medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024.

The copyright holder for this preprint

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

1 Title: Safety and efficacy of rapamycin on healthspan metrics after one year: PEARL Trial
2 Results
3
4 Authors: Girish Harinath1,2, Virginia Lee1,2, Andy Nyquist1, Mauricio Moel1, Jesper Hagemeier1,
5 Stefanie L. Morgan1,2, *, Anar Isman1, Sajad Zalzala1
6
7 Affiliations:
1.
8 AgelessRx, Ann Arbor MI, USA
2.
9 Division of Research and Applied Sciences, AgelessRx, Ann Arbor MI, USA
3.
10 Data and Analytics Division, AgelessRx, Ann Arbor MI, USA
11 * Corresponding author
12

13 Corresponding author: Stefanie L Morgan, [email protected]

14
15 Abstract: Rapamycin has been shown to have longevity-enhancing effects in murine models,
16 but clinical data on its gerotherapeutic effects in humans remains limited. We performed a 48-
17 week double-blinded, randomized, and placebo-controlled decentralized study (Participatory
18 Evaluation of Aging with Rapamycin for Longevity [PEARL]; NCT04488601; registration date
19 2020-07-28) to evaluate the safety and efficacy of rapamycin in mitigating clinical signs of aging
20 in a normative aging cohort. Participants received 5 or 10 mg / week of compounded
21 rapamycin, or placebo for 48 weeks. Safety, adverse events (AEs) and blood biomarkers were
22 collected. Efficacy was assessed using DEXA scan-based measures and standardized surveys
23 assessing quality of life (QoL) and frailty. We did not detect significant differences in safety
24 blood biomarkers, or moderate to severe AEs between the rapamycin treatment groups and
25 placebo after 48 weeks. We detected dose-dependent (10 mg group) and sex-specific
26 improvements in lean tissue mass, pain, social functioning, overall QoL, and overall
27 osteoarthritis score in females, and in bone mineral content in men. Additionally, some
28 individuals receiving rapamycin experienced significant improvements in body composition
29 metrics that were associated with beneficial changes in gut health and lipid metabolism. We

NOTE: This preprint reports new research that has not been certified by peer review and should not be used to guide clinical practice. 1
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

30 conclude that low-dose, intermittent rapamycin administration over the course of 48 weeks is
31 safe and induces sex-specific improvements in multiple aspects of healthspan, with the most
32 robust improvements in lean tissue mass in women taking 10 mg rapamycin/week. Future work
33 will aim to identify biometric signatures of clinical effectiveness to inform personalized
34 treatment strategies that more broadly maximize efficacy and minimize side effects.
35
36 Keywords: rapamycin, aging, healthspan, longevity

2
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

37 Introduction
38 Aging is the biggest risk factor for all major chronic diseases, accounting for nearly 70%
39 of human mortality [1-3]. While advancements in medical technologies and public health
40 practices over the past 150 years have led to longer lifespans less shaped by natural selection,
41 the period of disease and disability-free life often referred to as “healthspan” has not kept pace
42 [4]. In conjunction with an epidemic of poor lifestyle habits, this has collectively led to a
43 growing chasm between lifespan and healthspan known as the healthspan gap, which in the
44 United States lasts several decades and is characterized by a high burden of functional disability
45 and age-related diseases (such as type 2 diabetes, osteoarthritis, and Alzheimer’s) that often
46 coexist as multi-morbidities [5]. While significant research has historically focused on treating
47 these diseases individually, a growing body of work within translational geroscience explores
48 developing gerotherapeutics that slow the aging process and delay the onset or prevent age-
49 related disease altogether [6].
50 The field of translational geroscience has made rapid advancements in recent years, due
51 in large part to strategic utilization of interventions already approved for other conditions by
52 the US Food and Drug Administration (FDA) [7]. By repurposing such drugs for their potential to
53 target the biology of aging and extend healthy longevity, clinical validation is fast-tracked to
54 permit more immediate collection of application-specific efficacy data. Notable among these is
55 rapamycin, which is widely used for its purported longevity and healthspan benefits within the
56 pro-longevity community [8]. While evidence supports a role for rapamycin in improving life-
57 and healthspans in preclinical studies [9], little data exists on its clinical efficacy in normative
58 aging humans.
59 As an FDA-approved small molecule drug, rapamycin is an evolutionarily conserved
60 inhibitor of the mammalian target of rapamycin serine/threonine kinase complex 1 (mTORc1), a
61 known regulator of aging processes [10, 11]. Hyperactivity of mTORC1 has been linked to
62 multiple chronic disease processes in nearly every organ system, while mTORC1 inhibition
63 induced by caloric restriction and rapamycin or its derivatives (rapalogs) has been shown to
64 reliably extend lifespan and healthspan across organisms from yeast to non-human primates
65 [12-21]. Rapamycin-mediated mTORC1 inhibition has demonstrated particular efficacy as a

3
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

66 geroprotective intervention in mice, extending lifespan in heterogeneous genetic backgrounds

67 across multiple studies from independent labs at multiple dosages, dosing periods, and
68 regimens, even in elderly animals [14, 16, 18, 21, 22].
69 While rapamycin has demonstrated a more pronounced lifespan-enhancing effect in
70 female than male mice, mitigation or reversal of age-associated changes in multiple organ
71 tissues and the immune system have been observed for both sexes [23-25]. These findings have
72 been reproduced in companion dogs and marmosets, however, clinical data on rapamycin’s
73 gerotherapeutic effects in humans remains limited [9, 12, 17, 26]. Given the substantial promise
74 in preclinical data, it is essential to obtain a deeper understanding and clearly define the clinical
75 benefits of rapamycin use for improving healthy aging in the generally healthy, adult
76 population. In particular, it will be important to understand how rapamycin may impact the
77 phenotypes of the biological aging process that substantially increases the risk of age-related
78 disease and mortality, such as an accumulation of visceral adipose tissue (VAT) and a loss of
79 lean muscle tissue and bone mass [27-32]. These are some of the most salient and functionally
80 consequential measures of biological aging, which frequently lead to reduced quality of life
81 (QoL), increased pain, and limited mobility, particularly for post-menopausal women [33, 34].
82 Collectively, this contributes to the marked decline in health that often occurs during this time
83 period [33]. It has been suggested that use of low-dose rapamycin may mitigate these features
84 of the aging process, enhancing healthspan [25, 35, 36].
85 Widespread adoption of rapamycin as a gerotherapeutic has historically been limited by
86 concerns regarding its known impact on immunosuppression, hyperlipidemia, and
87 hyperglycemia [37]. However, the vast majority of these effects stem from high, chronically
88 administered doses utilized in severely ill organ transplant or cancer patients, where the clinical
89 aim is inhibition of the immune system or anti-tumorigenic effects. In contrast, as a
90 gerotherapeutic for normative aging populations, low-dose, intermittent rapamycin is revealing
91 promise for minimizing side effects while still mitigating aspects of age-related decline [38-40].
92 For example, Mannick et al. demonstrated that healthy elderly individuals taking 0.5 mg of a
93 rapalog daily or 5 mg/ week for 6 weeks mitigated age-related immune decline by enhancing
94 the adaptive immune system’s response to vaccination [41]. This supports our recent findings

4
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

95 from a study of 333 low-dose rapamycin users indicating a high perceived QoL and improved
96 health outcomes compared to non-users [8]. Preliminary data suggest that these effects stem
97 from a partial inhibition of mTORc1 in low-dose rapamycin (rather than the complete inhibition
98 observed at higher doses), which has been demonstrated to mitigate age-related decline by
99 stimulating autophagy, reducing senescent burden, and reducing pro-inflammatory cytokines
100 that drive chronic inflammation [9, 15, 24, 42, 43]. While promising findings such as these have
101 encouraged some physicians to prescribe off-label rapamycin as a therapy to maintain
102 healthspan, there are many open questions that require further study, particularly in a clinical
103 setting.
104 An important gap in the clinical understanding of rapamycin for longevity is that to date,
105 no long-term randomized controlled trials (RCT) have been conducted to explore low-dose,
106 intermittent rapamycin regimens for improving multiple healthspan metrics in normative aging
107 cohorts. The current study, the Participatory Evaluation of Aging with Rapamycin for Longevity
108 (PEARL) trial, aimed to address this gap. PEARL represents the largest and longest clinical study
109 of rapamycin use for healthy aging performed to date. This 48-week double-blinded,
110 randomized placebo-controlled decentralized study evaluated the safety and efficacy of
111 rapamycin in mitigating clinical signs of aging in a generally healthy cohort. Here, we present
112 the first evidence to date, to our knowledge, that low-dose rapamycin administration is safe
113 over the course of 48 weeks and induces sex-specific improvements in lean muscle mass, bone
114 mineral content, aspects of age-related frailty, and overall QoL. Further, we highlight additional
115 benefits observed specifically in rapamycin responders that may offer mechanistic insights into
116 key biological drivers of clinical effectiveness that will be investigated in greater depth in future
117 studies.
118
119 Methods
120 Study design
121 The PEARL study was a decentralized, single-center, prospective, double-blind, placebo-
122 controlled trial assessing rapamycin in healthy individuals aged 50-85 years, to determine the
123 safety and efficacy in mitigating aging-related decline (Supplementary Figure S1).

5
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

124

125 The study was conducted in accordance with the standards of Good Clinical Practice, as
126 defined by the International Conference on Harmonisation and all applicable federal and local
127 regulations. The study protocol was approved by the institutional review board of the Institute
128 of Regenerative and Cellular Medicine in May 2020 (IRCM; approval number IRCM-2020-252).
129
130 Study endpoints
131 The primary endpoints of this study included safety, which included the collection of
132 adverse events (AEs), safety blood biomarkers, and visceral adiposity. Secondary endpoints
133 included efficacy blood biomarkers, DEXA scan-determined lean tissue mass and bone mineral
134 content, and standardized surveys assessing quality of life and frailty.
135

136 Study population

137 Participants were recruited via the AgelessRx online medical platform. Patients were
138 screened for eligibility, and if deemed eligible, informed consent was obtained for participation
139 in the study.
140 Participants were eligible for the study if they were aged between 50 and 85 years at
141 the start of the study, were interested in taking rapamycin off-label, were willing to undergo
142 minimally invasive tests, and were in good health or had well-managed clinically-stable chronic
143 diseases. Participants were excluded from the study if they had anemia, neutropenia, or
144 thrombocytopenia, were premenopausal, were scheduled to undergo major surgery in next 12
145 months, were undergoing or were scheduled to undergo chemotherapy, were scheduled for
146 immunosuppressant therapy for an organ transplant, had impaired wound healing or history of
147 chronic open wound, untreated dyslipidemia, impaired hepatic function, chronic infections
148 requiring ongoing treatment or monitoring (e.g. human immunodeficiency virus/acquired
149 immunodeficiency syndrome, chronic Lyme disease), allergy to rapamycin, clinically-relevant
150 primary or secondary immune dysfunction or deficiency, chronic oral corticosteroid or
151 immunosuppressant medication use, fibromyalgia, chronic fatigue syndrome/myalgic
152 encephalitis, breast implant illness, congestive heart failure, impaired renal function, poorly

