Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Jan 10;42(1):e70121. doi: 10.1002/dmrr.70121

Earlier Menopause and Risk of Metabolic Dysfunction‐Associated Steatotic Liver Disease: A Global Cohort Study

Joshua Stokar 1,2,✉, Rivka Dresner‐Pollak 1,2
PMCID: PMC12790266  PMID: 41518227

ABSTRACT

Objectives

Metabolic dysfunction‐associated steatotic liver disease (MASLD) is a growing public health concern that contributes to liver and cardiovascular complications. The prevalence of MASLD in women increases sharply around age 50 years, but the relationship between an earlier age at natural menopause and MASLD is unknown.

Methods

Using the TriNetX global federated network, we identified women with earlier menopause (< 50 years). The control cohort consisted of similarly aged pre‐menopausal women. Cases of premature (< 40 years) or surgical menopause, non‐MASLD causes of steatotic liver disease (SLD), or sex‐hormone therapy were excluded. Propensity‐score matching adjusted for baseline characteristics and metabolic risk factors, resulting in two matched cohorts of 20,979 women (total n = 41,958) in the final analysis. Outcomes included new diagnoses of MASLD (metabolic dysfunction‐associated steatohepatitis) and the MASLD metabolic factors: pre‐diabetes/diabetes, hypertension, dyslipidaemia, and overweight/obesity over 5 years of follow‐up.

Results

Earlier menopause was associated with an increased risk of developing MASLD (HR 1.322, 95% CI 1.170–1.492), new‐onset dyslipidaemia (1.083; 1.045–1.122) and pre‐diabetes (1.130; 1.060–1.205). Findings were consistent across stratified analyses by pre‐existing metabolic risk factors (HR 95% CI for MASLD with pre‐existing dysglycaemia 1.370, 1.042–1.800; dyslipidaemia 1.340, 1.053–1.705; hypertension 1.230, 0.998–1.516; overweight 1.280, 1.086–1.510).

Conclusions

Risk of MASLD is increased following menopause before age 50. Further studies should assess the incorporation of menopause timing into female‐specific cardiometabolic risk assessment.

Keywords: age, menopause, metabolic dysfunction‐associated steatotic liver disease

Abbreviations

ALT

alanine aminotransferase

BMI

body mass index

CI

confidence interval

CV

cardiovascular

DPP‐4

dipeptidyl peptidase‐4

EHR

electronic health records

FSH

follicle‐stimulating hormone

GLP‐1RA

glucagon‐like peptide‐1 receptor agonist

HbA1c

haemoglobin A1c

HCO

Healthcare Organisation

HR

hazard ratio

HRT

hormone replacement therapy

ICD‐10

International Classification of Diseases, 10th Revision

MASH

metabolic dysfunction‐associated steatohepatitis

MASLD

metabolic dysfunction‐associated steatotic liver disease

NAFLD

non‐alcoholic fatty liver disease

NCO

negative control outcome

POI

premature ovarian insufficiency

PSM

propensity score matching

SD

standard deviation

SGLT‐2

sodium‐glucose co‐transporter‐2

SH

steatohepatitis

SLD

steatotic liver disease

SMD

standardised mean difference

T2DM

type 2 diabetes mellitus

Tx

treatment

1. Introduction

Metabolic dysfunction‐associated steatotic liver disease (MASLD) is a major public health concern, affecting millions worldwide and increasing the risk of both liver related and cardiovascular complications [1]. Previously termed non‐alcoholic fatty liver disease (NAFLD), MASLD is defined by the presence of steatotic liver disease (SLD) with at least one of the metabolic risk factors: dysglycaemia, hypertension, overweight/obesity, dyslipidaemia, and no other causes of SLD [2]. MASLD exhibits a sexually dimorphic pattern, with men having a 1.6‐fold higher incidence rate than women overall; however, MASLD incidence in women rises sharply around age 50, coinciding with the typical onset of natural menopause [3, 4]. Both animal models [5, 6] and epidemiological studies in humans [7, 8] suggest that pre‐menopausal oestrogens exert protective effects against MASLD. Conditions of abrupt or very early loss of endogenous oestrogens, such as surgical menopause or premature ovarian insufficiency (POI; menopause before age 40 years), are associated with higher rates of SLD as well as steatohepatitis and hepatic fibrosis [8, 9, 10]. Although early natural menopause (< 45 years) has been associated with higher risks of type 2 diabetes and hypertension [11, 12, 13], its relationship with MASLD remains poorly studied. Previous studies have yielded conflicting results, likely due to small sample sizes and failure to distinguish between natural and surgical menopause [14, 15, 16].

Importantly, post‐menopausal women with MASLD are more likely to develop severe fibrosis compared to pre‐menopausal women with MASLD, particularly those of normal weight [17]. With the recent introduction of effective therapeutics for MASLD [18, 19, 20], early identification and screening of subjects at risk is key. Despite growing evidence linking oestrogen loss to MASLD progression, current screening guidelines do not consider menopause alone as a sufficient risk factor for MASLD screening [2]. Moreover, women and female‐specific risk factors have historically been underrepresented in cardiovascular research [21]. We aimed to determine whether earlier natural menopause independently predicts MASLD after adjusting for key metabolic confounders, thereby identifying a female‐specific factor for targeted screening and treatment.

2. Methods

2.1. Data Source

We used TriNetX (Cambridge, USA), a global federated health network that aggregates electronic health records (EHRs) from multiple large healthcare organisations (HCOs). TriNetX standardises data across institutions through automated preprocessing and mapping to a unified clinical data model, ensuring consistency and reducing missing data bias [22]. Though TriNetX does not provide researchers with direct access to raw data, the harmonised database has been externally validated and widely used in a wide range of research areas [23, 24, 25], including studies relating to menopause as well as MASLD [26, 27, 28]. This study was conducted on the Global Collaborative Network, which at the time of analysis (07/2025) encompassed 151 HCOs, comprising 173,505,848 individuals across 18 countries.

2.2. Inclusion Criteria

Women aged 40–49 years, with a documented post‐menopausal or pre‐menopausal status in their EHR, were included in the analysis. The index date for analysis was defined as that of the first post‐menopausal or pre‐menopausal status, respectively. To diagnose menopause, we required a combination of a menopause‐related ICD‐10 code (N95.1, E28.3) with a measurement of serum follicle stimulating hormone (FSH) ≥ 20 mIU/mL. This FSH threshold is consistent with prior large‐scale EHR‐based menopause studies, recognising that individual variation exists during the perimenopausal transition [29]. Pre‐menopausal status was defined based on FSH < 20mIU/mL and no menopause‐related ICD‐10 code during the study period.

