Metabolic dysfunction-associated steatotic liver disease (MASLD), previously known as non-alcoholic fatty liver disease (NAFLD), represents the hepatic manifestation of metabolic syndrome and has reached pandemic proportions, with a global prevalence exceeding 30% (1). Recent studies indicate a significant increase in global MASLD prevalence: from 25.26% in 1990–2006 to 38.00% in 2016–2019. Geographically, prevalence peaks in Latin America (44.37%), followed by the Middle East and North Africa (36.53%), South Asia (33.83%), South-East Asia (33.07%), North America (31.20%), East Asia (29.71%), and Asia Pacific (28.02%), while Western Europe reports the lowest burden (25.10%) (2). The recent nomenclature shift to MASLD underscores its strong association with cardiometabolic dysregulation (3, 4), including obesity, insulin resistance, and dyslipidemia (4–7). The increased prevalence of MASLD is parallel to the increasing rates of obesity and type 2 diabetes (T2D) (8, 9). Approximately 10–30% of individuals with isolated steatosis progress to metabolic dysfunction-associated steatohepatitis (MASH) and advanced liver disease; however, concurrent T2D elevates this risk to 65% (10). Emerging data indicate that MASLD patients have an elevated risk of cardiovascular mortality, which now surpasses liver-related complications as the primary cause of death (11, 12). Additionally, MASLD elevates risks of extrahepatic conditions: chronic kidney disease (13), and several extrahepatic cancers (78% higher for uterine cancer, 38% higher for colorectal cancer, 2.5-fold for bladder cancer, and 2-fold for kidney cancer) (14). The economic burden of MASLD is substantial and growing. In 2016, direct annual healthcare costs in the USA reached $103 billion, exceeding the combined costs of Germany, France, and Italy (€27.7 billion) and the UK (£5.24 billion). Over the next decade, this burden is projected to rise to $1.005 trillion in the USA and €334 billion in Europe (15).

Despite its substantial clinical burden and the projected healthcare costs, therapeutic options are largely limited to lifestyle modifications, and the efficacy of pharmacological interventions remains unsatisfactory. Current pharmacological management of MASLD faces significant unmet needs. Among metabolic agents, glucagon-like peptide-1 receptor (GLP-1R) agonists (e.g., semaglutide) reduce hepatic fat through weight loss, but evidence for fibrosis improvement remains insufficient (16); peroxisome proliferator-activated receptor (PPAR) agonists (e.g., pioglitazone) improve histological features of steatohepatitis, yet adverse effects (weight gain, edema, potential heart failure) constrain clinical utility (17); sodium-dependent glucose transporters-2 (SGLT-2) inhibitors ameliorate glucolipid metabolism but lack controlled trials assessing histological endpoints. For liver-targeted therapies, resmetirom (a thyroid hormone receptor β agonist) is the only MASH-targeting drug with positive results from a registrational phase III clinical trial (18), though long-term safety (e.g., thyroid/gonadal effects) and hard endpoint benefits require further validation; farnesoid X receptor (FXR) agonists (e.g., obeticholic acid) failed approval due to hepatotoxicity and unfavorable risk–benefit profiles. In patients with cirrhosis or hepatocellular carcinoma, comprehensive strategies such as chemotherapy, resection, or transplantation are warranted. Collectively, limitations in histological response rates, safety profiles, and long-term outcomes underscore the imperative to explore adjunctive approaches for halting disease progression and mitigating cardiovascular risk. Nutritional interventions, such as low-carbohydrate diets (LCDs), intermittent energy restriction, and calorie-restricted diets, are considered beneficial for MASLD (19).

Among nutritional strategies, LCDs have garnered particular attention due to their dual potential to ameliorate both hepatic steatosis and associated metabolic derangements (20). Mechanistic studies in animal models demonstrate that LCDs reduce hepatic diacylglycerol (DAG) accumulation, shift ceramide synthesis toward beneficial very long-chain species, and upregulate mitochondrial fatty acid oxidation genes. These effects collectively suppress lipogenesis, exert anti-steatotic effects, and preserve insulin signaling (21). Clinical evidence further supports their efficacy: LCDs promote weight loss while simultaneously enhancing insulin sensitivity, preserving β-cell function, and optimizing glucometabolic parameters (22). Currently, a unified definition of LCDs is lacking. It is generally accepted that diets with carbohydrate intake <20 g/day (or <10% of total calories from carbohydrates) are considered ketogenic, <130 g/day (or <26% of total calories) are regarded as low carbohydrate, and diets with <45% of total calories from carbohydrates are classified as mild low carbohydrate diets (23, 24).

