Semaglutide for metabolic dysfunction–associated steatohepatitis (MASH): Estimating eligibility from the 2021–2023 National Health and Nutrition Examination Survey (NHANES)

INTRODUCTION There has been considerable interest in the potential of GLP-1 agonists for treating chronic liver disease.1 Following the positive results of the Phase 3 ESSENCE trial,2 on August 15, 2025, the FDA-approved Wegovy (semaglutide) for the treatment of metabolic dysfunction–associated steatohepatitis (MASH).3 Recently, the eligibility for semaglutide across all current indications in the United States, including diabetes management, weight management, and secondary prevention of cardiovascular disease (CVD), was estimated to be nearly 137 million individuals based on the National Health and Nutrition Examination Survey (NHANES, 2015–2020).4 Although weight management subsumed all eligibility for secondary prevention of CVD, as well as all but 7.6 million who were eligible for diabetes management, insurance may not cover semaglutide for weight management alone. MASH was omitted from those eligibility estimates, as it had not yet been approved for this indication. In this journal, using the 2017–2020 NHANES cohort, it was reported that up to 8.3 million adults would be eligible for resmetirom, the first FDA-approved drug for MASH,5 and practice guidelines for resmetirom therapy were published in Hepatology.6 Inspired by these reports and the FDA approval of semaglutide for MASH, we analyzed the most recent NHANES cohort (2021–2023) to estimate the number of people in the United States who would be eligible for semaglutide for MASH and compared this to eligibility for diabetes. For comparison, we used vibration-controlled transient elastography (VCTE) cutoffs similar to those of the report on resmetirom,5 as well as more conservative ranges for MASH and moderate-to-advanced liver fibrosis (stages F2–F3) validated against liver biopsy.7 METHODS We analyzed the 2021–2023 NHANES dataset to estimate eligibility using broader5 and more conservative7 criteria for MASH. We included non-pregnant adults aged 18 and above with a valid VCTE assessment and a body mass index (BMI) measurement. Broader eligibility was defined as a controlled attenuation parameter (CAP) score >285 dB/m,5 BMI ≥25, and clinically significant fibrosis (CSF) evidenced as liver stiffness measurement (LSM) 8.0–15.0 kPa, excluding possible cirrhosis (LSM >15 kPa),5 excessive alcohol use (>20/30 g/d in women/men), and low platelet count (<150×103 cells/µL). More conservative eligibility was defined as CAP >310 dB/m, BMI ≥25, and LSM 8.2–13.6 kPa.7 Diabetes was defined by the American Diabetes Association (ADA) criteria (fasting A1c ≥6.5 or fasting blood glucose ≥126 mg/dL) and did not include self-reported diabetes in those who do not meet ADA criteria.8 Analyses accounted for the complex sampling procedures (ie, stratum, cluster, and weight) and were conducted using IBM SPSS Statistics (Version 29; Armonk, NY). For brevity, additional methodological details are in the Digital Supplement, https://links.lww.com/HC9/C198 (eg, Figure 1 depicts the determination of the final sample for analysis).FIGURE 1: *Reasons for exclusion applied after a valid VCTE measurement are not mutually exclusive. Analyses excluded individuals with no valid VCTE, no BMI value, heavy alcohol use, and a low platelet count (<150×103 cells/µL). Abbreviations: BMI, body mass index; NHANES, National Health and Nutrition Examination Survey; VCTE, vibration-controlled transient elastography.RESULTS The 2021–2023 NHANES file includes 8112 non-pregnant adults, of whom 5526 had valid VCTE values, resulting in 4773 individuals remaining in the cohort after exclusions (eg, no BMI measure, heavy alcohol use). Table 1 presents weighted sample characteristics for patients eligible for either MASH or diabetes, diabetes, and MASH (broadly and conservatively defined: Note that categories overlap). TABLE 1 - Sample characteristics Variables Full eligible sample (MASH or diabetes)a (n=787) Diabetes-eligible (n=602) MASH-eligible, broader estimate (n=273) MASH-eligible, conservative estimate (n=142) Eligible persons N, million (95% CI) 27.8 (24.4–31.1) 20.4 (17.8–23.1) 10.3 (9.0–11.7) 7.2 (6.0–8.4) % (95% CI) 14.5 (13.0–16.1) 10.7 (9.4–12.1) 5.4 (4.8–6.1) 3.8 (3.1–4.6) Age, years (M±SE) 57.0±0.75 59.2±0.75 52.3±1.23 52.1±1.2 Male sex, % (95% CI) 55.3 (50.6–60.0) 55.1 (50.0–61.1) 55.6 (49.9–61.2) 52.4 (46.9–57.8) Race/ethnicity, % (95% CI) NH Black 14.1 (8.9–21.5) 15.6 (9–25.7) 9.3 (5.1–16.3) 8.6 (4.8–14.8) NH White 55.5 (49.6–61.3) 52.8 (47.2–58.3) 64.0 (51.8–74.6) 62.7 (49.4–74.4) Mexican American 9.4 (3.7–21.7) 9.8 (4.0–22.0) 8.4 (3.3–20.1) 9.9 (3.5–24.6) Other HISP 8.3 (4.9–13.8) 8.6 (4.7–15.1) 7.2 (3.8–13.4) 5.4 (2.7–10.3) Other 12.7 (9.5–16.8) 13.3 (9.3–18.7) 11.1 (6.9–17.2) 13.5 (7.4–23.2) Education, % (95% CI) Less than college 42.2 (35.7–49.0) 45.7 (37.8–53.8) 39.4 (30.5–49.0) 37.0 (27.0–48.1) Some college 33.0 (27.8–38.6) 31.2 (27.0–35.7) 34.8 (25.5–45.4) 32.5 (27.3–48.9) College degree or higher 24.8 (18.0–33.1) 23.1 (16.2–31.8) 25.8 (17.4–36.4) 25.6 (15.8–31.0) Married, % (95% CI) 62.9 (55.8–69.4) 61.5 (52.7–69.5) 66.2 (60.2–71.7) 64.1 (54.2–72.9) Ratio of family income to poverty (M±SE) (n=4264) 2.91±0.12 2.78±0.14 3.12±0.11 3.01±0.18 Smoker, % (95% CI) 13.1 (10.1–16.7) 13.1 (10.1–16.8) 12.7 (8.6–18.5) 15.4 (10.2–22.5) BMI (M±SE) 34.3±.0.36 33.2±0.33 37.3±0.58 37.5±.66 VCTE values (M±SE) CAP, dB/m 315.8±3.10 307.75±3.78 339.9±3.25 351.6±2.55 LSM, kPa 8.9±0.28 8.4±0.37 10.2±0.17 10.0±0.14 Laboratory values (M±SE) HbA1C (%) 7.2±0.10 7.8±0.10 6.3±0.11 6.2±0.10 Platelet count (103 cells/μL) 259.7±2.90 259.4±2.98 264.0±6.26 264.2±9.23 Awareness (n=4687), % Yes (95% CI) Ever told they have liver disease 9.2 (7.5–11.2) 8.7 (7.3–10.3) 10.7 (7.1–15.8) 11.8 (7.0–19.1) Doctor has told them they have diabetes 52.7 (48.5–56.9) 68.1 (64.5–71.6) 25.8 (19.0–34.1) 24.4 (15.3–36.6) Note: Categories overlap.aThe full eligible sample (MASH or diabetes) is based on the broader eligibility for MASH (CAP >285 dB/m, BMI ≥25, LSM 8.0–15.0 kPa, excluding excessive alcohol use and low platelet count).Abbreviations: BMI, body mass index; CAP, controlled attenuation parameter; HISP, Hispanic; LSM, liver stiffness measurement; MASH, metabolic dysfunction–associated steatohepatitis; NH, non-Hispanic; VCTE, vibration-controlled transient elastograph. The weighted population size estimate after exclusion criteria was 191,335,008 million (95% CI: 167–216 million), of whom 5.4% (95% CI: 4.8–6.1), or 10.3 million (95% CI: 9.0-11.7 million) were presumed eligible for semaglutide on the basis of MASH using the broader criteria. Using a more conservative estimate, 3.8% (95% CI: 3.1–4.6), or 7.2 million (95% CI: 6.0–8.4 million) were presumed eligible for semaglutide for F2–F3 MASH. In contrast, 10.7% (CI: 9.4–12.1), or ~20.4 million (95% CI: 17.8–23.0 million), were eligible based on diabetes. Of those eligible based on MASH, 27.7% (conservative estimate, 95% CI: 21.2%–35.3%) and 28.6% (broad estimate, 95% CI: 23.8%–34.0%) had diabetes, but 72.3% (conservative estimate, 95% CI: 64.7%–78.8%) and 71.4% (broad estimate, 95% CI: 66.0–76.2) did not have diabetes. Thus, between 5.2 million and 7.4 million additional Americans could be eligible for semaglutide for MASH, not including those with comorbid diabetes. DISCUSSION The addition of semaglutide to the armamentarium of pharmacological interventions for MASH will result in millions of additional Americans being eligible for treatment. This patient population overlaps substantially with those eligible for resmetirom, although the mechanistic differences would provide more options, or could even prove complementary (eg, resmetirom is liver-directed7 and semaglutide targets weight loss and cardiometabolic risk2). Whereas eligibility for MASH is nearly entirely subsumed by eligibility for weight management, as only 2.9% (unweighted) of individuals eligible for MASH in this sample had a BMI of <27, medical insurance often does not cover semaglutide for weight management alone. Limitations. These estimates using the broader criteria represent upper-bound estimates, given that a more individualized clinical evaluation could reveal other potential contraindications (eg, other diagnoses of chronic liver disease, such as hepatitis B), which may be based on examination findings and laboratory values not available in the 2021–2023 NHANES dataset (eg, AST, ALT). Additional consideration of exclusions is provided in the Digital Supplement, https://links.lww.com/HC9/C198. Furthermore, the relatively smaller unweighted sample sizes produced relatively large confidence intervals. All individuals with diabetes were presumed to be eligible for semaglutide, which again represents an upper-bound estimate, and NHANES does not differentiate between type 1 and type 2 diabetes. Finally, as millions of Americans become newly eligible for semaglutide on the basis of MASH, successful implementation of prescribing, care pathways, and real-world medication adherence will influence whether these population-level estimates translate into meaningful reductions in morbidity and mortality. There will be new and continued challenges, such as the need for expanded non-invasive screening and equity in access, as MASH disproportionately affects older adults, rural patients, and certain racial and ethnic groups.6 There will be a need for an adaptation in the current care system among specialties such as internal medicine, endocrinology, gastroenterology, and hepatology to meet the increased clinical burden, particularly with respect to the challenge of medication adherence and persistence in real-world settings outside clinical trials.9 Furthermore, awareness of liver disease was poor, as only 1 in 10 individuals eligible for semaglutide on the basis of MASH reported ever having been told that they have a liver disease, which contrasts with 68.1% of patients with diabetes reporting that a doctor has told them that they have diabetes. With FDA approval of semaglutide for MASH, a first step might be increasing awareness of chronic liver disease among the millions of unaware patients. Given that MASH mortality increased from 2006 to 2023 and is expected to continue increasing over the next 20 years,10 this analysis from the 2021–2023 NHANES cohort demonstrates a potential opportunity to mitigate a looming public health crisis with new pharmaceutical options available for treating MASH.