Is Acanthosis Nigricans a Marker for Metabolic Syndrome? A Systematic Review

Abstract

Background/Objective

Certain skin conditions are proven indicators of underlying systemic disease. Specifically, acanthosis nigricans (AN) is commonly associated with type 2 diabetes and insulin resistance. However, AN may also be a marker of metabolic syndrome (MS). MS is linked to several health conditions leading to death. Therefore, early identification of MS is crucial in preventing disease progression. While AN may provide a visual cue for MS risk, its utility as a clinical screening tool remains underexplored. To address this, we conducted a systematic review to evaluate the association between AN and MS.

Methods

We systematically searched PubMed, Embase, and Cochrane Review using the term “acanthosis nigricans,” identifying 656 articles, including 322 case reports, that reported AN in association with at least one MS criteria, as defined by the American Heart Association. Article quality was determined using the Strength of Recommendation Taxonomy (SORT) grading system.

Results

Across the 656 included articles, the unweighted mean proportion of study participants with AN was 70.8% (standard deviation of 34.9%), reflecting that many studies evaluated broader patient populations in which AN was one of several conditions assessed. Of all MS criteria, insulin resistance showed the highest mean prevalence of AN at 72.6%. Within all articles, 65.9% met more than one MS criterion, with co-occurrence of insulin resistance and increased body mass being the most common combination (25.5%). In a subanalysis of non-case-report studies in which all participants had AN, 67.4% met the full criteria for MS, with increased body mass being the most frequently reported individual criterion (89.7% of cases).

Conclusions

This review provides a better estimate of AN observed in individuals with MS or in those at risk for MS. Integrating dermatologic examination for AN into primary care and preventive health may improve early detection of MS, reduce costs and unnecessary testing, and promote targeted interventions. Educational efforts are needed to increase clinician awareness of the association between AN and MS.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s40257-026-01058-7.

Key Points

Acanthosis nigricans is a skin sign that is a reliable indicator of metabolic syndrome. Out of the metabolic syndrome criteria, it is associated the most with insulin resistance and increased body mass.

Introduction

The potential of using acanthosis nigricans (AN) as a marker of metabolic syndrome (MS) has been suggested in dermatology, endocrinology, and pediatrics [1–3]. However, these recommendations rely largely on case reports, case series, and a limited number of original articles. Few systematic reviews exist on the relationship between AN and MS, and none have quantified the degree of this association with individual MS criteria. This systematic review is a comprehensive evaluation of the literature to best elucidate the degree of association between MS and AN, and investigates whether AN may serve as a predictive tool for identifying patients at risk for MS.

Metabolic Syndrome: Definition, Diagnosis, and Benefits of Early Detection

MS is a cluster of metabolic abnormalities that significantly increase the risk of heart disease, stroke, and type II diabetes mellitus (T2DM). These abnormalities include insulin resistance, dyslipidemia, central obesity, and hypertension [4]. The concept of MS first emerged in 1988, when Gerald Reaven observed that insulin resistance frequently co-occurred with cardiovascular diseases and T2DM, coining the term “Syndrome X” [5]. The qualifier “metabolic” was later added to differentiate it from unrelated cardiac conditions with the same name [5]. To formally diagnose MS, five key indices are evaluated: (1) waist circumference and/or body mass, (2) fasting glucose levels, (3) triglyceride levels, (4) high-density lipoprotein (HDL) cholesterol levels, and (5) blood pressure (BP) [5]. A diagnosis of MS is made if three or more of these criteria are found to be abnormal; however, some sources define elevated triglycerides as greater than 150 mg/dL (World Health Organization [WHO] 1998), while others define it as greater than 177 mg/dL (European Group for the Study of Insulin Resistance [EGIR] 1999) [6, 7]. Standardized definitions remain an ongoing area of debate. The pathophysiology of MS also remains incompletely understood, with some investigators proposing that each component is a distinct pathology, while others suggest it is a single disease process [4]. It is likely that the pathogenesis involves both genetic and acquired factors via mechanisms involving inflammation and metabolic dysfunction [8]. Acquired risk factors for the development of MS include high caloric diets, sedentary lifestyles, and overeating, which promote central adiposity [9]. At the cellular level, chronic inflammation, neurohormonal changes, and insulin resistance appear to drive progression of MS and its evolution into other chronic disorders [10]. MS produces a pro-inflammatory state that can result in atherosclerotic damage [11], cardiovascular disease, acute cardiovascular events [12, 13], and T2DM [14]. MS affects roughly one in five individuals across the USA, with an estimated 53.9% increase in economic burden for those with MS compared with those without [11, 12]. Therefore, early detection, before the onset of additional symptoms and end organ damage, has the potential to substantially reduce the health and economic impact of MS within our communities.

