Introduction

Androgenetic alopecia (AGA) is the most prevalent hair loss type, affecting up to 50% of men and approximately 30% of middle-aged women1, with a different hair loss pattern. In men, it starts with a receding hairline and the formation of a “M” shape, followed by hair loss on the crown of the head, ultimately resulting in complete baldness on the crown. In women, it is characterized by thinning and reduced hair density in the central scalp areas2. In total, AGA influences over 50% of adults3. It shows varying prevalence across different ethnic groups and tends to become more common with increasing age. White men are particularly influenced, with as many as 80% showing AGA signs by the time they are of 704. Unfortunately, without intervention, AGA is irreversible1. This adversely affects a person’s appearance, self-image, self-confidence, and overall quality of life5. It can also contribute to the emergence of psychological disorders and heighten the odds of non-communicable diseases1. AGA is determined shrinkage of genetically susceptible hair follicles due to the influence of androgen hormones6. Additionally, many additional elements, including persistent low-grade inflammation and oxidative damage, contribute to the development of this condition. AGA, influenced by chronic inflammation, has been associated with health issues consisting of cancer, coronary heart disease, and metabolic syndrome7.

Nonalcoholic fatty liver disease (NAFLD), recently recognized as a liver-related component of metabolic syndrome, is defined by the buildup of fat within liver cells exceeding 5% in individuals without considerable alcohol consumption or other secondary causes. It is linked to cardiometabolic comorbidities including dyslipidemia, type 2 diabetes, and hypertension8,9. NAFLD influences a remarkable percentage of the global population, with recent studies estimating its prevalence at approximately 38% among adults10. Its incidence continues to rise globally, driven by factors such as obesity, reduced physical activity, and the adoption of unhealthy dietary habits11. Furthermore, it ranks as the second most common indication for liver transplantation12. The designation ‘Metabolic dysfunction-associated steatotic liver disease (MASLD)’ has recently been introduced as a replacement for NAFLD, highlighting the central role of metabolic dysfunction in its pathogenesis13. Prior research has shown that elevated androgen levels are linked to a higher risk of developing and advancing MASLD, indicating that androgens may contribute to the disease’s onset and progression14,15. Given the well-established hormonal influence on MASLD pathogenesis, particularly the role of androgens, it is worth exploring conditions in which androgens are also implicated, such as AGA.

Previous research has suggested a link between AGA and metabolic syndrome16. Moreover, metabolic syndrome has been shown to potentially mediate the association between the inflammatory potential of diet and AGA17. Similarly, insulin resistance in the related condition of MASLD may lower sex hormone-binding globulin (SHBG) levels, resulting in increased free androgen concentrations. These androgens may contribute to hair follicle miniaturization, a key feature of AGA pathogenesis18,19. Despite overlapping pathophysiological mechanisms and shared metabolic risk profiles, the direct association between MASLD and AGA remains underexplored. This study intends to evaluate the potential relationship between MASLD and AGA, hypothesizing that shared hormonal and metabolic pathways may underlie both conditions.

Materials and methods

Study population and exclusion criteria

This cross-sectional study was conducted on the FACS population (n = 10,138), a subset of the PERSIAN cohort, which is a longitudinal study focusing on non-communicable diseases20. Participants included adults aged 35–70 years. During the initial phase of FACS, comprehensive questionnaires were administered, and physical examinations were conducted. The study population consisted of individuals from the rural region of Sheshdeh, Fasa, Iran, and its 24 surrounding villages. Data collection occurred between 2015 and 2016, with all participants providing informed written consent before enrollment. All data were securely stored online using specialized software. The study variables were extracted from this database, following the FACS data collection protocol20. Participants were excluded if they lacked laboratory data required for FLI and MASLD calculation, had missing information on AGA or alcohol consumption, or reported extreme calorie intakes (< 800 kcal/day or > 5500 kcal/day). Additionally, individuals with conditions that could significantly impact AGA, including systemic corticosteroid users, those with anemia, thyroid disorders, cancer, along with subjects with hepatitis B or C, or pregnancy, were excluded (Fig. 1). We aimed to exclude participants with significant alcohol consumption, typically defined as > 20 g/day for women and > 30 g/day for men. We had access to detailed data on quantitative alcohol consumption per day (in grams). However, none of the participants met the criteria for significant alcohol consumption and were thus not excluded based on this factor. This was consistent with the MASLD diagnostic criteria, which were applied to the study population.

