Swedish Study Links Mediterranean-Style Diet Scores to Folate and GFAP

On September 15, 2026, BMC Medicine published “Biological signature of dietary patterns in older adults,” an analysis using baseline data from the Swedish National Study on Aging and Care in Kungsholmen (SNAC-K). Researchers examined dietary scores and 54 blood-based biomarkers in 1,769 adults aged 60 and older.

What did the study find about healthy aging?

The authors identified dietary patterns associated with blood-marker differences that they interpret as potentially relevant to healthy aging. Higher Alternative Mediterranean Diet (AMED) scores were associated with higher folic acid and lower GFAP. Higher Alternative Healthy Eating Index (AHEI) scores were associated with higher folic acid and lower GDF15. Higher scores on the reverse-transformed Empirical Dietary Inflammatory Index (rEDII) were associated with lower levels of several markers, including insulin and leptin as well as GDF15.

Was this research conducted in humans?

This was an observational study in humans, not an experiment in cells or animal models.

The researchers used baseline information from older adults participating in SNAC-K. Their biomarker panel covered metabolic and inflammatory processes alongside vascular and neurodegenerative processes. Rather than focusing on a single nutrient or blood measurement, the analysis placed several dietary scores beside a broad set of biological signals.

That distinction helps explain the study’s purpose. The researchers were looking for a biological signature associated with eating patterns: which blood measurements differed according to how participants scored on dietary questionnaires. The central comparison was between reported dietary patterns and measured blood markers within this group of older adults.

How were the dietary patterns assessed?

According to the paper, participants completed a validated food-frequency questionnaire. Researchers then used their reported intake to calculate the AMED and AHEI scores as well as rEDII. These were different ways of scoring dietary patterns rather than interchangeable names for the same measurement.

AMED represented the Mediterranean-style pattern in this analysis. AHEI represented the alternative healthy-eating pattern. The rEDII score captured dietary inflammatory potential in a reversed form: a higher score indicated lower inflammatory potential.

Keeping those distinctions visible is important when reading the results. A statement about the Mediterranean-style score should remain attached to AMED. Likewise, a finding associated with rEDII should be described as a result for that score rather than assigned to every pattern considered healthier.

The study belongs to a broader approach that examines whole dietary patterns alongside biological measurements. For related context, AgeAmaze’s coverage of dietary patterns and blood-marker research addresses the same general type of evidence. Here, the useful starting point is the specific score and the specific marker paired with it.

Which blood-marker findings belonged to each score?

The associations highlighted by the authors remained after correction for multiple comparisons. That detail matters because the researchers assessed a broad biomarker panel. The reported findings therefore reflect the paper’s statistical correction rather than an unfiltered list of every association examined.

Folic acid was a shared finding for AMED and AHEI: higher scores on either pattern were associated with higher blood levels. Their other highlighted associations differed. AMED was associated with lower GFAP, while AHEI was associated with lower GDF15.

The rEDII findings had a different emphasis. Higher scores were associated with lower levels of several markers, including insulin and leptin as well as GDF15. The overall picture was therefore a set of overlapping but distinct associations rather than a single blood signature shared by every dietary score.

GFAP is a blood protein studied in relation to glial activity and neurodegenerative processes. Its presence in this analysis gives the Mediterranean-style finding particular relevance to brain-health research. The precise result to retain is the association between higher AMED scores and lower measured GFAP.

How were the Mediterranean-score associations reported?

For each standard-deviation increase in AMED score, the authors reported an association of β = 0.066 for folic acid and β = −0.018 for GFAP at the median biomarker level. The positive estimate described the association with higher folic acid. The negative estimate described the association with lower GFAP.

These details keep the numerical findings tied to their original comparison. The dietary measure was AMED score, and the outcomes were blood-marker measurements. A careful account preserves those pairings rather than turning the results into a general statement that every healthier dietary score had the same biological associations.

What limits the interpretation of these findings?

The authors identify several weaknesses that affect how the results should be understood:

• The analysis cannot establish cause or time order. Diet and biomarkers were assessed cross-sectionally, so the researchers could not determine which came first. Reverse causality remains possible: existing health differences could influence eating patterns rather than dietary patterns producing the observed blood-marker differences. The authors explicitly caution against causal conclusions.
• Dietary intake was self-reported. Participants described their food intake over the previous year. Recall error could affect those answers, while social-desirability bias could influence what people reported. Using a validated questionnaire supports a consistent assessment method, but it does not remove the reporting limitations acknowledged in the paper.
• Other health factors may contribute to the associations. The authors state that residual confounding from chronic disease and other factors cannot be fully excluded. This leaves uncertainty about how much of the observed relationship belongs to diet itself rather than to other differences among participants.
• The same score could represent different food combinations. Participants with similar dietary scores did not necessarily eat the same foods. That makes a score useful for examining a broad pattern while limiting the ability to identify a particular food combination responsible for an association. The findings should remain attached to the assessed patterns.
• Supplement-use information was unavailable. Folate can come from food or supplements. Without supplement data, the researchers could not attribute measured folic-acid levels solely to food intake. This study also did not test supplements or establish that taking folate supplements would reproduce the dietary-score association.
• The sample limits generalizability. Participants were predominantly urban and highly educated as well as affluent. Those characteristics restrict how confidently the findings can be extended to older adults with different backgrounds. The authors therefore identify the need for research in larger and more diverse groups.
• Missing data may affect the study picture. People with greater disease burden were more likely to have missing information. The authors flag this because the available data may not capture participants with different health circumstances equally well. That issue belongs alongside the sample’s social characteristics when judging how broadly the findings apply.
• Blood-marker associations were not demonstrated health benefits. The analysis did not test whether a diet prevented dementia or extended lifespan. It also did not establish improved clinical outcomes. The reported coefficients are study estimates rather than percentages of benefit, and lower GFAP in this comparison is not evidence that brain disease was prevented.

For readers, the practical response is to keep the finding at the level measured: an early biological association between dietary scores and blood markers. AgeAmaze’s explanation of connecting biomarker signals with meaningful health outcomes provides a framework for making that distinction. Personal decisions about dietary changes or supplements should be discussed with a qualified healthcare professional rather than based on these associations alone.

What evidence should come next?

The next step is the larger and more diverse longitudinal and intervention research requested by the authors, followed by testing whether any diet-related biomarker changes correspond to meaningful improvements in health.

How AgeAmaze helps

Treating a lower GFAP reading as proof of better brain health skips the central evidence question in this dietary study: whether a blood-marker association translates into a meaningful benefit. AgeAmaze explains longevity research and biomarkers for research-minded adults while keeping established findings separate from early interpretations.

Read the research

Sources

1. Biological signature of dietary patterns in older adults