However, the paradigm of increased dietary fat consumption inherent to LCDs introduces critical clinical uncertainties. Controversies persist regarding the role of dietary fat in cardiometabolic health (25). This cardiovascular vulnerability raises concerns about LCD-induced alterations in lipid profiles, particularly elevated LDL-cholesterol levels observed in some trials (26, 27). Consequently, uncertainty regarding the cardiovascular impact of LCDs limits their use in MASLD patient populations. Given that international multidisciplinary expert consensus has endorsed LCDs as a key dietary strategy for MASLD management (19), investigating their impact on cardiovascular risk factors in MASLD patients holds clear theoretical and clinical significance.

To address this critical evidence gap, we conducted a systematic meta-analysis evaluating the cardiovascular safety profile of LCDs interventions in adults with MASLD. This study aims to support evidence-based, personalized dietary strategies that integrate hepatic recovery with cardiovascular protection in MASLD management.

This meta-analysis demonstrates that LCDs significantly improve key cardiometabolic parameters among patients with MASLD, including reductions in HbA1c, TG, body weight, and BMI. While the pooled analysis showed no statistically significant change in LDL-C (SMD, −0.12; 95% CI, −0.41 to 0.17, p = 0.41), the high heterogeneity (I² = 72%) indicates substantial inconsistency between studies. Similarly, the nonsignificant results for HDL-C (I² = 85%) and TC (I² = 51%) must be interpreted with caution. Although LCDs were defined broadly as <45% carbohydrate energy, the included trials encompassed three subtypes: ketogenic (<10%; n = 2), strict low-carbohydrate (<26%; n = 3), and mild low-carbohydrate (26–45%; n = 11) diets. Importantly, subgroup analyses using the three-tier classification were precluded for 9 of 13 outcomes (SBP, DBP, HOMA-IR, HbA1c, LDL-C, HDL-C, body weight, WC, WHR) due to insufficient studies (<2 trials per subgroup). This limitation necessitated our dichotomous approach (<26% vs. ≥26% carbohydrate energy) to ensure statistical robustness. Moreover, stricter carbohydrate restriction (<26% total energy) was associated with additional clinical benefits, including significant improvements in blood pressure (systolic/diastolic), insulin sensitivity (HOMA-IR), and reductions in waist circumference. However, the lack of WHR data in this subgroup represents a critical evidence gap. Given that WHR more accurately reflects central adiposity distribution than waist circumference alone, future trials should prioritize standardized WHR measurement, especially when evaluating interventions targeting visceral fat reduction. Short-term LCDs interventions (<24 weeks) were particularly effective for glycemic control and lipid profile improvement. Our findings suggest that while low carbohydrate diets may be effective in reducing body weight and improving metabolic parameters in MASLD patients, their impact on cardiovascular risk factors is more complex and may vary depending on various factors, such as the duration and intensity of the diet.

Improving insulin resistance—a central pathophysiological link between MASLD and cardiovascular disease—is a cornerstone of LCDs efficacy (6, 44). Consistent with prior evidence (22), our findings demonstrate that LCDs significantly reduce insulin resistance. By reducing postprandial glycemic excursions and insulin secretion, carbohydrate restriction attenuates hepatic de novo lipogenesis (DNL) via suppression of carbohydrate-responsive element-binding protein (ChREBP) (45), while enhancing peripheral glucose uptake through AMP-activated protein kinase (AMPK)-mediated GLUT4 translocation (46). This dual mechanism not only reduces HOMA-IR but also disrupts the vicious cycle of ectopic lipid accumulation of MASLD. Furthermore, LCDs-induced reductions in dietary fructose intake mitigate hepatic lipid accumulation by downregulating lipogenic enzymes such as sterol regulatory element-binding protein 1c (SREBP-1c), thereby suppressing fatty acid synthesis and promoting mitochondrial β-oxidation (47).

The effects of LCDs on lipid profiles remain controversial (48). Studies in non-MASLD populations have reported elevated TC, LDL-C, and HDL-C levels with LCD interventions (49). However, our analysis of MASLD patients demonstrated TG was significantly reduced, while TC, LDL-C, and HDL-C remained unchanged. This divergence may reflect that MASLD patients characterized by insulin resistance, likely derive greater benefits from LCD-induced suppression of de novo lipogenesis and enhanced fatty acid oxidation.

