The association of ultra-processed food intake on age-related muscle conditions: a systematic review and dose–response meta-analysis with meta-regression

Abstract

Objectives

Chronic excessive intake of ultra-processed foods (UPFs) has been linked to various metabolic conditions; however, its impact on skeletal muscle mass and function in older adults remains unclear. Therefore, we conducted this study to examine the association between UPF intake and age-related muscle outcomes, including frailty, sarcopenia, low muscle mass (LMM), and/or low muscle strength (LMS).

Methods

A systematic search was conducted in ISI Web of Science, LILACS, PubMed/MEDLINE, and Scopus without restrictions up to November 1, 2024. Relative risks (RRs) and 95% confidence intervals (CIs) were pooled using a random-effects model. Study quality and the presence of publication bias were assessed using the Newcastle–Ottawa Scale, Egger’s regression asymmetry test, and Begg’s rank correlation test.

Results

Data from 29 studies were included. Cohort studies showed that higher UPF intake was significantly associated with an increased risk of frailty (RR = 1.40; 95% CI 1.25–1.58; I² = 83.0%; p < 0.001; n = 11), but not with LMS. In contrast, cross-sectional studies indicated that higher UPF intake was significantly associated with an increased risk of LMS (RR = 1.13; 95% CI 1.06–1.20; I² = 0.0%; p < 0.001; n = 5), but not with frailty, sarcopenia, or LMM. Furthermore, a 100 g increase in UPF intake was associated with a 3% higher risk of frailty (RR = 1.03; 95% CI 1.01–1.06; I² = 85.1%; p = 0.016; n = 5). Non-linear dose–response analysis showed a positive linear association between UPF intake and frailty risk (P_non-linearity = 0.807; P_dose-response < 0.001; n = 5).

Conclusion

Higher UPF intake was associated with an increased risk of frailty in cohort studies and with low muscle strength in cross-sectional studies. These findings suggest that regular consumption of UPFs may negatively affect muscle health, potentially impairing quality of life and independence in older adults.

Supplementary Information

The online version contains supplementary material available at 10.1186/s41043-025-00986-0.

Introduction

The global increase in population aging presents both societal and individual challenges. Sarcopenia and its components, such as low muscle mass, reduced muscle strength, and impaired physical performance have emerged as major public health concerns due to their strong associations with adverse outcomes and premature death in older adults [1, 2]. Muscle mass and strength begin to decline as early as the fourth decade of life and become more evident around the age of 50. These declines occur at estimated annual rates of 0.8–1% for muscle mass and 2–3% for muscle strength [3, 4]. In geriatric medicine, frailty is defined as a progressive decline in physiological systems and functional capacity, which is associated with increased risks of disability, falls, hospitalizations, long-term care, depression, cardiovascular disease, and mortality [5]. Frailty, often resulting from the combined effects of multiple health conditions, can lead to functional limitations, malnutrition, and unintentional weight loss [6].

Nutritional imbalances are increasingly recognized as key contributors to the decline in age-related muscle mass [7, 8]. Impaired nutrient absorption and utilization, especially of protein along with reduced appetite and diminished hunger and thirst cues, can lead to decreased food intake, unintentional weight loss, and muscle wasting in older adults [9, 10]. Recent studies have implicated pro-inflammatory foods, including processed meats, organ meats, sugar-sweetened beverages (e.g., soda), and refined grains (e.g., white bread) in disrupting protein balance, leading to reduced muscle protein synthesis and, ultimately, lower muscle mass and strength [11, 12]. Processed and ultra-processed foods (UPFs) may contribute to inflammation by altering the gut microbiome [13]. UPFs are typically ready-to-eat or ready-to-heat products made from industrially processed ingredients [14]. They are often energy-dense and characterized by an obesogenic nutrient profile high in saturated fats, added sugars, trans fats, and sodium [15]. Compared to minimally processed or whole foods, UPFs generally contain less protein, and their consumption is often associated with lower overall protein intake [16]. Consequently, a high UPF intake may displace nutrient-dense foods in the diet, thereby increasing the risk of various chronic conditions, including obesity, diabetes, mental health and neurodegenerative disorders, adverse pregnancy outcomes, inflammatory bowel disease, renal dysfunction, and mortality [17–24]. Growing evidence suggests that higher UPF intake is associated with an increased risk of muscle-related conditions in older adults, including sarcopenia [25], frailty [26], and reduced hand grip strength [27], although findings remain inconsistent [28–30]. Furthermore, Western dietary patterns often characterized by high UPF intake have been linked to elevated inflammation and oxidative stress, which may further impair muscle function and accelerate age-related muscle decline [31]. Several studies have demonstrated a significant detrimental association between Western dietary patterns, including high intake of sweets and sugar-sweetened beverages and age-related muscle conditions such as sarcopenia and frailty [26, 32]. However, not all studies have found consistent associations [30, 33].

Although a 2020 meta-analysis found that higher UPF intake was associated with an increased risk of frailty [34], no study to date has comprehensively evaluated the relationship between UPF intake and age-related declines in muscle mass and function. Therefore, we conducted a systematic review and dose–response meta-analysis with meta-regression to examine the association between UPF intake and age-related muscle outcomes, including sarcopenia, low muscle mass, low muscle strength, and frailty.

Methods

This systematic review and dose–response meta-analysis was conducted in accordance with the 2020 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [35]. The study protocol was registered with the international prospective register of systematic reviews database (PROSPERO) under the registration number CRD42023434881.

Literature search and selection

A comprehensive systematic literature search was conducted in ISI Web of Science, Scopus, PubMed/MEDLINE, and LILACS databases without any restrictions up to November 1, 2024. The scientific keywords and search strategy are presented in Supplementary Table 1. Data from grey literature sources, including reviews, conference abstracts, case reports, notes, short surveys, letters, and reports were retrieved through manual searches of reference lists from original research articles in the included databases.

Inclusion and exclusion criteria

Inclusion criteria were as follows: a) observational studies (cohort, cross-sectional, or case–control) in adult participants (≥ 18 years) that reported data on the association between UPF intake and the risk of age-related muscle conditions including frailty [36] (meeting three out of five phenotypic criteria indicating compromised energetics: low grip strength, low energy, slowed walking speed, low physical activity, and/or unintentional weight loss), sarcopenia[37, 38] (age-related loss of muscle mass, plus low muscle strength, and/or low physical performance), low muscle strength [37, 38] (handgrip strength for men: < 26 kg and women: < 18 kg), and/or low muscle mass [37, 38] (skeletal muscle index for men: ≤ 7.23 kg/m² and women: ≤ 5.67 kg/m²). The included studies reported effect estimates as odds ratios (ORs), relative risks (RRs), or hazard ratios (HRs), with corresponding 95% confidence intervals (CIs). Exclusion criteria included: a) studies undertaken in children and adolescents (< 18 years), b) studies without sufficient data, and c) research with no relevant exposure. Article titles and abstracts and, subsequently, full-text reviews obtained from database searches meeting the inclusion criteria were evaluated by two investigators (SM and F-HA). Any disagreements of opinion regarding study inclusion/exclusion criteria were decided by consensus following discussion. The PICOS tool for individual studies was displayed in Supplementary Table 2.