6
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

153 controlled diabetes, type I or insulin-dependent type II diabetes, untreated or treated within
154 the last five years for substance abuse disorder, and untreated or poorly controlled mental
155 health disorder. Further, those who had taken metformin, rapamycin, or rapalogs were
156 excluded unless the participant agreed to a 6 month washout period prior to the start of the
157 trial.
158
159 Treatments
160 Study participants were randomized into three groups: receiving 5 mg of rapamycin, 10
161 mg of rapamycin, or placebo, once per week. Placebo capsules were compounded to look
162 similar and were taken orally (Belmar Pharma Solutions (Golden, CO, USA)). Study participants
163 were prescribed the drug for 48 weeks upon enrollment in the study and were dispensed
164 supplies for 12 weeks at a time.
165
166 Assessments
167 All assessments were performed at baseline, after 24 weeks, and after 48 weeks of
168 rapamycin treatment. Blood testing was performed two additional times, at 2 weeks and 4
169 weeks of treatment, to evaluate safety. All blood testing was performed by local Quest
170 Diagnostics of LabCorp laboratories, and included complete blood count (CBC), comprehensive
171 metabolic panel, liver function tests, renal function tests, and lipid panels. DEXA body
172 composition scans were performed primarily by DexaFit or Fitnescity locations, based on local
173 availability, though some participants utilized third party locations based on site availability.
174 Regardless of site, the outcome measures obtained by DEXA scans were body mass index (BMI),
175 visceral fat, bone mineral content, bone mineral density, and lean tissue mass. Gut health was
176 evaluated using the Thorne Gut Health test.
177 Summary health scores were calculated as a sum of responses on baseline survey
178 responses of health for questions of frequency of use of smoking, consuming caffeinated
179 beverages, consuming alcoholic beverages, eating sugary, salty, or “junk” foods, exercise
180 (vigorous and moderate intensity), and weight. Benefit Score was calculated by ranking the
181 percent change on each DEXA measure evenly into high, medium, and low benefit, and

7
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

182 summing scores into a meta score. VAT measures were reverse-scored, as decrease in VAT raw
183 values corresponds with improvement on this measure. The summary meta score was then
184 again divided evenly into a high, medium, and low ranking. High scores were considered
185 “improved”, and low scores were considered “decline”, while in-between scores were
186 considered “stasis”.
187 Health-related QoL was assessed by the short-form 36 (SF-36) survey, which consists of
188 36 questions covering eight health domains: physical functioning, role limitations due to
189 physical health, bodily pain, general health perceptions, vitality (energy levels), social
190 functioning, role limitations due to emotional problems, and emotional health [44]. The
191 responses are scored and summarized to provide a profile of an individual's perceived health
192 status. Pain, fitness, and functional limitations were assessed using the Western Ontario and
193 McMaster Universities Osteoarthritis (WOMAC) index, which is a questionnaire that is
194 commonly used to assess the health status of individuals with osteoarthritis of the hip and knee
195 [45]. It consists of 24 items divided into three subscales: pain, stiffness, and physical function.
196 The responses are scored to provide quantitative assessments of the severity of symptoms and
197 functional limitations associated with osteoarthritis.
198
199 Adverse Events
200 Adverse events (AEs) were obtained through weekly monitoring forms sent out to
201 participants. Clinical trial staff reviewed and documented AEs. Both expected (already known)
202 and unexpected AEs were reported and documented. A full list is presented in Table 2.
203
204 Statistical analyses
205 All initial analyses were performed by researchers blinded to the study treatment.
206 Demographic and baseline characteristics were summarized using descriptive statistics. All data
207 were tested for normality using visual inspection of histograms and Q-Q plots, and Shapiro-
208 Wilks tests were performed. Data were summarized using mean and standard deviation (SD)
209 and where appropriate, counts and percentages were reported. Baseline data were compared
210 between the three treatment arms using Analysis of Variance (ANOVA), independent t-test or

8
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

211 χ2 test. To compare the difference in mean change from baseline to most recent follow-up (48
212 weeks) between treatment arms, we employed ANOVA and BMI-adjusted marginal mixed-
213 effect models. Following this, full age-, sex-, and BMI-adjusted linear mixed-effects models with
214 a repeated measures structure were conducted. Interaction terms (time by treatment arm, as
215 well as time by treatment arm by sex) were included to investigate group and gender effects
216 and their trajectory over time. Where appropriate, stratified post-hoc analyses were conducted
217 within sex and between treatment arms (10 mg vs. placebo and 5 mg vs. placebo). All analyses
218 were conducted using SPSS 28 (IBM, Armonk, NY, USA) and Python 3.10 (Python Software
219 Foundation). A p-value of p<0.05 was considered significant.
220
221 Results
222
223 Participant demographics
224 A total of 115 participants were included in this study, of whom 40 received 5 mg/week
225 of rapamycin, 36 received 10 mg/week of rapamycin, and 39 received placebo (Supplementary
226 Fig S1a). At baseline, all participant groups were comparable across measures of age, sex,
227 height, weight, BMI, and blood safety markers, with the exception of the mean HbA1C, which
228 was lower in the 5 mg group than placebo (5mg mean = 5.2%, SD = 0.2; placebo mean = 5.3%,
229 SD = 0.3; t(72)=2.556, p=0.013; Table 1). Overall, participants were in exceptionally good health
230 at baseline, as evaluated by self-reports of health (Supplementary Fig S1b).
231
232 Safety: Adverse events
233 AEs were collected throughout the study as reported by participants on periodic surveys
234 or through direct contact with investigators (Supplementary Fig 2a; see Table 2 for all AEs). The
235 most commonly reported AEs were cold/flu/sinus symptoms, musculoskeletal pain, mouth
236 sores, and malaise, with a greater number of AEs reported in the placebo group relative to
237 treatment conditions (Supplementary Fig 2b, Table 2). The most commonly reported
238 rapamycin specific AEs were gastrointestinal issues. Surprisingly, mouth sores, which are often
239 associated with rapamycin treatment [46], were most frequently reported in the placebo group

9
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

240 (placebo: n=27, 5 mg: n=4, 10 mg: n=14; Table 2), though the same number of individuals
241 reported mouth sores in the 10 mg and placebo group (n = 6).
242 Notably, throughout the length of the trial, participants in the rapamycin treatment
243 groups reported increased instances of improvements in chronic ailments (chronic issue flare)
244 that were present when they first entered the trial compared to placebo (placebo: n = 14, 5 mg:
245 n = 24, 10 mg: n = 28). Conversely, participants in the placebo group reported increased
246 instances of worsening of their chronic ailments (chronic issue severity) compared to treatment
247 groups (placebo: n = 12, 5 mg: n = 4, 10 mg: n = 6; Table 2). No significant safety-related issues
248 were detected in the blood work results of any participants during periodic check-ins, or after
249 the 48 week study period for the 105 participants who completed the entirety of the study
250 (Supplementary Table S1).
251 Of the 123 total participants enrolled in the study, Serious AEs (SAEs) were reported for
252 four rapamycin users (three in 5mg group and one in 10mg group) and two placebo users.
253 Specifically, in the 5 mg group, one participant was hospitalized during the study period due to
254 respiratory symptoms that seemed related to anemia, which resolved with a blood transfusion.
255 A second participant in the 5 mg group had a severe infection requiring treatment from their
256 healthcare provider, which resolved after medication use (prednisone, doxycycline, and
257 albuterol). The third participant in the 5mg group had a hamstring tear prior to the study, that
258 healed slowly during the trial. In the 10mg group, a participant reported a sore throat, but
259 tested positive for group A strep and was prescribed amoxicillin, which resolved the symptoms.
260 One participant in the placebo group died of myocardial infarction during the study period. A
261 second participant in the placebo group reported a stomach virus that required attention from
262 a medical provider, and was diagnosed with small urinary infection. Symptoms resolved after
263 treatment.
264 While not an intended endpoint for this study, the study period was during the COVID-
265 19 pandemic, and a confirmed COVID-19 infection was reported for 25 participants: 10 in the
266 placebo group (29%), 8 in the 5 mg treatment group (21%), and 7 in the 10 mg treatment group
267 (21%). Among these, no differences in infection rates, symptom severity, duration of illness, or

10
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

268 length of recovery were noted between groups. Importantly, those with SAEs (described above)
269 were not noted to have COVID-19 upon physician testing.
270
271 Efficacy of rapamycin on healthspan metrics
272 DEXA Scan results
273 Participants underwent DEXA scans to assess visceral adipose tissue (VAT) mass, bone
274 mineral content (BMC), and lean tissue mass (LTM) at three timepoints (baseline, 24, and 48
275 weeks of treatment) to explore the effects of rapamycin on measures crucial for healthy aging
276 and the prevention of frailty and disease. At baseline, VAT mass, BMC, and LTM were similar
277 between the three treatment groups (Supplementary Table S2). Over the study period, VAT
278 mass increased most in the placebo group (mean = +35.34 g, SD = 290.76 g) and decreased in
279 the group taking 5 mg rapamycin (mean = -39.85 g, SD = 185.18 g), while BMC and LTM
280 increased only in the 10 mg rapamycin treatment group (BMC: mean = +19.41 g, SD = 104.34;
281 LTM = 517.87 g, SD = 2440.18). While notable trends of change in body composition measures
282 from baseline to 48 weeks were observed between treatment groups, none were statistically
283 significant (linear mixed-effects models interaction: VAT: F(2, 101.408) = 0.978,
284 p=0.379; BMC: F(2,104.415) = 1.403, p=0.250; LTM: F(2,126.996) = 1.115, p=0.331; BMD:
285 F(2,166.294) = 1.596, p=0.552) (Figure 1a, Supplementary Table S2). While some changes to
286 some body composition metrics (e.g. visceral adipose) in rapamycin treatment groups were
287 observed at 24 weeks, these effects stabilized to measures reported for 48 weeks by the end of
288 the trial (Supplementary Fig S3a, Supplementary Table S2).
289 As rapamycin has been suggested to have differential effects in males and females, we
290 explored this possibility by evaluating changes in DEXA scan data across sex (Table 3). Analysis
291 with linear mixed-effects models revealed statistically significant differences in BMC change
292 between groups across sex (interaction F(2, 193.245) = 4.489, p = 0.012, Table 3), which
293 subsequent analysis revealed was specifically driven by improvements in the male 10 mg
294 rapamycin treatment group vs placebo (interaction F(1, 129.224) = 4.563, p = 0.035, (Figure 1b).
295 This was consistent with changes of bone mineral density (interaction F(1, 131.924) = 4.956, p =
296 .028).

