Mediterranean Diet Mitochondrial Proteins Pilot Shows Modest Link in Cardiac Patients

October 26, 2026 · Science

A cross-sectional pilot study of 49 older atrial fibrillation patients found modestly higher humanin and SHMOOSE in those who followed a Mediterranean diet most closely. Both are mitochondrial microproteins, and diet and protein levels were measured at the same moment. The investigation, conducted by researchers at the USC Leonard Davis School of Gerontology and Sapienza University of Rome, measured dietary scores and circulating proteins at a single timepoint.

Because both variables were recorded simultaneously, this observational design cannot show that dietary choices altered protein production. Public reporting described the cohort as older adults who closely followed the diet. The underlying clinical data describe elderly cardiac patients under active treatment for non-valvular atrial fibrillation. The findings provide preliminary hypothesis-generating data rather than clinical evidence that dietary changes modify circulating mitochondrial peptides.

What This Study Found and What It Cannot Show

A cross-sectional evaluation of forty-nine atrial fibrillation patients demonstrated that high Mediterranean diet adherence was associated with modest elevations in circulating humanin and SHMOOSE. Both statistical associations sat near the standard threshold for significance, with humanin reaching p = 0.045 and SHMOOSE reaching p = 0.046. The single-timepoint measurement prevents any determination of directionality or cause.

The pilot study enrolled individuals with a mean age of 78.4 years who were actively receiving care for non-valvular atrial fibrillation. These participants were drawn from an existing clinical registry rather than the community at large. The data cannot confirm whether eating a Mediterranean diet raises microprotein concentrations in healthy adults or improves long-term cardiovascular outcomes.

Study Cohort Characteristics and Atrial Fibrillation Registry

The study cohort comprised forty-nine patients drawn from a single-centre registry of individuals diagnosed with non-valvular atrial fibrillation. Investigators at Sapienza University of Rome selected this convenience sample from their active registry population. Study protocol required all enrolled patients to be at least sixty years old.

The final sample had a 57 percent female composition. The mean age was 78.4 years with a standard deviation of 8.7 years. Participants carried a moderate burden of chronic disease.

Adherence was scored using a validated nine-item questionnaire with possible results spanning zero to nine. Researchers split the cohort into two cohorts based on these survey totals. The high-adherence group contained twenty patients scoring between seven and nine, whereas the low-to-medium-adherence group contained twenty-nine patients scoring between zero and six.

Humanin and SHMOOSE Measurements Across Diet Adherence Groups

Humanin and SHMOOSE were measured from stored plasma samples using in-house sandwich ELISA protocols, and reported only as natural-log-transformed values. Those values carry no concentration units. The paper labels them pg/mL, but the log transformation strips those mass units from the reported figure.

Mean log-transformed SHMOOSE reached 7.331 plus or minus 0.53 in the high-adherence cohort compared to 7.067 plus or minus 0.36 in the lower-adherence cohort, producing a p-value of 0.046. Log-transformed humanin averaged 6.833 plus or minus 0.23 among high-adherence patients versus 6.666 plus or minus 0.33 in the lower-adherence group, yielding a p-value of 0.045.

Individual food item analysis linked specific intake patterns to each microprotein. Higher intake of olive oil, fish, and legumes was associated with higher humanin values. Higher olive oil intake and lower consumption of refined carbohydrates tracked with higher SHMOOSE numbers. The publication did not report exact quantities, gram weights, or daily portion counts for any of these food categories.

Oxidative Stress Marker Correlations and Null Findings

Researchers examined circulating markers of oxidative stress to investigate biological mechanisms connecting dietary habits, microproteins, and cardiovascular protection. The investigation measured sNox2-dp alongside 8-iso-PGF2-alpha. Nox2 is the enzyme the authors propose humanin interacts with.

Correlations between the microproteins and these oxidative markers yielded mixed results across the cohort. Humanin levels showed a statistically significant inverse correlation with sNox2-dp, yielding a Spearman rho of -0.335 with p = 0.019. Following multivariable adjustment for patient sex and body mass index, this relationship remained statistically significant with an unstandardised regression coefficient B of -0.010, a standardised beta of -0.302, and p = 0.040.