Data extraction

Two investigators (SM and ST) independently extracted the following data from articles meeting the inclusion criteria: a) first author’s name, year of publication, and country of origin; b) study characteristics (design, follow-up period, and source of health data); c) participant characteristics (number of participants/cases, age, and sex); d) UPF assessment method; e) muscle health-reported outcomes; f) main study results (outcomes); and g) co-variates utilized for adjustments in multivariate analyses. Any disagreements regarding data extraction were resolved through discussion and consensus.

Quality assessment

Two researchers (SM and M-GH) assessed the quality of each included study using the Newcastle–Ottawa Scale (NOS) [39]. The NOS was developed to evaluate the risk of bias in non-randomized studies included in systematic reviews and meta-analyses, and it allocates a maximum of 9 points across three domains: study group selection (four points), study group comparability (two points), and exposure and outcomes for case–control or cohort studies (three points). Studies scoring 7–9 are deemed high quality/low risk of bias, whereas a score of 0–3 indicates a high risk of bias. The final consensus on NOS quality assessments for the selected studies is reported in Table 1.

Table 1.

Characteristics of included studies

Author (year; location)	Study design / Follow up (years) / Source of data/ Health status	Population/ Age/(women/men)	Muscle health outcomes Critria (muscle mass and muscle streng)	Ultra-processed foods assessment method	Main results	Adjusted variables		Quality score
Kuczmarski et al. (2015, Spain)	Cross-sectional study /-/ The HANDLS study/adults	N = 2176/ Age = 69.2 ±0.3years/ (1231/ 945)	Sarcopenia ( Laurentani et al. 2003)/ (DXA/hand dynamometer)	24-h dietary recall /Desert pizza, sandwich and sweet drink	The higher desert pizza, sandwich, and sweet drink intake was not associated with the risk of sarcopenia	Age, sex, race, and socioeconomic status		+7/9
León-Muñoz et al. (2015, Spain)	Cohort study /12 years/ The Bordeaux sample of the Three-City Study. / older adults	N = 1872/ Age = 69.2 ±0.3years/ (165/ 135)	Frailty (Fried et al. 2001)/ (-/hand dynamometer)	Diet history /Western dietary pattern	Adherence to Western dietary patterns was associated with a risk of frailty	Number of frailty components at baseline, sex, age, educational level, occupation, smoking, body mass index, energy intake, cardiovascular disease, diabetes mellitus, cancer, asthma or chronic bronchitis, osteomuscular disease, depression requiring treatment, number of drug treatments, and score on the Mini-Mental State Examination.		+8/9
Hashemi et al. (2015, Iran)	Cross-sectional study /-/ -/ Menopausal women 53 years old or older.	N = 300/ Age = 66.6±7.8 rs years/ (165/ 135)	Sarcopenia (EWGSOP)/(DXA/hand dynamometer)	FFQ /Western dietary pattern	Adherence to Western dietary patterns was associated with a risk of sarcopenia	Age, sex, energy intake, physical activity, smoking, alcohol consumption, drug consumption, and positive history of disease		+7/9
Pilleron et al. (2016, France)	Cohort study /12 years/ The Bordeaux sample of the Three-City Study. / older adults	N = 972 / Age = 73.2 ±4.6 years/ (636/ 336)	Frailty (Fried et al. 2001)/ (-/hand dynamometer)	24-h dietary recall/ Pizza, snacks, sandwiches, and biscuits	The higher pizza, snack, sandwich, and biscuits intake was associated with a risk of frailty	Marital status, education level, and income. adjusted for multimorbidity, BMI, and depressive symptomatology, and using the Mini-Mental State Examination (MMSE)		+8/9
Barrea et al. (2018, Italy)	Cross-sectional study /-/The PREDIMED project (PREvencion con DIeta MEDiterr anea) / elderly women aged 60-85 years.	N = 84/ Age = 71.7 ± 5.5 years/ (84/ 0)	Low muscle strength ( Roberts et al. 2011)/(-/hand dynamometer)	24-h dietary recall/ Soda drink, butter, cream, and margarine intake	The higher soda drink, butter, cream, and margarine intake was associated with a risk of low muscle strength	BMI		+5/9
Mohseni et al. (2018, Iran)	Cross-sectional study /-/ -/ Menopausal women 45 years old or older.	N = 250/ Age = 57.6 ±6.2 years/ (250/ 0)	Sarcopenia, low muscle strength, low muscle mass (EWGSOP)/( Bioelectrical impedance /hand dynamometer)	FFQ /Western dietary pattern	The adherence to the Western dietary pattern was not associated with a risk of sarcopenia, low muscle strength, or low muscle mass	Age, physical activity, BMI, menopause duration, hypothyroidism, hormone replacement therapy, ACEi use, statin use, vitamin D		+7/9
Laclaustra et al. (2020, Spain)	Cohort study /3 years/ the Seniors-ENRICA cohort/individuals aged 60 years and over	N =1973/ Age =68.5 ± 6.3 years/ ( 1007 / 996)	Frailty (Fried et al. 2001)/(-/hand dynamometer)	Diet history /added sugars	The consumption of added sugars in the diet of older people was associated with frailty, mainly when present in processed foods.	Age, sex, education, smoking status, BMI, energy intake, comorbidities, MEDAS (excluding sweetened drinks and pastries), time spent watching TV, and leisure-time physical activity		+9/9
Marcos-Pardo et al. (2020, Europ)	Cross-sectional study /-/ The LifeAge study / middle-aged and older adults	N = 629/ Age = 65.41 ± 8.54 years/ (395/ 234)	Sarcopenia, low muscle strength, low muscle mass (EWGSOP)/(DXA/hand dynamometer)	Diet history /Hamburgers, sausages, or cold cuts	Higher hamburgers, sausages, or cold cuts intake was inversely associated with a risk of low muscle mass	Age and sex		+6/9
Sandoval-Insausti et al. (2020, Spain)	Cohort study /3.5 years/ EPIC (European Prospective Investigation into Cancer and Nutrition) / individuals aged 60 years and over	N =18222 / Age =68.6 ± 6.6 years/ ( 935 / 887)	Frailty (Fried et al. 2001)/(-/hand dynamometer)	FFQ /NOVA classification	Consumption of ultra-processed foods is strongly associated with frailty risk in older adults.	Age, sex, level of education, marital status, tobacco consumption, former-drinker status, chronic respiratory disease, coronary disease, stroke, osteoarthritis/arthritis, cancer, depression requiring treatment, and number of medications used		+9/9
Struijk et al. (2020, USA)	Cohort study /22 years/ Nurses’ Health Study (NHS) / women aged 60 years and over	N =71,935 / Age = 63 years/ ( 71,935 / 0)	Frailty (Fried et al. 2001)/(-/hand dynamometer)	FFQ / sugar-sweetened beverages (SSBs), artificially sweetened beverages (ASBs)	The higher SSBs and ASBs intake was associated with a risk of frailty	Age, calendar time, BMI, smoking status, alcohol intake, energy intake, physical activity, medication use, Alternate Healthy Eating Index, cancer, heart disease, and diabetes.		+9/9
Zhang et al. (2020, China)	Cross-sectional study /-/ Tianjin Chronic Low-grade Systemic Inflammation and Health (TCLSIH) / adults	N = 35,175 / Age = 40.1 (40.0, 40.2) years/ (16,138/ 19,037)	Low muscle strength (CAMRY-Digital Hand Dynamometer critria)	FFQ /Sweets intake	The sweets intake was associated with a risk of low muscle strength	Age, sex, height, smoking status, drinking status, education levels, employment status, household income, physical activity, family history of diseases, metabolic syndromes, and each other dietary pattern		+8/9
Lorraine O’Connell et al. (2021, Ireland)	Cross-sectional study /-/ community-dwelling volunteers irland/ adults aged ≥60 years	N = 142 / Age = 74.1 ± 6.80 years/ (81 / 61 )	Frailty (Fried et al. 2001)/(-/hand dynamometer)	FFQ / Sugars, preserves, and snacks intake	The sugar, preserves, and snacks intake was not associated with the risk of frailty	Age, sex, energy intake, body mass index, and use of dietary supplements		+5/9