11
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

297 Similarly, statistically significant differences in LTM were observed between groups
298 when assessed across sex (F(2, 206.447) = 5.439, p = 0.005, Table 3), which subsequent
299 analyses revealed were primarily driven by differences between sex in the 10mg rapamycin
300 treatment group (F(1, 134.554) = 8.880, p = 0.003, Figure 1c). Post-hoc within sex analyses
301 confirmed that gains in LTM over 48 weeks were distinct to female participants taking 10 mg
302 rapamycin compared to placebo (female 10mg = +1,939.7g (SD = 1,998.9) versus female
303 placebo = -298.1g (SD = 1,191.1); F(2, 41.242) = 6.423; β = 55.5, SE = 16.9, p = 0.002). This
304 highlights that the average improvement in LTM in female participants was particularly robust
305 at an estimated rate of 55.5g per week faster than those on placebo. No significant changes in
306 BMC and LTM were observed in all other groups, nor were any significant changes in VAT
307 observed upon subsequent analyses (Supplementary Fig S3b).
308 Notably, marked variability in measures of VAT, LMT, and BMC was observed for all
309 rapamycin users independent of dose (Table 3, Supplementary Table S2). This resulted in
310 dramatically different impacts for the highest and lowest responding individuals across all
311 measures for both males and females (Supplementary Figure S4). Exploration of raw values of
312 response for individual participants revealed categories of individuals with improvement and
313 decline for each body composition measure (Supplementary Table S2, Figure 2a). Comparing
314 across groups, we observed that a greater percentage of individuals in the improvement group
315 were rapamycin users, with no specific bias in response across all measures by sex (Figure 2b,
316 c). Upon further investigation of individual body composition measures, sex-agnostic rapamycin
317 associated improvement held true for VAT (Figure 2d), however, the percent of participants
318 showing improvement in BMC and LTM differed by gender (Figure 2e, f). This is particularly
319 notable for BMC, where two thirds of 10mg males showed improvement (Figure 2e), and in
320 LTM, where all of the 10mg female participants demonstrated improvements (Figure 2f). Upon
321 further investigation, we noted that improvements in VAT correlated with
322 improvements on BMC and LTM, as well as with improved weight and BMI (Figure 3a).
323 Given these findings, we computed a “Benefit Score” for users based on summary
324 scores of improvement or decline on each body composition measure evaluated by DEXA.
325 Benefit score was significantly associated with rapamycin use (t(44.41) = 2.175, p = 0.035),

12
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

326 though some placebo users also saw global improvements and subsequently high Benefit
327 Scores (Figure 3b). As our bioavailability study has previously suggested a 3.5x reduction in
328 potency for compounded rapamycin relative to generic, we reasoned that the 10mg group was
329 more likely to be in a relevant bioavailable therapeutic dose range, and thus further
330 investigated Benefit Score in the 10mg vs Placebo groups to explore what might be driving
331 differences in rapamycin response. Importantly, separation of Benefit Scores by rapamycin dose
332 and sex in this manner revealed a pronounced female benefit bias, with all female rapamycin
333 users showing only improvement on the Benefit Score (Figure 3c). Subsequent analysis of gut
334 health and blood biomarkers by benefit score revealed significant changes in measures of gut
335 dysbiosis (F(2,54) = 3.215, p = 0.04; t(54) = -2.150, p = 0.036), immune readiness (F(2, 54) =
336 3.157, p = 0.05; t(54) = 2.517, p = 0.015; Figure 3d), and LDL levels (F(2, 102) = 4.262, p = 0.017;
337 t(74) = -2.505, p = 0.014; Figure 3e) between high and low Benefit Scores, with rapamycin users
338 showing greater magnitude of beneficial change than placebo.
339
340 Blood Biomarkers
341 In addition to completing routine blood work for safety monitoring, blood biomarkers of
342 health, such as LDL, hemoglobinA1c, triglycerides, and insulin levels were collected throughout
343 the study. While there were no significant differences between groups in biomarkers that were
344 outside the healthy reference range after 48 weeks, significant differences between treatment
345 and control groups were observed for mean change in corpuscular hemoglobin (MCH: 5mg: -0.1
346 (0.9), 10mg: -0.4 (1.1), Placebo: 0.2 (1.2); F(2, 101) = 3.232, p = 0.040) and mean platelet
347 volume (MPV: 5mg: -0.1 (0.4), 10mg: -0.1 (0.4), Placebo: 0.1 (0.4); F(2, 101 = 4.802, p = 0.10).
348 Post-hoc testing between treatment group pairs revealed additional changes for participants in
349 the 10 mg rapamycin treatment group in chloride levels (10mg: +0.6 (1.5), Placebo: -0.3 (2.0);
350 t(65) = 2.125, p = 0.037), carbon dioxide levels (10mg: -1.2 (1.8), Placebo: -0.4 (2.2); t(65) =
351 1.786, p = 0.079), and mean corpuscular volume (MCV: 10mg: -1.0 (2.2), Placebo: 0.4 (2.8);
352 t(65)=2.258, p = 0.027), and patients in the 5 mg rapamycin treatment group in platelets
353 (Supplementary Figure S5a-b, Supplementary Table S1). Stratifying by sex did not reveal any
354 additional patterns of significant change between dosing groups and controls over time.

13
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

355

356 Quality of life

357 Evaluation of rapamycin’s effects on self-reported quality of life (QoL) was conducted
358 using the SF-36 scale. Baseline measures revealed generally high QoL scores across the study
359 cohort, with no notable significant differences between groups (overall QoL scores of 83.7
360 (10.4), 88.1 (7.0), and 84.5 (12.0) for the 5 mg, 10 mg, and placebo groups, respectively, (F(2,
361 110) = 1.987, p = 0.142; Supplementary Table S3). After 48 weeks of treatment most SF-36
362 scores improved in all groups, however, with the exception of general health (5mg: +5.6 (10.3),
363 10mg: +3.2 (10.0, Placebo: +0.8 (12.0); F(2, 87) = 4.488, p = 0.014), no statistically significant
364 changes were detected using marginal linear mixed-effect models between treatment groups
365 across time (Supplementary Figure S6a, Supplementary Table S3).
366 As we observed sex-specific differences in DEXA measures, we similarly evaluated
367 whether QoL measures differed across sex between the treatment groups over time (mean QoL
368 scores stratified by sex are presented in Table 4). Linear mixed-effect models were employed to
369 further explore the interaction of sex and treatment groups with change in QoL over time.
370 Significant longitudinal differences in pain between groups and sex were observed (interaction
371 F(2, 201.308) = 6.815, p = 0.001), which was driven by a sex difference specifically in the 10 mg
372 rapamycin treatment group (10mg vs Placebo: F(1, 130.926) = 16.403, p < 0.001, Figure 4a,b).
373 This was confirmed in post-hoc analysis, where specifically females taking 10 mg rapamycin
374 (+14.7 SD = 13.9) saw improvements in pain compared to placebo (β = 0.332, SE = 0.110, F(1,
375 6.928) = 9.121, p = 0.020), indicating that similar to gains in LTM, among females,
376 improvements in pain are another robust effect of 48 weeks of rapamycin treatment.
377 Within the 10 mg rapamycin treatment group, the rate of improvement in social
378 functioning and overall quality of life also showed group and sex differences (Supplementary
379 Table S3), with both social functioning (+12.5 SD = 24.1) and QoL (+5.8 SD = 4.6) improving
380 most in the 10mg female group (group by sex interaction F(2, 188.862) = 2.701, p = 0.070 and
381 F(2, 200.291) = 2.028, p = 0.134, respectively). When limiting analyses to 10mg compared to the
382 placebo group, differences in the rate of improvements in social functioning and quality of life
383 between group and sex reached significance (interaction F(1, 132.742) = 5.551, p = 0.020 and

14
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

384 F(1, 123.268) = 6.753, p = 0.010, respectively), which, as with pain, appeared to be specifically
385 driven by females within the 10 mg rapamycin treatment group (Figure 4a,b).
386
387 Frailty
388 Finally, we measured symptoms of osteoarthritis (OA) using the WOMAC survey as a
389 marker for frailty, another important factor defining the length of an individual's healthspan.
390 Baseline scores were not significantly different across groups (Supplementary Table S3), and
391 linear mixed-effects models again revealed no significant patterns of change in scores across all
392 groups over the 48-week study period (Supplementary Figure S6b). However, given previous
393 sex-specific effects in other outcome measures, analyses of WOMAC scores were repeated by
394 exploring differences across sex and groups (Table 4). This revealed statistically significant
395 differences in the WOMAC pain score between groups and sex (interaction F(2, 194.531) =
396 3.516, p=0.032), with improvements observed particularly for female participants in the 5 mg
397 and 10 mg rapamycin group (-0.9 SD = 1.8; -0.8 SD = 2.9, respectively; Figure 4c,d), and trends
398 of improvement for the overall frailty score in the same groups (Supplementary Table S3;
399 group by sex interaction F(2, 202.257) = 2.172, p = .117). Evaluation of changes across sex
400 limited to the 10 mg rapamycin group versus placebo revealed a significant improvement in
401 both measures (interaction F(1, 120.251) = 5.791, p = 0.018 and F(1, 111.480) = 5,925, p =
402 0.017, respectively). This highlights that once again, female participants in the 10 mg group
403 appear to improve most (Figure 4c,d). No other significant differences were observed for this
404 measure.
405
406 Discussion
407 Few clinical trials to date have evaluated the effects of rapamycin and its derivatives in
408 generally healthy individuals, and those that have been conducted are often challenged by a
409 small cohort size, short-term follow-up, or both. While the most robust of these studies have
410 suggested improvements in age-related immune decline in healthy elderly individuals
411 administered low-dose everolimus for 6 to 16 weeks [41], no study to date has definitively
412 demonstrated the efficacy of rapamycin for supporting healthy aging in normative aging

15
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

413 individuals. The PEARL trial represents one of the most robust and comprehensive efforts
414 evaluating the long-term safety and efficacy of low-dose rapamycin for addressing multiple
415 aspects of age-related decline in a normative aging cohort.
416 In the current study, no significant differences in markers of metabolic health, liver, and
417 kidney function, or moderate to severe AEs in the rapamycin treatment groups were reported
418 compared to placebo after 48 weeks. This is consistent with previous reports summarized in a
419 review by Maier’s group that healthy individuals are not likely to experience serious side effects
420 from low dose rapamycin, and suggests that concerns over negative effects of rapamycin
421 stemming from studies of high-dose, daily usage in chronically ill individuals may have limited
422 applicability in the context of rapamycin use for healthy aging [47]. Indeed, we observed more
423 instances of individuals reporting improvements in chronic ailments in the rapamycin treatment
424 groups than in placebo, and more instances reported of worsening ailments in the placebo
425 group compared to treatment groups after 48 weeks. This effect will be important to
426 investigate further in longitudinal follow-up with PEARL participants.
427 While among PEARL participants, those taking 10 mg of rapamycin per week did initially
428 report more gastrointestinal and neurological AEs (e.g. diarrhea and headaches) compared to
429 those taking placebo, all (except for one) were acute AEs that resolved on their own, or were
430 not severe enough to prevent continuation of the rapamycin regimen. Interestingly, despite
431 well-known impacts of rapamycin on causing mouth sores and a strong association with its
432 bioavailability and potency [46, 48], participants in the placebo group reported increased
433 incidents of mouth sores than those in either of the treatment groups, despite the same
434 number of individuals reporting mouth sores in the 10 mg rapamycin group and placebo (n = 6).
435 This may be a unique property of compounded rapamycin, though follow-up studies are
436 currently underway to understand this more comprehensively.
437 Of note, the PEARL trial was ongoing during the COVID-19 pandemic and rapamycin has
438 been associated with favorable outcomes in response to SARS-CoV2 [8, 49, 50]. Consistent with
439 this, we did observe a decreased incidence of COVID-19 reports in the rapamycin treatment
440 groups (19.7%) compared to placebo (25.6%). Clinical trials are underway that are specifically

