Cosmos Study

New 2026 Hyperlipidemia Guideline Expand Treatment Eligibility from 41% to 55% of Adults

August 5, 2026

Team A:Kersten Bartelt, RNEmily Higgs

Team B:Ryan Stolcpart, PharmDEric Barkley

Key Findings

Under the 2026 cholesterol treatment guideline, 54.6% of adults aged 20–79 without established cardiovascular disease were eligible for lipid-lowering therapy, compared with 41.3% under the 2018 guideline, a 13.3 percentage-point increase in the treatment-eligible share.
Overall, 13.9% of patients gained eligibility for treatment under the 2026 guideline, while 0.6% lost eligibility.
Among patients who gained eligibility under the 2026 guideline, women and older adults were over-represented. Women made up 66.4% of the population that gained eligibility but account for only 57.9% of the overall study population. Likewise, patients aged 70–79 made up 36.1% of the eligibility gainers but account for only 15.8% of the overall study population. Among the much smaller group who lost eligibility, men, Black patients, and adults aged 50–59 were over-represented (men: 79.6% vs. 42.1% overall; Black patients: 47.5% vs. 11.6%; aged 50–59: 50.1% vs. 21.3%).
Despite these compositional differences, treatment eligibility net increased within every demographic subgroup examined, including men (+10 percentage points), Black patients (+8 percentage points), and adults aged 50–59 (+10 percentage points).

High cholesterol (also known as hyperlipidemia) is a major cause of heart attacks and strokes. Lowering cholesterol with medications like statins is one of the most common ways adults are treated to prevent a first cardiovascular event. National guidelines help clinicians decide who benefits from starting these medications, and when those guidelines change, millions of people can move into or out of a treatment recommendation. The 2018 American Heart Association (AHA) and American College of Cardiology (ACC) guideline recommended statin therapy for adults aged 40–75 based on the Pooled Cohort Equations (PCE), a risk calculator that estimates a person’s 10-year chance of a cardiovascular event using age, sex, race, blood pressure, cholesterol, smoking status, and diabetes status.¹ In March 2026, the AHA and ACC released an updated hyperlipidemia guideline that retired the PCE in favor of the newer PREVENT-ASCVD (Predicting Risk of Cardiovascular Disease EVENTs-Atherosclerotic Cardiovascular Disease) equations, which incorporate kidney function and statin use and do not use race as an input.² The 2026 guideline also lowered the risk threshold for considering therapy (3% over 10 years, down from 7.5%), expanded eligibility for younger patients and those aged 76 to 79, and elevated chronic kidney disease (CKD) and HIV from risk modifiers to standalone qualifying conditions.³ The PREVENT equations produce different risk estimates than the PCE across demographic groups,⁴ but the practical effect of the new guideline on who is offered treatment, and whether some groups gain or lose eligibility, has not yet been characterized in a large real-world population.

We studied more than 21 million U.S. adults aged 20–79 with an outpatient visit between March 2025 and February 2026, established prior care, total cholesterol results, and HDL results, who did not have prior atherosclerotic cardiovascular disease (ASCVD). Each patient was evaluated under both the 2018 and 2026 treatment frameworks using lab results, diagnoses, medications, and demographic information from their chart. Patients qualified for treatment if they met any pathway under a given guideline (for example, an LDL cholesterol of 190 mg/dL or higher, a diagnosis of diabetes, or a calculated 10-year ASCVD risk above the relevant threshold). Patients were then placed into one of four mutually exclusive categories: eligible under both guidelines, newly eligible under the 2026 guideline only, no longer eligible under the 2026 guideline, or ineligible under both.

Treatment eligibility expanded substantially under the 2026 guideline. Across all studied patients, 54.6% met the 2026 criteria compared with 41.3% under the 2018 criteria, as shown in Figure 1. Of all studied patients, 40.7% were eligible under both guidelines, 13.9% were newly eligible (eligible under the 2026 guideline but not the 2018 guideline), and 0.6% were no longer eligible (eligible under the 2018 guideline but not the 2026 guideline); the remaining 44.8% were ineligible under either framework. The newly eligible group was roughly 23 times larger than the no longer eligible group, indicating that the 2026 guideline operates almost entirely as an expansion rather than a substitution of the prior treatment population.