Acanthosis Nigricans as a Diagnostic Indicator of Metabolic Syndrome

AN is a physical sign that can aid in the early detection of MS. It is strongly associated with hyperinsulinemia, a key feature of MS [13], and it presents as a velvety, darkened skin lesion with pathological features of hyperkeratosis and papillomatosis [14]. It usually occurs in intertriginous areas, most commonly appearing on the posterior neck but also frequently found in the axilla, groin, elbows, and knees [15]. At a cellular level, insulin regulates the growth of keratinocytes, endothelial cells, and fibroblasts [16]. In a hyper-insulinemic state, there is increased proliferation, migration, and extracellular matrix secretion by these cells, leading to the development of AN [16]. Therefore, AN may serve as a valuable tool for identifying individuals with hyperinsulinemia who are at risk of developing T2DM [17], permitting early intervention. Early intervention is advantageous from multiple standpoints, including cost, with one study citing that the average lifetime cost per person was projected to be $13,240 for diagnosed diabetes and only $500 for prediabetes [18]. The use of AN as a sign of MS may be a low-cost, noninvasive screening sign that permits early intervention before the development of downstream consequences. Despite its clinical relevance, AN’s sensitivity and specificity as a marker of MS remains uncertain [13]. While some studies support AN as a predictor of increased insulin levels, few have systematically quantified its association with MS components [14, 19]. One systematic review found that, in children/adolescents with obesity, AN was consistently associated with T2DM, insulin resistance, obesity, and T2DM risk [20]. Building upon this, our systematic review aims to quantify the prevalence of AN in individuals with MS, or its diagnostic criteria, to further evaluate its utility as a diagnostic tool.

Methods

Defining Metabolic Syndrome Criteria

We used the MS diagnostic criteria from the American Heart Association (AHA) and National Heart, Lung, and Blood Institute (NHLBI) as a uniform method to determine the criteria met by subjects in the included studies [21]. The criteria require at least three of the following for a diagnosis of MS: (1) a waist circumference of at least 40 inches (102 cm) in men or 35 inches (89 cm) in women, measured at the top of the iliac crest at the end of a normal expiration, (2) triglyceride levels of at least 150 mg/dL (1.70 mmol/L) or pharmacologic therapy for elevated triglyceride levels, (3) HDL-cholesterol levels of less than 40 mg/dL (1.05 mmol/L) in men or less than 50 mg/dL (1.30 mmol/L) in women or pharmacologic therapy for reduced HDL cholesterol levels, (4) a systolic blood pressure of at least 130 mmHg or diastolic blood pressure of at least 85 mmHg or pharmacologic therapy for hypertension, and (5) a fasting glucose level of at least 100 mg/dL (5.6 mmol/L) or pharmacologic therapy for elevated fasting glucose levels.

Study Selection

This review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) 2020 guidelines [22]. PubMed, Embase, and Cochrane Review were searched in October 2023 to identify all articles with the term “acanthosis nigricans,” encompassing studies of all languages and years up to the search date. The review protocol was not registered in the International Prospective Register of Systematic Reviews (PROSPERO). After removing duplicates, an initial screening of all abstracts was conducted by six medical students (J.G., K.D., J.R.M., T.T., E.P.M., S.K.). Two students screened abstracts for inclusion for full-text review, and a third student resolved conflicts. Articles were included if they included a study population with AN (or case report or series) and reported both an AN prevalence (defined in the section “Methods”) and at least one MS criteria. To maximize inclusivity, articles were not required to report on all of the MS criteria, resulting in missing data. Exclusion criteria included lack of full text, inaccessibility, or publication as a commentary. After exclusion of 1857 abstracts, full-text articles were assessed for eligibility by eight medical students and one research fellow (J.G., K.D., J.R.M., T.T., E.P.M., J.N., K.C., K.N., M.J.). Studies that were found to not meet criteria were excluded, and data were extracted from the articles included. A re-review of all articles included by the first team was performed by three medical students (K.D., J.R.M., J.G.) to validate selection. Studies were included for final analysis if they discussed AN in the context of MS or at least one MS criteria, were published in a peer-reviewed journal, and included human subjects.