Fig. 1
figure 1

Study flowchart.

Full size image

Covariates

The study incorporated various covariates from FACS data, including marital status, educational attainment (university degree or not), and employment status (defined as working at least 8 h per week, including seasonal work such as farming). Socioeconomic status was assessed using the Asset Index21. Other variables included daily calorie intake (kcal/day), physical activity (measured as metabolic equivalent of tasks [MET] over 24 h), smoking status (current regular smoker or not), and regular alcohol consumption (defined as consuming approximately 200 ml of beer or 45 ml of liquor once per week for at least six months)22. This definition was used to separate participants with regular alcohol consumption and was utilized as a differentiating variable from daily alcohol intake, which was measured using a food frequency questionnaire. The aim of using this regular consumption threshold was to distinguish habitual alcohol consumption from sporadic drinking patterns, which might have a different impact on metabolic health.

Additional factors potentially influencing AGA were also considered, such as heart disease, depression and other psychiatric disorders, use of antidepressants or beta-blockers, night sleep duration (calculated as wake-up time minus bedtime), and daily intake of dietary protein (grams/day), iron, magnesium, phosphorus, zinc (milligrams/day), and vitamins A and D (micrograms/day). Kidney function was evaluated using the glomerular filtration rate (GFR), calculated via the MDRD formula. The Systemic Inflammation Response Index (SIRI) was defined as neutrophil count × monocyte/lymphocyte count, while the Triglyceride-glucose Index (TyG index) was computed as Ln (TG [mg/dL] × glucose [mg/dL]/2). Lipid-lowering medication use was classified as taking any statins, fibrates, or PCSK9 inhibitors. Dietary intake of total carbohydrates, proteins, lipids, and fiber was assessed using a validated 125-item Food Frequency Questionnaire (FFQ), categorized according to PERSIAN cohort standards.

Definition of MASLD and AGA

The Fatty Liver Index (FLI) was employed to identify steatotic liver disease (SLD). This index, previously validated against sonography in the studied population, provides a reliable estimate of hepatic steatosis when sonography is unavailable23. The FLI is calculated using gamma-glutamyl transferase (GGT), waist circumference (WC), serum triglyceride (TG) levels, and body mass index (BMI)24:

$$\:{F}{L}{I}=\frac{{e}\:0.953{*}{l}{o}{g}{e}\left({T}{G}\right)+0.139{*}{B}{M}{I}+0.718{*}{l}{o}{g}{e}\left({G}{G}{T}\right)+0.053{*}{W}{C}-15.745}{1+{e}0.953{*}{l}{o}{g}{e}\left({T}{G}\right)+0.139{*}{B}{M}{I}+0.718{*}{l}{o}{g}{e}\left({G}{G}{T}\right)+0.053{*}{W}{C}-15.745}{*}100$$

For the Iranian population, the FLI cutoff values for diagnosing SLD are 46.9 in men (sensitivity: 82.4%, specificity: 76.8%) and 53.8 in women (sensitivity: 82.3%, specificity: 76.5%)25. Individuals with scores exceeding these thresholds are considered at risk for SLD, with the FLI having been validated across multiple population studies26,27,28. MASLD was defined based on the adult cardiometabolic criteria. Specifically, MASLD was diagnosed in individuals with evidence of SLD in the presence of at least one of the following five metabolic risk factors29:

  1. 1.

