A diet high in ultra-processed foods may do more than affect weight. It could be quietly reshaping muscle health, with MRI evidence revealing increased fat infiltration in those at risk of knee osteoarthritis.
Study: Ultra-processed Foods and Muscle Fat Infiltration at Thigh MRI: Data from the Osteoarthritis Initiative. Image credit: Thenongphoto/Shutterstock.com
A recent Radiology study examined how ultra-processed food (UPF) intake relates to thigh muscle fat infiltration (MFI) among individuals at risk of knee osteoarthritis (KOA), using thigh MRI as the primary imaging measure.
UPFs and musculoskeletal health
UPFs are industrially engineered formulations distinguished by the presence of additives, including emulsifiers, preservatives, and synthetic sweeteners, that are not typically used in conventional home cooking. Nutritionally, they are energy-dense yet poor in micronutrients, delivering disproportionately high levels of refined fats, sodium, and free sugars.
A high UPF diet has been associated with a broad spectrum of metabolic and systemic harm, including abdominal obesity, metabolic syndrome, cardiometabolic disease, frailty, depression, cancer, and elevated all-cause mortality. The global rise in obesity is closely tied to the growing dominance of UPFs in modern diets.
Obesity is a well-established risk factor for KOA, the most frequently affected joint, and one of the most disabling and economically burdensome conditions globally. Thigh muscle function is integral to knee stability, and its decline operates bidirectionally, both driving and resulting from reduced mobility. Notably, increased intramuscular fat infiltration in the knee extensors may stem from disuse atrophy secondary to knee pain, or from broader muscle impairment and systemic inflammation in older, high-BMI individuals.
Higher UPF consumption has been linked to greater knee pain, thinner cartilage, slower gait speed, lower skeletal muscle mass, and higher body fat, independent of sociodemographic factors. Yet the specific relationship between UPF-heavy diets and skeletal muscle quality remains underexplored, particularly in populations where early intervention may be most impactful.
Assessing the link between UPF intake and MRI-quantified MFI in KOA-susceptible adults
This cross-sectional secondary analysis examined the relationship between UPF consumption and thigh MFI, quantified via MRI, in adults at risk of KOA. It used baseline data from the Osteoarthritis Initiative (OAI), a multicenter, prospective cohort study evaluating KOA biomarkers (2004–2015).
To minimize confounding from disuse muscle atrophy, participants without radiographic disease or pain in the knee or hip joints (defined as no established radiographic osteoarthritis, Kellgren–Lawrence grade ≤1, though some participants had doubtful early radiographic changes) and without systemic comorbidities at baseline were included. Any participant with radiographic KOA, definite hip osteoarthritis, inflammatory arthritis, systemic diseases, such as stroke, diabetes, or cancer, or missing thigh MRI or dietary data was excluded.
Participants completed the Block Brief 2000 food frequency questionnaire (FFQ), which provided information about dietary habits over the preceding 12 months. However, the FFQ was not specifically designed to capture food processing levels, and classification using the NOVA system may introduce misclassification and relies on retrospective self-report, which is subject to recall bias. Consumption frequency was reported on an eight-tier scale, and quantity was documented using standard portion sizes.
Thigh MRI scans were obtained using a 3.0-T Magnetom Trio scanner at all centers. A slab of 15 contiguous axial sections covering the middle third of the femur was acquired, and all 15 images were used to assess MFI to account for variation in thigh muscle coverage across participants. A Goutallier grade was assigned to each muscle observed in at least 8 of the 15 axial images. Sum scores were calculated for knee extensors, knee flexors, thigh adductors, and bilateral thigh muscles overall.
Higher UPF consumption is significantly associated with greater thigh MFI
A total of 615 participants, including 340 women and 275 men, met the eligibility criteria. The mean age of the participants was 59.5 years, and their mean BMI was 27. The cohort was predominantly White and non-Hispanic. Among the participants, 64.6 % were overweight, and 24.2 % were obese. Abdominal obesity, defined by WHO waist circumference thresholds, was present in about 43 % of men and 81 % of women.
The majority of participants reported above-median income, and over half had no history of smoking. On average, UPFs made up about 41 % of participants’ daily diet. Men consumed both more total calories and a higher proportion of UPFs compared to women.
Among the thigh muscle groups, flexors showed the highest MFI score and extensors the lowest, as measured by Goutallier grading. Women consistently showed greater MFI than men across all muscle groups. The relatively lower degree of fat infiltration in extensors may partly explain their weaker observed association with UPF intake, rather than indicating a muscle-specific absence of effect. While Goutallier grading showed good inter- and intraobserver agreement in this study, it remains a semiquantitative method with only moderate reproducibility and does not provide fully quantitative measures of intramuscular fat.
The relationship between UPF consumption and MFI was linear across all muscle groups. In BMI-adjusted models, statistically significant associations were observed for all thigh muscles, flexors, and adductors, but not extensors, whereas abdominal-circumference-adjusted models showed significant associations across all muscle groups and generally showed stronger associations than BMI-adjusted models.
Both regression models, one adjusting for BMI and the other for abdominal circumference, showed a statistically significant association between higher UPF intake and higher Goutallier grades across most thigh muscles. Each standard deviation increase in UPF proportion corresponded to more than a 0.1 SD rise in muscle fat infiltration.
Among individual muscle groups, the adductors showed the strongest association with UPF intake, followed by the extensors, both of which reached statistical significance in abdominal-circumference-adjusted models. Representative MRI scans visually illustrate differences in MFI between participants with low and high proportions of UPF in their diets.
Sex did not significantly modify the relationship between UPF consumption and MFI in any muscle group, suggesting the association holds equally for both men and women.
Sensitivity analyses adjusting for total dietary fat yielded consistent results, reinforcing the primary findings across muscle groups. When stratified by radiographic knee osteoarthritis status, the UPF-MFI association was stronger in knees with doubtful early radiographic changes (KL grade 1) than in those with no radiographic involvement, although this finding was based on sensitivity analyses and should be interpreted cautiously.
Conclusions
The current study suggests that higher UPF consumption is associated with poorer muscle quality, evidenced by greater intramuscular fat infiltration on thigh MRI, independent of sex, among individuals at risk for knee osteoarthritis.
Because this was a cross-sectional analysis relying on self-reported dietary data and semiquantitative imaging, the findings do not establish causality and should be considered hypothesis-generating. Additionally, the study population consisted primarily of older adults at risk for KOA, which may limit generalizability to younger or healthier populations.
The potential role of diet quality in preserving muscle health warrants further investigation, with UPF reduction representing a promising modifiable target in this population.
Download your PDF copy by clicking here.
Exercise cuts 'chemo brain' and fatigue in cancer patients