Samadi et al. (2021, Iran)	Cross-sectional study /-/ Ravansar Non-Communicable Chronic Disease (RaNCD) cohort/adults	N = 2781 / Age = 68.85 ± 2.64 years/ (1785 / 996 )	Low muscle strength (Lauretani et al., 2003)/(-/hand dynamometer)	FFQ /Unhealthy dietary pattern	Adherence to unhealthy dietary patterns was associated with a risk of low muscle strength	Age, sex, education level, economic status, physical activity, and energy intake.		+8/9
Wang et al. (2022, China)	Cross-sectional study /-/ Shanghai Suburban Adult Cohort and Biobank (SSACB) survey / Community-dwelling older	N = 780 / Age = 68.85 ± 2.64 years/ (445 / 335 )	Frailty (Fried et al. 2001)/(-/hand dynamometer)	FFQ /Sugar, oil, and condiments	The higher sugar, oil, and condiments intake was not associated with a risk of frailty	Age, sex, BMI, energy intake, educational level, marital status, household income, smoking status, alcohol use, and doing housework		+7/9
Zhang et al. (2022, China)	Cohort study /5 years / The Tianjin Chronic Low-grade Systemic Inflammation and Health (TCLSIH) / adults aged 40 years and over	N = 5409 / Age = 48.3 (44.1, 54.5) years/ (2093 / 3316 )	Low muscle strength (CAMRY-Digital Hand Dynamometer critria)	FFQ /NOVA classification	The higher ultra-processed foods intake was associated with a risk of faster grip strength decline	Age, sex, height, smoking status, drinking status, education levels, employment status, household income, physical activity, family history of diseases, metabolic syndromes, and each other dietary pattern		+9/9
Hao et al. (2022, USA)	Cross-sectional study /-/ National Health and Nutrition Examination Survey (NHANES) / Community-dwelling older	N = 2329 / Age = 70.25 ± 0.28 years/ (1148/ 1181)	Frailty (Fried et al. 2001)/(-/hand dynamometer)	24-h dietary recall/ NOVA classification	The ultra-processed foods intake was positively associated with the frailty risk in underweight-normal weight and overweight people	Age, sex, marital status, BMI, race, education level, poverty-income ratio, waist, diabetes, hypertension, angina, arthritis, congestive heart disease, myocardial infarction, stroke, and total energy intake		+8/9
Vohra et al. (2022, USA)	Cross-sectional study /-/National Health and Nutrition Examination Survey (NHANES) Study/adults aged 40 years and over	N = 350 / Age = 69.6 ± 0.71 years/ (203/ 147)	Frailty (Fried et al. 2001)/(-/hand dynamometer)	24-h dietary recall/ Carbs and fats dietary pattern	The adherence to carbs and fats dietary pattern was associated with a risk of frailty	Age, sex, race, BMI, income-to-poverty ratio, energy intake, and smoking status		+7/9
Almeida et al. (2022, Brazil)	Cross-sectional study /-/- / Community-dwelling older adults aged ≥60 years	N = 118 / Age = 74.6 ± 9.3 years/ (53/ 65)	Sarcopenia (Cruz-Jentoft et al. 2019)	FFQ /Ultra-processed foods	The higher ultra-processed foods intake was associated with a risk of sarcopenia	Age, sex		+5/9
Maroto-Rodriguez et al. (2022, USA)	Cohort study /3.3 years / The Seniors-ENRICA-1 / Community-dwelling older adults aged ≥60 years	N = 1880 / Age = 68.65 ±6.38 years/(971/ 909)	Frailty (Fried et al. 2001)/(-/hand dynamometer)	Diet history/ Processed meat	Higher processed meat intake was associated with a risk of frailty	Age, sex, education, smoking status, BMI, energy intake, and prevalent conditions		+8/9
Chuy et al. (2022, France)	Cohort study /15 years / Three-City-Bordeaux cohort / Community-dwelling older adults	N = 1210 / Age = 75.6 ± 4.8 years /(752/ 458)	Frailty (Fried et al. 2001)/(-/hand dynamometer)	24-h dietary recall/ Simple carbs intake	The higher simple carbs intake was associated with a risk of frailty	Age, sex, protein intake, total energy intake, smoking status, alcohol consumption, depressive symptomatology, and global cognitive performances		+9/9
Zupo et al. (2023, Italy)	Cross-sectional study /-/Salus in Apulia Study / Community-dwelling older	N = 2185 / Age = 73.48±6.24 years/ (1211/ 974)	Frailty (Fried et al. 2001)/(-/hand dynamometer)	FFQ /NOVA classification	The higher ultra-processed food intake was associated with a risk of frailty	Age, gender, education, inflammatory cytokines, alcohol intake, protein/energy ratio, energy intake, and multimorbidity		+8/9
Topan et al. (2023, Romania)	Cross-sectional study / Patients with liver cirrhosis.	N = 206 / Age = 61.6 ± 9.4 years/ (74 / 127)	Sarcopenia (EWGSOP2)	24 h diet recall/Sweets intake	The sweets intake was not associated with the risk of sarcopenia	Age, sex, and weight		+5/9
Bai et al. (2023, China)	Cross-sectional study / the Chinese Longitudinal Healthy Longevity Survey / community-dwelling oldest old individuals	N = 7267/ Age = 97.5 ± 7.7 years/ ( 4278 / 2989 )	Sarcopenia (SARC-CalF)/(questions/ questions)	24 h diet recall/Sugar intake	The sugar intake was not associated with the risk of sarcopenia	Age, sex, residence, and educational back-ground, BMI, exercise status, current smoking status, and current alcohol drinking status and activities of daily living disability and number of comorbidities.		+8/9
Inoshita et al. (2023, China)	Cross-sectional study / Patients with chronic kidney diseases.	N = 441/ Age = 79.7 ± 5.9 years/ ( 242 / 198 )	Sarcopenia (AWGS)/(DXA/hand dynamometer)	Dietary variety score/ Sweets intake	The sweets intake was not associated with the risk of sarcopenia	Un-adjusted		+5/9
Wang et al. (2023, China)	Cross-sectional study / Community-dwelling older	N = 1051/ Age = 67.92 ± 7.62 years/ ( 678 / 244 )	Sarcopenia  (AWGS)/(DXA/hand dynamometer)	FFQ/ Sweets intake	The sweets intake was not associated with the risk of sarcopenia	Age and sex		+6/9
Shateri et al. (2024, Iran)	Cross-sectional study / Patients with chronic kidney diseases.	N = 110/ Age = 66.0 ± 21.0 years/ ( 50 / 60 )	Sarcopenia  (AWGS)/ (Bioelectrical impedance /hand dynamometer)	FFQ /NOVA classification	Higher intake of UPF is associated with a higher risk of sarcopenia	Age, sex, smoking, and energy intake		+5/9
Kong et al. (2024, USA)	Cross-sectional study / the National Health and Nutritional Examination Survey (NHANES)/ Community-dwelling adults	N = 10,255/ Age = 39.27 ± 0.26 years/ ( 5106/ 5149)	Low muscle mass ( Foundation for the National Institutes of Health definition)/(DXA/hand dynamometer)	FFQ /NOVA classification	Higher intake of UPF is associated with a higher risk of low muscle mass	Age, ethnicity, PIR, marital status, home status, education, physical activty, smoke status, drinks, eGFR, UACR, hypertension, DM, angina, coronary heart disease, congestive heart failure, heart attack, stroke, cancer, energy, and protein (g).		+8/9
Fung et al. (2024, USA)	Cohort study /≥26 years / the Nurses’ Health Study (cohort study) / Community-dwelling older adults	N = 63,743/ Age = 60 years and older/ ( NR/ NR)	Frailty (Fried et al. 2001)/(-/hand dynamometer)	FFQ /NOVA classification	Higher intake of UPF is associated with a higher risk of frailty in older females	Age, energy intake, alcohol intake, baseline BMI, physical activity, smoking, highest academic degree obtained, postmenopausal hormone use and drug use		+9/9
Clayton-Chubb et al. (2024, Australia)	Cross-sectional study /-/ the ASPirin in Reducing Events in the Elderly (ASPREE) randomised trial and the ASPREE Longitudinal Study of Older Persons (ALSOP) cohor/ Community-dwelling adults	N = 12,416/ Age = 77.0 (74.7–80.5) years/ ( 6751/ 5665)	Frailty (Fried e al. 2001)/(-/hand dynamometer)	FFQ /NOVA classification	The higher ultra-processed food intake was associated with a risk of frailty	Age, sex, BMI, educational attainment, living situation, and current smoking and alcohol drinking status,		+8/9