2.3. Exclusion Criteria

We excluded women with specific conditions that could confound the diagnosis of an earlier age at natural menopause, including surgical menopause (ICD‐10: Z90.72), Turner syndrome (Q96), primary adrenal insufficiency (E27.1), hypopituitarism (E23), hyperprolactinemia (E22.1), anorexia nervosa (F50.5), oncology treatments (TNX:1001–1005), chronic kidney disease (N18) or liver diseases (K70‐77). Women with documented use of sex hormone‐modifying medications during the year up to the index date or during follow‐up (contraceptives VA HS200, VA HS800, ATC G03C; gonadotropins VA HS400, RxNorm 25357, 134404; hormone replacement therapy, VA HS100; ICD‐10 Z79.890) were also excluded.

Women with potential non‐MASLD causes of SLD during the follow‐up period were excluded including: alcohol abuse or dependence (F10.1‐2, Z71.4), alcoholic liver disease (K70), chronic viral hepatitis (B18), Wilson's disease (E83.01), long‐term exposure to corticosteroids (E24, Z79.52), autoimmune hepatitis (K75.4), toxic or drug induced hepatitis (K71.6), as well as women with any history of lipid storage diseases (E88.1, E75.5, E78.6), alpha‐1 antitrypsin deficiency (E88.01) or hereditary haemochromatosis (E83.110).

We did not exclude users of anti‐diabetic agents (including insulin), as their initiation typically follows metabolic deterioration; thus, exclusion could introduce collider bias by preferentially removing higher‐risk women. However, the low baseline prevalence of these therapies limits their confounding effect. As glucagon‐like peptide 1 receptor agonists (GLP1‐RA) have recently been shown to effectively treat and potentially reverse MASLD, we conducted a separate analysis excluding prior users of GLP1‐RA.

2.4. Propensity Score Matching

For the propensity score, we used baseline characteristics from the year leading up to the index date including demographics, social determinants of health and healthcare access, past diagnosis of metabolic criteria, medications (anti‐diabetic, anti‐hypertensive, lipid‐lowering, anti‐obesity) and blood test results. TriNetX categorises race as: Asian, American Indian or Alaska Native, Black or African American, Native Hawaiian or Other, White, and unknown. TriNetX categorises ethnicity as: Hispanic or Latino, non‐Hispanic or Latino, and unknown. Baseline characteristics were compared between cohorts in each age category, with a standardised mean difference (SMD) ≥ 0.1 considered significant [30].

Propensity score matching (PSM) was employed using all the baseline characteristics listed in Table 1 using TriNetX's built‐in algorithm, using logistic regression implemented by the function LogisticRegression of the scikit‐learn package in Python version 3.7. The algorithm did not implement any specific imputation to account for missing data. The covariate matrix was randomised before matching to eliminate the influence of record ordering and based on the greedy nearest neighbour matching approach, with a calliper distance of 0.1 pooled SDs of the logit of the propensity score.

TABLE 1.

Baseline characteristics used in the propensity score.

