Individuals with Long COVID symptoms show subtle impairments in glucose tolerance, serum leptin, lean mass and enhanced T-cell DPP-4 activity, with metabolic changes not driven by SARS-CoV-2 virus in subcutaneous adipose tissue.
Key Findings
Results
T-cell oxidative phosphorylation did not differ between Long COVID positive and Indeterminate groups nor by PROMIS symptom scores.
This was the primary outcome of the study
Comparisons were made between LC+ (n=15) and Indeterminate (n=39) groups using the LCRI classification
PROMIS symptom score subgroups included High Symptom Burden (HSB) >15: n=25, Intermediate Symptom Burden (ISB) 10-15: n=14, and Low Symptom Burden (LSB) <10: n=15
No significant difference in T cell mitochondrial respiration was detected across any grouping
Results
T cell glycolytic activity was greater in Long COVID positive participants compared to Indeterminate participants.
This was a secondary outcome finding
The difference was observed between LC+ (n=15) and Indeterminate (n=39) groups classified by the Long COVID RECOVER Index (LCRI)
BMI and fat mass did not differ between groups, suggesting glycolytic differences were not attributable to body composition
Oxidative phosphorylation was unchanged, indicating a selective shift toward glycolysis rather than overall mitochondrial dysfunction
Results
HOMA-IR was higher and HOMA-B was lower in LC+ participants compared to Indeterminate participants, indicating subtle impairments in glucose metabolism.
Prevalence of Type 2 diabetes did not differ by symptom scores between groups
HOMA-IR elevation indicates greater insulin resistance in LC+ individuals
HOMA-B reduction indicates lower beta-cell function in LC+ individuals
These findings suggest subclinical glucose metabolic impairment in Long COVID even in the absence of frank diabetes
Results
SARS-CoV-2 viral RNA was not detectable in subcutaneous adipose tissue biopsies from Long COVID participants.
Subcutaneous adipose tissue biopsies were obtained as a secondary outcome measure
Viral RNA was undetectable in all biopsy samples tested
Mean time from initial infection to study participation was 894 days
This finding suggests that persistent SARS-CoV-2 in subcutaneous adipose tissue does not drive the metabolic changes observed in Long COVID participants
Results
CD26+ T cell number and DPP-4 (CD26) activity were higher in the High Symptom Burden group and correlated significantly with symptom scores.
Correlation between CD26+ T cell number/DPP-4 activity and symptom scores was r=0.52, p<0.01
DPP-4 is expressed on T cell surfaces (CD26) and has known roles in immune regulation and glucose metabolism
This association was observed across both symptom classification systems (PROMIS and LCRI)
Plasma cytokines and stimulated T-cell cytokines did not differ by symptom group in either classification system
Results
Lean mass and femoral bone mineral density (BMD) trended lower in the High Symptom Burden group compared to the Low Symptom Burden group.
This trend was observed across both classification systems (PROMIS scores and LCRI)
The differences did not reach statistical significance and were described as trends
BMI and fat mass did not differ between symptom burden groups
Mean BMI of the full cohort was 31, and 80% of participants were female with mean age 51.7 years
Methods
The study cohort was predominantly female with a mean BMI in the overweight/obese range and was assessed approximately 2.5 years after initial infection.
54 total participants enrolled; 80% female
Mean age was 51.7 years; mean BMI was 31
Mean time from initial SARS-CoV-2 infection was 894 days (~2.4 years)
Participants were drawn from the RECOVER study across Maine, Louisiana, and Kentucky
Groups were matched by age, sex, and BMI
Results
Serum leptin levels showed subtle differences associated with Long COVID symptom burden.
Serum leptin was identified as one of the subtle impairments observed in individuals with Long COVID symptoms
This finding was noted alongside glucose tolerance and lean mass differences
Specific leptin values and p-values were not reported in the abstract
What This Means
This research suggests that people with Long COVID — defined as persistent symptoms at least 3 months after SARS-CoV-2 infection — show subtle but measurable differences in how their bodies process energy and regulate the immune system, compared to people who recovered from COVID-19 without lasting symptoms. Specifically, those with greater Long COVID symptom burden had signs of insulin resistance (higher HOMA-IR) and reduced insulin-producing cell function (lower HOMA-B), as well as higher activity of an enzyme called DPP-4 on their T cells, which was strongly linked to symptom severity. Notably, the study found no detectable SARS-CoV-2 virus in fat tissue sampled from participants, even though many were still experiencing symptoms nearly 2.5 years after their original infection.
The study also found that people with more Long COVID symptoms tended to have less lean body mass and lower bone density in the femur, though these differences did not reach statistical significance. A key immune finding — that T cell mitochondrial energy production (oxidative phosphorylation) would be altered — was not confirmed, but T cells in the Long COVID group did show increased glycolytic activity, suggesting a shift in how these immune cells generate energy. Inflammatory markers in the blood and T cell cytokine responses did not differ between groups.
This research suggests that Long COVID is associated with subtle metabolic changes — particularly around blood sugar regulation and immune cell behavior — but that these changes are probably not caused by the virus hiding in fat tissue. The strong correlation between DPP-4 enzyme activity and symptom severity is particularly noteworthy, since DPP-4 inhibitors are already used as diabetes medications, which raises questions about potential overlapping mechanisms between Long COVID and metabolic disease. The study was small (54 participants) and cross-sectional, so it cannot establish cause and effect, but it points toward metabolic and immune pathways worth investigating further.
Check Your Own Numbers
Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.
Emery I, Sahagun S, Zunica E, Axelrod C, Beyl R, Cho E, et al.. (2026). Metabolic, viral, and immune phenotypes in long COVID.. Frontiers in endocrinology. https://doi.org/10.3389/fendo.2026.1899115