Data Extraction

The following data were extracted from each study: affiliated institutions of first authors, publication journal, publication year, article type, study type, total number of study participants, MS criteria measured in the study (low HDL cholesterol, elevated triglycerides, insulin resistance, increased body mass, cardiovascular dysfunction), whether the study discussed and/or had subjects that met full (three or more) criteria for MS diagnosis, medical conditions causing MS criterion, medications causing MS criterion, and the prevalence of AN. We defined AN prevalence as the total number of study participants with AN divided by the total number of study participants (in case reports, prevalence was defined as 100%). In cases where a study did not explicitly provide data for a given criterion, to avoid any overestimation of the association between AN and those features, we interpreted this as indicating the absence of that criterion and categorized it accordingly. Following this, three medical students (K.D., J.R.M., J.G.) re-extracted data to perform a subanalysis, focusing on study populations with 100% prevalence of AN. They recorded the number of individuals who met each individual MS criterion, as well as those who met the full diagnostic criteria for MS.

Quality Assessment

To classify the level of evidence, each article was assessed using the Strength of Recommendation Table (SORT) scale [23]. SORT was selected because it allows for evaluation of clinical relevance and evidence quality across a diverse range of study types, including case reports and case studies, for which other commonly used tools such as the Newcastle–Ottawa scale are not applicable [24]. SORT grading levels were based on quality and evidence type. Level 1 signified good quality and patient-oriented evidence. Level 2 signified limited quality and patient-oriented evidence. Level 3 signified consensus guidelines in absence of data, opinion, disease-oriented evidence, case reports/series, and low-quality evidence.

Data Analysis

Descriptive statistics were used to summarize the characteristics of the articles included in the analysis. Counts and percentages were used to summarize the articles that met the MS criteria; UpSet plots (Ahlmann-Eltze) from the ggupset package visually show the frequencies of the criteria and the combinations of criteria met [25]. The mean prevalence of AN and other summary statistics were calculated as unweighted averages across studies. In addition, sample-size-weighted prevalence estimates were calculated to account for differences in study population size. A linear regression model was used to evaluate the relationship between an article’s AN prevalence and SORT levels. Linear mixed-effects models were used to assess the relationship between articles’ AN prevalence and the MS criteria met, with a random intercept for articles to account for variability across articles. Two-sample t-test compared AN prevalence among articles meeting fewer than three MS criteria versus those meeting three or more. A predefined subanalysis was conducted including studies with 100% prevalence of AN. A subgroup analysis of case reports was also done. All statistical analyses were performed using R: A Language and Environment for Statistical Computing [26].

Results

Article Selection

The initial database search across PubMed, Embase, and Cochrane Review yielded 5074 articles, from which 2108 duplicates were removed, yielding a total of 2966 articles (Fig. 1). Following abstract screening, 1857 articles were excluded per criteria above (“Method” section; Fig. 1). After full-text assessment of the remaining 1109 articles, an additional 453 were excluded on the basis of our predefined criteria. To ensure accuracy, a second reviewer independently verified the extracted data (“Methods” section).

Fig. 1.

PRISMA flow diagram illustrating the process of article selection, data extraction, and analysis. Subanalysis was limited to studies reporting 100% AN prevalence within their study populations. AN acanthosis nigricans, SORT Strength of Recommendation Taxonomy

Characteristics of Included Articles

In total, we excluded 2310 articles that did not include AN in an appropriate population or did not include the necessary MS elements, leaving 656 articles included for analysis (Fig. 1). These studies comprised a cohort of 318,300 total individuals. Demographic characteristics are summarized in Table 1 and Fig. 2. Study designs varied, with cross-sectional reports and case series most represented. The average SORT score, excluding case reports, was 2.4, on a scale of 1 (best quality) to 3 (worst). A detailed analysis of study quality and its relationship with metabolic syndrome (MS) criteria is presented in the section “Study Quality and Prevalence of MS Criteria by SORT Level” (Section 3.8). For a full reference list of included articles, and their details, see Supplemental Tables S1–3.

Table 1.

