chatGPT:(5.6maxpaid)
This is a useful study showing that rapamycin’s effect on lifespan can reverse with dietary conditions. The evidence for substantial harm under some conditions is convincing; the explanation for that harm remains speculative.
Jackson et al.’s paper, “Diet-dependent, beneficial and adverse effects of rapamycin on life span of Drosophila melanogaster”, examines why their laboratory previously found lifespan shortening when other laboratories usually reported extension.
Summary. The researchers conducted four adult longevity experiments involving five fly strains and both sexes. Each treatment group contained approximately 100 flies distributed across four vials. Adult flies received 200 µM rapamycin in their food continuously from two days after emergence.
The main comparison involved a cornmeal–torula yeast diet and a brewer’s yeast–sucrose diet. Using the authors’ classification, rapamycin was harmful in 19 of 26 comparisons on the torula diet, beneficial in one, and without a statistically significant effect in six. On the brewer’s diet, eight comparisons were classified as beneficial and 14 showed no significant effect. These counts include some results that were significant only before correction for multiple comparisons.
Examples from the simultaneous comparison in Experiment 4 illustrate the differences:
| Fly strain and sex | Cornmeal–torula diet | Brewer’s yeast diet |
|---|---|---|
| w¹¹¹⁸ males | 78 → 38 days: −51.3% | 68.5 → 69 days: +0.7% |
| wDah females | 58 → 43 days: −25.9% | 56 → 59 days: +5.4% |
| w¹¹¹⁸ females | 93 → 95 days: +2.2% | 72 → 75 days: +4.2% |
Values are median lifespan, shown as control → rapamycin. Statistical support differs between these changes; notably, the +5.4% increase did not survive correction for multiple comparisons.
A particularly informative observation is that untreated flies generally lived longer on the torula diet. Consequently, the adverse results occurred on a diet capable of supporting long lifespans, extending the concern beyond overtly poor nutritional conditions.
Other findings were:
- Food storage mattered. In one male strain, rapamycin reduced median lifespan by approximately 20–23% when food was used 1–3 days after preparation, versus 12–13% after 4–9 days.
- Solvent, lighting and mating history did not provide a general explanation for the contrasting results.
- Removing preservatives generally maintained or increased control lifespan. This frustrated the proposed explanation that rapamycin benefited flies principally by suppressing harmful microbial growth.
- Development was highly sensitive to rapamycin. At the tested concentrations of 10 and 200 µM, offspring failed to reach adulthood while remaining on the drug-containing food. At 0.1–1 µM, development was delayed. At 0.1 µM, the proportion reaching adulthood was not significantly reduced. Some arrested larvae resumed development after transfer to untreated food.
The authors propose that rapamycin might reduce excessive nutrient utilisation on the richer diet, while pushing flies on the leaner diet towards nutritional insufficiency. They appropriately acknowledge that this explanation is conjectural.
Novelty. The general finding that rapamycin can shorten fly lifespan, and that nutrition modifies its effects, was already established:
- Villa-Cuesta et al. (2014) reported dose-dependent lifespan shortening under low-nutrient conditions and no reproducible benefit on their balanced diet. Original study
- Rohde et al. (2021) demonstrated substantial genetic variation in responses to rapamycin. Original study
- Bearden et al. (2024), from the present authors’ laboratory, had already reported lifespan shortening in male y w flies. Original study
The present paper’s contribution is to reproduce and investigate these contrasting outcomes systematically within one laboratory, across several strains and both sexes. Demonstrating adverse effects under conditions supporting long control lifespans strengthens the earlier evidence. Testing food storage, solvents, mating and preservatives also helps identify which experimental details deserve attention.
I would therefore rate its novelty as moderate, with substantial value for reproducibility and experimental design. It does not identify a new molecular mechanism.
Critique. Its strengths include the use of several genetic backgrounds, both sexes, matched solvent controls, repeat experiments and conditions previously associated with lifespan extension. The authors also use survival models that account for flies sharing vials and explicitly distinguish nominal from corrected statistical significance.
The main limitations are:
-
The dietary comparison does not identify the responsible dietary factor.
These are complex recipes differing in yeast species and quantity, sucrose, cornmeal, agar and preservatives. They are not matched diets varying a single nutrient. Cornmeal also contributes nutrients, so the difference cannot be reduced to yeast or sugar concentration.Moreover, the directly compared stocks had been maintained on their respective diets for at least five months. The comparison therefore includes developmental and parental dietary history, alongside adult diet. The authors’ nutritional-insufficiency explanation would require defined diets, measurements of nutrient intake and physiological evidence of deficiency. Egg-laying preference provides little information about actual nutrient consumption.