BMI, body mass index; FFQ, food-frequency questionnaire; DXA, Dual-energy x-ray absorptiometry; AWGS, Asian Working Group for Sarcopenia; EWGSOP, The European Working Group on Sarcopenia in Older People

Statistical analyses and data synthesis

Statistical analyses were undertaken using STATA version 14.0 (StataCorp, College Station, TX, USA) and SPSS version 25.0 (IBM, Armonk, NY, USA). In this meta-analysis, the overall effect sizes were established as relative risks (RRs) with corresponding 95% confidence intervals (CIs). Effect estimates from the included studies were pooled to provide a comprehensive summary of the association between UPF intake and muscle-related health outcomes (e.g., frailty, sarcopenia) [40]. The synthesized effect estimates for this research were reported as pooled relative risk (RR) with 95% CI. Due to anticipated heterogeneity between studies, effect estimates were calculated using the DerSimonian-Laird weighted random-effects model [41]. A pairwise meta-analysis was first conducted by comparing the highest and lowest categories of UPF intake to evaluate associations with muscle-related conditions. The primary outcomes were pooled using RR and 95% CI values derived from these comparisons. Heterogeneity was assessed using Cochran’s Q and I² statistics, where I² was calculated as [(Q–df)/Q × 100%], with Q representing the χ² value and df the degrees of freedom. Between-study heterogeneity was considered significant when the Cochran’s Q statistic yielded a p-value < 0.01 or I² exceeded 50%. Specifically, I² values < 25%, 25–50%, 50–75%, and > 75% were interpreted as indicating low, moderate, high, and extreme heterogeneity, respectively. Additionally, to examine potential sources of heterogeneity, subgroup analyses were conducted based on age (< 60 or ≥ 60 years), UPF classification method (NOVA classification, Western diet pattern, fast food, or sweets consumption), geographic region (USA, Europe, Asia), participant sex (male, female, both), number of participants (< 1000 or ≥ 1000), number of “cases” (participants diagnosed with a muscle-related condition, such as frailty or sarcopenia; categorized as < 500 or ≥ 500), dietary assessment method (food frequency questionnaire, 24-h recall, or brief diet history questionnaire), and adjustments for key covariates. Furthermore, meta-regression analyses were performed when 10 or more study arms were available [42] and aimed to examine whether sex, body mass index (BMI), smoking status, physical activity, alcohol intake, and energy intake modified the relationship between UPFs intake and age-related muscle outcomes.

Publication bias was assessed through visual inspection of funnel plots, Egger’s regression asymmetry test, and Begg’s rank correlation test [43, 44]. If publication bias was suspected, adjustment for funnel plot asymmetry was done by imputing missing study data using the Duval and Tweedle trim-and-fill method. A p-value < 0.05 was considered statistically significant. A dose–response meta-analysis was conducted to estimate the RRs per 100 g increment in UPFs intake based on the method introduced by Greenland and colleagues [45, 46]. Studies eligible for dose–response analysis included those reporting cases and non-cases (or person-years) and providing at least three distinct categories of UPF intake. We used a one-stage linear mixed-effects model to combine study-specific dose–response slopes into an overall average slope, thus estimating the linear relationship between UPF intake increments and muscle-related health outcomes.

Quality of evidence

The general certainty of evidence across studies was rated using the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) working group guidelines. The GRADE methodology aims to enhance transparency, consistency, and rigor in evaluating evidence and informing clinical decision-making. Based on the GRADE assessment criteria, the quality of evidence was categorized into four levels: high, moderate, low, and very low [47]. Factors influencing the certainty of evidence, such as study design, risk of bias, inconsistency, indirectness, imprecision, and publication bias were considered in determining the overall confidence in the findings.

Results

Study characteristics

A total of 7,383 records were identified through database searches and reference lists. After removing duplicates, 5,662 articles remained (Fig. 1). After screening titles and abstracts, 5,601 articles were excluded. The remaining 61 full-text articles were assessed, and 32 were subsequently excluded for the following reasons: ten articles were without sufficient data [48–56], 17 studies reported not relevant outcomes [57–73], three studies were conducted on non-relevant exposure [28, 74], and two studies were performed on other study designs (Supplemental Table 3) [75, 76]. Ultimately, 29 studies met our inclusion criteria and were included in the present work [25, 33, 77–102].

Fig. 1.

Flow chart of the process of the study selection

The general characteristics of the included studies are summarized in Table 1. Nine were cohort studies [78, 82, 83, 85, 87, 89, 90, 93, 97] and twenty were cross-sectional [25, 33, 77, 79–81, 84, 86, 88, 91–96, 98–102]. All of the included articles were published between 2015 and 2023 and were conducted in USA [81, 85, 90, 91, 97, 99], China [92, 93, 95, 98, 102], Spain [80, 82, 83, 89], Brazil [79], France [78, 87], Italy [77, 94], Ireland [33], Romania [101], Australia [96], and Iran [25, 86, 88, 100]. Fully adjusted relative risks (RRs) were reported for 246,261 participants across the included studies and were pooled for meta-analysis to assess the associations between UPF intake and the risk of frailty, sarcopenia, low muscle mass, or low muscle strength. Regarding study outcomes, frailty risk was reported in 13 articles [33, 78, 81–83, 85, 87, 89–92, 94, 96, 97]; 10 articles documented sarcopenia [25, 79, 80, 84, 86, 95, 98, 100–102]; five reported low muscle strength [77, 84, 86, 88, 93]; and three reported the risk of low muscle mass [84, 86, 99]. For body composition assessment, two studies used bioelectrical impedance [30, 100], six used DEXA[25, 29, 84, 92, 98, 99], and one used computed tomography images [101]. Muscle strength was most commonly assessed using a handgrip dynamometer in 32 studies [25–27, 29, 30, 33, 78, 81, 84, 85, 87–89, 91–94, 96–103]. The Newcastle–Ottawa grade was applied for the quality assessment of selected studies and indicated seventeen studies of high quality [25, 78, 80–83, 85–97, 99] and four of medium quality [33, 77, 79, 84, 98, 100–102], (Supplemental Table 4). Furthermore, the results indicated a high level of inter-rater agreement for data extraction and quality assessment (Kappa = 0.866).

