Identifying Familial Hypercholesterolemia Using a Blood Donor Screening Program With More Than 1 Million Volunteer Donors

Key Points

Question  What is the prevalence of familial hypercholesterolemia in a population of blood donors?

Findings  In this cohort study of 1 178 102 individual blood donors, 3473 individuals (or 1 of every 339) met criteria for familial hypercholesterolemia.

Meaning  Blood donation may serve as a novel method of identifying individuals with familial hypercholesterolemia.

Importance  Familial hypercholesterolemia is an autosomal-dominant disorder that often causes premature coronary artery disease. Unfortunately, familial hypercholesterolemia remains largely undiagnosed.

Objective  To estimate the prevalence of familial hypercholesterolemia in a population of blood donors.

Design  This analysis of deidentified data from blood donors 16 years and older who donated to Carter BloodCare, one of the largest independent blood programs in the United States, between January 2002 and December 2016. Carter BloodCare, which serves a population of about 8 million in Texas, routinely measures total nonfasting serum cholesterol levels as part of a donor health screening program. Data analysis occurred from October 2017 to March 2019.

Exposure  Blood donation.

Main Outcomes and Measures  Familial hypercholesterolemia was defined using the Make Early Diagnosis to Prevent Early Death general population criteria, with total nonfasting serum cholesterol thresholds of 270, 290, 340, and 360 mg/dL for donors younger than 20 years, 20 to 29 years, 30 to 39 years, and 40 years or older, respectively (to convert cholesterol values to mmol/L, multiply by 0.0259). For repeated donors, the maximum observed total cholesterol level was used for analyses.

Results  The study included 1 178 102 individual donors with a total of 3 038 420 blood donations. Of all individual donors (median total cholesterol level, 183 [interquartile range (IQR), 157-212] mg/dL; median age, 32 [IQR, 19-47] years; 619 583 [52.6%] women), a total of 3473 individuals (or 1 in every 339) met criteria for familial hypercholesterolemia. This group had a median (IQR) total cholesterol of 332 (297-377) mg/dL. Estimated prevalence was higher at younger ages (<30 years: 1:257) compared with older ages (≥30 years: 1:469; P < .001) and in men (1:327) compared with women (1:351; P = .03). Among 2219 repeated donors who met familial hypercholesterolemia criteria at least once, 3116 of 10 833 total donations (28.8%) met FH criteria.

Conclusions and Relevance  The prevalence of familial hypercholesterolemia using the Make Early Diagnosis to Prevent Early Death criteria in a large cohort of blood donors was similar to the estimated prevalence of this disorder in the general population. The blood donor screening program could be a novel strategy to detect and notify individuals with potential familial hypercholesterolemia, particularly younger individuals in whom early detection and treatment is especially helpful, as well as guide cascade screening.

Familial hypercholesterolemia (FH) is an autosomal-dominant disorder characterized by premature coronary artery disease (CAD) owing to lifelong elevated low-density lipoprotein cholesterol levels.^1,2 The prevalence of heterozygous FH in the general population is around 1 in 250, with more than 20 million individuals affected worldwide.¹ Unfortunately, FH remains underdiagnosed and undertreated,^2,3 with diagnosis crucial for prompting initiation of cholesterol-lowering medications to reduce CAD risk.^4,5 Presently, no coordinated FH screening system exists in the United States, but novel strategies have been proposed.^6,7

Blood donation may serve a broader public health role through disease screening. Beyond testing donors for transmissible infections, screening can extend to noninfectious conditions or risks.⁸ For example, many blood donation centers provide total cholesterol (TC) screening at no charge,⁹ and some measure glycated hemoglobin to screen for diabetes.^10,11 Thus, we sought to estimate the prevalence of FH in a very large database of more than 1 million blood donors as a preliminary step in assessing blood donation as a mechanism for FH screening and intervention.

Donors 16 years or older who voluntarily donated blood to Carter BloodCare between January 2002 and December 2016 were included. Donors missing data for age or TC were excluded. All blood donors consented to the donation process and cholesterol testing, and donors who were 16 years of age required signed parental consent. The institutional review board of the University of Texas Southwestern Medical Center approved this study, and requirement for informed consent was waived.

Deidentified data were obtained from the Carter BloodCare database. Demographic data, including age at the time of donation, sex, and race/ethnicity, were routinely collected, and TC was measured for each donation. Donors were provided a unique identification code at each donation for online retrieval of TC results.

To classify FH, the Make Early Diagnosis to Prevent Early Death (MEDPED) criteria were used, with TC thresholds of 270, 290, 340, and 360 mg/dL for donors younger than 20 years, 20 to 29 years, 30 to 39 years, and 40 years or older, respectively¹² (to convert cholesterol values to millimoles per liter, multiply by 0.0259). For repeated donors, the maximum TC value was used for FH classification.

