*No significant differences at baseline between arm 1 and arm 2 (FMD), with exception of total cholesterol (P = 0.014) and LDL (P = 0.024). †P values comparing within-group changes were calculated using paired two-tailed Student’s t test. ‡Plus-minus values are means ± SD rounded to the nearest tenth. §Between-arm comparison was calculated using two-tailed two-sample equal variance t tests. Using the Benjamini-Hochberg method for controlling the FDR of 0.05. ||P values indicate that the null hypothesis of no difference between the control diet (arm 1) to FMD (arm 2) can be rejected. ¶The BMI is the weight in kilograms divided by the square of the height in meters. **Analyzed by DEXA.
and risk factors during the first FMD cycle (at day 5 of the FMD and before refeeding; table S3) and after completion of three FMD cycles (5 to 7 days after completing the third FMD cycle; table S2).
At the end of the first FMD cycle and before resuming the normal diet, body weight (P < 0.0001), BMI (P < 0.0001), absolute lean body mass (P < 0.0001), waist circumference (P < 0.0001), fasting glucose (P < 0.0001), IGF-1 (P < 0.0001), diastolic blood pressure (P < 0.0003), triglycerides (P < 0.0001), and LDL (P < 0.0026) were significantly reduced compared to baseline. In contrast, relative lean body mass (P = 0.02), b-hydroxybutyrate (P < 0.0001), and IGFBP-1 (P < 0.0001) were increased. Both absolute and relative total body fat (P = 0.075 and P = 0.047, respectively), systolic blood pressure (P = 0.076), as well as CRP (P = 0.75) were not significantly changed after completion of the first FMD cycle compared to baseline (table S3). These results indicate that subjects did follow the dietary changes imposed by the FMD and responded to them as anticipated.In subjects who completed three FMD cycles (combining both FMD arms) and who returned to the normal diet for 5 to 7 days, body weight (P < 0.0001; n = 71), BMI (P < 0.0001; n = 71), total body fat (absolute, P < 0.0001; relative, P < 0.0001; n = 70), trunk fat (absolute, P < 0.001; relative, P = 0.0002; n = 70), absolute lean body mass (P =0.0001; n = 70), waist circumference (P < 0.0001; n = 52), IGF-1 (P < 0.0001; n = 69), systolic and diastolic blood pressure (P < 0.0001 and P < 0.0004, respectively; n = 70), total cholesterol (P = 0.004; n = 55), LDL (P < 0.0011; n = 55), and HDL (P = 0.02; n = 55) were significantly reduced, and relative lean body mass (P = 0.0002; n = 70) was increased. Fasting glucose (P = 0.28; n = 66), b-hydroxybutyrate (P = 0.23; n = 69), IGFBP-1 (P = 0.84; n = 69), triglycerides (P = 0.16; n = 55), and CRP (P = 0.052; n = 69) were not significantly changed (table S2). In summary, the combined FMD groups from arms 1 and 2 confirmed that the FMD cycles promoted potent effects on many metabolic markers and disease risk factors, which are maintained after subjects return to their normal diet.
FMD effects stratified by baseline risk factor values: A post hoc observational pre-post FMD comparison
Age-related physiological changes that lead to increased risk factors occur before diseases can be diagnosed (20, 21). We used the aggregated FMD data of both study arms and performed a post hoc analysis of the FMD effect on risk factors for CVD and metabolic syndrome, defined as three of five of the following conditions: abdominal obesity, elevated fasting glucose, elevated blood pressure, high serum triglycerides, and low HDL cholesterol (1). We selected clinically relevant cutoffs and compared normal and at-risk subjects for each risk factor: total cholesterol >199 mg/dl and LDL cholesterol levels >130 mg/dl are associated with an increased risk for CVD (22), a fasting glucose >99 mg/dl indicates impaired fasting glucose/prediabetes (23), and triglyceride levels >100 mg/dl (24) as well as CRP >1 mg/liter are associated with increased risk for CVD (25). For serum IGF-1, no clinically relevant risk level has been established, but a number of epidemiological studies have associated IGF-1 levels above 200 ng/ml with various cancers (17, 26). We therefore compared the effect of FMD cycles on subjects in the highest quartile of IGF-1 expression (>225 ng/ml) with that on subjects with IGF-1 levels ≤225 ng/ml.