Demographics of all articles

Category	Subcategory	Number of articles	Percentage of total
Study type	Basic science	7	1.1%
	Case–control study	21	3.2%
	Case report	322	49.0%
	Case series	17	2.6%
	Clinical trial	7	1.1%
	Cohort study	43	6.6%
	Cross-sectional report	148	22.6%
	Randomized control trial	24	3.6%
	Other	67	10.2%
SORT level	1	108	16.5%
	2	167	25.4%
	3	381	58.1%
Publication year	Before 1980	25	3.8%
	1980–1989	47	7.2%
	1990–1999	56	8.5%
	2000–2009	104	15.9%
	2010–2019	272	41.5%
	2020 and later	152	23.1%
Journal types	Dermatology	105	16.0%
	Endocrine	229	34.9%
	Other	322	49.1%
Publishing location	Africa	11	1.7%
	Asia	171	26.1%
	Europe	137	20.9%
	Middle East	77	11.7%
	North America	204	31.1%
	Oceania	9	1.4%
	South America	47	7.1%

Descriptive statistics of the 656 articles included in the analysis. “Other” includes journals focused on obstetrics and gynecology, allergy, neurology, clinical research, pharmaceuticals, gastroenterology, general medicine, pediatrics, nursing, surgery, nutrition, ophthalmology, oncology, primary care, cardiology, sleep medicine, public health, genetics, infectious disease, psychiatry, and epidemiology. SORT Strength of Recommendation Taxonomy

Fig. 2.

AN as a worldwide phenomenon. Representation of the publishing location (determined using the affiliation of the first author), and how many articles they produced, using a world map. AN acanthosis nigricans

Representation of MS Criteria and Combinations of Criteria

All 656 included articles were analyzed to assess the relationship between AN and each MS criterion. The most frequently reported criteria were insulin resistance (83.8%, n = 550), increased body mass (75.0%, n = 492), and elevated triglycerides (29.9%, n = 196), while low HDL (22.0%, n = 144) and cardiovascular dysfunction (28.5%, n = 187) were the least observed criteria. Overall, 65.9% (432) of studies reported study populations meeting more than one MS criterion, with co-occurrence of insulin resistance and increased body mass being the most common combination (25.5%, n = 167). The least common combinations were low HDL + increased body mass, and cardiovascular dysfunction + increased body mass + triglycerides. Notably, 37.8% (n = 248) of the articles reported study populations that met the definition of MS (at least three MS criteria), and 7% (n = 46) reported populations that met all five criteria (Fig. 3).

Fig. 3.

AN is most commonly observed with insulin resistance and increased body mass. An UpSet plot showing the different combinations of MS criteria that were reported in all articles. Connected and filled dots within rows represent the corresponding set of MS criteria. Empty dots represent the MS criteria not involved. Each row represents an individual measure, with articles reporting multiple measures represented by multiple rows. MS criteria may be analyzed in isolation or in combination with others. AN acanthosis nigricans, MS metabolic syndrome, HDL high-density lipoprotein

Relationships between AN Prevalence and MS Criteria

The average unweighted prevalence of AN across all articles was 70.8%, while the weighted prevalence was 17.3%. The unweighted mean and median prevalence values reflect study-level estimates, which are strongly influenced by the large number of small studies conducted in enriched or high-risk clinical populations where AN prevalence is often very high (80–100%). In contrast, the sample-size-weighted prevalence reflects the underlying patient population and is driven by the largest epidemiologic studies, which typically enrolled thousands to over 100,000 participants and reported substantially lower prevalence values. Because these large studies represent the majority of all patients in the dataset, the weighted pooled prevalence (17.3%) is much lower than the unweighted mean (70.8%). This discrepancy is expected and reflects differences in study design and population sampling rather than methodological error.

The highest AN prevalence was observed in participants with insulin resistance at 72.6% (Table 2). Interestingly, AN prevalence was higher in articles that met fewer than three criteria (74.6%) compared with those that met three or more (66.7%) (Table 2).

Table 2.

Stratification of AN prevalence by MS criteria and SORT level in all articles

Category	Subgroup	Mean (SD)
Prevalence of AN	Unweighted Weighted	70.8 (34.9) 17.3
Prevalence of AN within each SORT criteria	1	45.9 (32.1)
	2	47.3 (30.3)
	3	88.4 (26.0)
Prevalence of AN within each MS criterion	Low HDL	64.7 (34.4)
	Elevated triglycerides	70.9 (33.4)
	Insulin resistance	72.6 (34.2)
	Increased body mass	66.7 (34.7)
	Cardiovascular dysfunction	61.9 (36.2)
Prevalence of AN that met MS diagnosis	Less than 3 MS criteria met	74.6 (34.2)
	3 or more MS criteria	66.7 (34.4)

Average AN prevalence for all articles, within each SORT criterion, and for each individual MS criterion

AN acanthosis nigricans, MS metabolic syndrome, HDL high-density lipoprotein, SORT Strength of Recommendation Taxonomy, SD standard deviation

Modifying Factors Potentially Contributing to the Presence of AN

In addition to reporting the absolute prevalence of AN in the articles, we collected data on modifying factors, such as concomitant diseases present in individuals with AN (Table 3). Endocrine disorders were the most frequently reported comorbidities (53.1%), though 33.4% of articles did not report any associated diseases. Among reported medications, insulin was the most frequently cited drug found in association with AN development, with three total cases (Table 3). No causality assessments were performed to determine whether the medications directly contributed to AN development, though.