16
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

441 designed to more robustly evaluate the efficacy of rapamycin and its derivatives on COVID
442 outcomes [28-32, 51].
443 Over the 48 weeks of the study, weekly rapamycin use demonstrated dose-dependent
444 and sex-specific improvements in multiple functional healthspan metrics compared to placebo,
445 including lean tissue mass, bone mineral content, pain (SF-36, WOMAC), social functioning (SF-
446 36), overall quality of life (SF-36), and overall osteoarthritis (WOMAC) score. All statistically
447 significant benefits were observed in participants taking 10 mg rapamycin per week and
448 consistent with preclinical reports of a female-benefit bias in mice, female participants
449 demonstrated significant benefits across all the outcome measures except bone mineral
450 content [14, 18]. Notably, the most robust improvements observed after 48 weeks were
451 increases in lean muscle mass within this group. However, as all groups of rapamycin users
452 (across gender and dose) had individuals that showed large improvements in body composition
453 measures over time, a summary improvement score accounting for improvement or decline
454 across all DEXA body composition measures was computed as a Benefit Score. Importantly, all
455 women taking 10mg of rapamycin had improvements on the Benefit Score, and showed no
456 declines, whereas men were represented in both groups. Across all participants, those with
457 high benefit scores had significant improvements in gut health measures of gut dysbiosis and
458 immune readiness, and in LDL levels relative to those with low benefit scores. Greater
459 responses on these measures were observed for rapamycin users than placebo users. Taken
460 together, these findings suggest that digestive health, diet, and/or microbiome may play a
461 significant role in rapamycin response and subsequent likelihood of benefits, and will be a focus
462 of our future work.
463 Findings from this study that male participants in the 10 mg rapamycin group gained an
464 average of 1.4% BMC over 48 weeks and female participants in the 10 mg group gained an
465 average LTM of 4.5% hold significant promise for rapamycin in reducing risks of age-related
466 disease and mortality. This is particularly true for the highest rapamycin responding individuals,
467 who exhibit greater improvements in body composition than typical for individuals of this trial
468 age cohort. For example, one female participant in the 10 mg rapamycin group exhibited a 19%
469 improvement in lean muscle mass over the 48 week period, and another saw a 15% increase in

17
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

470 BMC over two years of taking rapamycin consistently during and post-trial. Given the expected
471 worsening of body composition metrics each year after the age of 60, which have a
472 compounding influence on the risk of age-related adverse outcomes, rapamycin clearly has an
473 important role as a gerotherapeutic in the future of healthy longevity medicine
474 While this trial extended notably longer than other human trials of rapamycin use for
475 healthy longevity to date, it is likely that even greater effects would be observed with an
476 increased observation period, a broader (specifically higher) range of doses, as well as a larger
477 study cohort. Further, while it was outside the scope of this study to determine all sources of
478 variability, there are several lifestyle factors (such as diet, alcohol consumption, sleep quality,
479 and activity levels) that may influence VAT, LTM, and BMC changes over time [52, 53], and
480 which may have yet-to-be-identified impacts on rapamycin bioavailability that would add value
481 to future studies on this topic.
482 Across all measures in this study, a remarkable level of variability in response was
483 observed for all rapamycin users, regardless of dose. Given our recent work on the variability of
484 rapamycin bioavailability in individuals, we expect that it played a meaningful role in the results
485 observed here, though this trial concluded prior to our findings on bioavailability. Further, we
486 have discovered since the conclusion of this trial that compounded rapamycin is approximately
487 3.5x less bioavailable than commercially available formulations, suggesting that the 5mg and
488 10mg rapamycin groups received an average equivalent of 1.4mg and 2.9mg respectively [48].
489 Although both doses are relatively low, making the observation of benefits in the treatment
490 group more striking, the 10mg rapamycin cohort in this study was more firmly in the range of
491 what is thought to be an optimal longevity dose range for rapamycin [8, 41, 47, 54]. While
492 future studies will be necessary to draw clear conclusions on whether even higher doses lead to
493 improved effects, we expect that this is a key factor in the often stronger and more reliable
494 patterns of improvement seen for the 10mg rapamycin cohort relative to the 5mg cohort. We
495 highlight this to emphasize the importance of personalized rapamycin dosing and continual
496 routine monitoring of blood rapamycin levels in users to ensure maximal benefits, until such
497 time as the optimal longevity dosing dynamics for rapamycin are more clearly understood.

18
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

498 Beyond physical measures of rapamycin impacts explored in this study, self-reported
499 measures of perceived health and quality of life also showed improvement for all groups. This is
500 especially important in a normative aging population, still in their healthspan, as evaluating
501 changes in QoL captures multiple dimensions of health and functioning that both precede and
502 go beyond disease progression alone. As maintaining an active, pain-free, high quality of life
503 (QoL) in aging is one of the most important aspects of healthspan, findings from the current
504 study that rapamycin users had improved scores on multiple measures of the well-established
505 SF-36 and WOMAC scales are particularly promising. While a female-benefit bias was noted for
506 both measures, significant improvements were observed only for the female 10mg rapamycin
507 group. This is consistent with previous findings in a cohort of 333 adults using off-label
508 rapamycin for healthy aging in which rapamycin was associated with increased reports of
509 happiness, brain function, feelings of youthfulness, calm, reduced anxiety and pain compared
510 to non-users in that same timeframe (up to 10 years for some participants) [8].
511 Of particular note, with PEARL, we are the first to show that 10 mg rapamycin/week for
512 48 weeks has a modest but significant benefit on osteoarthritis symptoms, specifically in
513 women, based on the WOMAC clinical OA assessment. As our study was within a normative
514 aging population, small improvements in early age-related progression of OA could lead to
515 substantial cumulative benefits over time on frailty outcomes [55, 56]. Despite this, follow-up
516 studies are required to determine why males did not demonstrate similar OA associated
517 benefits in this study, and what additional adjustments or considerations would be required for
518 men to derive similar OA related benefits.
519
520 Conclusion
521 The PEARL trial is the largest and longest decentralized RCT evaluating the safety and efficacy of
522 low, intermittent “longevity doses” of rapamycin on healthy aging through the measurement of
523 clinically relevant healthspan metrics. Our findings provide evidence that low dose, intermittent
524 rapamycin regimens are well tolerated and do not lead to any serious adverse events or clinical
525 abnormalities when administered for at least one year within normative aging individuals.
526 Beyond this, we observed benefits for rapamycin users in the PEARL trial, as measured by

19
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

527 significant improvements in lean muscle mass, pain, social functioning, frailty outcomes, and
528 overall quality of life in females, and improved bone mineral density in males. Further, there
529 were some individuals in the rapamycin treatment groups that experienced large
530 improvements in body composition metrics after 48 weeks, which were further associated with
531 beneficial changes in gut health and LDL levels. Future work will more comprehensively explore
532 what drives high response rates to rapamycin, and will aim to identify biometric signatures that
533 can be used to predict clinical effectiveness and inform personalized treatment strategies.
534 As our previous investigation on the bioavailability of rapamycin suggests no significant
535 differences between men and women, further investigation in to mediators of sex specific
536 effects of rapamycin are warranted to elucidate what role factors such as differences in
537 hormone signaling, growth factor signaling, mTOR signaling activity, genetic polymorphisms, or
538 other factors meaningfully contribute to the sex-specific effects we observed in this study, and
539 particularly the female-benefit bias we observed for most measures. While the current report
540 details the primary endpoints for the PEARL trial, many PEARL participants continue to take
541 longevity doses of rapamycin, and have consented to share ongoing information. Reports on
542 these longitudinal studies will be released publicly as available, in keeping with our strong
543 conviction that freely permitting combinatory result analyses will allow for better
544 understanding of optimal dosing, safety, changes in aging biomarkers, and overlapping
545 outcomes for this important gerotherapeutic intervention.
546 Collectively, the PEARL clinical trial is one of the first to provide evidence that low-dose
547 rapamycin may be safely administered for extended periods of time and may counteract
548 several aspects of age-related decline. Further, it represents the first study indicating dose-
549 dependent, sex specific efficacy for rapamycin in improving healthspan metrics in humans. As
550 such, our collective evidence suggests that well beyond merely clinical measures of health
551 improvements, rapamycin may promote essential, comprehensive well-being associated with
552 “adding life to years, not just years to life”.
553
554
555

20
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

556 Statements and Declarations:

557
558 Acknowledgements
559 The authors would like to thank the participants who took part in this study. Financial support
560 from our generous donors, particularly Vitalik Buterin, Micah Zoltu, Brad Armstrong, and
561 anonymous gifts, administrative support, and article publishing charges were provided by
562 AgelessRx. Clinical trial oversight was conducted by PHAGE corporation. We extend special
563 thanks to Mikhail Blagosklonny, James Watson, Alan Green and Thorne for their participation
564 and support in conducting this trial.
565
566 Competing interests
567 GH, VL, AN, MM, SM, AI, and SZ are employees and shareholders of AgelessRx. JH has received
568 financial compensation from AgelessRx for their contributions.
569
570 Author contributions
571 Virginia Lee, Andy Nyquist, Anar Isman, and Sajad Zalzala designed and implemented the study.
572 Girish Harinath, Jesper Hagemeier, and Stefanie Morgan performed data analysis. Girish
573 Harinath, Virginia Lee, Andy Nyquist, Mauricio Moel, Stefanie Morgan, Anar Isman, and Sajad
574 Zalzala wrote and edited the manuscript. All work was supervised by Stefanie Morgan, Anar
575 Isman, and Sajad Zalzala. Corresponding author is Stefanie Morgan.
576
577 References
578 1. Al-Maskari F. LIFESTYLE DISEASES: An Economic Burden on the Health Services. UN
579 Chronicle [Internet]. [cited 2024 August]. Available from:
580 https://www.un.org/en/chronicle/article/lifestyle-diseases-economic-burden-health-
581 services#:~:text=By%202030%2C%20the%20proportion%20of,African%20and%20Eastern%20M
582 editerranean%20regions.
583 2. Ford ES, Bergmann Mm Fau - Kröger J, Kröger J Fau - Schienkiewitz A, Schienkiewitz A
584 Fau - Weikert C, Weikert C Fau - Boeing H, Boeing H. Healthy living is the best revenge: findings

