL, Baker-Smith CM, Beaton AZ, Boehme AK, Buxton AE, Commodore-Mensah Y, Elkind MSV, Evenson KR, Eze-Nliam C, Fugar S, Generoso G, Heard DG, Hiremath S, Ho JE, Kalani R, Kazi DS, Ko D, Levine DA, Liu J, Ma J, Magnani JW, Michos ED, Mussolino ME, Navaneethan SD, Parikh NI, Poudel R, Rezk-Hanna M, Roth GA, Shah NS, St-Onge M-P, Thacker EL, Virani SS, Voeks JH, Wang N-Y, Wong ND, Wong SS, Yaffe K, Martin SS. Heart Disease and Stroke Statistics—2023 Update: A Report From the American Heart Association. Circulation. 2023;147:e93–e621. [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.

Supplementary Materials

Supplemental Publication Material
NIHMS1941443-supplement-Supplemental_Publication_Material.pdf (606.2KB, pdf)

RESOURCES