Feast and famine: in vitro techniques to assess longevity diets

Since it’s impractical to study lifespan as an endpoint in CR trials, we need biomarkers of aging so that we can measure the progress of aging in real time and thereby assess the efficacy of the treatment.

To this end, Allard et al. collected serum from participants enrolled in one of two dietary restriction studies: FEAST (alternate-day fasting, ADF) or CALERIE (CR and CREX), and used this serum to culture human hepatoma (HepG2) cells. This technique was first described by the de Cabo lab in 2003 as a useful in vitro technique for assessing the phenotypic effects of a CR diet in rats and Rhesus monkeys. The researchers selected HepG2 cells for this work because these cells express many of target genes of interest, and to facilitate interspecies comparison (their original paper used rat serum on the rat hepatoma cell line Fao). These prior studies in animals suggested that many components of serum (insulin, glucose, IGF-1, and fatty acids) could modulate gene expression pathways involved in longevity. This is the first study using human serum in a cell culture system.

In vitro cellular adaptations of indicators of longevity in response to treatment with serum collected from humans on calorie restricted diets.

Two pilot studies were undertaken to examine the effects of ADF and CR on indicators of health and longevity in humans. In this study, we used sera collected from those studies to culture human hepatoma cells and assessed the effects on growth, stress resistance and gene expression. Cells cultured in serum collected at the end of the dieting period were compared to cells cultured in serum collected at baseline (before the dieting period). Cells cultured in serum from ADF participants, showed a 20% increase in Sirt1 protein which correlated with reduced triglyceride levels. ADF serum also induced a 9% decrease in proliferation and a 25% increase in heat resistance. Cells cultured in serum from CR participants induced an increase in Sirt1 protein levels by 17% and a 30% increase in PGC-1α mRNA levels.

This study adds support to Sirt1 as a biomarker for longevity, and found that expression of PGC-1α is variable between intervention diets.

I was concerned by the lack of phenotypic benefits in cells treated with CR sera, as compared to cells treated with ADF sera: The CR-treated cells exhibited no increase in heat-shock resistance, no decrease in cell proliferation, and no correlation of Sirt1 expression to reduced triglyceride levels.

In interpreting this finding, the authors suggest that the increased effects seen in the ADF participant serum could be due to the “short, regular intervals of complete caloric deprivation,” which might provide a more potent serum profile capable of producing the observed in vitro changes. Indeed, in studies of rodents , the ADF stress response phenotype can exceed the effects seen in rodents following a CR diet. Understanding serum differences between an ADF regimen and CR could elucidate the critical factors in this protective phenotype.

Alternatively, I think that these data could be a result of the cell type used in this assay (a tumor line); it would be interesting to see this experiment repeated on primary cells or a non-cancerous transformed cell line.

While these potentially exciting preliminary results shed some positive light on ADF, is this a realistic intervention for human longevity? So far, there have only been short (3 wks) studies in humans and none of the published data address participant satisfaction. If people think vegan and CR diets are difficult to maintain- this may be the least popular diet yet.



  1. Thanks for a very interesting post. I agree with your point regarding that it would have been nice if they ran the same experimental protocol but using primary cells (maybe in the future).
    On a side note I can tell you that ADF is surprisingly relatively easy to maintain (fasting period usually in the 20-22 hour range).

  2. “These prior studies in animals suggested that many components of serum (insulin, glucose, IGF-1, and fatty acids) could modulate gene expression pathways”

    Surprising that they didn’t directly monitor for changes in these levels in the test serum, or did they?

  3. They did look at serum triglycerides and insulin in FEAST participants, but only reported the data as they correlated with Sirt1 gene expression.
    I do agree that direct analysis of the serum could enhance our understanding of these data.

  4. “If people think vegan and CR diets are difficult to maintain- this may be the least popular diet yet.”
    It really doesn’t have to be. The physiological effects of consuming very high proportion of NEFA in the diet is not unlike fasting. Just severely pare down the carb and protein content to the necessary levels.

Comments are closed.