    Body mass index (BMI) ≥ 25 kg/m² or waist circumference (WC) > 94 cm in males and > 80 cm in females.

  2. 2.

    Fasting serum glucose ≥ 100 mg/dL, 2-hour post-load glucose ≥ 140 mg/dL, glycated hemoglobin (HbA1c) ≥ 5.7%, or a diagnosis/treatment of type 2 diabetes.

  3. 3.

    Blood pressure ≥ 130/85 mmHg or current use of antihypertensive medications.

  4. 4.

    Plasma triglycerides ≥ 150 mg/dL or use of lipid-lowering medications.

  5. 5.

    Plasma high-density lipoprotein cholesterol (HDL-C) ≤ 40 mg/dL in males or ≤ 50 mg/dL in females, or use of lipid-lowering medications.

Androgenic Alopecia (AGA) was assessed by a specialist dermatologist involved in the FACS data collection. Classification was conducted using the Hamilton-Norwood scale for men and the Ludwig scale for women.

Statistical analysis

Descriptive statistics were used to summarize data, with means and standard deviations for continuous variables and frequencies and percentages for categorical variables. For univariate analysis, the independent t-test was applied to continuous variables, while the chi-square test was used for categorical variables to assess associations with AGA.

Multivariate analyses were conducted using logistic regression to calculate odds ratios (ORs) and 95% confidence intervals (CIs), adjusting for key confounding variables identified in the literature and from univariate analyses (p < 0.2). Study findings were stratified by subgroups potentially influencing MASLD, including physical activity, carbohydrate intake, total lipid intake, fiber intake, protein intake, total calorie intake, insulin resistance (assessed via TyG index), inflammation (assessed via SIRI), and lipid-lowering medication use. These variables were categorized based on the population median.

All statistical analyses were performed using SPSS software version 23 (IBM Corp., Armonk, N.Y., USA), with a significance threshold set at p < 0.05.

Results

Population characteristics

This study included a total of 7,993 participants, comprising 3,926 males and 4,067 females, with a mean age of 48.62 ± 9.63 years. MASLD was identified in 2,736 individuals, while AGA was present in 6,004 participants (Table 1).

Table 1 Clinical characteristics of the population studied according to the presence or absence of androgenic alopecia (AGA).
Full size table

Univariate analysis results

The univariate analysis revealed several significant associations between demographic, lifestyle, and metabolic factors and the presence of AGA. The analysis showed that individuals with AGA were generally older (mean age: 49.47 ± 9.63 years vs. 46.07 ± 9.18 years, p < 0.001), with a higher body mass index (BMI) (25.82 ± 4.80 vs. 24.83 ± 4.74, p < 0.001), and were less physically active (41.42 ± 10.97 METs vs. 42.90 ± 13.08 METs, p < 0.001). Furthermore, AGA was associated with shorter night sleep duration (6.50 ± 1.96 h vs. 6.90 ± 1.79 h, p < 0.001) and lower renal function (GFR: 79.68 ± 20.53 ml/min/1.73 m² vs. 84.67 ± 20.47 ml/min/1.73 m², p < 0.001).

In terms of dietary intake, individuals with AGA consumed significantly less protein (90.46 ± 32.94 g/day vs. 99.00 ± 35.25 g/day, p < 0.001), iron (22.70 ± 8.97 milligrams/day vs. 24.40 ± 9.61 milligrams/day, p < 0.001), magnesium (381.88 ± 128.56 milligrams/day vs. 413.93 ± 140.63 milligrams/day, p < 0.001), and phosphorus (1334.71 ± 462.37 milligrams/day vs. 1449.10 ± 498.98 milligrams/day, p < 0.001). Additionally, AGA was significantly associated with gender (p < 0.001), marital status (p < 0.001), education level (p = 0.017), and employment status (p < 0.001). Specifically, women were more likely to have AGA than men (55.1% vs. 44.9%, p < 0.001), and individuals who were married or employed had higher odds of developing AGA. Furthermore, significant associations were found with smoking (p < 0.01), alcohol consumption (p < 0.001), and ischemic heart disease (p < 0.001), as well as the presence of MASLD (p < 0.001) (Table 1).