-
One adult concentration cannot establish the full relationship between diet and treatment.
All adult longevity experiments used 200 µM. This concentration has precedent in successful fly studies, making it a reasonable replication choice, but it cannot establish whether lower concentrations would avoid harm or produce benefits on torula food.Crucially, equal concentration in food does not establish equal exposure inside the fly. Food consumption, absorption, drug stability and tissue rapamycin concentrations were not measured. The apparent dietary interaction could therefore involve altered drug exposure, altered biological response, or both. Likewise, weaker effects after food storage suggest a change in exposure but do not demonstrate chemical degradation.
-
The beneficial effects are small and less robust than the abstract’s tally suggests.
The eight positive brewer’s-diet comparisons include three that were only nominally significant. Counting the tables’ own corrected-significance markings, five survived Holm correction.The largest reported median increase, +5.4%, represents 56 versus 59 days, with an uncorrected p value of 0.026. It did not remain significant after adjustment. These are worthwhile signals, but the evidence for a consistent benefit is appreciably weaker than the evidence for the largest adverse effects. The comparison counts also should not be treated as independent replications by different laboratories.
-
Some survival-model interpretation needs correction.
In Supplement 5, the interaction model’s rapamycin hazard ratio of approximately 1.98 applies specifically to the reference group: female y w flies on torula medium. It is not an overall treatment effect across every strain, sex and diet. Comparing its significance with the common treatment coefficient from a model without interactions does not, by itself, demonstrate contradictory results: those coefficients estimate different quantities. Supplement 5In addition, some survival curves cross: rapamycin increases earlier mortality while improving later survival. A median or single hazard ratio can conceal that trade-off. Age-specific mortality and average survival over a defined follow-up period would make the consequences clearer.
-
The study measures survival without establishing how ageing changed.
There are no measurements of TOR inhibition, autophagic flux, energy reserves, tissue pathology or functional healthspan. It therefore cannot determine whether harm arose through impaired translation, excessive suppression of growth or repair, nutritional insufficiency, or another process.Longer control lifespan also does not establish that a diet is biologically optimal in every respect. The observations support effects on survival; claims about the underlying rate of ageing require additional evidence.
-
Developmental and human implications require careful separation.
The developmental experiments show growth arrest or delayed development during direct exposure. They do not establish permanent adult infertility. The ability of some larvae to resume development after drug removal is relevant here.For humans, this paper generates a credible hypothesis that diet could modify benefit and harm. It does not identify which human diets would do so, quantify the risk, or establish effects of intermittent treatment. A concentration in fly food cannot be translated directly into a human weekly dose.
The most informative follow-up would combine chemically defined diets with several rapamycin concentrations, measured food intake and tissue drug levels, and assays of TOR activity and autophagy. Starting adult treatment groups from the same rearing population, measuring functional health alongside survival, and replicating the findings in another laboratory would substantially strengthen the conclusions.
Q2: how would the concentrations in fly food in this study be estimated to compare to human dosing
We can estimate how much rapamycin a fly swallowed, but converting that into an equivalent human dose requires assumptions about absorption and clearance. The study did not measure food consumption or internal drug concentrations.
The adult treatment was 200 µM continuously in food. Using rapamycin’s molecular weight of 914.2 g/mol, this equals 182.8 mg/L, or 183 nanograms per microlitre of food. Rapamune prescribing information
| Concentration in this study | Rapamycin per litre of food | Experiments using it |
|---|---|---|
| 0.1 µM | 0.0914 mg/L | Development |
| 1 µM | 0.914 mg/L | Development |
| 10 µM | 9.14 mg/L | Development |
| 200 µM | 182.8 mg/L | Adult lifespan and development |
Estimating the swallowed dose. For illustration, assume an adult fly weighs 1 mg and consumes 1 µL of food daily:
$$
\text{Daily dose}
=\frac{0.000183\ \mathrm{mg}}{0.000001\ \mathrm{kg}}
\approx183\ \mathrm{mg/kg/day}.
$$
Food consumption of this order is plausible: one established liquid-feeding experiment measured approximately 1.5 µL per fly per day. However, consumption of the solid diets in Jackson’s study could differ substantially. Ja et al., 2007
The sensitivity to assumed consumption is:
| Assumed daily intake by a 1 mg adult fly | Swallowed dose at 200 µM |
|---|---|
| 0.5 µL | 91 mg/kg/day |
| 1 µL | 183 mg/kg/day |
| 2 µL | 366 mg/kg/day |
These are illustrative scenarios, not measured doses or a confidence interval.