21
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

585 from the European Prospective Investigation Into Cancer and Nutrition-Potsdam study. (1538-
586 3679 (Electronic))
587 3. Kvaavik E, Batty Gd Fau - Ursin G, Ursin G Fau - Huxley R, Huxley R Fau - Gale CR, Gale
588 CR. Influence of individual and combined health behaviors on total and cause-specific mortality
589 in men and women: the United Kingdom health and lifestyle survey. (1538-3679 (Electronic))
590 4. Coe E, Dewhurst M, Hartenstein L, Hextall A, Latkovic T. At least six years of higher
591 quality life for everyone is within reach. Adding years to life and life to years [Internet]. 2022
592 [cited 2024 August]. Available from: https://www.mckinsey.com/mhi/our-insights/adding-
593 years-to-life-and-life-to-years.
594 5. Garmany A, Yamada S, Terzic A. Longevity leap: mind the healthspan gap. NPJ Regen
595 Med. 2021;6(1):57 https://doi.org/10.1038/s41536-021-00169-5.
596 6. Sierra F. The Emergence of Geroscience as an Interdisciplinary Approach to the
597 Enhancement of Health Span and Life Span. (2157-1422 (Electronic))
598 7. Kulkarni AS, Aleksic S, Berger DM, Sierra F, Kuchel GA, Barzilai N. Geroscience-guided
599 repurposing of FDA-approved drugs to target aging: a proposed process and prioritization.
600 Aging Cell. 2022;21(4):e13596 https://doi.org/10.1111/acel.13596.
601 8. Kaeberlein TL, Green AS, Haddad G, Hudson J, Isman A, Nyquist A, et al. Evaluation of
602 off-label rapamycin use to promote healthspan in 333 adults. Geroscience. 2023;45(5):2757-68
603 https://doi.org/10.1007/s11357-023-00818-1.
604 9. Mannick JB, Lamming DW. Targeting the biology of aging with mTOR inhibitors. Nat
605 Aging. 2023;3(6):642-60 https://doi.org/10.1038/s43587-023-00416-y.
606 10. Chung J, Kuo Cj Fau - Crabtree GR, Crabtree Gr Fau - Blenis J, Blenis J. Rapamycin-FKBP
607 specifically blocks growth-dependent activation of and signaling by the 70 kd S6 protein
608 kinases. (0092-8674 (Print))
609 11. Castilho RM, Squarize Ch Fau - Chodosh LA, Chodosh La Fau - Williams BO, Williams Bo
610 Fau - Gutkind JS, Gutkind JS. mTOR mediates Wnt-induced epidermal stem cell exhaustion and
611 aging. (1875-9777 (Electronic))

22
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

612 12. Urfer SR, Kaeberlein TL, Mailheau S, Bergman PJ, Creevy KE, Promislow DEL, et al. A
613 randomized controlled trial to establish effects of short-term rapamycin treatment in 24
614 middle-aged companion dogs. GeroScience. 2017;39(2):117-27 10.1007/s11357-017-9972-z.
615 13. Robida-Stubbs S, Glover-Cutter K Fau - Lamming DW, Lamming Dw Fau - Mizunuma M,
616 Mizunuma M Fau - Narasimhan SD, Narasimhan Sd Fau - Neumann-Haefelin E, Neumann-
617 Haefelin E Fau - Sabatini DM, et al. TOR signaling and rapamycin influence longevity by
618 regulating SKN-1/Nrf and DAF-16/FoxO. (1932-7420 (Electronic))
619 14. Miller RA, Harrison De Fau - Astle CM, Astle Cm Fau - Fernandez E, Fernandez E Fau -
620 Flurkey K, Flurkey K Fau - Han M, Han M Fau - Javors MA, et al. Rapamycin-mediated lifespan
621 increase in mice is dose and sex dependent and metabolically distinct from dietary restriction.
622 (1474-9726 (Electronic))
623 15. Juricic P, Lu Y-X, Leech T, Drews LF, Paulitz J, Lu J, et al. Long-lasting geroprotection from
624 brief rapamycin treatment in early adulthood by persistently increased intestinal autophagy.
625 Nature Aging. 2022;2(9):824-36 10.1038/s43587-022-00278-w.
626 16. Bitto A, Ito TK, Pineda VV, LeTexier NJ, Huang HZ, Sutlief E, et al. Transient rapamycin
627 treatment can increase lifespan and healthspan in middle-aged mice. eLife. 2016;5:e16351
628 10.7554/eLife.16351.
629 17. Barnett BG, Wesselowski SR, Gordon SG, Saunders AB, Promislow DEL, Schwartz SM, et
630 al. A masked, placebo-controlled, randomized clinical trial evaluating safety and the effect on
631 cardiac function of low-dose rapamycin in 17 healthy client-owned dogs. (2297-1769 (Print))
632 18. Arriola Apelo SI, Pumper CP, Baar EL, Cummings NE, Lamming DW. Intermittent
633 Administration of Rapamycin Extends the Life Span of Female C57BL/6J Mice. (1758-535X
634 (Electronic))
635 19. Powers RW, 3rd, Kaeberlein M Fau - Caldwell SD, Caldwell Sd Fau - Kennedy BK,
636 Kennedy Bk Fau - Fields S, Fields S. Extension of chronological life span in yeast by decreased
637 TOR pathway signaling. (0890-9369 (Print))
638 20. Colman RJ, Anderson Rm Fau - Johnson SC, Johnson Sc Fau - Kastman EK, Kastman Ek
639 Fau - Kosmatka KJ, Kosmatka Kj Fau - Beasley TM, Beasley Tm Fau - Allison DB, et al. Caloric
640 restriction delays disease onset and mortality in rhesus monkeys. (1095-9203 (Electronic))

23
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

641 21. Harrison DE, Strong R Fau - Sharp ZD, Sharp Zd Fau - Nelson JF, Nelson Jf Fau - Astle CM,
642 Astle Cm Fau - Flurkey K, Flurkey K Fau - Nadon NL, et al. Rapamycin fed late in life extends
643 lifespan in genetically heterogeneous mice. (1476-4687 (Electronic))
644 22. Miller RA, Harrison DE, Astle CM, Baur JA, Boyd AR, de Cabo R, et al. Rapamycin, But Not
645 Resveratrol or Simvastatin, Extends Life Span of Genetically Heterogeneous Mice. The Journals
646 of Gerontology: Series A. 2011;66A(2):191-201 10.1093/gerona/glq178.
647 23. Flynn JM, O'Leary Mn Fau - Zambataro CA, Zambataro Ca Fau - Academia EC, Academia
648 Ec Fau - Presley MP, Presley Mp Fau - Garrett BJ, Garrett Bj Fau - Zykovich A, et al. Late-life
649 rapamycin treatment reverses age-related heart dysfunction. (1474-9726 (Electronic))
650 24. Selvarani R, Mohammed S, Richardson A. Effect of rapamycin on aging and age-related
651 diseases-past and future. Geroscience. 2021;43(3):1135-58 https://doi.org/10.1007/s11357-
652 020-00274-1.
653 25. Zhang Y, Zhang J, Wang S. The role of rapamycin in healthspan extension via the delay of
654 organ aging. Ageing Res Rev. 2021;70:101376 https://doi.org/10.1016/j.arr.2021.101376.
655 26. RapAdmin. Breaking: 15% Healthy Lifespan improvement via Rapamycin seen in
656 Marmosets 2024 [Available from: https://www.rapamycin.news/t/breaking-15-healthy-
657 lifespan-improvement-via-rapamycin-seen-in-marmosets/14325.
658 27. Weston AD, Grossardt BR, Garner HW, Kline TL, Chamberlain AM, Allen AM, et al.
659 Abdominal body composition reference ranges and association with chronic conditions in an
660 age- and sex-stratified representative sample of a geographically defined American population.
661 J Gerontol A Biol Sci Med Sci. 2024;79(4) https://doi.org/10.1093/gerona/glae055.
662 28. Hunter GR, Gower BA, Kane BL. Age related shift in visceral fat. Int J Body Compos Res.
663 2010;8(3):103-8
664 29. Volpi E, Nazemi R, Fujita S. Muscle tissue changes with aging. Curr Opin Clin Nutr Metab
665 Care. 2004;7(4):405-10 https://doi.org/10.1097/01.mco.0000134362.76653.b2.
666 30. Whitaker KM, Choh AC, Lee M, Towne B, Czerwinski SA, Demerath EW. Sex differences
667 in the rate of abdominal adipose accrual during adulthood: the Fels Longitudinal Study. Int J
668 Obes (Lond). 2016;40(8):1278-85 https://doi.org/10.1038/ijo.2016.48.

24
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

669 31. Berger C, Langsetmo L, Joseph L, Hanley DA, Davison KS, Josse R, et al. Change in bone
670 mineral density as a function of age in women and men and association with the use of
671 antiresorptive agents. CMAJ. 2008;178(13):1660-8 https://doi.org/10.1503/cmaj.071416.
672 32. Alswat KA. Gender disparities in osteoporosis. J Clin Med Res. 2017;9(5):382-7
673 https://doi.org/10.14740/jocmr2970w.
674 33. Mikkola TM, Kautiainen H, von Bonsdorff MB, Salonen MK, Wasenius N, Kajantie E, et al.
675 Body composition and changes in health-related quality of life in older age: a 10-year follow-up
676 of the Helsinki Birth Cohort Study. Qual Life Res. 2020;29(8):2039-50
677 https://doi.org/10.1007/s11136-020-02453-1.
678 34. Chlabicz M, Dubatowka M, Jamiolkowski J, Sowa P, Lapinska M, Raczkowski A, et al.
679 Subjective well-being in non-obese individuals depends strongly on body composition. Sci Rep.
680 2021;11(1):21797 https://doi.org/10.1038/s41598-021-01205-6.
681 35. Lin H, Salech F, Lim A, Vogrin S, Duque G. The effect of rapamycin and its analogues on
682 age-related musculoskeletal diseases: a systematic review. Aging Clin Exp Res.
683 2022;34(10):2317-33 https://doi.org/10.1007/s40520-022-02190-0.
684 36. den Hartigh LJ, Goodspeed L, Wang SA, Kenerson HL, Omer M, O'Brien KD, et al. Chronic
685 oral rapamycin decreases adiposity, hepatic triglycerides and insulin resistance in male mice fed
686 a diet high in sucrose and saturated fat. Exp Physiol. 2018;103(11):1469-80
687 https://doi.org/10.1113/EP087207.
688 37. Konopka AR, Lamming DW, Investigators RP, Investigators E. Blazing a trail for the
689 clinical use of rapamycin as a geroprotecTOR. Geroscience. 2023;45(5):2769-83
690 https://doi.org/10.1007/s11357-023-00935-x.
691 38. Mannick JB, Del Giudice G, Lattanzi M, Valiante NM, Praestgaard J, Huang B, et al. mTOR
692 inhibition improves immune function in the elderly. (1946-6242 (Electronic))
693 39. Mannick JB, Morris M, Hockey H-UP, Roma G, Beibel M, Kulmatycki K, et al. TORC1
694 inhibition enhances immune function and reduces infections in the elderly. Science
695 Translational Medicine. 2018;10(449):eaaq1564 10.1126/scitranslmed.aaq1564.