The remaining three oxidative stress analyses failed to maintain significance. Humanin initially demonstrated a weak inverse correlation with 8-iso-PGF2-alpha at rho = -0.283 and p = 0.049, but this association vanished after adjusting for sex and body mass index. SHMOOSE demonstrated no meaningful correlation with either oxidative marker, showing null relationships with sNox2-dp at p = 0.346 and with 8-iso-PGF2-alpha at p = 0.614.

Methodological Constraints of Cross-Sectional Pilot Data

A single-timepoint design cannot evaluate temporal progression. Evaluating dietary intake and microprotein levels simultaneously leaves the sequence of events undetermined. Confounding variables or baseline metabolic differences could explain the observed variance.

The authors explicitly stated that their findings cannot prove causality and should be viewed as preliminary observations. Multiple unmeasured variables were absent from the statistical models. Physical activity, metabolic parameters, and other confounders were not accounted for in the primary analyses.

Study limitations extended to sample size and diagnostic tools. The sample of forty-nine patients lacked statistical power for stratified subgroup analyses. Dietary adherence was captured via a brief nine-item questionnaire that may not capture the full Mediterranean diet pattern. Stored plasma volume constraints also restricted the number of microprotein targets the laboratory could evaluate.

Clinical Evaluation Standards for Mitochondrial Microprotein Biomarkers

Humanin and SHMOOSE are small peptides encoded within the mitochondrial genome rather than the cell nucleus. Across published human studies, clinical literature reveals complex and conflicting patterns. Circulating humanin levels do not correlate uniformly with health status across different human patient groups.

In a study published in Discover Oncology by Mohamed MH and colleagues, circulating humanin was significantly higher in 75 breast cancer patients compared to 70 age-matched control subjects (p < 0.001). Conversely, an investigation in Antioxidants by Suliman ME and co-authors found that humanin was significantly lower in 160 adults with advanced chronic kidney failure than in 80 controls. A clinical study in Cardiorenal Medicine by Bolignano D and colleagues evaluated 83 chronic haemodialysis patients of mean age 67. Both unusually low and unusually high humanin levels tracked with increased cardiovascular events.

These conflicting directions demonstrate that medical science has not established an optimal reference range or clinical target for humanin. No target level exists. ClinicalTrials.gov lists no interventional clinical trials administering humanin to human patients; existing registry entries evaluate the peptide solely as an observational biomarker. No commercial test for humanin or SHMOOSE is named in available sources, and the pilot study relied on an in-house laboratory ELISA protocol. Our analysis of the best biological age test platforms explains that reproducible reference standards are required before markers warrant consumer testing.

Research Boundaries and Unresolved Clinical Questions

This pilot study establishes neither a dietary intervention protocol nor a clinical biomarker for general longevity. The participants were elderly patients managing chronic cardiac conditions, meaning results cannot be extrapolated to healthy middle-aged populations. The investigation did not assess clinical outcome data.

Food intake records lacked exact serving sizes, preventing the formulation of quantitative nutritional recommendations. The statistical signals for both microproteins sat just beneath the p = 0.05 threshold in a small cohort, leaving them vulnerable to sample variability. Readers managing a cardiac condition should take dietary questions to their own clinician. Further details on how lifestyle changes interact with cellular biomarkers appear in our guide to anti-inflammatory diet mechanisms and our assessment of when to test biological age markers.

• Study Cohort: The sample comprised forty-nine elderly cardiac patients treated for atrial fibrillation, not healthy community-dwelling adults.
• Study Design: Cross-sectional data collection prevents drawing conclusions regarding dietary cause and effect.
• Measurement Units: Microprotein values were reported as mathematical natural logs rather than standard blood concentrations.
• Mechanistic Support: Only one of four oxidative-stress correlations survived adjustment for sex and BMI. Two were null and one lost significance.
• Testing Utility: There are no standardised commercial tests, reference ranges, or therapeutic targets for circulating humanin or SHMOOSE.