Multivariate analysis results

Multivariate logistic regression analysis revealed that individuals with MASLD had 31% higher odds of developing AGA (OR = 1.31, 95% CI: 1.13–1.52, p < 0.001). Notably, among females with MASLD, the likelihood of having AGA was 62% higher (OR = 1.62, 95% CI: 1.30–2.03, p < 0.001), whereas this association was not statistically significant in males (Table 2).

Table 2 Multiple logistic regression analysis for the association between MASLD and AGA.
Full size table

Subgroup analyses

The subgroup analyses presented in Table 3 further examined how the MASLD-AGA relationship differs across levels of specific factors, such as carbohydrate intake, physical activity, and lipid-lowering medication use.

Table 3 Multiple logistic regression models examining the association between MASLD and AGA stratified by various subgroups potentially affecting MASLD.
Full size table

In individuals with carbohydrate intake above the population median, those with MASLD demonstrated a significantly increased risk of AGA by 45% (OR = 1.45, 95% CI: 1.19–1.77, p < 0.001). Similarly, among individuals with physical activity levels below the median, those with MASLD had 48% higher odds of developing AGA (OR = 1.48, 95% CI: 1.22–1.80, p < 0.001). Additionally, in individuals not using lipid-lowering medications, MASLD was associated with a 30% increased likelihood of AGA (OR = 1.30, 95% CI: 1.12–1.51, p < 0.001).

Conversely, no significant association between MASLD and AGA was observed in individuals with low carbohydrate intake, high physical activity levels, or those taking lipid-lowering medications. Furthermore, the relationship between MASLD and AGA was independent of dietary lipid, protein, and fiber intake, total daily calorie consumption, insulin resistance, and inflammatory markers (Table 3).

Discussion

Our findings imply that MASLD is independently related with increased odds of AGA, particularly among women, and that modifiable lifestyle factors, including carbohydrate intake and physical activity, and lipid-lowering medication use may influence this association. These results underscore the interplay between metabolic liver disease and dermatologic manifestations. To our knowledge, this is the first study to directly investigate the adjusted and pure link between MASLD and AGA. Thus, we identified MASLD as an independent correlate of AGA, particularly in females.

MASLD as a novel independent correlate for AGA, especially in females

Previous studies have shown that insulin resistance and metabolic dysfunction play a part in AGA. In 2023, Vinay, K., et al. discovered that insulin resistance and changes in hormone levels were present in one-third of males with early-onset AGA. These conditions were more common in people with severe alopecia30. Similarly, Bakry et al. (2014) observed that AGA was notably linked with metabolic syndrome and insulin resistance31. These studies highlight the metabolic component of AGA in men. Moreover, our results suggest a potential relationship between MASLD and AGA specifically in women. Estrogen plays a protective role in lipid metabolism and upregulates sex hormone binding globulin (SHBG), which binds circulating androgens. In women with MASLD, lower SHBG levels may result in a higher free androgen burden, potentially worsening AGA32,33. Our findings suggest that gender-specific approaches may be beneficial when evaluating patients with metabolic dysfunction. However, further studies are required to confirm whether MASLD is an independent risk factor for AGA. Taken together, insulin resistance-induced SHBG suppression, hepatic inflammation, and oxidative stress may represent a common mechanistic axis linking MASLD and AGA. Lifestyle factors like diet and exercise modulate these pathways, offering potential therapeutic leverage points.

Dietary carbohydrates as a modifier of the MASLD–AGA association

Another novel aspect of our results was investigating the contribution of carbohydrate intake to the MASLD-AGA association. Our findings show that the link between MASLD and AGA is stronger when people eat higher levels of carbohydrates.