Table 3.

Observed disease categories and drugs found in association with AN development

Category	Subcategory	Number of articles	Percentage of total (%)
Disease categories	Dermatologic	8	1.2%
	Ears, nose, and throat	1	0.2%
	Endocrine	349	53.1%
	Gastrointestinal	5	0.7%
	Genetic	35	5.3%
	Gynecologic	1	0.2%
	Multidisease multisystem	17	2.6%
	Musculoskeletal	3	0.5%
	Neurologic	1	0.2%
	Oncologic	14	2.1%
	Rheumatologic	3	0.5%
	N/A	219	33.4%
Pharmaceuticals	Growth hormone	2	0.3%
	Insulin	3	0.5%
	Oral contraceptive pill	1	0.2%
	Oral prednisone	2	0.3%
	Protease inhibitor	1	0.2%
	Risperidone	1	0.2%
	Nicotinic acid	1	0.2%
	Phentermine	1	0.2%
	N/A	644	97.9%

Diseases co-occurring with AN were recorded and sorted into body system categories. Additionally, medications found in association with the development of AN were also recorded. AN acanthosis nigricans, N/A not available

Case Report Analysis

Although limited by their anecdotal nature, limited generalizability, and susceptibility to bias, 322 case reports were included in our analysis to broaden estimates of prevalence. The benefit of looking at a large number of AN case reports is that it allows us to evaluate additional features present in AN beyond simply identifying associations with other factors.

Among these, insulin resistance was the most commonly reported MS criterion (84.8%, n = 273), followed by increased body mass (57.8%, n = 186), elevated triglycerides (29.5%, n = 95), low HDL (16.8%, n = 54), and cardiovascular dysfunction (21.7%, n = 70). Insulin resistance in isolation was the most commonly observed criterion, reported in 25.5% (n = 82) of the case reports. The most observed overlapping combination of criteria were insulin resistance and increased body mass (19.3%, n = 62). In total, 33.2% (n = 107) of case reports met the definition of MS, and 1.6% (n = 5) met all five MS criteria (Fig. 4).

Fig. 4.

AN is most commonly observed with insulin resistance and body mass in case reports. An UpSet plot showing the different combinations of MS criteria that was reported in case reports only. Connected and filled dots within rows represent the corresponding set of MS criteria. Empty dots represent the MS criteria not involved. Each row represents an individual measure, including multiple rows for articles reporting multiple measures. AN acanthosis nigricans, MS metabolic syndrome, HDL high-density lipoprotein

Analyzing Prevalence of MS Criteria in a Subpopulation of Individuals with AN

To evaluate potential bias from case report inclusion, we conducted a subanalysis excluding case reports of 213 articles reporting MS criteria among individuals with AN. In this cohort, the MS criteria were observed (most common to least common) as follows: increased body mass (89.7%), insulin resistance (76.0%), low HDL (71.2%), elevated triglycerides (64.7%), and cardiovascular dysfunction (53.8%). Additionally, the prevalence of individuals that met full criteria for MS in individuals with AN was 67.4% (Table 4). These results were consistent with the findings from the comprehensive analysis, suggesting minimal bias from case report inclusion; therefore, no additional tests were conducted to further assess bias. A statistically significant association was found between MS criteria and the AN prevalence reported by articles (p < 0.001, linear mixed effect model with random effect for article).

Table 4.

MS criteria representation in patients with AN

Category	Subgroup	Number of articles	Mean (SD)
Prevalence of each MS criterion	Low HDL	17	71.2% (24.6%)
	Elevated triglycerides	15	64.7% (24.7%)
	Insulin resistance	123	76.0% (31.9%)
	Increased body mass	140	89.7% (22.5%)
	Cardiovascular	23	53.8% (32.6%)
Prevalence of MS	Three or more MS criteria	30	67.4% (28.2%)

Average prevalence of each MS criteria, and MS itself, in articles that included study populations with 100% of individuals having AN. AN acanthosis nigricans, MS metabolic syndrome, HDL high-density lipoprotein, SD standard deviation

Study Quality and Prevalence of MS Criteria by SORT Level

Given the limitation of case reports, we restricted our SORT-based quality analysis to 334 non-case-report studies to assess how study quality may have affected observed associations between AN and MS. The average SORT score among these studies was 2.4 (1 = high quality, 3 = low quality), with 101 studies classified as SORT level 1, 170 as level 2, and 63 as level 3. Among these, 152 studies reported full MS diagnoses. The most frequently reported MS components were insulin resistance or high glucose (n = 277), followed by increased body mass (n = 306). Table 5 presents the distribution of each MS component across SORT levels. Notably, studies rated SORT level 3 were less likely to report full MS diagnoses compared with level 2, despite similar reporting of individual criteria. This may reflect differences in study rigor, methodology, or screening practices, and suggests that lower-quality studies may overestimate or underreport MS prevalence.