Publication History and Source Trial Documentation

The primary research paper was authored by Vicinanza R, Cammisotto V, Wan J, Yen K, Violi F, Pignatelli P and Cohen P, with Roberto Vicinanza as lead author and Pinchas Cohen as senior author. It was published in Frontiers in Nutrition on 10 March 2026 under the title 'Mediterranean diet adherence is associated with mitochondrial microproteins Humanin and SHMOOSE; potential role of the Humanin-Nox2 interaction in cardioprotection.' The study is indexed under DOI 10.3389/fnut.2025.1727012 and PMID 41883858 as an open-access report.

Media coverage emerged seven months later when ScienceDaily syndicated an institutional press release on 24 October 2026 based on materials supplied by the University of Southern California. The syndication simplified the trial description by referring to participants as older adults while omitting their underlying atrial fibrillation diagnosis. All source materials and registry records were accessed in October 2026.

Frequently Asked Questions

Does the Mediterranean diet increase lifespan?

This pilot study did not track participant survival or measure lifespan extension. It evaluated forty-nine older adults with non-valvular atrial fibrillation at a single timepoint to observe associations with mitochondrial peptides. Observational correlations in cardiac patients cannot demonstrate that dietary choices extend life.

Is the Mediterranean diet good for longevity?

In the March 2026 pilot study from Rome, high adherence scores tracked with modestly higher log-transformed levels of humanin and SHMOOSE. Because the pilot trial was cross-sectional and included only forty-nine cardiac patients, it cannot confirm longevity benefits directly.

What are mitochondrial microproteins?

Mitochondrial microproteins are small peptides that are encoded directly within mitochondrial DNA rather than nuclear DNA. Humanin and SHMOOSE belong to this biological class. However, their precise physiological regulation in humans remains under active scientific investigation, with conflicting blood levels observed across different clinical conditions.

What is humanin?

Humanin is a mitochondrial-derived microprotein encoded in mitochondrial DNA rather than the nuclear genome. Researchers propose humanin interacts with Nox2 in a potential cardioprotective pathway. Clinical human data present a mixed picture: circulating humanin is elevated in breast cancer patients, reduced in advanced kidney failure, and shows a curvilinear association with cardiovascular events in haemodialysis patients. It lacks an established clinical reference range.

Does the Mediterranean diet raise humanin levels?

The available scientific literature does not prove that eating a Mediterranean diet raises humanin levels. In the pilot study of forty-nine atrial fibrillation patients, individuals with higher diet scores had slightly higher log-transformed humanin values than those with lower scores (6.833 versus 6.666, p = 0.045). Because the study measured diet and protein levels simultaneously without an interventional dietary change, it cannot demonstrate that the food caused the higher protein levels.

Can I get my humanin or SHMOOSE levels tested?

No validated commercial laboratory tests currently exist for humanin or SHMOOSE. The researchers evaluated plasma samples using in-house sandwich ELISA protocols rather than validated assays. Furthermore, because humanin levels vary unpredictably across different diseases, scientists have not defined what constitutes a normal, deficient, or optimal reading. Measuring these microproteins outside an approved clinical study offers no actionable medical insight.

Are peptides good for anti-aging?

On clinical registries, humanin is studied strictly as an observational biomarker; no registered trial administers it to people as an anti-aging supplement.

What is bad about the Mediterranean diet?

This study does not address that question. Its only diet-related limitation is methodological: adherence was captured with a brief nine-item questionnaire that may not capture the full Mediterranean diet pattern, and the authors list physical activity and metabolic parameters among the confounders they could not account for.

Related

• Humanin Peptide Evidence: Clinical Research and Biomarker Review
• MOTS-c Evidence: Mitochondrial Peptides and Metabolic Health
• Best Biological Age Test: Diagnostic Accuracy and Platforms
• What Is Biological Age? Biomarkers and Measurement Methods
• When to Test Biological Age: Clinical Timing and Utility
• Anti-Inflammatory Diet: Scientific Evidence and Mechanisms