EHR code Before PSM After PSM
Post‐menopause Pre‐menopause SMD Post‐menopause Pre‐menopause SMD
Total number (n =) 24,521 125,838 20,979 20,979
Age at index (years, mean ± SD) TNX AI 46.4 ± 2.37 43.8 ± 2.81 0.9965 46.2 ± 2.39 46.2 ± 2.4 0.0160
Current age (years, mean ± SD) TNX age 52.8 ± 5.6 50.1 ± 6.03 0.4730 52.6 ± 5.62 52.6 ± 5.76 0.0106
Race: n = (%)
American Indian/Alaska Native TNX 1002‐5 85 (0.367%) 278 (0.238%) 0.0236 75 (0.701%) 80 (0.644%) 0.0039
Asian TNX 2028‐9 1864 (8.057%) 6735 (5.759%) 0.0907 1731 (8.251%) 1890 (9.009%) 0.0270
Black/African American TNX 2054‐5 3403 (14.709%) 11,991 (10.254%) 0.1351 2983 (14.219%) 3057 (14.572%) 0.0100
Other TNX 2131‐1 830 (3.587%) 5007 (4.282%) 0.0357 754 (3.594%) 714 (3.403%) 0.0104
Unknown race UNK 2225 (9.617%) 24,231 (20.721%) 0.3133 2142 (10.210%) 2110 (10.058%) 0.0051
White TNX 2106‐3 14,557 (62.919%) 68,255 (58.368%) 0.0933 13,147 (62.667%) 12,993 (61.933%) 0.0151
Ethnicity n = (%):
Hispanic/Latino TNX 2135‐2 2232 (9.647%) 5733 (4.903%) 0.1835 1905 (9.081%) 1827 (8.709%) 0.0131
Not Hispanic/Latino TNX 2186‐5 15,140 (65.439%) 44,720 (38.242%) 0.5657 13,399 (63.869%) 13,732 (65.456%) 0.0332
Native Hawaiian/Pacific Islander TNX 2076‐8 172 (0.743%) 442 (0.378%) 0.0490 147 (0.701%) 135 (0.644%) 0.0070
Unknown ethnicity TNX UN 5764 (24.914%) 66,486 (56.855%) 0.6870 5675 (27.051%) 5420 (25.835%) 0.0276
Marital status n = (%):
Divorced TNX D 1843 (7.966%) 6990 (5.977%) 0.0781 1602 (7.636%) 1580 (7.531%) 0.0040
Domestic partner TNX T 186 (0.804%) 969 (0.829%) 0.0027 166 (0.791%) 174 (0.829%) 0.0043
Married TNX M 9380 (40.543%) 40,031 (34.232%) 0.1307 8526 (40.641%) 8642 (41.194%) 0.0112
Never married TNX S 4251 (18.374%) 19,669 (16.82%) 0.0408 3773 (17.985%) 3673 (17.508%) 0.0125
Separated TNX L 275 (1.189%) 811 (0.694%) 0.0513 232 (1.106%) 224 (1.068%) 0.0037
Widowed TNX W 635 (2.745%) 1620 (1.385%) 0.0957 524 (2.498%) 536 (2.555%) 0.0036
Socioeconomic determinants of health n = (%): ICD‐10 Z55‐Z65 430 (1.859%) 927 (0.793%) 0.0933 338 (1.611%) 322 (1.535%) 0.0061
Education ICD‐10 Z55 17 (0.073%) 14 (0.012%) 0.0298 11 (0.052%) 10 (0.048%) 0.0021
Unemployment ICD‐10 Z56 48 (0.207%) 92 (0.079%) 0.0341 37 (0.176%) 33 (0.157%) 0.0047
Occupational hazards ICD‐10 Z57 10 (0.043%) 23 (0.02%) 0.0133 10 (0.048%) 10 (0.048%) < 0.0001
Housing & economic ICD‐10 Z59 124 (0.536%) 258 (0.221%) 0.0514 95 (0.453%) 92 (0.439%) 0.0021
Social environment ICD‐10 Z60 107 (0.462%) 279 (0.239%) 0.0379 89 (0.424%) 92 (0.439%) 0.0022
Upbringing ICD‐10 Z62 11 (0.048%) 36 (0.031%) 0.0085 10 (0.048%) 10 (0.048%) < 0.0001
Family circumstances ICD‐10 Z63 125 (0.54%) 273 (0.233%) 0.0494 100 (0.477%) 101 (0.481%) 0.0007
Psychosocial circumstances ICD‐10 Z65 71 (0.307%) 142 (0.121%) 0.0401 51 (0.243%) 45 (0.215%) 0.0060
Access to health care n = (%):
Office visit TNX AMB
Emergency department services 1013711 2721 (11.761%) 6952 (5.945%) 0.2058 2276 (10.849%) 2208 (10.525%) 0.0105
Dietary counselling ICD‐10 Z71.3 348 (1.504%) 751 (0.642%) 0.0837 284 (1.354%) 250 (1.192%) 0.0145
Exercise counselling ICD‐10 Z71.82 84 (0.363%) 168 (0.144%) 0.0437 68 (0.324%) 58 (0.276%) 0.0087
Lifestyle n = (%):
Alcohol use ICD‐10 F10.9 50 (0.216%) 104 (0.089%) 0.0326 38 (0.181%) 34 (0.162%) 0.0046
Alcohol detected in blood ICD‐10 Y90 10 (0.043%) 15 (0.013%) 0.0182 10 (0.048%) 10 (0.048%) < 0.0001
Smoking ICD‐10 F17 1584 (6.846%) 2575 (2.202%) 0.2249 1190 (5.672%) 1141 (5.439%) 0.0102
Lack of exercise ICD‐10 Z72.3 10 (0.043%) 15 (0.013%) 0.0182 10 (0.048%) 10 (0.048%) < 0.0001
Physical activity codes ICD‐10 Y93 90 (0.389%) 159 (0.136%) 0.0495 74 (0.353%) 51 (0.243%) 0.0201
Inappropriate diet ICD‐10 Z72.4 15 (0.065%) 31 (0.027%) 0.0179 12 (0.057%) 16 (0.076%) 0.0074
Pre‐existing medical conditions n = (%):
Hypertension ICD‐10 I10 3825 (16.533%) 8389 (7.174%) 0.2926 3161 (15.067%) 3099 (14.772%) 0.0083
Impaired fasting glucose R73.01 322 (1.392%) 710 (0.607%) 0.0789 265 (1.263%) 259 (1.235%) 0.0026
Impaired glucose tolerance ICD‐10 R73.02 65 (0.281%) 160 (0.137%) 0.0316 54 (0.257%) 64 (0.305%) 0.0090
Metabolic syndrome ICD‐10 E88.81 170 (0.735%) 419 (0.358%) 0.0511 138 (0.658%) 140 (0.667%) 0.0012
Prediabetes R73.03 874 (3.778%) 1727 (1.477%) 0.1442 706 (3.365%) 677 (3.227%) 0.0077
Type 1 diabetes mellitus ICD‐10 E10 165 (0.713%) 351 (0.3%) 0.0582 132 (0.629%) 124 (0.591%) 0.0049
Type 2 diabetes mellitus ICD‐10 E11 1395 (6.03%) 3552 (3.037%) 0.1442 1160 (5.529%) 1174 (5.596%) 0.0029
Family history of diabetes ICD‐a10 Z83.3 320 (1.383%) 799 (0.683%) 0.0693 265 (1.263%) 248 (1.182%) 0.0074
Dyslipidaemia ICD‐10 E78 3998 (17.28%) 7776 (6.65%) 0.3320 3189 (15.201%) 3145 (14.991%) 0.0059
Overweight & obesity ICD‐10 E66 3313 (14.32%) 9140 (7.816%) 0.2084 2768 (13.194%) 2782 (13.261%) 0.0020