In a post hoc analysis, we tested how the changes in the FMD normal and at-risk subgroups compared to those in the control diet normal and at-risk subgroups, as defined by their baseline levels of various risk factors (Table 3). We observed a benefit of the FMD, but not in the control arm, on BMI in all BMI subgroups (P value for interaction = 0.03), although the FMD was particularly beneficial among subjects
Fig. 2. Change analysis of metabolic variables during the randomization.
Effects on aging/disease markers and risk factors in all subjects who completed the randomized analysis in either the control arm or the FMD arm (5 to 7 days after the third cycle of FMD). (A) Body weight, (B) BMI, (C) total body fat, (D) trunk fat, (E) lean body mass, (F) waist circumference, (G) serum glucose level, (H) insulin-like growth factor 1, (I) systolic blood pressure, (J) diastolic blood pressure, (K) triglycerides, (L) total cholesterol, (M) LDL, (N) HDL, and (O) CRP were measured in both cohorts as described. The D change represents a comparison to baseline. All data are means ± SD. Betweenarm comparisons were calculated using two-tailed two-sample equal variance t tests. For some of the 100 enrolled participants, the nurses were unable to collect all the samples/measurements from all subjects. We therefore excluded subjects with incomplete measurements from a particular marker group (see Table 2 for details). Abs, absolute; Rel, relative; BP, blood pressure.
who were obese (BMI >30) at baseline. The FMD-dependent reduction in IGF-1 was also larger in participants with baseline IGF-1 ≥225 ng/ml (P value for interaction = 0.018).
Next, we evaluated the effect size, that is, efficacy in normal and atrisk subjects (Table 4) in subjects stratified by risk factor. Subjects with a BMI of greater than 30 (obese) experienced a greater reduction in BMI by the end of the three FMD cycles than those with a BMI of less than 25 (P = 0.011 between groups) and BMI of 25 to 30 (P = 0.0011 between groups). Systolic blood pressure was reduced by 2.4 ± 6.3 mmHg in subjects with baseline systolic blood pressure ≤120 mmHg but by 6.7 ± 6.9 mmHg in subjects with systolic blood pressure >120 mmHg (P = 0.013 between groups), and diastolic blood pressure was reduced by 1.5 ± 5.1 mmHg in subjects with diastolic blood pressure ≤80 mmHg but by 5.5 ± 6.4 mmHg in those with baseline levels above 80 mmHg (P = 0.01 between groups). Fasting glucose did not change for participants with baseline levels ≤99 mg/dl but was reduced by 11.8 ± 6.9 mg/dl in participants with baseline fasting glucose >99 mg/dl (P < 0.0001 between groups); notably, this reduction brought glucose in these subjects within the healthy range. IGF-1 levels in subjects with baseline levels above 225 ng/ml were reduced by 55.1 ± 45.6 ng/ml, nearly four times more (P < 0.001 between groups) than the 14.1 ± 39.9 ng/ml reduction observed in participants with IGF-1 concentrations below 225 ng/ml. Triglyceride levels were reduced more in participants with baseline levels >100 mg/dl (P = 0.0035 between groups). Total cholesterol was reduced significantly more in participants with total cholesterol higher than 199 mg/dl at baseline (P = 0.015 between groups). LDL was reduced by 14.9 ± 21.7 mg/dl in those with total cholesterol baseline levels above 199 mg/dl but was not reduced by FMD cycles in normal-range subjects (P = 0.013 between groups). There was no reduction (P = 0.094 between groups) in HDL levels for those study participants with HDL levels below or above 50 mg/ dl at baseline. CRP was not reduced for subjects with levels below 1 mg/liter but was reduced by 1.6 ± 1.3 mg/liter and returned to thenormal levels in most subjects with baseline CRP higher than 1 mg/liter (P = 0.0003 between groups). A graphical summary of these data is presented in Fig. 3; before-after dot plots of individual subjects in the control cohort as well as in normal and at-risk subjects in the FMD cohort are presented in fig. S3.
This post hoc analysis indicates that the FMD had more pronounced effects in at-risk participants than in those subjects with risk factor values within the normal range, with the exception of HDL. Larger randomized trials are necessary to confirm the results on the efficacy of the FMD in the treatment of patients at risk for diseases.