In a recent network meta-analysis, it was indicated that among various dietary patterns, low-carbohydrate diet has the most positive effect on weight loss and body fat reduction17. Moreover, previous research has identified a link between high carbohydrate intake and MASLD development34. According to Basaranoglu et al., excessive carbohydrate consumption causes stress in the endoplasmic reticulum, turns on the stress-related kinase, breaks down mitochondria, and boosts apoptotic activity in the liver35. Additionally, some other studies found the role of excessive carbohydrate intake in AGA through polyol pathways that impact glucose metabolism and contribute to conditions like MetS, which are shown to be linked to AGA36,37,38, but our research is the first to specifically identify carbohydrate consumption as a key dietary determinant associated with the relationship between MASLD and AGA. These results underscore the likely role of dietary modifications in preventing AGA, particularly in individuals with MASLD, though further research is needed to confirm their clinical impact.

Physical activity and its protective role

Our findings suggest that physical activity may mitigate the association between MASLD and AGA, though additional investigations are mandatory to determine the fundamental mechanisms. In individuals with low physical activity, MASLD was significantly associated with AGA, whereas in those with high physical activity, the relationship was no longer significant. A new study by Chen et al. mentioned that exercise speeds up the release of neuregulin 4 (Nrg4), an important adiponectin, from fat tissue. Because Nrg4 changes the cGAS-STING signaling pathway in hepatocytes39, the liver has less steatosis and inflammation. On the other hand, a study by Bucarey et al. investigated the role of oxidative stress in the link between physical activity and MASLD; they showed that physical activity enhances myokine release, bioactive peptides released by skeletal muscle, thereby preventing the progression of MASLD through improved hepatic antioxidant defenses40. Besides that, research has shown that increasing insulin sensitivity through exercise is a key part of protecting against MASLD41. Regular physical activity could impact androgen level and reduce oxidative stress and inflammatory responses, factors associated with AGA42. Our study is the first to show that physical activity can help protect against AGA in people with MASLD. This shows how important it is to change people’s lifestyles as a way to help mitigate AGA risk.

Our analysis indicates that lipid-lowering medication use may mitigate the association between MASLD and AGA. Specifically, among individuals receiving these medications, the MASLD–AGA link was not significant, whereas it remained significant in those not on such therapy. Statins, the primary class of lipid-lowering agents, have demonstrated efficacy in reducing hepatic steatosis and inflammation in MASLD patients43. The mechanisms underlying these benefits include improvement of endothelial function, reduction of oxidative stress, and modulation of inflammatory responses44. Additionally, statins may influence androgen metabolism by upregulating hepatic SHBG production, leading to decreased free androgen levels45. Given that elevated androgen levels contribute to AGA pathogenesis, the SHBG-mediated reduction in bioavailable androgens might explain the diminished MASLD–AGA association observed in statin users. These findings suggest that lipid-lowering therapies could offer dermatologic benefits in addition to their established cardiovascular and hepatic advantages.

Mechanistic insight: insulin resistance and inflammation

Insulin resistance and inflammation both increase each other with different mechanisms46,47. Insulin resistance dysregulates SHGB, leading to higher androgen bioavailability, which plays a crucial role in AGA and hepatic fat accumulation in MASLD48,49. On the other hand, inflammatory cytokines promote hepatic steatosis and can lead to fibrosis, advancing to more severe liver disease50. Additionally, a recent study carried out by Plante and colleagues demonstrated the role of perifollicular inflammation in AGA51. However, higher or lower insulin resistance and inflammation levels did not mediate or modulate the MASLD-AGA association. Either way, we adjusted for their pathogenesis role in the MASLD-AGA association to get more accurate and definite results. Insulin resistance and systemic inflammation are central to both MASLD and AGA pathogenesis, contributing to hepatic fat accumulation and androgen dysregulation. While these factors are implicated in the observed association, they did not mediate the effect in our model.