Fatty liver ICD‐10 K76.0 254 (1.098%) 302 (0.258%) 0.1024 165 (0.787%) 146 (0.696%) 0.0106
Polycystic ovarian syndrome ICD‐10 E28.2 176 (0.761%) 1901 (1.626%) 0.0797 173 (0.825%) 161 (0.767%) 0.0064
Sleep apnoea ICD‐10 G47.3 920 (3.976%) 2257 (1.93%) 0.1211 734 (3.499%) 723 (3.446%) 0.0029
Bariatric surgery ICD‐10 Z98.84 318 (1.374%) 775 (0.663%) 0.0709 269 (1.282%) 266 (1.268%) 0.0013
Knee osteoarthritis ICD‐10 M17 554 (2.395%) 1032 (0.883%) 0.1193 442 (2.107%) 413 (1.969%) 0.0098
Ischaemic heart disease ICD‐10 I20‐I25 299 (1.292%) 578 (0.494%) 0.0849 227 (1.082%) 219 (1.044%) 0.0037
Hypothyroidism ICD‐10 E03 2502 (10.814%) 5890 (5.037%) 0.2151 2093 (9.977%) 2003 (9.548%) 0.0145
Pregnancy related
Gestational diabetes ICD‐10 O24.4 19 (0.082%) 127 (0.109%) 0.0086 16 (0.076%) 22 (0.105%) 0.0095
Hypertensive disorders of pregnancy ICD‐10 O10‐O16 16 (0.069%) 118 (0.101%) 0.0109 16 (0.076%) 11 (0.052%) 0.0094
Obesity complicating childbirth ICD‐10 O99.21 10 (0.043%) 58 (0.05%) 0.0030 10 (0.048%) 10 (0.048%) < 0.0001
Laboratory, mean ± SD:
BMI, kg/m2 TNX 9083 29.6 ± 7.34 29.9 ± 8.01 0.0424 29.6 ± 7.43 29.9 ± 7.64 0.0378
BP, systolic (mm Hg) TNX 9085 121 ± 15.4 121 ± 15.9 0.0170 121 ± 15.4 121 ± 15.5 0.0434
BP, diastolic (mm Hg) TNX 9086 76.1 ± 10.9 75.8 ± 11.2 0.0260 76.1 ± 10.9 76.1 ± 10.8 0.0049
Glomerular filtration rate by MDRD (mL/min/1.73 m2) TNX 8001 85.9 ± 19.9 87.6 ± 19.1 0.0879 86.1 ± 19.9 86.9 ± 20.2 0.0376
Urine albumin/creatinine (mg/g) TNX LG37542‐4 10.2 ± 47.9 17.6 ± 235 0.0440 10 ± 49.8 10.7 ± 50.4 0.0134
Triglycerides (mg/dL) TNX 9002 119 ± 83.1 107 ± 93.5 0.1286 118 ± 82.1 118 ± 99.7 0.0044
HDL cholesterol (mg/dL) TNX 9001 59.1 ± 17.3 57.9 ± 16 0.0771 59.2 ± 17.4 58.2 ± 16.2 0.0598
LDL cholesterol (mg/dL) TNX 9002 118 ± 35.5 111 ± 31.9 0.1966 117 ± 35 116 ± 33.9 0.0324
Glucose (mg/dL) TNX 9025 99 ± 32.6 94.9 ± 26.7 0.1359 98.6 ± 32.4 99.3 ± 33.7 0.0205
Haemoglobin A1c (%) TNX 9037 5.81 ± 1.25 5.47 ± 1.01 0.2964 5.79 ± 1.26 5.79 ± 1.22 0.0066
Alanine aminotransferase (IU/L) TNX 9044 23.9 ± 23.6 20.8 ± 21.1 0.1364 23.5 ± 23.2 22.5 ± 19.8 0.0466
Thyroid stimulating hormone (mIU/L) TNX 9040 4.23 ± 41 3.86 ± 36.7 0.0096 4.21 ± 41.1 4.03 ± 39.1 0.0047
25‐Hydroxyvitamin D (ng/mL) TNX 9034 32.5 ± 16.5 29.9 ± 15.3 0.1678 32.3 ± 16.4 32.3 ± 16.8 0.0003
Medications n = (%):
Lipid modifying ATC C10 1642 (7.097%) 3348 (2.863%) 0.1956 1299 (6.192%) 1274 (6.073%) 0.0050
Antihypertensive ATC C02 510 (2.204%) 1252 (1.071%) 0.0894 422 (2.012%) 413 (1.969%) 0.0031
ACE‐I/ARB ATC C09 1699 (7.344%) 4132 (3.533%) 0.1686 1446 (6.893%) 1441 (6.869%)
Beta‐blockers ATC C07 1557 (6.73%) 4243 (3.628%) 0.1403 1333 (6.354%) 1337 (6.373%) 0.0008
Calcium channel blockers ATC C08 976 (4.219%) 2491 (2.13%) 0.1193 825 (3.933%) 822 (3.918%) 0.0007
Diuretics ATC C03 1826 (7.892%) 5019 (4.292%) 0.1510 1538 (7.331%) 1517 (7.231%) 0.0039
Vasodilators ATC C04 93 (0.402%) 222 (0.19%) 0.0391 71 (0.338%) 67 (0.319%) 0.0033
Aspirin ICD‐10 Z79.82 163 (0.705%) 286 (0.245%) 0.0670 125 (0.596%) 118 (0.562%) 0.0044
Thyroid hormone VA HS851 1863 (8.052%) 4860 (4.156%) 0.1633 1575 (7.508%) 1534 (7.312%) 0.0075
Anti‐diabetic
Metformin RxNorm 6809 805 (3.479%) 2537 (2.17%) 0.0791 685 (3.265%) 706 (3.365%) 0.0056
Insulin ICD‐10 Z79.4 205 (0.886%) 510 (0.436%) 0.0555 169 (0.806%) 163 (0.777%) 0.0032
SGLT2‐I ATC A10BK 109 (0.471%) 306 (0.262%) 0.0347 95 (0.453%) 92 (0.439%) 0.0021
DPP4‐i ATC A10BH 82 (0.354%) 243 (0.208%) 0.0277 75 (0.358%) 70 (0.334%) 0.0041
Sulfonylureas ATC A10BB 179 (0.774%) 452 (0.387%) 0.0510 150 (0.715%) 146 (0.696%) 0.0023
Meglitinides RxNorm 73044 10 (0.043%) 10 (0.009%) 0.0216 10 (0.048%) 10 (0.048%) < 0.0001
GLP‐1 analogues ATC A10BJ 521 (2.252%) 1710 (1.462%) 0.0585 449 (2.14%) 458 (2.183%) 0.0029
Dulaglutide RxNorm 1551291 116 (0.501%) 330 (0.282%) 0.0351 98 (0.467%) 82 (0.391%) 0.0117
Liraglutide RxNorm 475968 93 (0.402%) 313 (0.268%) 0.0233 78 (0.372%) 93 (0.443%) 0.0112
Semaglutide RxNorm 1991302 349 (1.508%) 1142 (0.977%) 0.0480 303 (1.444%) 298 (1.42%) 0.0020
Tirzepatide RxNorm 2601723 157 (0.679%) 542 (0.463%) 0.0286 140 (0.667%) 143 (0.682%) 0.0017
Anti‐obesity
Orlistat RxNorm 37925 10 (0.043%) 24 (0.021%) 0.0127 10 (0.048%) 10 (0.048%) < 0.0001
Phentermine RxNorm 8152 314 (1.357%) 926 (0.792%) 0.0549 267 (1.273%) 264 (1.258%) 0.0013
Bupropion RxNorm 42347 976 (4.219%) 2741 (2.344%) 0.1054 836 (3.985%) 817 (3.894%) 0.0047
Vitamins VA VT000 2529 (10.931%) 6896 (5.897%) 0.1821 2102 (10.02%) 2092 (9.972%) 0.0016
Vitamin E VA VT600 67 (0.29%) 214 (0.183%) 0.0220 55 (0.262%) 50 (0.238%) 0.0048
Glucocorticoids VA HS051 5209 (22.515%) 14,120 (12.075%) 0.2787 4384 (20.897%) 4398 (20.964%) 0.0016
Open in a new tab