Nonfasting TC was measured from 2002 through 2009 using the Abbott Aeroset System (Abbott Laboratories) and from 2010 through 2016 using the Beckman Coulter AU680 Chemistry Analyzer (Beckman Coulter Diagnostics). Both assays have a total coefficient of variation of less than 3%.

Results are reported as median (interquartile range [IQR]) for continuous variables and number (percentage) for categorical variables. The proportions of FH were compared across demographic categories with the χ² test for 2-sample comparisons and the χ² trend test for more than 2 categories. Statistical comparisons for TC values across age categories were performed using the Jonckheere-Terpstra trend test. All statistical analyses were performed using SAS version 9.4 (SAS Institute) with a 2-sided P value of .05 used to define statistical significance.

The study included data from 1 178 102 donors (3 038 420 total blood donations). Baseline characteristics are displayed in Table 1. The median (IQR) age was 32 (19-47) years, 619 583 participants (52.6%) were women, and 763 713 participants (64.8%) were white. The median TC was 183 (IQR, 157-212) mg/dL, and the median number of donations was 1 (IQR, 1-3).

A total of 3473 donors met MEDPED criteria for FH, a prevalence of 1 of every 339 donors (Table 2). The prevalence was greater for donors younger than 30 years (1:257 [2127 of 546 619]) vs older donors 30 years or older (1:469 [1346/631 483]; P < .001) and men (1:327 [1708/558 462]) vs women (1:351 [1765/619 583]; P = .03). There was no significant difference by race, except for Asian donors, who had a slightly higher prevalence (1:248 for Asian individuals [n = 128 of 31 773] vs 1:342 for people of all other ethnicities [n = 3345 of 1 146 329]; P < .001) than other groups. In sensitivity analyses, estimated prevalence of FH was not significantly different with different cholesterol assays (2002-2009, 1:329 [1622/533 784]; 2010-2016, 1:348 [1852/644 318]; P = .10).

Among donors who met FH criteria, 1254 (36.1%) donated only once, and 2219 (63.9%) were repeated donors. Among the repeated donors meeting FH criteria at least once, 3116 of their 10 833 donations (28.8%) met FH criteria (Figure). This included 779 of 1629 donations (47.8%) meeting less stringent MEDPED TC criteria (TC >270 mg/dL; age <20 years), compared with 954 of 4808 donations (19.8%) meeting the most stringent criteria (TC >360 mg/dL; age >40 years). In the second group, a larger proportion of repeated TC values were close to normal levels (TC <210 mg/dL) (1508 of 4808 [31.4%] vs 276 of 1629 [16.9%]).

On a donor level, we evaluated the differences in maximum and minimum TC levels for different groups. Among repeated donors who met FH criteria only once, the median difference between maximum and minimum TC values between their donations was 109 (interquartile range, 62-177) mg/dL (a 33% difference). In contrast, the difference in TC for repeated donors who never met FH criteria was only 28 (14-47) mg/dL (a 14% difference).

In a cohort of more than 1 million blood donors, 3473 met MEDPED criteria for FH, a prevalence of 1:339; this is comparable with the US population estimate of 1:250.¹³ Additionally, prevalence was higher at younger ages (<30 years), highlighting the potential for early diagnosis and intervention. Among repeated donors who met FH criteria at least once, achieving FH criteria persisted over multiple donations in a considerable proportion. These data suggest that blood donation could represent a novel approach to screen for FH, particularly in the younger population.

Despite significant advances in identifying mutations and molecular therapeutic targets related to FH, it remains underdiagnosed and undertreated.^2,3 Blood donation programs represent a unique opportunity as a public health portal to screen for diseases such as FH. Approximately 6.4 million people in the United States donate blood every year, and 29% are first-time donors.¹⁴ This study demonstrated that blood donors tend to be relatively young (median age, 32 [IQR, 19-47] years), with the prevalence of FH greater at younger ages. Several explanations could account for this observation, including lower MEDPED TC thresholds for younger individuals, survival bias owing to individuals with FH dying of premature CAD, and lower TC levels in older individuals affected by lipid-lowering treatment.