Ultra-processed food intake and adult muscle health

Table 2 displays the association between dietary ultra-processed food intake and muscle health disorders. Results from cohort studies revealed that higher UPF intake was significantly associated with a greater risk of frailty (RR = 1.40; 95% CI 1.25–1.58; I² = 83.0%; p < 0.001; n = 11, Fig. 2), but not with low muscle strength (RR = 0.81; 95% CI 0.30–2.21; I² = 98.7%; p = 0.628; n = 2, Fig. 3). However, cross-sectional studies suggested that higher UPF intake was significantly associated with a greater risk of low muscle strength (RR = 1.13; 95% CI 1.06–1.20; I² = 0.0%; p < 0.001; n = 5, Fig. 4). In addition, cross-sectional studies suggested that UPF intake was not associated with the risk of frailty (RR = 1.32; 95% CI 1.18–1.48; I² = 83.6%; p < 0.001; n = 16, Fig. 5), sarcopenia (RR = 1.17; 95% CI 0.85–1.18; I² = 67.5%; p = 0.337; n = 13, Fig. 6), or low muscle mass (RR = 0.82; 95% CI 0.35–1.94; I² = 89.7%; p = 0.657; n = 3, Fig. 7; Table 2). A high degree of heterogeneity was observed among the included studies. Subgroup analyses were conducted to identify potential sources of heterogeneity. Heterogeneity in frailty outcomes was reduced in subgroup analyses based on region, disease ascertainment, dietary assessment method, and adjustment for energy intake (Table 3). Heterogeneity observed for sarcopenia was not reduced in any subgroup analysis (Table 4).

Table 2.

Dietary ultra-processed food and the risk of muscle health disorders

	Highest vs. lowest category meta-analysis					Dose–response meta-analysis
	Studies, n	RR (95% CI)	P value	I2, %	P heterogeneity	Dose, unit	Studies, n	RR (95% CI)	P value		I2, %	P heterogeneity		Quality of evidence
Cohort studies
Frailty	11	1.40 (1.25, 1.58)	< 0.001	83.0	< 0.001	100 g	5	1.03 (1.01, 1.06)	0.016	85.1		< 0.001	⊕ ⊕ ◯◯ Low
Low muscle strength	2	0.81 (0.30, 2.21)	0.682	98.7	< 0.001	–	–	–	–	–		–	⊕ ⊕ ◯◯ Low
Cross-Sectinal studies
Frailty	5	1.20 (0.86, 1.67)	0.280	81.3	< 0.001	–	–	–	–	–		–	⊕ ⊕ ◯◯ Low
Sarcopenia	13	1.17 (0.85, 1.60)	0.337	67.5	< 0.001	–	–	–	–	–		–	⊕ ⊕ ◯◯ Low
Low muscle strength	5	1.13 (1.06, 1.20)	< 0.001	0.0	0.530	–	–	–	–	–		–	⊕ ⊕ ◯◯ Low
Low muscle mass	3	0.82 (0.35, 1.94)	0.657	89.7	< 0.001	–	–	–	–	–		–	⊕ ⊕ ◯◯ Low

RR; Relative Risk, CI; Confidence Interval

Fig. 2.

Forest plots demonstrating RR and 95% CI of pooled results from the random-effects models to evaluate the relationship between ultra-processed foods consumption and risk of frailty among cohort studies

Fig. 3.

Forest plots demonstrating RR and 95% CI of pooled results from the random-effects models to evaluate the relationship between ultra-processed foods consumption and risk of low muscle strength in cohort studies

Fig. 4.

Forest plots demonstrating RR and 95% CI of pooled results from the random-effects models to evaluate the relationship between ultra-processed foods consumption and risk of low muscle strength among cross-sectional studies

Fig. 5.

Forest plots demonstrating RR and 95% CI of pooled results from the random-effects models to evaluate the relationship between ultra-processed foods consumption and risk of frailty among cross-sectional studies

Fig. 6.

Forest plots demonstrating RR and 95% CI of pooled results from the random-effects models to evaluate the relationship between ultra-processed foods consumption and risk of sarcopenia in cross-sectional

Fig. 7.

Forest plots demonstrating RR and 95% CI of pooled results from the random-effects models to evaluate the relationship between ultra-processed foods consumption and risk of low muscle mass in cross-sectional

Table 3.

Subgroup analyses of ultra-processed food intake and the risk of frailty (Highest vs. lowest category meta-analysis) in cohort studies

Sub-groups	Number of effect sizes	Relative Ratio (95%CI), Pvalue	I2 (%), Pheterogeneity	P between
Overall	11	1.40 (1.25, 1.58), < 0.001	83.0, < 0.001
Ultra-processed food assessment method				0.168
NOVA food classification	2	1.23 (1.00, 1.51), 0.048	83.0, < 0.001
Western diet pattern	1	1.65 (0.87, 1.12), 0.123	0.0, < 0.001
Fast-food	2	1.70 (1.39, 2.08), < 0.001	0.0, 0.617
Sweets consumption	6	1.42 (1.23, 1.64), < 0.001	36.6, 0.163
Region				< 0.001
USA	4	1.37 (1.26, 1.50), < 0.001	57.7, 0.069
Europe	6	1.78 (1.41, 2.25), < 0.001	0.0, 0.907
Australia	1	1.11 (1.07, 1.64), < 0.001	0.0, < 0.001
Sex				0.602
Male	1	1.35 (0.43, 4.24), 0.607	0.0, < 0.001
Female	5	1.34 (1.28, 1.41), < 0.001	0.0, 0.484
Both	5	1.59 (1.15, 2.20), 0.005	85.5, < 0.001
Number of participants				0.340
< 1000	3	1.66 (1.17, 2.34), 0.004	0.0, 0.804
> 1000	8	1.38 (1.22, 1.57), < 0.001	83.7, < 0.001
Number of cases				0.210
< 500	8	1.59 (1.22, 2.07), 0.001	75.8, < 0.001
> 500	3	1.34 (1.27, 1.40), < 0.001	0.0, 0.450
Dietary assessment method				0.012
FFQ	7	1.29 (1.15, 1.46), < 0.001	84.6, < 0.001
24 h Recall	1	1.74 (1.05, 2.89), < 0.001	0.0, < 0.001
Diet history	3	1.79 (1.48, 2.15), < 0.001	0.0, 0.602
Disease ascertainment				0.001
Fried Frailty Scale	8	1.75 (1.50, 2.05), < 0.001	0.0, 0.953
FRAIL scale	3	1.34 (1.27, 1.40), < 0.001	0.0, 0.450
Adjustments
Body mass index				0.748
Yes	5	1.43 (1.24, 1.65), 0.001	51.4, 0.084
No	6	1.38 (1.15, 1.66), < 0.001	87.9, < 0.001
Smoking status				0.312
Yes	6	1.31 (1.21, 1.41), < 0.001	92.2, < 0.001
No	5	1.45 (1.20, 1.75), < 0.001	0.0, 0.704
Physical activity				0.312
Yes	6	1.31 (1.21, 1.41), < 0.001	92.2, < 0.001
No	5	1.45 (1.20, 1.75), < 0.001	0.0, 0.704
Alcohol intake				0.229
Yes	8	1.49 (1.29, 1.71), < 0.001	45.7, 0.075
No	3	1.28 (1.05, 1.56), < 0.001	94.0, < 0.001
Energy intake				0.049
Yes	7	1.62 (1.31, 2.00), < 0.001	64.9, 0.009
No	4	1.27 (1.12, 1.43), < 0.001	64.5, 0.038

A p-value less than 0.05 is generally considered statistically significant

¹Calculated by Random-effects model

Table 4.