Figure 1

Treatment Eligibility Under the 2018 and 2026 Hyperlipidemia Guidelines

Figure 1. The distribution of patients aged 20–79 without prior atherosclerotic cardiovascular disease who qualify for lipid-lowering therapy under the 2018 guideline, the 2026 guideline, both, or neither.

The patients who changed eligibility status under the 2026 guidelines were not demographically representative of the overall study population, as shown in Figure 2. Among those who gained eligibility, 66.4% were women (while women accounted for 57.9% of the overall study population) and 36.1% were aged 70–79 (while that age group made up 15.8% of the overall study population), reflecting the extension of the eligible age range from 75 to 79 and the addition of CKD as a standalone qualifying condition. The much smaller group who lost eligibility skewed in the opposite direction: 79.6% were men (vs. 42.1% overall), 47.5% were Black (vs. 11.6% overall), and 50.1% were aged 50–59 (vs. 21.3% overall). This composition is directionally consistent with prior work showing that the PREVENT equations estimate lower 10-year ASCVD risk than the PCE for some demographic groups, particularly Black adults, in part because PREVENT does not use race as an input.⁴

Figure 2

Demographic Distribution of Patients Who Gain and Lose Eligibility

Figure 2. The distribution of patients aged 20–79 without prior atherosclerotic cardiovascular disease who gained eligibility and lost eligibility for hyperlipidemia treatment.

Although the group that lost eligibility was disproportionately male, Black, and middle-aged, these patients were far outnumbered by their counterparts who gained eligibility, so every subgroup gained treatment eligibility on net, as shown in Figure 3. Eligibility rose by 16 percentage points among women and 10 points among men. Gains increased with age, from less than a percentage point among adults aged 20–29 to a 32 percentage-point jump among those aged 70–79. One population that stands out from that pattern is patients aged 30–39 who had a 15 percentage point jump, surpassing the 3 percentage-point jump seen for those aged 40–49. This might be explained by the screening age being lowered from 40 to 30. Every race and ethnicity group also gained, ranging from an 8 percentage-point increase among Black patients to 15 points among White patients. In other words, the demographic skew among those who lost eligibility reflects which patients were most affected by the change to the PREVENT equations, not a net reduction in treatment eligibility for any group.

Figure 3

Rate of Eligibility Change Under Hyperlipidemia Guidelines by Patient Demographics

Figure 3. The rate of eligibility change under the 2018 and 2026 hyperlipidemia guidelines by sex, age, and race/ethnicity.

These data come from Cosmos, a dataset created in collaboration with a community of Epic health systems representing more than 307 million patient records from 2,000 hospitals and more than 49,000 clinics from all 50 U.S. states, Canada, Lebanon, and Saudi Arabia. This study was completed by two teams that worked independently, each composed of a clinician and research scientist. The two teams came to similar conclusions. Graphics by Brian Olson.

Grundy SM, Stone NJ, Bailey AL, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. 2019;139(25):e1082-e1143. doi:10.1161/CIR.0000000000000625
Khan SS, Matsushita K, Sang Y, et al. Development and validation of the American Heart Association’s PREVENT equations. Circulation. 2024;149(6):430-449. doi:10.1161/CIRCULATIONAHA.123.067626
Blumenthal RS, Morris PB, Gaudino M, et al. 2026 ACC/AHA/AACVPR/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of dyslipidemia: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation. Published online March 13, 2026. doi:10.1161/CIR.0000000000001423
Anderson TS, Wilson LM, Sussman JB. Atherosclerotic cardiovascular disease risk estimates using the Predicting Risk of Cardiovascular Disease Events equations. JAMA Intern Med. 2024;184(8):963-970. doi:10.1001/jamainternmed.2024.1302

Data Definitions

Study period: 3/1/2025 to 2/28/2026

Study population: inclusion

Adults aged 20–79 with:

A known sex of female or male
Residence in the U.S.
At least 1 outpatient face-to-face visit within 18 months before index date
Total cholesterol (TC) and HDL result on the same day (defines index date)
For score-based pathway classification, the following labs are required within 18 months before index date: eGFR, systolic blood pressure (SBP)

