The 67 participants recruited to participate in the small trial study of rapamycin bioavailability included 24 females and 43 males who were routinely taking either compounded (n = 23 at 5, 10, or 15 mg) or commercial rapamycin (n = 44 at 2, 4, 6, or 8 mg), with the specific dose per individual being determined by a partnership between patient and physician to optimize the balance between healthy aging benefits and side effect tolerance. While age was statistically higher for individuals taking compounded rapamycin (compounded mean = 61.7 years, SD = 9.1; commercial mean = 56.8 years, SD = 9.3; t(65) =  − 2.050, p = 0.044), and gender distributions had a non-significantly higher number of males than females in both groups (compounded: males = 17 (74%), females = 6 (26%); commercial: males = 26 (59%), females = 18 (41%); \(\chi\) 2 (1, N = 67) = 1.443, p = 0.230), BMI was similar across both groups (compounded mean = 23.8 kg/m2, SD = 2.9; commercial mean = 25.1 kg/m2, SD = 4.8; t(65) = 1.158, p = 0.251; Supplementary Table S1, Supplementary Figure S1). Participant distribution across doses is summarized in Supplementary Table S2 (Fig. 1a).

For all rapamycin formulations and doses, blood rapamycin levels were evaluated 24 h after dosing. Stratifying blood rapamycin levels by dose suggested that levels across the dosing ranges used for each formulation were somewhat similar (Fig. 1b), though clearly not equivalent, by dose. This was not fully explained by differences in potency of the two formulations (compounded was independently validated to be approximately 75% as potent as commercial per milligram), and no significant association was observed between blood rapamycin levels and BMI, gender, length of time taking rapamycin, self-reported activity level, pre-existing health conditions, or other medications taken simultaneously for either formulation (see Supplementary Table S3 for a summary). However, given the differences in blood rapamycin levels per dose for each formulation, we evaluated the average rapamycin blood level per mg of rapamycin taken to better understand differences in bioavailability for each formulation (Fig. 1c, Table 1). To obtain an estimate of comparative efficacy, the per mg bioavailability for each formulation was examined both as an average for each dose as well as an average across all doses (Fig. 1d, Table 1). From this, we estimated blood rapamycin levels of compounded to be 0.27 ng/mL per 1 mg dose, and commercial to be 0.87 ng/mL per 1 mg dose, resulting in a 31.03% bioavailability of compounded rapamycin relative to commercial rapamycin.

We next explored the linearity of rapamycin dose to blood concentration for the study cohort using mixed effect models. Within both compounded (B = 0.173, SE = 0.071, t(25.6) = 2.442, p = 0.022; 95% CI = 0.027–0.318) and commercial (B = 0.697, SE = 0.111, t(38.5) = 6.269, p < 0.001, 95% CI = 0.472–0.922) groups, significant associations were observed between dose and blood rapamycin levels (Fig. 2a, Supplementary Table S4.1), though the slope of the dose-to-blood curve differed significantly between the two formulations, (B =  − 0.524; SE = 0.130, t(78.6) =  − 4.026, p < 0.001; 95% CI =  − 0.783 to − 0.265; Fig. 2a, Supplementary Table S4.2). This suggests that while both formulations had linear dose-to-blood level relationships, commercial rapamycin elicits a significantly stronger response in blood rapamycin levels as dosage is increased compared to compounded.

Importantly, however, we noted substantial inter-individual variability in blood rapamycin level concentrations at each dose of compounded and commercial rapamycin administered, suggesting differences in bioavailability between people taking the same rapamycin dose independent of formulation (Fig. 2a). Thus, to more fully evaluate whether this variability is inherently characteristic of rapamycin overall or is rather a facet of an individual response, we explored blood rapamycin levels in participants for whom two measurements were available within 30 days (compounded n = 15, 12 with increasing doses and 3 with same doses; commercial n = 6, 1 with increasing doses and 5 with the same doses; Supplementary Table S5). For those taking the same dose of rapamycin at both timepoints (compounded, n = 3; commercial, n = 5), blood rapamycin levels were similar between the first and second measurement of the same dose for most (though not all) participants, suggesting a relatively stable individual response (Fig. 2b). Similarly, for participants receiving two differing doses, the second (higher) dose tended to be followed by an expected higher blood concentration of rapamycin, though some variability in response was again observed (Fig. 2c). No meaningful differences were observed in trends between compounded and commercial formulations.