For comparison, 5–10 mg weekly in a hypothetical 70 kg human averages approximately 0.010–0.020 mg/kg/day swallowed. The fly estimates are therefore thousands of times larger on a simple body-weight basis. That ratio is not a ratio of biological effects or an appropriate human dose conversion. It leaves absorption, metabolism, distribution and dosing frequency unaccounted for.
An earlier study provides a more useful internal-exposure measurement. Bjedov et al. measured 3.3 ± 0.2 ng of rapamycin per mg of fly wet weight after feeding flies food containing the same 200 µM concentration. That equals 3.3 µg/g of whole-fly material—approximately 3.6 µM if one assumes a density of 1 g/mL. This measurement came from a different experiment and cannot be assumed to apply to Jackson’s flies or both diets. Bjedov et al., 2010
For human reference, product information reports the following mean blood peaks after single, fasted tablet doses in healthy volunteers:
| Human oral dose | Mean peak blood concentration | Molar concentration |
|---|---|---|
| 5 mg, as five 1 mg tablets | 10.3 ng/mL | Approximately 11 nM |
| 10 mg, as ten 1 mg tablets | 23.6 ng/mL | Approximately 26 nM |
These are study averages with substantial individual variability. Rapamune data sheet, tables 5–6
The whole-fly concentration is numerically much higher, but whole-fly material and human blood are different compartments. Whole-fly measurements combine tissues and potentially drug remaining in the gut. Human blood levels are themselves strongly affected by rapamycin’s partitioning into blood cells and binding to proteins. Consequently, dividing those concentrations does not establish a human-equivalent dose.
A defensible comparison would measure drug exposure over time in relevant tissues, preferably including the unbound fraction, and the magnitude and duration of TORC1 inhibition. Continuous feeding also differs from weekly administration: the human terminal half-life is around 62 hours in the population described in the prescribing information, while the corresponding clearance in these flies was not measured. Clinical pharmacology information
For interpreting this paper, 200 µM represents a large nominal swallowed dose, with earlier evidence of substantial internal accumulation. An equivalent figure in human mg/week cannot be reliably assigned from the available measurements.
Q3: make an estimate as to what range in terms of mg/week in humans would be equivalent
For the study’s 200 µM adult treatment, a deliberately crude allometric estimate gives approximately 100–2,000 mg/week for a 70 kg human. This is a mathematical extrapolation with very low confidence, not a validated equivalent exposure or a dose to take.
I would calculate it using these assumptions:
- Adult fly weight: 1 mg.
- Food consumption: 0.5–2 µL/day.
- Food concentration: 183 mg/L rapamycin.
- Resulting fly intake: 91–366 mg/kg/day.
- No separate correction for differences in oral bioavailability between species.
Then apply two conventional size-scaling assumptions:
| Assumed scaling of total dose with body mass | Human amount at 1 µL/day fly intake | Range allowing 0.5–2 µL/day |
|---|---|---|
| Body mass raised to ⅔, approximating surface-area scaling | 217 mg/week | 109–435 mg/week |
| Body mass raised to ¾, an alternative allometric exponent | 979 mg/week | 490–1,959 mg/week |
These exponents are discussed in interspecies dose-conversion guidance, but extending them from a 1 mg insect to a human is an unvalidated extrapolation. The guidance does not establish a rapamycin conversion for flies. FDA guidance, Appendix A
For reproducibility, the calculation is:
$$
D_{\mathrm{human,week}}
7\times70\times D_{\mathrm{fly}}
\left(\frac{10^{-6}}{70}\right)^{1-b},
$$
where (D_{\mathrm{fly}}) is in mg/kg/day and (b) is either ⅔ or ¾.
Thus, about 200–1,000 mg/week is the result at the central feeding assumption; about 100–2,000 mg/week includes the assumed feeding variation. That range captures sensitivity to those assumptions—it is not a confidence interval, and actual pharmacological equivalence could lie outside it.
Here, mg/week means seven days’ total input from a daily regimen. It does not mean that taking the entire amount once weekly would reproduce the flies’ exposure.
Conclusion: The research paper is a really high continuous dose.