Prolonged adherence may unmask the adverse effects of suboptimal fat composition. LCDs shift energy substrate utilization to fats, promoting adipocyte lipolysis (50), but excessive dietary fat intake is associated with an increased risk of CVD (51). Among dietary factors, saturated fatty acids exert the most significant influence on blood lipids (52). Replacing saturated fats with unsaturated fats not only attenuates LDL-C but also improves insulin sensitivity (53). Moreover, further research suggests that insulin resistance may confer a higher cardiovascular risk compared to LDL cholesterol (54). So to some extent, LCDs may provide significant benefits, including weight loss, increased insulin sensitivity, and a potentially greater reduction in cardiovascular risk (55). Given the critical roles of LDL-C and TC in cardiovascular risk (56, 57), LCDs that reduce carbohydrates and unsaturated fatty acids should be prioritized in MASLD management.

LCDs may lower blood pressure through multiple mechanisms, including reduced insulin levels, modulation of vascular and neuroendocrine systems, and weight loss. By limiting carbohydrate intake, LCDs decrease insulin-mediated sodium reabsorption in the kidneys, reducing sodium and water retention (58). Additionally, LCDs modulate vascular endothelial function, the hypothalamic-pituitary-adrenal (HPA) axis, the sympathetic nervous system (SNS), and the renin-angiotensin-aldosterone (RAA) system, all of which play critical roles in BP control (59–61). Weight loss, a common outcome of LCDs, further contributes to BP reduction through caloric restriction and improved body composition (62).

Consistent with previous studies, LCDs’ long-term efficacy appears less satisfactory. One potential explanation is that reduced glycogen stores associated with LCDs adoption may lead to decreased physical activity and increased fatigue, ultimately resulting in diminished energy expenditure (63). Sustained adherence to LCDs may pose significant challenges, and the therapeutic benefits of dietary interventions are largely constrained by participant compliance. To address these limitations, future long-term prospective studies should investigate the sustained effects of LCDs on cardiovascular outcomes.

Our findings may provide reassurance to healthcare providers and patients considering the use of LCDs for the management of MASLD, as not all high-fat diets necessarily result in a substantial increase in CVD risks. These findings advocate a phased clinical approach to LCDs implementation in MASLD. Short-term interventions with strict carbohydrate restriction may rapidly improve insulin sensitivity and dyslipidemia by leveraging ketosis-driven appetite suppression and increased energy expenditure (64, 65). Prolonged adherence should prioritize unsaturated fats and high-fiber foods to mitigate lipid risks while sustaining metabolic benefits.

Some limitations should be considered. First, participants were unlikely to be blinded due to the study design, and self-reported dietary data could be affected by recall inaccuracies. Second, evaluation of WHR was feasible in only one study within the strict LCD subgroup, preventing meaningful assessment of its clinical utility. Third, heterogeneity in dietary fat composition across trials (e.g., saturated vs. unsaturated fats) obscured whether observed effects stemmed from carbohydrate restriction or fat quality changes. Fourth, the high attrition rates (>20%) in 25% of included trials may introduce selection bias. Participants with suboptimal responses or poor dietary adherence are more likely to withdraw, potentially inflating efficacy estimates. Fifth, while a three-tier classification (ketogenic/strict/mild) is theoretically preferable, sparse data for ketogenic diets (<10%; only 2 trials) limited subgroup comparisons. Finally, short intervention durations precluded assessment of long-term cardiovascular outcomes.

Future research efforts would benefit from focusing on the following critical aspects: (1) Standardizing nutrient quantification in LCDs to include daily carbohydrate intake in absolute grams and percentage of total energy, alongside detailed characterization of fat subtypes—specifically ratios of saturated, monounsaturated, and polyunsaturated fatty acids with primary sources identified (e.g., coconut oil, olive oil). (2) Implementing real-time dietary monitoring via validated digital diaries to reduce recall bias, and introducing interventions (e.g., personalized counseling, peer support networks) to improve long-term compliance. (3) Conducting long-term trials with definitive cardiovascular endpoints (e.g., major adverse cardiovascular events, cardiovascular mortality, carotid plaque progression) to evaluate the cardiovascular efficacy and safety of LCDs.

In patients with MASLD, LCDs demonstrate clinically meaningful improvements in insulin sensitivity, adiposity, and dyslipidemia, particularly when carbohydrate intake is restricted below 26% of total energy. To reconcile hepatic and cardiovascular health, it is necessary to emphasize carbohydrate restriction. Future research should prioritize: standardized nutrient assessment, enhanced adherence strategies, and cardiovascular endpoint trials.