25
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

696 40. Mannick JB, Teo G, Bernardo P, Quinn D, Russell K, Klickstein L, et al. Targeting the
697 biology of ageing with mTOR inhibitors to improve immune function in older adults: phase 2b
698 and phase 3 randomised trials. (2666-7568 (Electronic))
699 41. Mannick JB, Morris M, Hockey HP, Roma G, Beibel M, Kulmatycki K, et al. TORC1
700 inhibition enhances immune function and reduces infections in the elderly. Sci Transl Med.
701 2018;10(449) https://doi.org/10.1126/scitranslmed.aaq1564.
702 42. Walters HE, Cox LS. mTORC inhibitors as broad-spectrum therapeutics for age-related
703 diseases. Int J Mol Sci. 2018;19(8) https://doi.org/10.3390/ijms19082325.
704 43. Viscomi MT, D’Amelio M, Cavallucci V, Latini L, Bisicchia E, Nazio F, et al. Stimulation of
705 autophagy by rapamycin protects neurons from remote degeneration after acute focal brain
706 damage. Autophagy. 2012;8(2):222-35 10.4161/auto.8.2.18599.
707 44. Ware JE, Jr., Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I.
708 Conceptual framework and item selection. (0025-7079 (Print))
709 45. Riddle DL, Perera RA. The WOMAC pain scale and crosstalk from co-occurring pain sites
710 in people with knee pain: A causal modeling study. Phys Ther. 2020;100(10):1872-81
711 https://doi.org/10.1093/ptj/pzaa098.
712 46. Peterson DE, O'Shaughnessy JA, Rugo HS, Elad S, Schubert MM, Viet CT, et al. Oral
713 mucosal injury caused by mammalian target of rapamycin inhibitors: emerging perspectives on
714 pathobiology and impact on clinical practice. Cancer Med. 2016;5(8):1897-907
715 https://doi.org/10.1002/cam4.761.
716 47. Lee DJW, Hodzic Kuerec A, Maier AB. Targeting ageing with rapamycin and its
717 derivatives in humans: a systematic review. Lancet Healthy Longev. 2024;5(2):e152-e62
718 https://doi.org/10.1016/S2666-7568(23)00258-1.
719 48. Harinath G, Lee V, Nyquist A, Moel M, Wouters M, Hagemeier J, et al. The bioavailability
720 of compounded and generic rapamycin in normative aging individuals: A retrospective study
721 and review with clinical implications. MedRxiv. 2024
722 https://doi.org/10.1101/2024.08.12.24311432.

26
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

723 49. Patocka J, Kuca K, Oleksak P, Nepovimova E, Valis M, Novotny M, et al. Rapamycin: Drug
724 Repurposing in SARS-CoV-2 Infection. Pharmaceuticals (Basel). 2021;14(3)
725 https://doi.org/10.3390/ph14030217.
726 50. Singla A, Harun N, Dilling DF, Merchant K, McMahan S, Ingledue R, et al. Safety and
727 efficacy of sirolimus in hospitalised patients with COVID-19 pneumonia. Respir Investig.
728 2024;62(2):216-22 https://doi.org/10.1016/j.resinv.2023.12.009.
729 51. Bischof E, Siow RC, Zhavoronkov A, Kaeberlein M. The potential of rapalogs to enhance
730 resilience against SARS-CoV-2 infection and reduce the severity of COVID-19. Lancet Healthy
731 Longev. 2021;2(2):e105-e11 https://doi.org/10.1016/S2666-7568(20)30068-4.
732 52. Lv X, Cai J, Li X, Wang X, Ma H, Heianza Y, et al. Body composition, lifestyle, and
733 depression: a prospective study in the UK biobank. BMC Public Health. 2024;24(1):393
734 https://doi.org/10.1186/s12889-024-17891-6.
735 53. United States. Public Health Service. Office of the Surgeon General. 6 Determinants of
736 bone health. Bone health and osteoporosis : a report of the Surgeon General. Rockville, Md.:
737 U.S. Dept. of Health and Human Services, Public Health Service, Office of the Surgeon General;
738 2004. p. xxxii, 404 p.
739 54. Blagosklonny MV. Rapamycin for longevity: opinion article. Aging (Albany NY).
740 2019;11(19):8048-67 https://doi.org/10.18632/aging.102355.
741 55. Puts MT, Toubasi S, Atkinson E, Ayala AP, Andrew M, Ashe MC, et al. Interventions to
742 prevent or reduce the level of frailty in community-dwelling older adults: a protocol for a
743 scoping review of the literature and international policies. BMJ Open. 2016;6(3):e010959
744 https://doi.org/10.1136/bmjopen-2015-010959.
745 56. Apostolo J, Cooke R, Bobrowicz-Campos E, Santana S, Marcucci M, Cano A, et al.
746 Effectiveness of interventions to prevent pre-frailty and frailty progression in older adults: a
747 systematic review. JBI Database System Rev Implement Rep. 2018;16(1):140-232
748 https://doi.org/10.11124/JBISRIR-2017-003382.
749
750
751

27
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

752 Figure Legends

753 Figure 1. Effects of rapamycin use on changes in measures of body composition. Notable
754 trends of change in body composition from baseline to 48 weeks were observed between
755 treatment groups for all subjects (a). Analysis of changes in body composition measures within
756 sex revealed that bone mineral content improved significantly for men taking 10mg rapamycin
757 (b) as did lean tissue mass for females taking 10mg rapamycin (c). *p <0.05, **p ≤ 0.005
758

759 Figure 2. Heterogeneity in rapamycin response reveals differential patterns of effects. Within
760 all study arms, individual participants showed marked improvement and decline for each
761 measure of body composition (a). A greater percentage of individuals in the improvement
762 group were rapamycin users (b), with no specific bias in response across all measures by sex (c).
763 Similar trends were observed for measures of VAT (d), however, BMC (e), and LTM (f) showed
764 sex-specific improvement and decline.
765
766 Figure 3. Rapamycin response categorized by Benefit Scores revealed biological change
767 differences by response type. Improvements in VAT correlated with improvements on BMC
768 and LTM, as well as with improved weight and BMI (a). A “Benefit Score” was computed for
769 users based on summary scores of improvement or decline on each body composition measure
770 evaluated by DEXA, which was significantly associated with rapamycin use (b). Separation of
771 Benefit Scores by rapamycin dose and sex revealed a pronounced female benefit bias, with only
772 improvement for all female rapamycin users (c). Subsequent analysis of gut health and blood
773 biomarkers by benefit score revealed significant changes in measures of gut dysbiosis, immune
774 readiness (d), and LDL levels (e) between high and low Benefit Scores, with greater magnitude
775 of beneficial change for rapamycin users than placebo.
776

777 Figure 4. Effects of rapamycin use on quality of life and frailty. Quality of life measures from
778 the SF-36 showed improvements across most measures for female rapamycin users (a), with
779 statistically significant improvements in scores of pain, social functioning, and overall quality of
780 life. Male rapamycin users showed only trends of improvement for many measures, particularly

28
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

781 for the 5mg cohort (b). Similar patterns were seen for frailty measures from the WOMAC scale,
782 with female rapamycin users showing improvements across most measures, and significant
783 improvements in pain and overall frailty (c). In contrast, male rapamycin users did not show
784 improvements on WOMAC scores relative to controls (d).
785

786 Supplementary Figure S1. PEARL trial design and participant health scores. Schematic of trial
787 enrollment, participant screening, randomization, and followup (a), and a summary of all
788 participants health scores (b) highlighting significant baseline health of participants in this
789 study.
790
791 Supplementary Figure S2. Adverse events and types for PEARL participants. A summary of
792 adverse event types by trial arm showed no rapamycin-specific adverse event patterns (a).
793 Further investigation of adverse event types revealed similar effects for all trial arms, with the
794 exception of a higher proportion of gastrointestinal events reported for rapamycin users than
795 controls (b).
796
797 Supplementary Figure S3. Effects of rapamycin use on sex-specific changes in measures of
798 body composition over time. Change in body composition measures at the study midpoint (24
799 weeks) showed evidence of some measure of change (a) that largely stabilized to the patterns
800 reported after 48 weeks. Change at 48 weeks of VAT, and of BMC in females and LTM in males
801 did not show statistically significant improvements for rapamycin users relative to controls (b).
802
803 Supplementary Figure S4. Heterogeneity of rapamycin response in body composition
804 measures. Within trial arms, notable differences in participant responses were observed across
805 all measures, with sex-specific differences in proportion of participants improving or declining
806 for most measures.
807
808 Supplementary Figure S5. Changes in blood biomarkers in response to rapamycin use. While
809 no significant differences in biomarkers were outside the healthy reference range after 48

29
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

810 weeks, significant differences between treatment and control groups were observed for mean
811 change in corpuscular hemoglobin (MCH) and mean platelet volume (MPV), and for the 10 mg
812 rapamycin group in chloride levels, carbon dioxide levels, and mean corpuscular volume (MCV),
813 and patients in the 5 mg rapamycin treatment group in platelets (a). No statistically significant
814 changes were seen for other biomarkers (b).
815
816 Supplementary Figure S6. Effects of rapamycin use on quality of life and frailty measures
817 across groups. After 48 weeks of treatment most SF-36 scores of perceived Quality of Life
818 improved in all groups (a), as did WOMAC scores of frailty (b).
819
820
821

822
823
824
825
826
827

828
829
830
831
832

833
834

835
836
837
838

30
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

839 Tables

Table 1: Baseline demographics and blood safety markers

5 mg 10 mg Placebo F(df1, df2) p-value
N=40 N=36 N=39
Mean (SD) Mean (SD) Mean (SD)
Age (years) 61.2 (7.6) 62.4 (9.1) 62.6 (6.6) 0.425(2, 112) 0.655
Sex (Female) N (%) 17 (42.5) 9 (25.0) 15 (38.5) χ2=2.732(2) 0.255
Height (cm) 171.9 (10.0) 176.2 (9.0) 175.0 (9.1) 1.940(2, 101) 0.149
Weight (kg) 75.6 (15.3) 80.1 (13.6) 75.2 (14.5) 1.263(2, 105) 0.287
BMI (kg/m2) 25.2 (3.5) 25.6 (3.3) 24.3 (3.0) 1.407 (2, 100) 0.250
Glucose (mg/dL) 90.9 (3.5) 92.1 (8.8) 91.5 (11.1) 0.160(2, 112) 0.852
LDL cholesterol (mg/dL) 107.8 (34.4) 104.3 (39.5) 115.1 (32.4) 0.921(2, 112) 0.401
Non-HDL cholesterol 119.4 (37.4) 121.6 (43.5) 134.2 (34.5) 1.683(2, 112) 0.191
(mg/dL)
Triglycerides (mg/dL) 86.2 (32.4) 90.0 (39.8) 92.8 (45.8) 0.280(2, 112) 0.756
HbA1c (%) 5.2 (0.2) 5.3 (0.2) 5.3 (0.3) 3.418(2, 106) 0.036*
Insulin (uIU/mL) 5.8 (5.2) 6.1 (4.2) 5.0 (3.2) 0.528(2, 110) 0.528
ALT (U/L) 22.1 (10.8) 23.4 (9.0) 22.4 (12.9) 0.141(2, 112) 0.869
AST (U/L) 22.7 (7.2) 23.4 (5.3) 22.6 (6.7) 0.154(2, 112) 0.857
BUN (mg/dL) 16.8 (4.9) 15.8 (2.9) 16.3 (4.4) 0.487(2, 112) 0.616
Creatinine (mg/dL) 0.9 (0.2) 1.0 (1.2) 1.0 (0.3) 1.045(2, 112) 0.355
eGFR (mL/min/1.73m2) 85.4 (14.1) 83.0 (10.8) 79.1 (15.3) 2.007(2, 108) 0.139
ALT: alanine aminotransferase; AST: aspartate aminotransferase; BMI: body mass index; BUN: blood urea
nitrogen; eGFR: estimated glomerular filtration rate; HbA1c: hemoglobin A1c; HDL: high-density lipoprotein;
LDL: low-density lipoprotein; SD: standard deviation.
Groups were compared using Analysis of Variance and post-hoc independent t-test. *p < 0.05.