Table 5.

MS criteria stratified by SORT level (excluding case reports)

MS criteria	SORT level 1 (n = 101)	SORT level 2 (n = 170)	SORT level 3 (n = 63)
Metabolic syndrome (full diagnosis)	43	90	19
Low HDL	25	54	11
High triglycerides	23	59	19
Insulin resistance/high glucose	71	148	58
Increased body mass	84	164	58
Cardiovascular/hypertension	31	68	18

Number of studies meeting each individual metabolic syndrome criterion (low HDL, high triglycerides, insulin resistance or high glucose, obesity or increased body mass, cardiovascular disease or hypertension) as well as full MS diagnosis, stratified by SORT evidence level. Articles may have met more than one MS criterion. HDL high-density lipoprotein, MS metabolic syndrome, SORT Strength of Recommendation Taxonomy

Discussion

Important Takeaways Regarding the Demographics of Included Articles

Our comprehensive systematic review began with a broad search of all articles mentioning “acanthosis nigricans” to identify all relevant literature. Of the 656 articles that met inclusion criteria, the majority were case reports (49.0%), followed by cross sectional studies, (22.6%) and cohort studies (6.6%). Most studies were classified as SORT level 3 (58.1%), likely reflecting the large amount of case reports. Regarding timeline, 23.1% of the articles were published after 2020, suggesting growing interest in MS and noninvasive screening tactics. Only 16.0% were published in dermatologic journals, while 34.9% were published in endocrine focused journals, with the remainder spanning other specialties. This indicates a shortage of research on this topic from a dermatological standpoint, emphasizing the necessity for more dermatology-focused studies on the use of AN as a clinical marker. Geographically, most articles were published from locations in North America (31.1%) and Asia (26.1%), reinforcing the global relevance of MS and its associated risks across populations of all ethnicities, races, and genders.

AN Prevalence in the Context of MS Criteria: Insulin Resistance Shows the Highest Prevalence of AN

Our initial analysis, which examined all articles, including case reports and non-case reports, referencing at least one of the five diagnostic criteria for MS in conjunction with the prevalence of AN (defined as the number of participants with AN divided by the total study population), showed an unweighted prevalence of 70.8% averaged across all articles and a weighted prevalence of 17.3%. The difference between unweighted (70.8%) and weighted (17.3%) prevalence estimates highlights the heterogeneity of the available literature. Small studies and case series frequently reported very high AN prevalence, whereas the largest population-based studies contributed substantially lower estimates. Because sample-size-weighted prevalence was virtually unchanged when all case reports were removed, this pooled estimate appears robust and provides a more accurate reflection of the underlying patient population. Together, the unweighted and weighted results offer complementary perspectives on study-level reporting versus patient-level prevalence.

Among the five MS criteria, insulin resistance showed the highest prevalence of AN at 72.6%, while cardiovascular dysfunction had the lowest at 61.9%. This could suggest that, of the five MS criteria, this may have the weakest association with AN. This is difficult to definitively determine, however, as the absence of reporting of this criteria in each article was not recorded during data collection, and the population did not meet diagnostic criteria. Further work is needed to elucidate the validity of this association, though 61.9% is still relatively high.

Additionally, AN prevalence was slightly higher in populations that did not meet up to three MS criteria (74.6%) compared with those that met three or more MS criteria (66.7%). Consequently, this data may indicate that AN is more frequently observed in populations with MS risk factors than among individuals who have MS, though for the aforementioned reasons, this claim needs to be further studied. If this finding is proven true, this could support that AN is a better marker of insulin resistance than MS, and the presence of this criterion alone may be sufficient for the development of AN without the need for additional MS criteria. Importantly, however, the cross-sectional nature of the majority of included studies prevents determination of whether AN precedes insulin resistance or develops as a consequence of it. The temporal and causal relationship between AN and MS-related metabolic abnormalities remains unclear and requires prospective longitudinal investigation.