Subgroup analyses of ultra-processed food intake and the risk of sarcopenia (Highest vs. lowest category meta-analysis) in cross-sectinal studies

Sub-groups	Number of effect sizes	Relative Risk (95%CI), Pvalue	I2 (%), Pheterogeneity	P between
Overall	13	1.17 (0.85, 1.60), 0.334	67.5, < 0.001
Age				0.813
< 60 years	5	1.19 (0.85, 1.68), 0.315	0.0, 0.734
≥ 60 years	8	1.24 (0.77, 2.00), 0.368	79.3, < 0.001
Ultra-processed food assessment method				0.113
NOVA food classification	2	9.20 (1.38, 61.22), 0.022	75.3, 0.044
Western diet pattern	2	0.75 (0.37, 1.54), 0.437	29.8, 0.233
Fast-food	3	0.88 (0.62, 1.25), 0.485	0.0, 0.391
Sweets consumption	6	1.04 (0.77, 1.39), 0.814	39.4, 0.143
Region				0.531
US	5	1.85 (0.81, 4.25), 0.146	78.2, 0.001
Europe	2	0.67 (0.47, 0.95), 0.024	0.0, 0.508
Asia	6	1.12 (0.79, 1.60), 0.337	39.5, 0.142
Sex				0.862
Female	1	1.06 (0.47, 2.38), 0.888	-
Both	12	1.19 (0.84, 1.67), 0.326	70.2, < 0.001
Number of participants				0.800
< 1000	7	1.38 (0.69, 2.76), 0.361	81.8, < 0.001
> 1000	6	1.03 (0.88, 1.20), 0.701	0.0, 0.610
Dietary assessment method				0.550
FFQ	4	2.42 (0.56, 10.53), 0.239	87.3, < 0.001
24 h Recall	5	0.99 (0.65, 1.50), 0.956	39.2, 0.160
Diet history	1	0.73 (0.47, 1.14), 0.168	-
Other	3	1.03 (0.85, 1.25), 0.767	3.3, 0.355
Disease ascertainment
EWGSOP guideline	13	1.17 (0.85, 1.60), 0.334	67.5, < 0.001
Adjustments
Body mass index				0.711
Yes	2	0.98 (0.83, 1.17), 0.847	0.0, 0.853
No	11	1.28 (0.82, 2.00), 0.280	72.7, < 0.001
Smoking status				0.780
Yes	3	1.06 (0.50, 2.27), 0.878	67.6, 0.046
No	10	1.26 (0.82, 1.94), 0.297	70.4, < 0.001
Physical activity				0.360
Yes	3	0.95 (0.79, 1.16), 0.639	3.0, 0.357
No	10	1.40 (0.87, 2.24), 0.161	73.5, < 0.001
Alcohol intake				0.355
Yes	2	0.82 (0.47, 1.45), 0.499	50.0, 0.157
No	11	1.35 (0.88, 2.07), 0.169	70.5, < 0.001
Energy intake				0.946
Yes	2	1.28 (0.19, 8.66), 0.798	83.8, 0.013
No	11	1.17 (0.84, 1.62), 0.349	67.5, < 0.001
Sex				0.449
Yes	9	1.03 (0.80, 1.31), 0.823	32.2, 0.160
No	4	1.92 (0.58, 6.30), 0.283	87.9, < 0.001

A p-value less than 0.05 is generally considered statistically significant

Table 3 presents subgroup analyses of ultra-processed food intake and the risk of frailty among cohort studies (highest vs. lowest category meta-analysis). Subgroup analysis showed that higher UPF intake was significantly associated with an increased risk of frailty across all subgroups, except among males and those following a Western dietary pattern (Table 3). Table 4 presents subgroup analyses of ultra-processed food intake and the risk of sarcopenia (highest vs. lowest category meta-analysis). Subgroup analyses revealed no significant associations between UPF intake and sarcopenia risk (Table 4). Because the number of studies available for the other outcomes was fewer than or equal to five (e.g., three studies for muscle mass), it was not possible to perform subgroup analyses for the associations between UPF intake and the risk of low muscle mass or low muscle strength.

Meta-regression analysis

Supplementary Fig. 2 displayed the results of meta-regression analyses. No significant impact on frailty risk was observed when sex (p = 0.612), BMI (p = 0.983), smoking status (p = 0.612), physical activity (p = 0.407), alcohol intake (p = 0.296), and energy intake (p = 0.106) were studied. Similarly, sex (p = 0.622), BMI (p = 0.794), smoking status (p = 0.771), physical activity (p = 0.473), alcohol intake (p = 0.495), and energy intake (p = 0.970) did not modify the association between UPFs intake and sarcopenia risk.

Linear dose–response analysis

The outcomes of the linear dose–response analysis are presented in Table 2 and Fig. 8. We observed that a 100 g increase in UPF intake was associated with a 3% higher risk of frailty (RR = 1.03; 95% CI 1.01–1.06; I² = 85.1%; p = 0.016; n = 5). Because of the limited number of studies, linear dose–response analysis was not performed for other outcomes.

Fig. 8.

Forest plots showing the linear dose–response meta-analysis of mortality risk for 100 g change in ultra-processed food consumption in daily intake and risk of frailty

Non-linear dose–response analysis

Non-linear dose–response associations indicated a positive linear relationship between UPF intake and the risk of frailty (P_nonlinearity = 0.807, P_{dose-response} < 0.001, Fig. 9). Due to the small number of studies, non-linear dose–response analysis was not performed for other outcomes.

Fig. 9.

Non-linear dose–response indicated associations between UPF intake and the risk of frailty

Sensitivity analyses

The sensitivity analysis for the highest versus lowest category meta-analyses on muscle health outcomes, including frailty, sarcopenia, low muscle strength, and low muscle mass indicated that effect sizes were not driven by any single study (Supplemental Fig. 1).

Publication bias

No evidence of publication bias was found in studies assessing the association with sarcopenia risk (p = 0.180, Egger’s test; p = 0.230, Begg’s test). However, evidence of publication bias was observed in studies assessing frailty risk (p = 0.037, Egger’s test), although Begg’s test did not indicate publication bias (p = 0.312). Additionally, the funnel plot appeared asymmetrical for the association between UPF intake and frailty risk (Fig. 10). To assess the presence of publication bias, the trim-and-fill method was applied. The funnel plot (Fig. 11, A) showed asymmetry, indicating potential publication bias for the association between UPF intake and frailty risk but not sarcopenia. The analysis suggested that five studies may be missing, likely due to selective reporting or small-study effects. These studies were imputed on the left side of the plot to restore symmetry. After adjustment, the overall effect size was slightly attenuated but remained statistically significant, suggesting that the main findings are generally robust despite potential bias.