Outpatient face-to-face visit: An encounter of type Emergency, Office Visit, Follow-up, Telemedicine, Urgent Care, Walk-In, Routine Prenatal, Postpartum Visit, Fetal Care Consult, or Hospital Outpatient Visit

Study population: exclusion: History of ASCVD on the chart on or before the end of the study period
Evidence of heart transplant before the end of the study period
Pregnancy: ICD-10-CM O* codes or SNOMED code 77386006 between six months prior to the study period and the end of the study period, OR a birth record anywhere in the study period, OR a pregnancy record between six months prior to the study period and the end of the study period
Patients classifiable through non-score pathways (LDL ≥190, LDL persistently ≥160, diabetes, CKD, HIV) do not require a complete set of score inputs. Patients who do not qualify through any non-score pathway and have missing score inputs are classified as not assessable and are excluded.

Index date: Date of a total cholesterol (TC) and HDL lab result on the same day within the study period. If multiple qualifying dates exist, use the earliest.

Exposures: Guideline framework applied: 2018 ACC/AHA Blood Cholesterol Guideline vs. 2026 ACC/AHA Dyslipidemia Guideline. Each patient is evaluated under both frameworks. Note: Patients aged 20–29 cannot be assessed by either risk score. Patients aged 30–39 are assessable under PREVENT-ASCVD but not PCE. Patients aged 76–79 are assessable under PREVENT-ASCVD but not PCE. When patients are outside of the prescribed ages for the evaluation, they are categorized as “Ineligible for Treatment.”

2018 guideline: treatment pathways: Pathway 1 — LDL ≥190 mg/dL or history of familial hypercholesterolemia (age 20–75)
Pathway 2 — Diabetes (age 40–75)
Pathway 3 — PCE 10-year ASCVD risk ≥7.5% (age 40–75)
Pathway 4 — PCE 5% to <7.5% + 2018 risk enhancers (age 40–75)
INELIGIBLE FOR TREATMENT
Age <40 or >75 without diabetes or LDL ≥190, OR PCE <7.5% without risk enhancers, diabetes, or LDL ≥190

2026 guideline: treatment pathways: Uses the CPR (Calculate–Personalize–Reclassify) framework with PREVENT 10-year ASCVD score.
Pathway 1 — LDL ≥190 mg/dL or history of familial hypercholesterolemia
Pathway 2 — LDL persistently ≥160 mg/dL (age 20–79). “Persistently” = 2+ results ≥160 mg/dL separated by ≥30 days.
Pathway 3 — Diabetes, CKD, or HIV (age 30–79). CKD and HIV elevated from risk enhancers to standalone pathways (new in 2026).
Pathway 4 — PREVENT 10-year ASCVD score ≥3% (age 30–79)
INELIGIBLE FOR TREATMENT
PREVENT 10-year ASCVD score <3% without diabetes/CKD/HIV and LDL <160.

Outcomes: Eligible under both: meets treatment criteria under both 2018 and 2026 frameworks
Newly eligible (2026 only): does NOT meet 2018 criteria but DOES meet 2026 criteria
No longer eligible (2018 only): meets 2018 criteria but does not qualify via any 2026 pathway
Ineligible under both: meets neither 2018 nor 2026 criteria

ASCVD: Any of the following diagnoses or procedures on the patient’s chart on or before the end of the study period: acute coronary syndrome / angina, myocardial infarction, stroke, TIA, peripheral arterial disease, revascularization (diagnosis), revascularization (procedure)

Heart transplant: Any of the following on the patient’s chart on or before the end of the study period: a procedure with CPT codes 33945, 33935, or 33929; a diagnosis with ICD-10-CM code Z94.1 or T86.2* or SNOMED code 32413006 or 429272001; or an episode for a heart transplant

Acute coronary syndrome / angina: Unstable angina: ICD-10-CM code I20.0 or SNOMED code 25106000. Other angina: ICD-10-CM code I20.8, I20.9 or SNOMED code 194828000. Acute ischemic heart disease, unspecified: ICD-10-CM code I24.9 or SNOMED code 414795007. ASHD of native coronary with angina: ICD-10-CM code I25.110, I25.11*, I25.118, I25.119