Voluntary follow-up 3 months after FMD
We invited participants to return on a voluntary basis about 3 months (actual mean follow-up time, 3.2 ± 1.3 months; n = 50) after their third and final FMD cycle. In these subjects, the FMD’s effects on body weight, BMI, waist circumference, glucose (in at-risk subjects), IGF-1, and systolic (in at-risk subjects) and diastolic blood pressure persisted for at least 3 months after the final FMD cycle (table S4). Subjects with low HDL levels at baseline displayed increased HDL levels at the 3-month follow-up, whereas CRP levels remained significantly lower in study participants with baseline CRP levels above 1 mg/liter. Notably, some of the at-risk groups include only a few subjects, and thus, larger studies are needed to establish long-term effects of the FMD on disease risk factors.
These results indicate that some of the beneficial effects of multiple cycles of the FMD may last for several months. Although subjects were not advised to change their diet or exercise regimen after the FMD cycles ended, we cannot rule out that some of the changes after the additional 3 months may be a result of lifestyle changes such as healthier diets and/or improved physical activity after the completion of this trial.
Table 3. Post hoc comparisons for changes in risk factors for age-related diseases and conditions by baseline subgroups.
(FMD − Control)
Mean (95% CI)
(FMD − Control)
Mean (95% CI)
|25||−0.6 (−1.2 to −0.05)||0.03||0.03|
|25–30||−0.8 (−1.4 to −0.3)||0.003|
|30||−1.9 (−2.6 to −1.1)||0.0009|
|Systolic blood pressure (mmHg)|
|120||−3.4 (−7.2 to 0.5)||0.086||0.80|
|≥120||−4.3 (−10.4 to 1.8)||0.17|
|Diastolic blood pressure (mmHg)|
|80||−2.5 (−5.3 to 0.3)||0.08||0.87|
|≥80||−3.0 (−8.2 to 2.3)||0.26|
|Fasting glucose (mg/dl)|
|99||−0.8 (−5.2 to 3.6)||0.72||0.12|
|≥99||−11.7 (−25.0 to 1.5)||0.08|
|225||−18.7 (−38.6 to 1.2)||0.065||0.018|
|≥225||−70.9 (−109.3 to −32.6)||0.0004|
|100||−4.6 (−24.1 to 15.0)||0.64||0.38|
|≥100||−19.1 (−45.8 to 7.6)||0.16|
|Total, 199||−1.8 (−12.6 to 9.0)||0.73||0.88|
|Total, ≥199||−0.2 (−18.2 to 17.7)||0.98|
|LDL, 199 total|
|1.0 (−8.8 to 10.8)||0.84||0.60|
|LDL, ≥199 total|
|6.2 (−11.2 to 23.6)||0.48|
|HDL, 50||−1.2 (−9.0 to 0.5)||0.75||0.70|
|HDL, ≥50||0.5 (−4.2 to 5.2)||0.83|
|1||−0.4 (−1.5 to 8.7)||0.47||0.59|
|≥1||−0.9 (−2.4 to 0.6)||0.24|
This randomized phase 2 trial indicates that three cycles of a 5-day FMD per month are feasible, safe, and effective in reducing body weight, waist circumference and BMI, absolute total body and trunk fat, systolic blood pressure, as well as IGF-1. Metabolic markers such as fasting glucose, triglycerides, and CRP, as well as total, HDL, and LDL cholesterol, which were within the normal range at baseline, were not significantly affected in the randomized comparison after three FMD cycles. After 3 months, subjects from the control arm were crossed over to the FMD intervention. Our post hoc analysis of the aggregated data from all 71 subjects who completed three FMD cycles confirmed the effects of the FMD on trunk and total body fat, blood pressure, and IGF-1. A post hoc analysis also allowed us to analyze subjects with elevated risk factors or metabolic markers associated with metabolic syndrome and age-related diseases, such as high BMI, blood pressure, fasting glucose, triglycerides, CRP, cholesterol, and IGF-1. The FMD had more pronounced effects on all these markers in at-risk participants than in those subjects who had risk factor values within the normal range. Some of these metabolic markers, namely, CRP, systolic/diastolic blood pressure, and serum lipids, have been proposed as markers of biological aging (27). However, other markers affected by the FMD, including IGF-1 and glucose, have been strongly implicated in aging and age-related diseases (5, 18, 28).
Study participants were instructed not to alter their lifestyle for the duration of the trial and were allowed to consume food of their choice during the normal diet periods, that is, subjects were not placed on a prespecified or calorie-restricted diet. We observed changes that were both positive (total cholesterol and LDL) and negative (HDL) in arm 1 subjects during the control diet period, potentially explained by dietary habit changes in anticipation for the FMD, despite no change in weight, BMI, body fat, or lean mass. Similarly, the persistent effects of the FMD observed 3 months after study completion may be affected by changes in dietary habits.