840

31
medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Table 2: Self-reported adverse events during study period

5 mg 10 mg Placebo
Cardiovascular
-Cardiological event - - 1 (2.6%) 1
-Discomfort/Tightness in chest 1 (2.5%) 1 - -
Gastrointestinal
-Lower gastrointestinal symptoms 4 (10%) 5 6 (16.7%) 9 3 (7.7%) 3
-Upper and lower gastrointestinal symptoms 3 (7.5%) 3 1 (2.8%) 2 1 (2.6%) 1
-Upper gastrointestinal symptoms 2 (5%) 5 3 (8.3%) 4 -
Musculoskeletal and Orthopedic
-Musculoskeletal pain 10 (25%) 19 12 (33.3%) 20 11 (28.2%) 20
-Deep tissue injury (closed) 1 (2.5%) 2 - 1 (2.6%) 1
-Musculoskeletal/Orthopedic injury (closed) 1 (2.5%) 1 - 1 (2.6%) 1
Neurological
-Altered spatial orientation 1 (2.5%) 1 3 (8.3%) 7 2 (5.1%) 3
-Cognitive decline - 1 (2.8%) 1 -
-Neuromuscular symptoms - 1 (2.8%) 1 2 (5.1%) 2
Oral and Dental
-Mouth sores 3 (7.5%) 4 6 (16.7%) 14 6 (15.4%) 27
-Dental symptoms 2 (5%) 3 2 (5.6%) 2 1 (2.6%) 1
Respiratory
-Cold/Flu/Sinus symptoms 23 (57.5%) 35 19 (52.8%) 29 22 (56.4%) 33
-Respiratory symptoms 1 (2.5%) 2 - -
Sensory
-Auditory perception symptoms 2 (5%) 2 3 (8.3%) 3 2 (5.1%) 2
-Visual perception symptoms 1 (2.5%) 1 1 (2.8%) 1 -
Skin and Wound
-Superficial wound/Skin irritation 6 (15%) 8 7 (19.4%) 11 9 (23.1%) 12
-Deep wound (opened) - - 2 (5.1%) 2
-Skin bacterial/fungal infection 1 (2.5%) 1 - 2 (5.1% ) 2
Urinary, Kidney, and Reproductive
-Kidney/Urinary symptoms 4 (10%) 5 2 (5.6%) 2 2 (5.1%) 2

32
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

-Menopausal symptoms 1 (2.5%) 1 1 (2.8%) 1 2 (5.1%) 2

Other
-Erectile dysfunction - 1 (2.8%) 1 -
-Malaise 3 (7.5%) 11 2 (5.6%) 2 1 (2.6%) 1
-Eye infection/disorder 1 (2.5%) 1 1 (2.8%) 1 1 (2.6%) 1
-Mental Health Symptoms 1 (2.5%) 1 1 (2.8%) 1 1 (2.6%) 1
-Pain (not otherwise specified) 1 (2.5%) 1 1 (2.8%) 1 1 (2.6%) 1
-Sleep disturbances - 1 (2.8%) 1 -
Chronic Issue Flare 12 (30%) 23 13 (36.1%) 28 10 (25.6%) 14
Chronic Issue Severity 4 (10%) 4 4 (11.1%) 6 4 (10.3%) 12
*
Includes all reported events by the cohort. Some participants had multiple events or experienced
the same events multiple times. Adverse events reported as: n participants (%) n events. Chronic
issue flare: participants entered the PEARL trial with a chronic ailment which improved or stayed the
same during the trial. Chronic issue severity: participants entered the PEARL trial with a chronic
ailment that worsened during the trial.
841
842
843
844
845

846
847
848
849
850

851
852
853
854

855
856
857

33
medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Table 3: Dexa baseline output and change after 48 weeks, stratified by sex

Baseline

Values for Females, Males Changes by sex Changes by group

5 mg 10 mg Placebo F(df1, df2)† p-value† F(df1, df2) ‡ p-value ‡

500.6 407.0 518.2

F(2, 37) = 0.231 0.795
Visceral adipose (449.7) (346.2) (321.5) F(5, 107) =
< 0.001
tissue mass (g) 1321.2 1203.7 1084.3 5.426
F(2, 70) = 0.470 0.627
(995.0) (810.4) (690.4)

2174.4 2422.7 2196.6

F(2, 35) = 1.885 0.167
Bone mineral (245.6) (490.6) (239.8) F(5, 101) =
< 0.001
content (g) 3059.9 3048.9 3243.7 22.304
F(2, 66) = 1.242 0.295
(492.9) (397.4) (529.9)

39313.1 41231.8 38983.8

F(2 ,37) = 0.488 0.618
Lean tissue (5098.6) (8112.7) (3769.5) F(5, 106) =
< 0.001
mass (g) 58290.8 58864.5 59950.8 44.559
F(2, 69) = 0.317 0.729
(6084.4) (7178.5) (8154.6)

1.1210 1.1951 1.1035

F(2, 35) = 1.537 0.229
Bone Mineral (0.1182) (0.1799) (0.0838) F(5, 103) =
< 0.001
Density (g/cm2) 1.2558 1.3006 1.3397 8.983
F(68) = 1.890 0.159
(0.1750) (0.1228) (0.1341)

Change from baseline – 48 weeks

-29.3 -6.1 F(2,34.546) =

50.7 (69.3) 0.287
Visceral adipose (170.6) (116.2) 1.294 F(2, 203.934)
0.427
tissue mass (g) -48.6 19.2 26.9 F(2. 127.416) = = 0.854
0.72
(201.0) (277.4) (362.6) 0.329

34
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

-50.6 F(2, 43.111) =

-19.8 (64.4) -23.1 (82.0) 0.779
Bone mineral (106.8) 0.251 F(2, 193.245)
0.012
content (g) 45.2 F(2, 70.240) = = 4.489
-41.7 (95.2) -0.7 (100.6) 0.113
(93.3) 2.249

-60.6 1939.7 -298.1 F(1,70.369) =

0.004
Lean tissue (1613.8) (1998.9) (1191.1) 1.022 F(2, 206.447)
0.005
mass (g) -68.6 -6.0 -525.1 F(2, 80.616) = = 5.439
0.894
(3034.7) (2421.2) (2138.6) 0.112

-0.0114 -0.0100 -0.0021 F(2,43.316) =

0.243
Bone Mineral (0.0195) (0.0325) (0.0332) 1.461 F(2, 163.750)
0.191
Density (g/cm2) 0.0463 0.0046 -0.0012 F(2, 109.733) = = 1.673
0.369
(0.2108) (0.0361) (0.0267) 1.006

858 Values provided as mean (standard deviation). Groups were compared using Analysis of Variance (baseline)
859 and linear mixed effects model (change from week 0 through 48) by sex and group.
860

861

35
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Table 4: Baseline SF-36 and WOMAC and absolute change at 48 weeks, stratified by sex
Women Men
5 mg 10 mg Placebo 5 mg 10 mg Placebo
Baseline
SF-36
- Physical functioning 95.3 (6.2) 96.1 (4.2) 93.0 (10.7) 94.6 (8.3) 95.6 (9.8) 95.5 (8.7)
94.4
- Physical role limitations 92.7 (14.7) 83.3 (32.3) 94.6 (21.3) 95.4 (12.1) 90.9 (23.8)
(11.0)
- Emotional role 100.0
90.2 (22.9) 97.8 (8.6) 85.5 (26.3) 97.5 (8.9) 89.4 (23.9)
limitations (0.0)
65.6
- Energy/fatigue 62.4 (21.3) 67.7 (22.4) 66.5 (16.9) 73.5 (12.7) 69.6 (13.3)
(14.5)
76.9
- Emotional well-being 76.5 (14.6) 82.4 (10.1) 75.7 (14.3) 85.9 (8.7) 80.4 (11.5)
(12.0)
84.7
- Social functioning 92.7 (11.7) 92.5 (11.4) 92.4 (15.0) 96.3 (8.4) 94.9 (10.7)
(27.1)
74.4
- Pain 82.9 (9.9) 82.3 (18.5) 84.7 (14.7) 87.7 (13.1) 79.4 (18.0)
(20.3)
75.6
- General health 77.4 (19.5) 76.3 (21.2) 75.7 (15.0) 85.0 (15.3) 76.1 (18.3)
(17.9)
- Physical health score 87.1 (9.5) 85.1 (9.9) 83.8 (18.0) 87.4 (12.0) 90.9 (7.8) 85.5 (14.1)
- Mental health score 80.4 (12.1) 81.8 (8.4) 85.1 (10.5) 80.0 (14.0) 88.3 (6.9) 83.6 (11.9)
- Aggregate score 83.7 (10.0) 83.5 (7.3) 84.4 (13.5) 83.7 (10.9) 89.6 (6.3) 84.5 (11.3)
WOMAC
- Total score (0-96) 28.6 (5.5) 28.7 (6.3) 31.9 (12.6) 29.0 (6.7) 28.2 (8.3) 30.4 (8.1)
- Pain score (0-20) 6.4 (1.8) 6.4 (2.8) 6.7 (2.4) 6.5 (2.2) 6.2 (2.3) 6.1 (1.6)
- Stiffness score (0-8) 3.4 (1.3) 3.2 (1.1) 3.5 (1.6) 2.8 (1.1) 2.7 (1.1) 3.5 (1.2)
- Physical function score
18.8 (3.3) 19.0 (4.6) 21.7 (9.0) 19.7 (4.1) 19.3 (5.4) 20.8 (5.9)
(0-68)