Fig. 10.

Funnel plot for evaluation publication bias, A: Frailty, B: Sarcopenia

Fig. 11.

Funnel plot for trim-and-fill analysis to estimate the effect of potential missing studies, A: Frailty, B: Sarcopenia

Quality of evidence

The GRADE approach was used to assess the quality of evidence across studies for frailty, sarcopenia, low muscle strength, and low muscle mass outcomes associated with UPF intake and the relative risk of muscle health disorders. The quality of evidence for the association between UPF intake and frailty risk in cohort studies was downgraded to low due to inconsistency and publication bias (Table 2). In addition, the evidence for associations between UPF intake and the risk of sarcopenia, frailty, and low muscle strength was downgraded to low due to inconsistency and imprecision. Finally, the quality of evidence regarding low muscle mass in both cohort and cross-sectional studies was also categorized as low, owing to inconsistency and imprecision (Table 2).

Discussion

With the rapid and projected growth of the older population, multidisciplinary approaches involving physical activity and high-quality nutrition are essential for preventing frailty and promoting skeletal muscle health. The persistent lack of adequate, healthy, and nutrient-rich food intake is a growing concern that may accelerate frailty with advancing age. Among the main findings of this meta-analysis, higher UPF intake was significantly associated with an increased risk of frailty in cohort studies. The association with low muscle strength was not statistically significant. However, due to the small number of studies and high heterogeneity, these results should be interpreted with caution. Moreover, cross-sectional studies showed that higher UPF intake was associated with an increased risk of low muscle strength, while no significant associations were found between UPF intake and frailty, sarcopenia, or low muscle mass. Additionally, the dose–response analysis demonstrated that a 100 g increase in UPF intake was associated with a 3% higher risk of frailty.

The results of the current study should be interpreted with caution. Due to the small number of studies assessing certain outcomes such as low muscle strength in cohort studies or low muscle mass in cross-sectional studies, the statistical power to detect associations was likely low. These non-significant findings should not be interpreted as evidence of an absence of association, and further studies are needed to confirm the results. In addition, while a 3% increased risk of frailty per 100 g of UPF intake may appear modest at the individual level, its impact on public health could be substantial. Given the rising global intake of UPFs, their cumulative effects may contribute to a clinically meaningful increase in frailty risk at the population level. Moreover, even modest reductions in frailty are meaningful, given its well-established associations with falls, hospitalizations, disability, and mortality [104, 105].

Both quantitative factors (e.g., energy intake) and qualitative factors (e.g., nutrient quality) may contribute to the prevalence of frailty syndrome [106, 107]. Observational studies have examined the association between UPF intake and aging biomarkers, including muscle health status [25–27, 29, 87]; however, their findings have been inconsistent, possibly due to variations in follow-up duration, methods used to assess UPF intake, and the types and quantities of UPFs consumed across study populations. In addition, inconsistent findings may also stem from differences in study design, such as cohort versus cross-sectional approaches. By stratifying the results by study design, we aimed to clarify how methodological differences may influence the observed associations between UPF intake and frailty, enabling more cautious interpretation of the findings. Notably, a positive association between UPF intake and frailty risk was observed in cohort studies but not in cross-sectional studies. Because cross-sectional designs do not permit causal inference, it is possible that individuals with frailty may alter their diets, which may result in reverse causality. The observed association between UPF intake and the risk of frailty and low muscle strength aligns with a previous meta-analysis, which showed that higher dietary inflammatory index scores reflecting greater intake of fried foods, processed meats, sweets, and refined grains were positively associated with increased risks of frailty and low muscle strength [108].

The association between higher dietary inflammatory potential and an increased risk of low muscle strength was also observed in another study [109]. In contrast, adherence to a Mediterranean-style diet, characterized by a high amount of olive oil, whole grains, fruits, vegetables, and other minimally processed or unprocessed foods, was negatively associated with the risk of frailty and pre-frailty among elderly individuals, according to the meta-analysis result [110]. Moreover, the findings from dietary pattern analysis demonstrated that unhealthy dietary patterns, including high intakes of refined carbohydrates and processed products such as smoked sausages and hot dogs, were associated with an increased risk of frailty syndrome among older adults, while dietary patterns characterized by frequent fruit and water intake reduced the risk of frailty [111]. In addition, a recent systematic review of individuals over 60 years revealed that the highest UPF intake was associated with frailty incidence [112].

In the current study, the results revealed heterogeneity in the relationship between UPF intake and age-related muscle conditions. Hence, we conducted a subgroup analysis based on age, UPF classification method, geographic region, participant sex, number of participants, number of “cases,” dietary assessment method, and adjustments for key covariates to assess the association between UPF intake and the risk of frailty and sarcopenia in cohort and cross-sectional studies, respectively. Since the number of studies included for the other endpoints is less than or equal to five, it was not possible to conduct a subgroup analysis for the association between UPF intake and the other age-related muscle conditions. The selected subgroup variables were informed by their potential influence on the outcomes, as well as by prior research findings [22, 113, 114].

In the current work, we observed a positive association between UPF intake and the risk of frailty in studies that assessed the intake of fast foods or sugary sweets, or used the NOVA classification versus a Western dietary pattern in cohort studies. A positive association was also observed between UPF intake, assessed by the NOVA food classification, and sarcopenia risk in cross-sectional studies. It is important to note that the NOVA food classification system categorizes foods based on their degree of processing, encompassing various UPFs, such as fast foods, cookies, sugary drinks, and others. In addition, it has been reported that frailty is more frequent in women than in men [115], which helps explain the positive association between UPF intake and frailty in subgroups of women in cohort studies. Additionally, the result of the subgroup analysis of cross-sectional studies revealed a negative association between UPF intake and the risk of sarcopenia in European regions versus the US or Asia regions. This finding may reflect regional differences in lifestyle factors and types of UPFs consumed, as well as the use of different assessment tools for defining sarcopenia among the included studies [116].

Additionally, approximately 58% of staple foods found in large stores across the United States are ultra-processed, which is 41% higher than the rate found in European stores. Furthermore, ultra-processed foods in the United States contain 41% more ultra-processing markers than those sold in the European Union [117]. Additionally, frailty and sarcopenia tend to increase with aging [5, 118]. Therefore, differences in participant age across studies may have contributed to the heterogeneity observed in the pooled results. However, the subgroup analysis did not find any significant association between UPF intake and sarcopenia risk in studies that included participants over 60 years old or under 60 years old. Despite conducting the subgroup analysis, there are several reasons for between-study heterogeneity in the relationship between UPF intake and the risk of age-related muscle conditions. Differences in regional dietary patterns are one of the important factors that should be considered. Traditional diets vary greatly across regions in terms of food preparation methods, food composition, and diet quality. Moreover, differences in comorbidity status among the included studies should be noted. The methodological differences in measuring and adjusting comorbidities between included studies may lead to heterogeneity when combining the results of the studies. Additionally, there was inconsistency in adjusting for potential confounding variables between the included studies, which can explain the heterogeneity when we pooled the study results.