The present data serve as preliminary evidence supporting the potential for using blood donor programs as methods for FH screening. Identification of FH in younger donors creates opportunities for early intervention to reduce the risk of premature CAD. Blood donors are only eligible to donate if they are free of cancer; free of significant cardiovascular, pulmonary, and infectious disease; and not taking certain medications. Thus, these young, healthy individuals are unlikely to regularly access the traditional medical system. Furthermore, the incremental cost of TC testing is negligible compared with other tests routinely performed on blood donations.⁹

Ultimately, a focused intervention will be required. All donors in the Carter BloodCare system can access cholesterol results online, but few do so.⁹ Those with TC greater than 300 mg/dL are sent a letter recommending follow-up with a physician. Nevertheless, even among those meeting the highest MEDPED TC threshold (>360 mg/dL), fewer than one-third of their subsequent donations had near-normal TC values (31% with TC <210 mg/dL), suggesting inadequate identification and treatment of potential FH.

Components of a focused program can include enhanced questionnaire data capture regarding family and personal medical history for FH diagnostic criteria, more directed communication about the potential for FH in appropriate donors, facilitating connections to health care professionals with FH expertise, and instituting a mechanism for family cascade screening. However, such a program will require partnerships between blood donor programs and the health care system and with academic, public health, and FH advocacy groups.

Among those meeting FH criteria at least once, repeated donations met criteria around 29% of the time. Failure to meet criteria on subsequent donations could represent variability in TC measurements, reversible secondary causes, or treatment effects. We observed larger difference between maximum and minimum TC in donors meeting FH criteria (33%) vs those not meeting criteria (14%). Furthermore, those with the most stringent MEDPED TC criteria (>360 mg/dL) also had the largest proportion of subsequent TC values in the near-normal range. The percentage difference in TC is larger than what might be expected owing to random TC measurement variability and is possibly because of treatment effects. Unfortunately, regulators and accreditors only require ascertainment of specific categories of drugs (eg, antibiotics, antiplatelets, anticoagulants, insulins), and cholesterol-lowering medication use data were not collected. However, even if cholesterol treatment were implemented in these individuals, the possibility of FH and implications for cascade screening may not have been acknowledged.

There are notable limitations of the study. First, MEDPED criteria were the only diagnostic measures that could be applied, because only TC levels were available; however, these criteria have not been extensively evaluated in multiethnic populations. Furthermore, additional data on blood donors, such as cholesterol medications, personal or family medical history, health insurance status, and access to health care, were not available. Thus, we were unable to determine whether there were any secondary causes of hypercholesterolemia.

Applying MEDPED criteria for FH in a study of more than 1 million blood donors, the estimated prevalence of FH was similar to other estimates in the general population. Therefore, blood donation represents a novel and potentially cost-effective approach to identifying FH, particularly in younger individuals who are not engaged in the medical system. Identifying possible cases of FH has important implications for early treatment as well as cascade screening of family members.

Accepted for Publication: March 21, 2019.

Corresponding Author: Amit Khera, MD, MSc, Division of Cardiology, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390-8830 (amit.khera@utsouthwestern.edu).

Published Online: May 22, 2019. doi:10.1001/jamacardio.2019.1518

Author Contributions: Dr Khera had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Jackson, Ahmad, McGuire, Sayers, Khera.

Acquisition, analysis, or interpretation of data: Jackson, Keeton, Eason, Ayers, Gore, McGuire, Sayers, Khera.

Drafting of the manuscript: Jackson, Keeton, Khera.

Critical revision of the manuscript for important intellectual content: Keeton, Eason, Ahmad, Ayers, Gore, McGuire, Sayers, Khera.

Statistical analysis: Jackson, Keeton, Ayers.

Obtained funding: Sayers.

Administrative, technical, or material support: Jackson, Keeton, Eason, Sayers, Khera.

Supervision: Ahmad, Khera.

Conflict of Interest Disclosures: Dr McGuire reports honoraria for clinical trial leadership from AstraZeneca, Sanofi Aventis, Janssen, Boehringer Ingelheim, Merck & Co, Novo Nordisk, Lexicon, Eisai Inc, GlaxoSmithKline, Esperion, and Lilly USA and honoraria for consultancy from AstraZeneca, Sanofi Aventis, Lilly USA, Boehringer Ingelheim, Merck & Co, Pfizer, Novo Nordisk, Metavant, and Applied Therapeutics. Dr Ahmad reports grants from Regeneron and personal fees from Akcea outside the submitted work, research funds from Regeneron, and honorarium for educational talks and advisory board meetings from Akcea. Dr Ayers reports grants and personal fees from the National Heart, Lung, and Blood Institute, outside the submitted work. Mr Eason and Dr Sayers are employed by but have no equity interests in Carter BloodCare. No other disclosures were reported.

Funding/Support: This study is supported by the National Heart, Lung, and Blood Institute (grant K2-HL131939 [Dr Gore]) for a project that includes screening blood donors for cardiovascular risk factors, and the Doris Duke Charitable Foundation/University of Colorado Fund to Retain Clinical Scientists (Dr Gore).

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.