Change baseline - 48
weeks

36
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

SF-36
- Physical functioning +0.6 (4.8) 0.0 (6.0) +0.4 (4.1) -3.3 (15.3) -1.6 (6.3) -0.5 (9.7)
- Physical role limitations +3.1 (15.5) +3.1 (8.8) +5.4 (35.6) -7.1 (28.7) -4.0 (22.5) +0.0 (30.9)
- Emotional role
+4.2 (16.7) -4.2 (11.8) +2.4 (8.9) +4.8 (33.8) 0.0 (9.6) +9.3 (25.1)
limitations
+9.4
- Energy/fatigue +10.0 (17.4) +2.9 (13.7) +3.0 (12.7) +2.6 (8.8) +0.3 (9.0)
(12.9)
- Emotional well-being +5.8 (14.2) +5.0 (9.5) +4.6 (9.1) +5.2 (10.4) +0.8 (5.8) +3.3 (7.6)
+12.5
- Social functioning +3.9 (10.9) +1.8 (12.8) +2.5 (15.5) +1.0 (11.4) +2.1 (8.8)
(24.1)
+14.7
- Pain +7.0 (11.9) +1.8 (9.1) -1.2 (17.8) -0.9 (8.0) +1.5 (15.2)
(13.9)
+5.6
- General health +5.3 (12.3) +3.9 (10.2) +5.8 (8.8) +2.4 (9.4) -1.7 (13.0)
(12.1)
- Physical health score +4.0 (8.0) +5.9 (5.7) +2.9 (11.5) -2.4 (17.9) -1.0 (7.6) -0.2 (13.7)
- Mental health score +6.0 (11.4) +5.7 (5.9) +2.9 (7.7) +4.8 (13.5) +1.1 (5.3) +3.7 (8.9)
- Aggregate score +5.0 (8.9) +5.8 (4.6) +2.9 (7.8) +2.9 (10.3) +0.0 (5.6) +1.8 (8.7)
WOMAC
- Total score (0-96) -2.3 (5.5) -3.0 (6.2) -1.1 (5.1) -1.0 (4.7) +2.0 (7.8) -1.1 (6.6)
- Pain score (0-20) -0.9 (1.8) -0.8 (2.9) +0.4 (2.7) -0.4 (1.7) +0.6 (2.4) -0.3 (1.1)
- Stiffness score (0-8) -0.7 (1.3) -0.6 (1.2) -0.5 (0.9) -0.2 (1.1) +0.2 (1.0) -0.1 (1.2)
- Physical function score
-0.6 (3.4) -1.6 (5.2) -1.0 (3.3) -0.5 (2.7) +1.2 (5.3) -0.7 (5.6)
(0-68)
SF-36: short form 36; WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index. Values provided as
mean (standard deviation).

862
863
864

865
866

867

37
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Supplementary Table S1: Week 48 safety-related blood work assessment

5 mg 10 mg Placebo F(df1, df2) p-value
N=38 N=33 N=34
Mean (SD) * Mean (SD) * Mean (SD) *

Glucose (mg/dL) 91.3 (9.6) 91.8 (8.1) 89.7 (9.5) F(2, 101) = 0.489 0.614

LDL cholesterol (mg/dL) 107.9 (34.8) 96.2 (34.1) 103.5 (30.7) F(2, 102) = 1.095 0.339

Non-HDL cholesterol (mg/dL) 126.0 (37.6) 116.1 (38.4) 121.3 (32.5) F(2, 101) = 0.647 0.526

Triglycerides (mg/dL) 87.1 (33.2) 95.3 (52.1) 84.3 (36.0) F(2, 102) = 0.659 0.520

HbA1c (%) 5.2 (0.3) 5.3 (0.3) 5.3 (0.3) F(2, 102) = 0.961 0.386

Insulin (uIU/mL) 6.0 (4.7) 7.2 (4.2) 5.5 (3.2) F(2, 100) = 1.566 0.214

ALT (U/L) 20.1 (10.4) 25.0 (10.3) 22.0 (10.7) F(2, 102) = 1.967 0.145

AST (U/L) 21.7 (6.4) 24.6 (7.4) 23.0 (8.5) F(2, 102) = 1.355 0.263

BUN (mg/dL) 17.1 (3.6) 17.2 (3.9) 16.6 (3.9) F(2, 102) = 0.277 0.759

Creatinine (mg/dL) 0.9 (0.2) 0.9 (0.2) 1.0 (0.3) F(2, 102) = 0.597 0.553

eGFR (mL/min/1.73m2) 86.3 (13.6) 84.6 (12.3) 81.5 (15.6) F(2, 98) = 1.026 0.362

ALT: alanine aminotransferase; AST: aspartate aminotransferase; BMI: body mass index; BUN: blood urea
nitrogen; eGFR: estimated glomerular filtration rate; HbA1c: hemoglobin A1c; HDL: high-density lipoprotein;
LDL: low-density lipoprotein; SD: standard deviation.
Groups were compared using Analysis of Variance. *p < 0.05

868
869

870
871

38
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

872

Supplementary Table S2: Baseline Dexa outcomes and absolute change after 48 weeks
Baseline
Dexa Variable 5 mg 10 mg Placebo F(df1, df2) p-value
Visceral adipose 972.4 1016.2 866.5 F(2, 110) = 0.356 0.701
tissue mass (g) (900.2) (800.4) (635.9)
Bone mineral 2680.4 2901.6 2830.4 F(2, 104) = 1.239 0.294
content (g) (598.6) (493.4) (676.6)
Lean tissue mass (g) 50018.5 54715.6 51886.5 F(2, 109) = 1.561 0.215
(11058.6) (10518.4) (12339.9)
Bone Mineral 1.1996 1.2765 1.2488 F(2,106) = 2.054 0.133
Density (g/cm2) (0.1662) (0.1422) (0.1644)
Dexa Variable Change baseline - 48 weeks
5 mg 10 mg Placebo F(df1, df2) p-value
Visceral adipose -39.9 +13.1 +35.3 F(2, 101.408) = 0.379
tissue mass (g) (185.2) (246.2) (290.8) 0.978
Bone mineral -31.9 +19.4 -8.7 (93.6) F(2,104.415) = 1.403 0.250
content (g) (8.22) (104.3)
Lean tissue mass (g) -66.0 +517.8 -444.6 F(2,126.996) = 1.115 0.331
(2456.0) (2440.2) (1838.9)
Bone Mineral 0.1593 0.0008 -0.0015 F(2,166.294) = 1.596 0.552
Density (g/cm2) (0.0097) 0.0352) (00289)
Groups were compared using Analysis of Variance (baseline) and linear mixed effects model (change
from week 0 through 48). Values provided as mean (standard deviation).

39
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

Supplementary Table S3: Baseline SF-36 and WOMAC and absolute change at 48 weeks
Baseline 48 weeks
5 mg 10 mg Placebo p-value 5 mg 10 mg Placebo p-value
SF-36
- Physical functioning 94.9 95.7 94.5 0.821 -1.6 -1.2 (6.1) -0.2 0.754
(7.4) (8.7) (9.5) (12.0) (7.7)
- Physical role limitations 93.8 95.1 87.8 0.263 -2.7 -2.3 +2.3 0.970
(18.6) (11.7) (27.4) (24.1) (20.1) (32.6)
- Emotional role 87.5 98.2 92.8 0.054 +4.5 -1.0 +6.3 0.401
limitations (24.7) (7.7) (19.5) (27.4) (10.2) (19.7)
- Energy/fatigue 64.8 71.5 68.8 0.209 +6.1 +4.2 +1.4 0.081
(18.7) (13.4) (17.3) (15.2) (10.2) (11.2)
- Emotional well-being 76.0 83.7 81.2 0.019 +5.4 +1.8 +3.9 0.479
(14.2) (10.2) (10.9) (12.0) (6.9) (8.2)
- Social functioning 92.5 93.4 93.9 0.896 +3.1 +3.8 +1.9 0.745
(13.5) (15.7) (10.9) (13.5) (15.8) (10.6)
- Pain 83.9 84.4 80.6 0.528 +2.4 +2.9 +1.6 0.682
(12.8) (16.0) (18.0) (15.9) (11.7) (12.7)
- General health 76.4 82.6 76.2 0.203 +5.6 +3.2 +0.8 0.014
(16.8) (16.2) (19.2) (10.3) (10.0) (12.0)
- Physical health score 87.2 89.5 84.7 0.254 +0.4 +0.6 +1.2 0.850
(10.9) (8.6) (15.5) (14.6) (7.7) (12.7)
- Mental health score 80.2 86.7 84.2 0.037 +5.3 +2.2 +3.4 0.385
(13.1) (7.7) (11.3) (12.5) (5.7) (8.3)
- Aggregate score 83.7 88.1 84.5 0.142 +3.8 +1.4 +2.3 0.335
(10.4) (7.0) (12.0) (9.6) (5.9) (8.2)
WOMAC
- Total score (0-96) 28.8 28.3 31.0 0.325 -1.6 (5.0) +0.8 -1.1 0.325
(6.2) (7.8) (10.0) (7.7) (5.9)
- Pain score (0-20) 6.5 (2.0) 6.3 (2.4) 6.4 (2.0) 0.900 -0.6 (1.8) +0.3 0.0 (1.9) 0.184
(2.5)
- Stiffness score (0-8) 3.1 (1.2) 2.9 (1.1) 3.5 (1.4) 0.101 -0.4 (1.2) 0.0 (1.1) -0.3 0.456
(1.1)

40
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

- Physical function score 19.3 19.2 21.2 0.220 -0.6 (3.0) +0.5 -0.8 0.614
(0-68) (3.7) (5.1) (7.2) (5.3) (4.6)
SF-36: short form 36; WOMAC: Western Ontario and McMaster Universities Osteoarthritis Index. Values provided as
mean (standard deviation).

873

41
Figure 1 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

a
It is made available under a CC-BY-NC-ND 4.0 International license .

Percent gram change

Percent gram change

Percent gram change

b c
Percent gram change
Percent gram change

*
**
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
Figure 2 (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

a c
It is made available under a CC-BY-NC-ND 4.0 International license .

b d

e f
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
Figure 3 (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

a d

c
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
Figure 4 (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
It is made available under a CC-BY-NC-ND 4.0 International license .

a SF-36 Score Change for Females b SF-36 Score Change for Males

c d
WOMAC Score Change for Females WOMAC Score Change for Males

*
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
Supplementary Figure S1 It is made available under a CC-BY-NC-ND 4.0 International license .
(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

b
Supplementary Figure S2by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified
It is made available under a CC-BY-NC-ND 4.0 International license .

b
Supplementary Figure S3by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
(which was not certified
It is made available under a CC-BY-NC-ND 4.0 International license .

b
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
Supplementary Figure
(which was S4by peerIt review)
not certified is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is made available under a CC-BY-NC-ND 4.0 International license .
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
Supplementary Figure
(which was S5by peerIt review)
not certified is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
is made available under a CC-BY-NC-ND 4.0 International license .

*
* *
*

b
 medRxiv preprint doi: https://doi.org/10.1101/2024.08.21.24312372; this version posted August 26, 2024. The copyright holder for this preprint
Supplementary Figure
(which was S6by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
not certified
It is made available under a CC-BY-NC-ND 4.0 International license .