Note: Bold values indicate an SMD greater than 0.1 is considered as a sign of a potentially significant imbalance between the cohorts.

Abbreviation: GA, genetic algorithms.

2.5. Outcomes

The ICD‐10 code K76.0 (‘fatty liver, not elsewhere classified’) was used as recommended by the recent Delphi consensus statement for MASLD in current ICD systems [31]. The MASLD metabolic risk factor outcomes were defined as:

  • Pre‐diabetes or diabetes (ICD‐10 R73.01–0.3, E11 or HbA1c ≥ 5.7% or glucose ≥ 200 mg/dL, or diabetes treatment VA HS500, ATC A10, Z79.4).

  • Overweight or obesity defined by ICD‐10 (E65‐68) or BMI ≥ 25kg/m2. Waist circumference was unavailable in most EHRs. As analyses were adjusted for race and ethnicity, ethnic differences in BMI cut‐offs were not applied.

  • Hypertension defined by ICD‐10 (I10‐I1A) or a measurement ≥ 135/80 mmHg or antihypertensive treatment (VA CV400 CV490; ATC C02–04, C07‐09).

  • Dyslipidaemia defined by ICD‐10 E78.1 or triglycerides ≥ 150mg/dL or HDL ≤ 50mg/d or lipid‐lowering treatment (ATC C10; VA CV350).

A negative control outcome (NCO), which is not expected to be affected by menopausal status, was a visual test (SNOMED:36228007) [32].

We compared outcomes between the propensity score matched cohorts over 5 years of follow‐up using TriNetX's built‐in ‘compare outcomes’ analysis tool, which implements Kaplan‐Meier analysis and the log‐rank test using R's survival package v3.2.3. The analysis censored participants at the first occurrence of a new outcome or after the last fact in their records. Hazard ratio (HR) is based on a proportional hazard model wherein the cohort to which the participant belonged is the independent variable. HRs and 95% confidence intervals (95% CI) were used to calculate E‐values for statistically significant results [33]. E‐value indicates the strength an unmeasured confounder would need to have with both the exposure and the outcome to explain away the observed exposure–outcome association [34].

3. Results

The participant selection and matching processes are outlined in Figure 1. The primary final analysis included 20,979 women in each propensity score matched cohort. FSH levels were consistent with post‐menopausal and pre‐menopausal states, respectively (64.3 ± 30.5 vs. 9.3 ± 5.89 mIU/mL p < 0.0001). Significant differences in baseline characteristics including metabolic risk factors between the post‐ and pre‐menopausal cohorts were all resolved following PSM (Table 1). The mean follow‐up time was 1120 ± 708 and 1032 ± 714 days for the post‐ and pre‐menopausal cohorts, respectively.

FIGURE 1.

FIGURE 1

Open in a new tab

Participant selection and matching (n =). *Exclusion criteria are not mutually exclusive. Tx: treatment.

Over five years of follow‐up, women with menopause before age 50 had a significantly higher risk of MASLD compared with pre‐menopausal women (Figure 2A). Menopause before age 50 was associated with an increased risk of a new diagnosis of MASLD (616 vs. 450 cases; HR 1.460, 95% CI 1.227–1.737; p < 0.0001), as well as the MASLD metabolic risk factors including pre‐diabetes (1996 vs. 1752 cases; HR 1.13 95% CI 1.06–1.205; p = 0.0002) and dyslipidaemia (6394 vs. 5811 cases; HR 1.083 95% CI 1.045–1.122; p < 0.0001) but not hypertension (2355 cases vs. 2622 cases; HR 0.976 95% CI 0.924–1.032; p = 0.4010), frank diabetes (1472 vs. 1412 cases; HR 0.995 95% CI 0.925–1.07 p = 0.8862), or overweight (11,014 vs. 10,934 cases; HR 0.991 95% CI 0.965–1.018; p = 0.5153).

FIGURE 2.

FIGURE 2

Open in a new tab

Menopause before age 50 leads to increased risk of MASLD and related outcomes (n = 20,979 per propensity score matched cohort) (A) 5‐year HR (95% CI) for MASLD outcomes following menopause before age 50 versus pre‐menopause. Number of participants with an outcome (E‐value for significant results): MASLD‐ 616 versus 450 (1.97); Prediabetes‐ 1996 versus 1753 (1.51); Diabetes‐ 1315 versus 1272; Hypertension‐ 7736 versus 8672; Dyslipidaemia‐ 6394 versus 5811; Overweight/obesity‐ 11,014 versus 10,934. (B) HR (95% Ci) for MASLD diagnosis over time in‐months following menopause before age 50 versus pre‐menopause. CI: confidence interval; HR: hazard ratio; MASH: metabolic dysfunction‐associated steatohepatitis; MASLD: metabolic dysfunction‐associated steatotic liver disease.

As expected, the risk of an NCO was not significantly different between the cohorts (HR 1.038 95% CI 0.695–1.549 p = 0.8558). A sensitivity analysis of the risk for development of MASLD over time revealed that the increase in MASLD risk appeared already at 1‐year post menopause and remained proportional over time, with year‐to‐year fluctuations that were not statistically significant (Figure 2B).

To further explore the interaction between age at natural menopause and cardiometabolic dysfunction, we conducted a separate propensity score‐matched analysis of women who already had at least one of the MASLD metabolic risk factors at baseline (Figure 3A). We found that the risk of developing MASLD increased with each of the underlying metabolic risk factors following earlier menopause compared to pre‐menopause. As underreporting of alcohol consumption may lead to misclassification of MetALD as MASLD, we conducted a separate analysis of women without any documentation of alcohol intake, revealing similar results. Additional analyses stratified by socio‐demographic factors, age category (40–44 years or 45–49 years), race, ethnicity, HCO type, or an analysis limited to HCOs in the USA. Results remained consistent across multiple sensitivity and stratified analyses, reinforcing the association between earlier menopause and MASLD, independent of baseline metabolic status (Figure 3B).

FIGURE 3.

FIGURE 3

Open in a new tab

Stratified analysis of MASLD risk following menopause before age 50 versus pre‐menopause (n = per propensity score matched cohort). (A) Stratified by metabolic risk factors at baseline. (B) Stratified by socio‐demographic factors. CI: confidence interval; HR: hazard ratio.

As GLP1‐RAs have recently been shown to be effective treatments for MASLD [19, 20], we conducted a separate sensitivity analysis excluding users of GLP1‐RA at baseline. Consistent with the main analysis, here too we found an increased risk of MASLD following menopause before age 50 (715 vs. 537 cases; HR 1.293, 95% CI 1.156–1.446; p < 0.0001).

4. Discussion

To our knowledge, this is the largest analysis to examine the longitudinal relationship between age at natural menopause and MASLD incidence. After rigorous adjustment for multiple cardiometabolic risk factors, we found that menopause before age 50 was associated with a significantly increased risk of developing MASLD and metabolic outcomes over the following 5 years. As close to half of all women experience menopause before age 50, our findings are relevant to a large population of women.