Myocardial infarction: Acute MI: ICD-10-CM code I21* (excluding I21.B), I22* or SNOMED code 22298006. Postinfarction angina: ICD-10-CM code I23.7. Old MI: ICD-10-CM code I25.2 or SNOMED code 399211009

Stroke: ICD-10-CM code I63* or SNOMED code 432504007

TIA: ICD-10-CM code G45* or Z86.73 or SNOMED code 266257000 or 275526006

Peripheral arterial disease: Atherosclerosis: ICD-10-CM code I70* (excluding I70.0 [aorta] and I70.1 [renal artery]) or SNOMED code 399957001. Peripheral vascular disease, unspecified: ICD-10-CM code I73.9 or SNOMED code 400047006

Revascularization (diagnosis): ASHD of bypass graft with angina: ICD-10-CM code I25.7* (excluding I25.75*) or SNOMED code 399208008. ASHD of bypass graft without angina: ICD-10-CM code I25.810, I25.812. Presence of coronary angioplasty implant/graft: ICD-10-CM code Z95.5. Mechanical complication of coronary bypass: ICD-10-CM code T82.21*. Graft complications: ICD-10-CM codes T82.310*–T82.332*, T82.390*–T82.392*. History of percutaneous coronary intervention: SNOMED code 429639007. Presence of coronary stent: SNOMED code 706882009

Revascularization (procedure): PCI: CPT 92920–92944; ICD-10-PCS 027*. CABG: CPT 33510–33536; ICD-10-PCS 0210*–0213*. Carotid revascularization: CPT 35301, 35501, 37215, 37216

Familial hypercholesterolemia: ICD-10-CM code E78.01 or SNOMED code 398036000

PCE 10-year ASCVD score: Pooled Cohort Equations. Valid for adults aged 40–75 without ASCVD. Uses race and ethnicity as an input (White/African American/Other). Inputs: age, sex, race and ethnicity, total cholesterol (TC), HDL-C, systolic blood pressure (SBP), anti-hypertensive medication status, diabetes status, current smoker status.
2018 risk categories: Low <5%; Borderline 5% to <7.5%; Intermediate 7.5% to <20%; High ≥20%.
Note: PCE is known to overestimate 10-year ASCVD risk by 40–50% compared to observed event rates in contemporary populations.

PREVENT 10-year ASCVD score: PREVENT-ASCVD equation (MI + stroke only; not total CVD). Valid for adults aged 30–79 without ASCVD. Does NOT include race and ethnicity as an input. Values used in score were from index date or within the last 18 months. If any were missing, the score was not calculated.
Risk = exp(log-Odds) / (1 + exp(log-Odds)). Sex-specific log-Odds equations use age, TC, HDL, SBP, diabetes, current smoker, eGFR, anti-hypertensive medication, and statin use, with interaction terms. TC and HDL values must be in mmol/L (mg/dL × 0.02586 = mmol/L). Full coefficient list available in the source citation.
2026 risk categories: Low <3%; Borderline 3% to <5%; Intermediate 5% to <10%; High ≥10%.

2018 risk enhancers: Used for patients at borderline (5–<7.5%) or intermediate (7.5–<20%) PCE risk.
Family history of premature ASCVD (men <55, women <65): ICD-10-CM code Z82.4*
Persistently elevated LDL ≥160 mg/dL (2+ results ≥160 separated by ≥30 days)
Metabolic syndrome (3+ of: waist circumference, TG ≥150, HDL <40 M/<50 F, SBP ≥130 or diastolic ≥85, fasting glucose ≥100)
CKD
History of preeclampsia: ICD-10-CM code O14*–O15* or SNOMED code 15938005
Premature menopause (<40 y): ICD-10-CM code E28.31* or SNOMED code 373717006
Chronic inflammatory disorders (RA, psoriasis, HIV): ICD-10-CM code M05*–M06*, L40*, or B20* or SNOMED code 69896004 or 9014002 (Note: in 2026, HIV is elevated to a standalone treatment pathway.)
Persistently elevated TG ≥175 mg/dL (2+ results ≥175 separated by ≥30 days)
If measured: hsCRP ≥2.0 mg/L; Lp(a) ≥50 mg/dL or ≥125 nmol/L; apoB ≥130 mg/dL; ABI <0.9