Table 4. Post-hoc analysis of risk factors for age-related diseases and conditions in at-risk subjects who completed the FMD trial.
|Baseline*||FMD: 5 days after third FMD|
|n||Mean ± SD||(95% CI)||Mean ± SD||(95% CI)||P*||D†||Efficacy|
(comparing D) P‡
|25||27||22.4 ± 1.7||(21.68–23.03)||21.9 ± 1.6||(21.24–22.52)||0.0014||−0.5 ± 0.7||0.0020|||
|25–30||30||27.1 ± 1.4||(26.58–27.60)||26.4 ± 1.6||(25.76–26.99)||0.0001||−0.7 ± 0.7||0.21|
|30||14||34.4 ± 3.5||(32.37–36.36)||33.0 ± 3.5||(30.95–34.95)||0.0001||−1.4 ± 1.0||0.011|
|Systolic blood pressure (mmHg)|
|120||49||110.3 ± 6.9||(108.5–112.4)||107.9 ± 7.7||(105. 7–110.1)||0.0089||−2.4 ± 6.3||0.013|
|120||21||133.4 ± 9.0||(129.3–137.5)||126.7 ± 11.3||(121.5–131.8)||0.0002||−6.7 ± 6.9|
|Diastolic blood pressure (mmHg)|
|80||53||71.8 ± 5.0||(70.43–73.21)||70.3 ± 5.9||(68.64–71.89)||0.032||−1.5 ± 5.1||0.010|
|80||17||87.7 ± 6.9||(84.20–91.27)||82.2 ± 10.0||(77.09–87.38)||0.0026||−5.5 ± 6.4|
|Fasting glucose (mg/dl)|
|99||53||87.7 ± 6.4||(85.97–89.50)||89.0 ± 8.2||(86. 75–91.25)||0.29||1.3 ± 8.6||0.0001|
|99||13||104.2 ± 4.4||(101.5–106.9)||92.4 ± 8.2||(87.42–97.35)||0.0001||−11.8 ± 6.9|
|225||52||145.1 ± 38.4||(134.4–155.8)||131.0 ± 46.8||(118.0–144.0)||0.014||−14.1 ± 39.9||0.00088|
|225||17||286.6 ± 47.0||(261.6–311.7)||231.5 ± 64.7||(197.0–265.9)||0.0002||−55.1 ± 45.6|
|100||34||65.8 ± 18.4||(59.45–72.09)||69.5 ± 28.9||(59.58–79.45)||0.33||3.7 ± 22.3||0.0035|
|100||21||149.3 ± 41.2||(127.9–160.8)||123.7 ± 52.7||(97.54 –148.0)||0.058||−25.6 ± 50.2|
|Total, 199||40||170.3 ± 18.3||(164.4–176.2)||164.2 ± 20.2||(157. 7–170.6)||0.016||−6.1 ± 15.3||0.015|
|Total, 199||15||228.1 ± 23.0||(215.4–240.9)||208.2 ± 26.3||(193.6–222.8)||0.0088||−19.9 ± 25.4|
|LDL, 199 total|
|40||91.4 ± 20.2||(85.0 –97.73)||88.8 ± 21.0||(82.13–95.38)||0.22||−2.6 ± 13.4||0.013|
|LDL, 199 total|
|15||141.1 ± 29.1||(125.0–157.2)||126.2 ± 24.7||(112.2–139.9)||0.018||−14.9 ± 21.7|
|HDL, 50||17||44.1 ± 4.1||(41.94–46.17)||44.2 ± 4.6||(41.87–46.6)||0.82||0.1 ± 3.1||0.094|
|HDL, 50||38||69.6 ± 14.1||(65.06–74.17)||65.6 ± 11.6||(61.81–69.32)||0.015 −||−4.0 ± 10.0|
|1||43||0.4 ± 0.3||(0.28–0.46)||0.6 ± 0.9||(0.27–0.85)||0.20||0.2 ± 1.0||0.0003|
|1||26||3.3 ± 2.8||(2.22–4.44)||1.6 ± 1.2||(1.06–2.15)||0.0085||−1.7 ± 3.1|
*P values comparing within-group changes were calculated using paired two-tailed Student’s t test. †Plus-minus values are means ± SD rounded to the nearest tenth. ‡Between-group comparison was calculated using two-tailed two-sample equal variance t tests. §The BMI is the weight in kilograms divided by the square of the height in meters. ||One-way analysis of variance for the BMI groups.