Our findings support the hypothesis that the pre‐menopausal state confers cardiometabolic protection independent of chronological age. Early menopause has been linked to higher risks of type 2 diabetes and hypertension [11, 12, 13], while previous studies evaluating menopause and MASLD have been constrained by smaller cohorts, cross‐sectional designs, and insufficient distinction between natural and surgical menopause [14, 15, 16]. Importantly, the established association between menopausal status and MASLD in cross‐sectional studies does not clarify directionality; metabolic dysfunction may contribute to both earlier menopause and subsequent MASLD. Indeed, in our cohort, prior to propensity score matching, earlier menopause was associated with a worse metabolic profile at baseline. Several studies suggest that diabetes in pre‐menopausal women is associated with earlier menopause [35, 36]. In our cohort, the increased incidence in prediabetes, but not overt diabetes, following earlier menopause, is likely attributable to the relatively short follow‐up time. Taken together with existing evidence, these findings suggest a bidirectional relationship in which early oestrogen decline and subclinical metabolic dysfunction may reinforce each other, rather than a purely unidirectional causal pathway. Our findings have potential implications for clinical practice. Given the recent developments in pharmacotherapies for MASLD [18, 19, 20], early identification of high‐risk populations is critical. Currently, MASLD screening is primarily recommended for individuals with pre‐existing metabolic risk factors. Our data suggest that earlier menopause may be considered as a female‐specific MASLD risk factor.

Although MASLD is increasingly recognised as an independent risk factor for cardiovascular morbidity and mortality, our study was not designed or powered to evaluate major adverse cardiovascular events (MACE) in relation to age at menopause. The relatively young age and low cardiometabolic burden of the study population resulted in very low event rates over the available follow‐up period, limiting the feasibility of detecting meaningful differences between groups. Nonetheless, given the strong bidirectional relationship between MASLD and cardiovascular disease, it is plausible that an earlier age at menopause, through its association with increased MASLD risk, may contribute to long‐term cardiovascular consequences. Future studies with larger event rates, longer follow‐up, and dedicated cardiovascular endpoints are needed to determine whether menopause timing should be incorporated into sex‐specific cardiovascular risk assessment and prevention strategies.

Our findings raise the question of whether HRT may mitigate MASLD risk in women with an earlier age at natural menopause. While some observational data suggest a potential benefit of HRT in reducing hepatic steatosis [37, 38], results have been inconsistent and confounded by variations in HRT formulations, administration routes, and treatment timing [39]. To reduce confounding effects, we chose to exclude from the analysis women with documented HRT use during the study period. Prospective studies, specifically examining HRT's impact on MASLD risk in women with early menopause, are required.

4.1. Limitations

This study has several limitations inherent to its EHR‐based design. Menopausal status was defined using specific ICD codes combined with elevated FSH, which may have misclassified some women in the peri‐menopausal transition [40]; however, this approach has shown adequate accuracy for population‐level analyses [29]. MASLD was identified using ICD‐10 code K76.0, which has been shown to have a high predictive value for non‐alcoholic fatty liver disease but may not capture all MASLD phenotypes and is underdiagnosed relative to its actual prevalence [41]. Alcohol use and waist circumference were incompletely recorded, although sensitivity analyses excluding documented alcohol users and adjustment for BMI helped mitigate these issues. Despite robust propensity score matching, granular lifestyle data (caloric intake, specific physical activity levels) and genetic predispositions are not fully captured in EHR data. Therefore, as with all observational studies, residual confounding cannot be entirely excluded. Finally, the relatively short follow‐up duration may have limited detection of later‐onset MASLD as well as more advanced liver disease.

5. Conclusions

Earlier menopause identifies women at increased risk of the development of MASLD beyond traditional metabolic factors. Prospective studies are warranted to assess the incorporation of menopause timing into female‐specific cardiometabolic risk assessment, as well as specific interventions to prevent and treat MASLD around peri‐menopause, including lifestyle modifications, HRT and emerging MASLD therapeutics.

Author Contributions

J.S. and R.D.P. conceived and designed the study, interpreted the data, drafted and revised the manuscript. All authors have read and approved the final manuscript.

Funding

Financial support and sponsorship: This study was supported by grants from the Hadassah Ofek grant to J.S. and an ISF grant 2139/21 to R.D.P.

Ethics Statement

Data acquired from TriNetX are anonymised in compliance with HIPAA. Informed consent was not necessary, and an exemption was obtained from the Hadassah Medical Center’s institutional ethics board.

Conflicts of Interest

The authors declare no conflicts of interest.

Acknowledgements

The authors have nothing to report.

Stokar, Joshua , and Dresner‐Pollak Rivka. 2026. “Earlier Menopause and Risk of Metabolic Dysfunction‐Associated Steatotic Liver Disease: A Global Cohort Study,” Diabetes/Metabolism Research and Reviews: e70121. 10.1002/dmrr.70121.

Data Availability Statement

The data that support the findings of this study are available from TriNetX. Restrictions apply to the availability of these data, which were used under licence for this study. Data are available at https://trinetx.com/ with the permission of TriNetX. Release and/or sharing of data are not covered under the Hadassah Medical Organization data use agreement with TriNetX. All required information to replicate the network queries is presented in the manuscript. A request can be made to gain access to TriNetX (join@trinetx.com).