Prevalence of MS Criteria in a Subpopulation of Individuals with AN: Stronger Association between Body Mass and AN

In lieu of the above findings, we attempted to further elucidate the relationship of MS criteria and AN with a non-case-report subanalysis, which could yield more reliable and translatable results. Among participants with AN in this subanalysis, 67.4% met the criteria for MS diagnosis. This subanalysis yielded a statistically significant association between AN and MS diagnosis, hinting at an association with other MS criteria that was not evident in the initial broad analysis, though further prospective research is needed to validate our interpretation of the data (“Analyzing Prevalence of MS Criteria in a Subpopulation of Individuals with AN” in the “Results” section; “AN Prevalence in the Context of MS Criteria: Insulin Resistance Shows the Highest Prevalence of AN” in the “Discussion” section). Contrary to our earlier findings, increased body mass emerged as the most observed MS criteria (89.7%), surpassing insulin resistance. This shift may reflect more consistent documentation of body mass in structured studies or point to a stronger physiological association with AN. As the causal relationships between various MS criteria are better elucidated, it may become more apparent why AN is more tightly associated with body mass or insulin resistance, or both.

Implications for AN as a Diagnostic Tool for MS Screening

The findings of our subanalysis have important implications for dermatologic screening practices. By further investigating the association between AN and the diagnosis of MS, when analyzed in a more targeted population, this study underscores the importance of using visual examination to aid in the screening for systemic conditions. Clinicians should consider AN as a cutaneous clue to underlying systemic conditions, prompting further investigation into the diagnostic criteria for MS, such as glucose and lipid testing as well as blood pressure monitoring. As with other established physical signs of underlying disease, such as dyspnea in chronic obstructive pulmonary disease (COPD) (> 40% prevalence) and jaundice in hepatic pathology (55%), AN may serve as a reliable indicator for MS [27, 28]. Given its comparable or even higher prevalence among at risk populations, AN warrants similar clinical consideration. Leveraging this accessible marker could reduce diagnostic delays and associated health care costs, especially in resource-limited settings.

Study Quality and Evidence Strength

The strength of our conclusions is inherently linked to the quality of the included studies. Over half (58.1%) were classified as low scientific quality (SORT level 3), partly due to the large number of case reports in the dataset, which are limited in scope and SORT grading. To account for this, we performed a secondary analysis excluding case reports, allowing for a more rigorous evaluation of the association between AN and MS across higher-quality study designs.

In this subset, notable differences in the distribution of MS criteria emerged across SORT levels. While insulin resistance and obesity remained the most consistently reported components, the proportion of studies reporting full MS diagnoses decreased with lower SORT quality. This trend may reflect methodological limitations in lower-quality studies, such as publication bias or underestimation of prevalence. Alternatively, it may suggest that even among high-quality studies, underscreening or underreporting of MS criteria persists, potentially leading to underestimation of the true association between AN and MS. Additionally, variation in how MS criteria were defined and measured across studies further limits our ability to evaluate the true relationship between AN and individual components of MS.

Limitations of this Systematic Review

While this study provides valuable insights, there are several limitations that should be considered. First, this review was not registered in PROSPERO, which represents a limitation as prospective registration helps ensure methodological transparency and minimize selective reporting. However, this review adhered to PRISMA 2020 guidelines, supporting its overall methodological rigor. Another limitation is that 93.2% of the articles did not measure all five MS criteria, with only 36.5% assessing three or more criteria necessary for a full MS diagnosis. When a criterion was not explicitly reported, we interpreted this as indicating the absence of that criterion. While this approach was applied consistently, it may have introduced bias if absence of reporting reflected non-measurement rather than true absence. This would likely bias our findings toward an underestimation of the association between AN and MS.

Additionally, 58.1% of the articles were of moderate to low scientific quality as classified by the SORT grading scale. Many included articles were case reports, which, while valuable for descriptive insights, are inherently limited in generalizability and SORT grading. Though we conducted a secondary analysis excluding case reports to address this, the heterogeneity of study quality remains a limitation. Inconsistent definitions and documentation of MS components further hindered our ability to evaluate individual associations with AN.