The observed positive association between UPF intake and the risk of frailty and low muscle strength in cohort and cross-sectional studies can be explained by several potential mechanisms. UPF intake may increase the risk of overweight and obesity, which in turn is associated with a greater risk of frailty, according to a recent meta-analysis study [119]. A second reason could be the association between UPF intake and an increased risk of inflammatory conditions [120, 121], which may contribute to an increased risk of frailty and decreased muscle strength [122, 123]. Chronic inflammation can induce muscle loss, decrease muscle mass, and reduce muscle strength [124] by increasing the availability of myostatin, a negative regulator of muscle growth, and suppressing the insulin-like growth factor − 1 (IGF-1) axis, which plays a role in promoting protein synthesis [125, 126]. Moreover, inflammatory conditions induce oxidative stress, which can act as a trigger for muscle protein degradation through the activation of the Ubiquitin–Proteasome System and the Autophagy–Lysosomal Pathway, impairing protein synthesis [127]. Additionally, the consumption of UPFs increases the risk of frailty and low muscle strength through its influence on alterations in gut microbiota composition [128, 129]. Modification of gut microbiota composition by increasing pathological microbiota species and decreasing beneficial ones is associated with frailty in older adults [130]. Additionally, exposure to chemical components such as phthalates and bisphenols, which are used in the packaging of UPFs, has been shown to be inversely related to grip strength and positively associated with an increased frailty risk [131, 132].

Our study had several strengths, including the pooling of all data assessing the relationship between UPF intake and muscle health, as well as a dose–response analysis. However, the following limitations should be considered: Self-reported data raise the risk of several biases, including social desirability bias and recall bias, which can influence the reliability of study results [133]. For instance, assessing food intake using self-reported data, such as FFQs, may introduce recall bias, potentially leading to misclassification of UPF intake across studies. Additionally, measurement errors in self-reported assessments may arise from cognitive impairment due to age. Moreover, no validated, standardized tool currently exists for assessing UPF intake. Furthermore, using various methods to assess UPF intake (e.g., Western-type diet, fast food, sweets consumption) may lead to inconsistent classifications of UPFs across studies. The lack of a standardized method for assessing UPF intake across studies may affect the comparability of results. Moreover, we should consider that there was inconsistency in the clinical definitions of outcomes such as muscle mass and low muscle strength between included studies, which may influence the comparability and reliability of the pooled estimates.

For example, in diagnosing sarcopenia, both the EWGSOP2 and AWGS2 criteria incorporate physical performance as a diagnostic factor. The EWGSOP2 framework uses some specific tests like the Timed Up and Go (TUG) and the Short Physical Performance Battery (SPPB) and AWGS2 uses the 6-m walk test, SPPB, and the five-time chair stand test to assess the individuals performance [134]. While there is some likeness between these two framework, a difference is that EWGSOP2 requires all of them to be positive to diagnose severe sarcopenia, whereas AWGS2 requires only one. Therefore, when assessing the prevalence of severe sarcopenia, the AWGS2 definition tends to yield higher rates compared to the EWGSOP2 criteria. The inconsistency among diagnostic algorithms makes it difficult to compare prevalence of sarcopenia and severe sarcopenia across populations [134]. In addition, while we emphasized outcomes related to muscle health, including frailty, sarcopenia, low muscle strength, and low muscle mass, we acknowledge that adding ‘disability’ can introduce variability, as it may encompass heterogeneous functional limitations.

Although studies have adjusted for a wide range of possible covariates, it is not possible to completely rule out residual confounding factors, including genetic factors. Additionally, it should be considered that factors such as physical activity, socioeconomic status, and protein intake can influence the study’s results and should be considered potential confounding variables. The protective effect of physical activity on the risk of frailty, sarcopenia, and muscle mass [135–137] may influence the results of the study. Additionally, it has been observed that frailty occurs at an earlier stage and progresses more rapidly in individuals with greater socioeconomic disadvantage [138]. Socioeconomic status also influences sarcopenia [139]. Furthermore, it has been observed that increased dietary protein intake leads to greater handgrip strength, and higher protein intake is associated with higher muscle mass, which should be taken into account [140, 141].

In addition, some of the included studies had a cross-sectional design, so it is not possible to interpret the observed relationships as causal associations. The other limitation of this study is due to the dose–response analysis. Although the dose–response analysis showed that a 100 g increase in UPF intake was associated with a 3% higher risk of frailty, only five studies were included, so the findings should be interpreted with caution, and the further studies are needed to confirm the dose–response association. Additionally, the potential for selective reporting within the included studies and the exclusion of non-English and grey literature may introduce bias, which should be considered when interpreting the findings. Additionally, a key limitation of this meta-analysis is the potential for publication bias, as indicated by the funnel plot and confirmed by the trim-and-fill analysis for association between UPFs intake and risk of frailty. The imputation of five potentially missing studies implies that some negative or null results may not have been published. Although the corrected effect estimate remained significant, this adjustment highlights the need for cautious interpretation. Furthermore, the trim-and-fill method has its own limitations. It assumes that the cause of funnel plot asymmetry is solely due to publication bias, which may not always be the case. As such, results should be interpreted in light of these analytical constraints. In addition, in this study we observed a considerable heterogeneity in the pooled estimates, particularly for frailty and sarcopenia. therefore we conducted meta-regression analyses for some covariates that we believed could be possibly associated with heterogeneity to test sources of heterogeneity including sex, BMI, smoking status, physical activity, alcohol and energy intake. In spite of this, none of these variables modified the associations between UPF intake and risk of sarcopenia and frailty. These null findings suggest that the observed heterogeneity could not be explained by the examined lifestyle and demographic factors and it is possible that unmeasured or variably reported variables, such as specific types of UPFs, duration of exposure, or dietary habits of individuals, could be the cause of the heterogeneity observed. Given the observed heterogeneity, the findings should be interpreted cautiously, underscoring the need for future studies using standardized methodologies. Additionally, the positive association between UPFs intake and risk of low muscle strength was observed only among cross-sectional studies that it makes difficulties to infer causality. Therefore, conducting longitudinal studies are needed.

Future research could benefit from including a broader range of sources to minimize such limitations.

Conclusion

The meta-analysis of cohort studies revealed that higher UPF intake was associated with an increased risk of frailty. However, no statistically significant correlation was established between UPF intake and low muscle strength in these studies. In contrast, the pooled analysis of cross-sectional studies indicated that greater UPF intake was associated with an increased risk of low muscle strength. No significant associations were found between UPF intake and the risks of frailty, sarcopenia, or low muscle mass. Furthermore, the dose–response analysis indicated that a 100 g increase in UPF intake was associated with a 3% rise in the risk of frailty. Further prospective studies are needed to clarify the relationship between UPF intake and features of age-related muscle decline, including muscle strength, muscle mass, sarcopenia, and frailty.

Author contributions

SM designed this research. MA-HK, MZ, PA, and FP contributed to the conduct of the search. SM and ST performed the statistical analysis and interpreted the outcomes. SM, SM-GH, Sanaz Merabani, FH-A and MA-HK wrote the initial manuscript. DC and RB critically revised the manuscript and contributed to the subsequent drafts of the manuscript. All authors approved the final version of the manuscript.

Funding

None.

Availability of data and materials

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This article does not contain any studies with human participants or animals performed by any of the authors.

Competing interests

The author(s) declare no potential conflicts of interest with respect to the research, authorship, and/or publication.