References

  • 1. Fu C. E., Ng C. H., Yong J. N., et al., “A Meta‐Analysis on Associated Risk of Mortality in Nonalcoholic Fatty Liver Disease,” Endocrine Practice 29, no. 1 (2023): 33–39, 10.1016/j.eprac.2022.10.007. [DOI] [PubMed] [Google Scholar]
  • 2. Tacke F., Horn P., Wong V. W.‐S., et al., “EASL–EASD–EASO Clinical Practice Guidelines on the Management of Metabolic Dysfunction‐Associated Steatotic Liver Disease (MASLD): Executive Summary,” Diabetologia 67, no. 11 (2024): 2375–2392, 10.1007/s00125-024-06196-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Estes C., Razavi H., Loomba R., Younossi Z., and Sanyal A. J., “Modeling the Epidemic of Nonalcoholic Fatty Liver Disease Demonstrates an Exponential Increase in Burden of Disease,” Hepatology 67, no. 1 (2018): 123–133, 10.1002/hep.29466. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Cherubini A., Della Torre S., Pelusi S., and Valenti L., “Sexual Dimorphism of Metabolic Dysfunction‐Associated Steatotic Liver Disease,” Trends in Molecular Medicine 30, no. 12 (2024): 1126–1136, 10.1016/j.molmed.2024.05.013. [DOI] [PubMed] [Google Scholar]
  • 5. Stokar J., Gurt I., Cohen‐Kfir E., et al., “Hepatic Adropin Is Regulated by Estrogen and Contributes to Adverse Metabolic Phenotypes in Ovariectomized Mice,” Molecular Metabolism 60 (2022): 101482, 10.1016/j.molmet.2022.101482. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Turola E., Petta S., Vanni E., et al., “Ovarian Senescence Increases Liver Fibrosis in Humans and Zebrafish With Steatosis,” Disease Models & Mechanisms 8, no. 9 (2015): 1037–1046, 10.1242/dmm.019950. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Jaroenlapnopparat A., Charoenngam N., Ponvilawan B., Mariano M., Thongpiya J., and Yingchoncharoen P., “Menopause Is Associated With Increased Prevalence of Nonalcoholic Fatty Liver Disease: A Systematic Review and Meta‐Analysis,” Menopause 30, no. 3 (2023): 348–354, 10.1097/gme.0000000000002133. [DOI] [PubMed] [Google Scholar]
  • 8. Klair J. S., Yang J. D., Abdelmalek M. F., et al., “A Longer Duration of Estrogen Deficiency Increases Fibrosis Risk Among Postmenopausal Women With Nonalcoholic Fatty Liver Disease,” Hepatology 64, no. 1 (2016): 85–91, 10.1002/hep.28514. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Wang J., Wu A. H., Stanczyk F. Z., et al., “Associations Between Reproductive and Hormone‐Related Factors and Risk of Nonalcoholic Fatty Liver Disease in a Multiethnic Population,” Clinical Gastroenterology and Hepatology 19, no. 6 (2021): 1258–1266.e1251, 10.1016/j.cgh.2020.08.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Jin S., Li S., Fang P., Pan C., and Huang S., “Association of Hysterectomy With Nonalcoholic Fatty Liver Disease Among US Women,” Lipids in Health and Disease 23, no. 1 (2024): 34, 10.1186/s12944-024-02020-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Anagnostis P., Christou K., Artzouchaltzi A. M., et al., “Early Menopause and Premature Ovarian Insufficiency Are Associated With Increased Risk of Type 2 Diabetes: A Systematic Review and Meta‐Analysis,” European Journal of Endocrinology 180, no. 1 (2019): 41–50, 10.1530/eje-18-0602. [DOI] [PubMed] [Google Scholar]
  • 12. Anagnostis P., Theocharis P., Lallas K., et al., “Early Menopause Is Associated With Increased Risk of Arterial Hypertension: A Systematic Review and meta‐analysis,” Maturitas 135 (2020): 74–79, 10.1016/j.maturitas.2020.03.006. [DOI] [PubMed] [Google Scholar]
  • 13. Ko B.‐J., Jung J.‐H., Han K., and Nam G. E., “Age at Menopause and Development of Type 2 Diabetes in Korea,” JAMA Network Open 8, no. 1 (2025): e2455388, 10.1001/jamanetworkopen.2024.55388. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Veronese N., Notarnicola M., Osella A. R., et al., “Menopause Does Not Affect Fatty Liver Severity in Women: A Population Study in a Mediterranean Area,” Endocrine, Metabolic & Immune Disorders – Drug Targets 18, no. 5 (2018): 513–521, 10.2174/1871530318666180423101755. [DOI] [PubMed] [Google Scholar]
  • 15. Kim E.‐Y., Lee Y.‐J., Kwon Y.‐J., and Lee J.‐W., “Age at Menopause and Risk of Metabolic Dysfunction‐Associated Fatty Liver Disease: A 14‐Year Cohort Study,” Digestive and Liver Disease 56, no. 11 (2024): 1880–1886, 10.1016/j.dld.2024.05.003. [DOI] [PubMed] [Google Scholar]
  • 16. Park S., Park Y., Lee J., et al., “Lack of Association Between Early Menopause and Non‐Alcoholic Fatty Liver Disease in Postmenopausal Women,” Climacteric 23, no. 2 (2020): 173–177, 10.1080/13697137.2019.1650018. [DOI] [PubMed] [Google Scholar]
  • 17. Yoneda M., Thomas E., Sumida Y., Eguchi Y., and Schiff E. R., “The Influence of Menopause on the Development of Hepatic Fibrosis in Nonobese Women With Nonalcoholic Fatty Liver Disease,” Hepatology 60, no. 5 (2014): 1792, 10.1002/hep.27097. [DOI] [PubMed] [Google Scholar]
  • 18. Harrison S. A., Bedossa P., Guy C. D., et al., “A Phase 3, Randomized, Controlled Trial of Resmetirom in NASH With Liver Fibrosis,” New England Journal of Medicine 390, no. 6 (2024): 497–509, 10.1056/nejmoa2309000. [DOI] [PubMed] [Google Scholar]
  • 19. Newsome P. N., Buchholtz K., Cusi K., et al., “A Placebo‐Controlled Trial of Subcutaneous Semaglutide in Nonalcoholic Steatohepatitis,” New England Journal of Medicine 384, no. 12 (2021): 1113–1124, 10.1056/nejmoa2028395. [DOI] [PubMed] [Google Scholar]
  • 20. Loomba R., Hartman M. L., Lawitz E. J., et al., “Tirzepatide for Metabolic Dysfunction‐Associated Steatohepatitis With Liver Fibrosis,” New England Journal of Medicine 391, no. 4 (2024): 299–310, 10.1056/nejmoa2401943. [DOI] [PubMed] [Google Scholar]