One limitation of using aggregated published data is the potential for overlapping patient populations across included studies, particularly among studies conducted within similar institutions or geographic regions. Because individual-level data were not available, this could not be assessed and may have resulted in duplication of some patients, which could bias prevalence estimates. Additionally, aggregated data introduce the potential for an ecological fallacy, when patterns observed at the group level do not necessarily apply to individuals. Aggregated data also lack information about individual variability, which limits our ability to make person-level predictions and introduces the potential for bias, including publication bias. However, in the absence of individual-level data, and given the practical challenges of collecting it, analyzing high-quality aggregated findings remains the best available approach. Encouragingly, studies comparing aggregated and individual-level analyses in other contexts have found that, while individual-level data allow for more detailed modeling, the overall direction and strength of associations are often similar [29, 30]. This suggests that our findings likely reflect a real and meaningful relationship between acanthosis nigricans and metabolic syndrome, even if they cannot directly support predictive conclusions. Additionally, reporting of AN morphology or severity varied across studies, limiting our ability to describe how AN was characterized or to incorporate existing classification systems.

How Future Research Can Further Investigate the Diagnostic Utility of AN as a Marker of MS

Future efforts should focus on conducting high-quality, prospective, longitudinal studies to accurately evaluate the association between AN and MS. Additionally, studies are needed to investigate the long-term health consequences faced by individuals with MS who developed AN before their diagnosis, which will strengthen the relevance of this research and highlight the need for effective screening. To improve the accuracy of future research, we recommend adopting standardized reporting practices that clearly differentiate between missing data and instances where criteria are not met. This would enhance clarity and facilitate more precise analysis and strong conclusions. Future research would benefit from incorporating standardized AN morphological or severity classifications, such as those proposed in recent literature, to improve consistency in reporting and comparability across studies [31]. Moving forward, our “Skin as a Window to Health” study aims to evaluate the cutaneous associations of MS in a large study population. We will do this by conducting skin examinations and assessing participants prospectively for AN and each MS criterion through lab tests, blood pressure reads, and body mass measurements.

Conclusions

This systematic review demonstrates that AN is a clinically meaningful cutaneous marker associated with MS. Insulin resistance showed the strongest association with AN overall, whereas subanalysis identified increased body mass as the most prominent correlate, reinforcing the link between AN and metabolic dysfunction. The consistent association between AN and MS supports its utility as a noninvasive, accessible screening tool that may aid in earlier identification of at-risk individuals. Skin examinations for AN may facilitate the early identification of MS features, emphasizing the importance of a thorough dermatologic evaluation. Integrating routine skin examinations for AN into clinical practice could enhance the personalization of MS screening, result in earlier diagnosis, and potentially prevent other health complications associated with these diseases. It is especially important for primary care physicians to integrate these skin exams into their practice, as many people do not have access to a dermatologist. Additionally, visual assessments can detect abnormalities at a fraction of the cost of traditional diagnostic tests, thus minimizing unnecessary testing for individuals at low risk and lower overall healthcare costs. This knowledge can also empower patients by teaching them to visually recognize signs of health issues, encouraging a proactive approach to their well-being. We propose that refining screening protocols for patients with AN to include assessments for each MS criteria could be beneficial, as early identification may allow for timely intervention and improved patient outcomes. Our findings highlight the necessity for educational initiatives aimed at increasing clinician awareness of the true associations between AN and MS.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

This is an unfunded project.

Declarations

Conflicts of Interest

In the interest of transparency, the authors declare the following: Dr Alex Ortega-Loayza is supported by National Institute of Arthritis and Musculoskeletal and Skin Disease (NIAMS) for NIAMS/National Institutes of Health (NIH) (R01AR083110-01). He is also a member of the editorial board for the American Journal of Clinical Dermatology. Alex Ortega-Loayza was not involved in the selection of peer reviewers for the manuscript nor any of the subsequent editorial decisions. All other authors have no financial relationships, activities, or interests with any for-profit or not-for-profit third parties whose interests may be affected by the content of this manuscript. No funding was received for conducting this study.

Ethics Approval

No ethics or institutional review board (IRB) approval was needed for this study.

Author Contributions:

All authors contributed to the study conception and design. Material preparation and data collection were performed by Jordan Gillespie, Jaclyn Roland-McGowan, Kyra Diehl, Tayler Tobey, Elena Paz-Munoz, Kira Champelli, Molly Joyce, Katie Norris, Jake Nelson, and Somya Khare. The analysis was performed by Emile Latour. The first draft of the manuscript was written by Jordan Gillespie, Jaclyn Roland-McGowan, and Kyra Diehl. All authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data Availability

Data supporting the findings of this study are not openly available due to the nature of systematic reviews, but are available upon reasonable request to the corresponding author.

Code Availability

Code used to analyze the dataset and produce the findings are not openly available but are available upon reasonable request to the corresponding author.

Consent to participate

N/A.

Consent for publication

N/A.