Given the inherent limitations of small datasets, we expanded our investigation on rapamycin bioavailability dynamics to our Observational Research Database, which collects real-world data from individuals who have opted-in to anonymized research participation. From this, we identified 572 additional blood rapamycin samples from 316 individuals (87 females, 229 males) that contained sufficient information for further evaluation in the context of rapamycin bioavailability and dynamics over time after a single dose. As this evaluation was conducted retrospectively in a real-world cohort, a substantial proportion of the individuals were taking 6-mg commercial rapamycin (252 individuals (79.7%) and 473 timepoints (82.7%)), and only 10-mg dose was represented for the compounded formulation (Fig. 3a, Supplementary Table S6). Additionally, times between rapamycin dose administration and blood testing spanned 7 days, and intervals between blood tests for the same individual were generally 3 months or more.

We first evaluated whether blood rapamycin levels remained similar across doses and formulations within the larger cohort. To standardize evaluations, and allow for more reasonable comparison to the smaller study cohort, we restricted analysis to blood values obtained within 48 h after reported dose administration. This resulted in 167 datapoints, of which 84.4% were from the 6-mg commercial formulation group. Consistent with previous findings, average blood rapamycin level for each dose revealed similar patterns of response distribution (Fig. 3b), with variability in individual dose response that was not explained by factors such as gender or other medications, including Metformin, LDN, Acarbose, or NAD (Supplementary Table S8). Interestingly, the trends of increasing bioavailability by dose of commercial rapamycin observed in the small trial study (Fig. 1b) were not as apparent in this less tightly controlled cohort, even when all dose-to-test time ranges were included (Supplementary Figure S2a). To determine if this reflected any differences in the comparative bioavailability of each formulation in this larger cohort, we again averaged the blood rapamycin levels per mg of rapamycin taken for each dose. The resulting estimated bioavailability for compounded relative to commercial rapamycin was 30.26%, notably similar to the 31.03% estimate from the smaller cohort.

While uneven distribution of participants across doses and the single dosage (10 mg) of compounded rapamycin represented in this cohort limited reliable further exploration of dose linearity, the availability of multiple post-dose blood draw timepoints and longitudinal nature of this dataset permitted a preliminary investigation of the dynamics of rapamycin blood levels over time. We first explored the variation in blood rapamycin levels over the course of 1 week post-dose administration in participants using 6 mg of rapamycin (given the abundance of this dose in our dataset) who self-reported that they obtained blood rapamycin level evaluations from 1 to 7 days after taking their rapamycin dose (Supplementary Table S7). Averaging blood rapamycin levels for each day post dosing suggests that rapamycin blood levels peak 2 days after dosing, and decline gradually thereafter (Fig. 3c). These results were consistent when looking across all doses (Supplementary Figure S2b), as well as when subsetting for the most recent dose or doses at other timepoints (for patients with multiple dose-test timepoints within the dataset).

We next explored whether blood rapamycin levels were consistent over longer periods of time, as a robust number of individuals in this cohort (n = 228) had more than one dose-test timepoint over intervals of > 90 days. When evaluating all individuals with two timepoints, we found a significant increase in rapamycin levels from the first to second tests for the entire cohort (time 1 mean = 1.345, time 2 mean = 3.243, t(227) = 7.105, p < 0.001), though it should be noted that this was an overall effect, as not all individuals had increases. This effect remained when limiting to individuals who had two timepoints with a dose-test interval of 1–3 days (n = 56, time 1 mean = 2.093, time 2 mean = 3.793, t(55) = 2.668, p = 0.005; Fig. 3d), as well as by sex (female n = 11, time 1 mean = 0.818, time 2 mean = 2.773, t(10) = 2.288, p = 0.045; male n = 45, time 1 mean = 2.404, time 2 mean = 4.042, t(44) = 2.131, p = 0.039; Fig. 3e). Extending this analysis further to explore all datapoints over the full 1.5 year span of available datapoints (with a maximum individual timespan of 492 days) using linear regression analysis of blood rapamycin levels over time produced similar results of increasing blood rapamycin values over time (F(1, 580) = 33.735, p < 0.001, β = 0.234, t = 5.808, p < 0.001), which remained consistent after restricting to the most recent datapoints from patients with two or more samples (F(1, 314) = 14.648, p < 0.001, β = 0.211, t = 3.827, p < 0.001, Fig. 3f). However, to reduce potential bias of low values from new rapamycin users, we restricted the analysis to individuals who had been on rapamycin for more than 6 months. This resulted in a non-significant relationship between blood rapamycin values and time (F(1, 73) = 0.838, p < 0.363, β = −0.017, t = −0.915, p < 0.363), however, this appears to stem from a stabilization of rapamycin levels in this period for most individuals (Fig. 3g), though further studies will be required to elucidate this more conclusively.