Diabetes: A diagnosis with ICD-10-CM code E08*–E13* or SNOMED code 73211009

HIV: A diagnosis with ICD-10-CM code B20* or SNOMED code 86406008

CKD: A diagnosis with ICD-10-CM code N18* or SNOMED code 709044004

Anti-hypertensive medication: An order (prescription, administration, or patient-reported) or dispense between 18 months prior to the index date and the index date with ATC code C08*, C07*, C09*, C03*, or C02*

Statin: An order (prescription, administration, or patient-reported) or dispense between 18 months prior to the index date and the index date with a pharmaceutical subclass containing HMG CoA Reductase Inhibitors (statins), including statin combination products. Used as a covariate in the PREVENT-ASCVD formula (statin-use interaction term), not as the outcome.

Current smoker: Most recent smoking status documented is one of: Light Smoker, Heavy Smoker, Every Day, Current Every Day Smoker, Light Tobacco Smoker, Current Some Day Smoker, Smoker Current, or Some Days; OR packs per day > 0. Patients with no recorded status or another status were considered not a current smoker.

eGFR: Calculated from the creatinine lab value using the CKD-EPI Creatinine Equation (2021).

Creatinine: A lab result with LOINC code 38483-4 or 2160-0 from within 18 months prior to the index date that does not overlap with an admission

HDL: A lab result with LOINC code 2085-9 or 18263-4 from within 18 months prior to the index date

Total cholesterol (TC): A lab result with LOINC code 2093-3 from within 18 months prior to the index date

LDL: A lab result with LOINC code 13457-7, 18262-6, or 2089-1 from within 18 months prior to the index date

Systolic blood pressure (SBP): Most recent SBP reading from within 18 months prior to the index date. If multiple blood pressure readings are taken on the same day, use the latest for that day.

Lab value quality filters: Exclude records with physiologically implausible values:
HDL outside [10, 170] mg/dL
Total cholesterol (TC) outside [80, 550] mg/dL
SBP outside [70, 270] mmHg
Total cholesterol (TC) ≤ HDL (biologically impossible)
LDL outside [30, 500] mg/dL

Race and ethnicity

Self-reported, mapped to standards. Each patient went into the first bucket they fit in:

Black
Hispanic
Multiracial
White
Other race
Unknown race

PCE (2018): uses race and ethnicity (White/African American/Other) as an input.
PREVENT-ASCVD (2026): does NOT include race and ethnicity.

Stratifications: Age: 20–29, 30–39, 40–49, 50–59, 60–69, 70–79
Sex: female, male
Race and ethnicity
Guideline(s) under which patient is eligible

Model specifications: Descriptive cross-sectional analysis. Distribution of patients across the four-category eligibility outcome under the 2018 vs. 2026 frameworks, with demographic benchmarking against the overall study population. No regression model.

Limitations: 2018 guidelines require a cholesterol panel to calculate PCE. Patients without complete panels who also lack a non-score pathway are unassessable.
Patients currently on statin therapy may have artificially lowered LDL. The PREVENT 10-year ASCVD score formula partially addresses this via the statin-use interaction term.
Family history of ASCVD (Z82.4*) does not specify age of onset (premature = men <55, women <65).
Higher-risk ancestry (South Asian, Filipino) is not available at sufficient granularity in Cosmos.
2026 risk enhancers labs (Lp(a), hsCRP, TG, apoB, ABI) depend on whether ordered. Missingness ≠ absence of risk factor.
Diabetes duration (≥10 y T2DM, ≥20 y T1DM) is difficult to operationalize precisely from diagnosis date alone.
BMI is a factor in some PREVENT sub-models but was not used in this study.
PCE overestimates 10-year ASCVD risk by 40–50% vs. observed event rates, which may inflate 2018 “eligible” counts relative to PREVENT 10-year ASCVD score-based counts.
Patients with no documented race or ethnicity information were categorized as non-Black for PCE scoring.
Cosmos does not contain waist circumference readings or fasting status for blood sugar, which affects the metabolic syndrome definition for the 2018 risk enhancers.