  • 21. Wenger N. K., Lloyd‐Jones D. M., Elkind M. S. V., et al., “Call to Action for Cardiovascular Disease in Women: Epidemiology, Awareness, Access, and Delivery of Equitable Health Care: A Presidential Advisory From the American Heart Association,” Circulation 145, no. 23 (2022): e1059–e1071, 10.1161/cir.0000000000001071. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Kahn M. G., Callahan T. J., Barnard J., et al., “A Harmonized Data Quality Assessment Terminology and Framework for the Secondary Use of Electronic Health Record Data,” eGEMs (Generating Evidence \& Methods to improve patient outcomes) 4, no. 1 (2016): 18, 10.13063/2327-9214.1244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Kerem L. and Stokar J., “Risk of Suicidal Ideation or Attempts in Adolescents With Obesity Treated With GLP1 Receptor Agonists,” JAMA Pediatr 178, no. 12 (2024): 1307–1315, 10.1001/jamapediatrics.2024.3812. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Wang W., Wang C.‐Y., Wang S.‐I., and Wei J. C.‐C., “Long‐Term Cardiovascular Outcomes in COVID‐19 Survivors Among Non‐Vaccinated Population: A Retrospective Cohort Study From the Trinetx US Collaborative Networks,” eClinicalMedicine 53 (2022): 101619, 10.1016/j.eclinm.2022.101619. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Hsieh T. Y. J., Chang R., Yong S.‐B., Liao P.‐L., Hung Y.‐M., and Wei J. C.‐C., “COVID‐19 Vaccination Prior to SARS‐CoV‐2 Infection Reduced Risk of Subsequent Diabetes Mellitus: A Real‐World Investigation Using U.S. Electronic Health Records,” Diabetes Care 46, no. 12 (2023): 2193–2200, 10.2337/dc23-0936. [DOI] [PubMed] [Google Scholar]
  • 26. Shih Y.‐H., Yang C.‐Y., Wang S.‐J., and Lung C.‐C., “Menopausal Hormone Therapy Decreases the Likelihood of Diabetes Development in Peri‐Menopausal Individuals With Prediabetes,” Diabetes & Metabolism 50, no. 4 (2024): 101546, 10.1016/j.diabet.2024.101546. [DOI] [PubMed] [Google Scholar]
  • 27. Stokar J. and Szalat A., “Cardiovascular Safety of Romosozumab Vs PTH Analogues for Osteoporosis Treatment: A propensity‐Score‐Matched Cohort Study,” Journal of Clinical Endocrinology & Metabolism (2024): dgae173. [DOI] [PubMed] [Google Scholar]
  • 28. Henney A. E., Riley D. R., Hydes T. J., et al., “Metabolic Syndrome Traits Differentially and Cumulatively Influence Micro‐ and Macrovascular Disease Risk in Patients With MASLD,” Liver Int 44, no. 11 (2024): 3031–3049, 10.1111/liv.16086. [DOI] [PubMed] [Google Scholar]
  • 29. Backer L. C., Rubin C. S., Marcus M., Kieszak S. M., and Schober S. E., “Serum Follicle‐Stimulating Hormone and Luteinizing Hormone Levels in Women Aged 35–60 in the U.S. Population: The Third National Health and Nutrition Examination Survey (NHANES III, 1988–1994),” Menopause 6, no. 1 (1999): 29–35, 10.1097/00042192-199906010-00007. [DOI] [PubMed] [Google Scholar]
  • 30. Zhang Z., Kim H. J., Lonjon G., and Zhu Y., “Balance Diagnostics After Propensity Score Matching,” Annals of Translational Medicine 7, no. 1 (2019): 16, 10.21037/atm.2018.12.10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Hagström H., Adams L. A., Allen A. M., et al., “The Future of International Classification of Diseases Coding in Steatotic Liver Disease: An Expert Panel Delphi Consensus Statement,” Hepatology Communications 8, no. 2 (2024): e0386, 10.1097/hc9.0000000000000386. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Levintow S. N., Nielson C. M., Hernandez R. K., et al., “Pragmatic Considerations for Negative Control Outcome Studies to Guide Non‐randomized Comparative Analyses: A Narrative Review,” Pharmacoepidemiology and Drug Safety 32, no. 6 (2023): 599–606, 10.1002/pds.5623. [DOI] [PubMed] [Google Scholar]
  • 33. Mathur M. B., Ding P., Riddell C. A., and VanderWeele T. J., “Web Site and R Package for Computing E‐values,” Epidemiology 29, no. 5 (2018): e45–e47, 10.1097/ede.0000000000000864. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. VanderWeele T. J. and Ding P., “Sensitivity Analysis in Observational Research: Introducing the E‐value,” Annals of Internal Medicine 167, no. 4 (2017): 268–274, 10.7326/m16-2607. [DOI] [PubMed] [Google Scholar]
  • 35. Monterrosa‐Castro A., Blümel J. E., Portela‐Buelvas K., et al., “Type II Diabetes Mellitus and Menopause: A Multinational Study,” Climacteric 16, no. 6 (2013): 663–672, 10.3109/13697137.2013.798272. [DOI] [PubMed] [Google Scholar]
  • 36. Gold E. B., Crawford S. L., Avis N. E., et al., “Factors Related to Age at Natural Menopause: Longitudinal Analyses From SWAN,” American Journal of Epidemiology 178, no. 1 (2013): 70–83, 10.1093/aje/kws421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. McKenzie J., Fisher B. M., Jaap A. J., Stanley A., Paterson K., and Sattar N., “Effects of HRT on Liver Enzyme Levels in Women With Type 2 Diabetes: A Randomized Placebo‐Controlled Trial,” Clinical endocrinology 65, no. 1 (2006): 40–44, 10.1111/j.1365-2265.2006.02543.x. [DOI] [PubMed] [Google Scholar]
  • 38. Yang J. D., Abdelmalek M. F., Guy C. D., et al., “Patient Sex, Reproductive Status, and Synthetic Hormone Use Associate With Histologic Severity of Nonalcoholic Steatohepatitis,” Clinical Gastroenterology and Hepatology 15, no. 1 (2017): 127–131.e122, 10.1016/j.cgh.2016.07.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Cho L., Kaunitz A. M., Faubion S. S., et al., “Rethinking Menopausal Hormone Therapy: For Whom, What, When, and How Long?,” Circulation 147, no. 7 (2023): 597–610, 10.1161/circulationaha.122.061559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Ambikairajah A., Walsh E., and Cherbuin N., “A Review of Menopause Nomenclature,” Reproductive Health 19, no. 1 (2022): 29, 10.1186/s12978-022-01336-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Hayward K. L., Johnson A. L., Horsfall L. U., Moser C., Valery P. C., and Powell E. E., “Detecting Non‐Alcoholic Fatty Liver Disease and Risk Factors in Health Databases: Accuracy and Limitations of the ICD‐10‐AM,” BMJ Open Gastroenterology 8, no. 1 (2021): e000572, 10.1136/bmjgast-2020-000572. [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.

Data Availability Statement

The data that support the findings of this study are available from TriNetX. Restrictions apply to the availability of these data, which were used under licence for this study. Data are available at https://trinetx.com/ with the permission of TriNetX. Release and/or sharing of data are not covered under the Hadassah Medical Organization data use agreement with TriNetX. All required information to replicate the network queries is presented in the manuscript. A request can be made to gain access to TriNetX (join@trinetx.com).

Articles from Diabetes/Metabolism Research and Reviews are provided here courtesy of Wiley

RESOURCES