Strengths and limitations

Strengths of this study consist of using of a large, well-characterized population-based cohort, which enhances the generalizability of our findings. The standardized data collection procedures and rigorous exclusion criteria helped reduce potential confounding and misclassification biases. Notably, the AGA diagnosis was conducted by trained dermatologists using established clinical grading systems (Hamilton-Norwood and Ludwig scales), increasing the validity of the outcome assessment. Using FLI as a non-invasive, validated measure of hepatic steatosis, calibrated specifically for the Iranian population, allowed for practical identification of MASLD in the absence of imaging or biopsy. Furthermore, the inclusion of a wide range of relevant covariates, including sociodemographic, dietary, metabolic, and lifestyle factors, enabled comprehensive multivariate adjustment and meaningful subgroup analyses that helped uncover potential effect modifiers such as sex, physical activity, and carbohydrate intake. However, several limitations must be acknowledged. First, due to the cross-sectional design, causal inferences cannot be drawn; the directionality of the relationship between MASLD and AGA remains unclear. Second, while the FLI has been validated, it remains an indirect surrogate marker of hepatic steatosis and may introduce misclassification, especially in individuals with borderline values. Third, residual confounding is possible, particularly from unmeasured variables such as genetic predisposition, androgen levels, and detailed hair care practices, which were not captured in the dataset. Additionally, alcohol consumption was self-reported and may have been underreported due to cultural and legal considerations in Iran, potentially affecting the accurate identification of MASLD. Moreover, the study focused solely on a rural population, which may limit the extrapolation of the findings to urban or more ethnically diverse populations. Finally, the lack of data on menopausal status among all the female participants may limit our ability to fully assess the relationship between MASLD and AGA in women. since menopause has been recognized as a significant factor in the development of AGA, with hormonal changes influencing hair loss patterns. Future studies should consider incorporating menopausal status as a key variable to better understand its potential impact on AGA in this population. One notable finding of our study was the high prevalence of AGA (75%), which exceeds typical estimates of up to 50% in men and 30% in middle-aged women. Several factors may explain this discrepancy. Our cohort was older than the typical general population (35–70 years, mean 48.62 ± 9.63), and both univariable and multivariable analyses confirmed that AGA was significantly associated with age, supporting its role as a key determinant. Moreover, AGA is particularly common in Iran52, with prior studies reporting higher rates than global averages, which likely contributes to our findings. Nonetheless, we observed high AGA prevalence even among the younger participants in our cohort, underscoring that factors beyond age also play an important role.

Conclusion

This research identifies a considerable link between MASLD and AGA, particularly among women, even after adjusting for multiple confounders. These results imply that MASLD may serve as a novel and independent risk factor for AGA. Moreover, our results underscore the modifying roles of lifestyle factors, particularly high carbohydrate intake, low physical activity, and absence of lipid-lowering therapy, in shaping this association. Given the high global prevalence of both MASLD and AGA, clinicians, especially dermatologists and hepatologists, should consider routine metabolic evaluations in patients presenting with early-onset or progressive AGA. Recognizing AGA as a potential clinical marker of underlying metabolic dysfunction may facilitate earlier detection and management of MASLD. Furthermore, emphasizing lifestyle interventions such as increased physical activity, carbohydrate moderation, and potential use of lipid-lowering agents may help mitigate the dermatologic impact of metabolic disease. Prospective cohort investigations are required to approve the temporal and causal nature of the MASLD–AGA association. Future research should investigate the underlying biological mechanisms with a focus on sex-specific hormonal and metabolic pathways. Randomized controlled trials assessing whether lifestyle modifications or pharmacological interventions targeting MASLD can prevent or reduce AGA progression would provide clinically actionable insights. Additionally, incorporating genetic and inflammatory markers may enhance understanding of individual susceptibility to MASLD-related hair loss.