and/or physical activity. The composition of the diet tested in this trial was based on the FMD that is known to extend healthspan in mice. Similarly to the study in mice (5), we expect the FMD effects to be mostly independent of an overall caloric restriction, because both groups likely consumed similar levels of calories per month: For example, estimating a 9200 kJ diet for each of the 25 to 26 nonrestricted days and about 19,200 kJ kcal for the
Fig. 3. Post hoc analysis of metabolic variables in subgroups identified by severity of risk factors.
Subjects from both study arms who completed three FMD cycles were post hoc stratified on the basis of being in either normal-risk or at-risk subgroups for factors associated with age-related diseases and conditions. The D change shown represents comparisons to baseline. All data are means ± SD. Between-arm comparisons were calculated using two-tailed two-sample equal variance t tests. One-way
analysis of variance was used for the BMI groups (see Table 4 for details).
5 days of FMD per month, the between-group difference in consumed calories is expected to be about 10%. In addition, this difference may be overestimated because it is likely that subjects have an elevated calorie intake after the FMD period, as we have shown for mice (5). Day 1 of the FMD supplies ~4600 kJ (11% protein, 46% fat, and 43% carbohydrate), whereas days 2 to 5 provide ~3000 kJ per day (9% protein, 44% fat, and 47% carbohydrate); thus, fat and complex carbohydrates are the major source of calories in the FMD.
Our studies in cells and mice indicate that both glucose and proteins interfere with the protective and regenerative effects of fasting (29). Because our previous data indicate that dietary composition can be equally or more important than calorie restriction, it will be important to test the effects of a similarly restricted diet that provides higher proportions of carbohydrates and/or proteins. It remains to be established whether part of the effects of FMD that we observed are mediated by stem cell–based regeneration or rejuvenation, as indicated by our mouse studies (5).
The FMD-induced reduction in serum glucose and IGF-1 is of interest given their role in pro-aging signaling pathways and cancer (17, 30–33). In addition to a marker for insulin resistance and a metabolic input for cancer cells, glucose is associated with cellular sensitization to toxins and senescence (28, 34, 35). Growth hormone receptor deficiency, resulting in reduced IGF-1 levels, is associated with a major reduction in pro-aging signaling, cancer, and diabetes in humans (18). The observed reduction in IGF-1 in our study, but not either after 6 months of intermittent energy restriction (IER) (36) or after 6 years of 20% caloric restriction (37), is probably related to the long-lasting effects of the low protein/amino acid content of the FMD (average 5 days of FMD; 11.5% versus 21% IER or 24% long-term CR). In fact, twenty eight vegans consuming a moderately protein-restricted (10%) diet for about 5 years had reduced IGF-1 levels compared to a group that consumed a chronic 20% calorie-restricted diet (37). We also previously showed that reduced IGF-1 levels and reduced cancer risk were associated with low protein consumption in participants of the National Health and Nutrition Examination Survey cohort (17). Specific ingredients, for example, high levels of unsaturated fats and micronutrients, may also positively contribute to some of the beneficial effects of the FMD.
Note that 25% of the subjects who tested the FMD dropped out of the trial, whereas 10% of the participants opted out of the control arm. This indicates that, despite our efforts to reduce the burden of low-calorie/protein diets, adherence to this dietary regimen requires committed study participants. Further, compared to the control diet arm, the FMD arm imposed an additional daylong visit to the clinic, which may have contributed to reduced compliance. Compliance with prescribed therapies, even placebo, may be an identifiable marker for an overall healthy behavior of study participants (38). Thus, this kind of volunteer, who is observing a benefit and thus not dropping out, could introduce potential bias into the analysis of our trial. The overall comparability at baseline between the control and both FMD arms, as well as the comparable response to the FMD (arm 2 and arm 1 after crossover) suggests no major differences in response for those subjects who completed the trial. Further, those subjects who dropped out of this trial were not different in age or BMI compared to those who completed the trial. It remains to be established why we experienced a gender difference (82% of dropouts were female). The 25% overall dropout rate (all causes) of study participants before the completion of the third FMD cycle is in the range observed in other trials aimed at evaluating dietary interventions in adults. For example, 16 weeks of dieting in combination with physical exercise yielded a discontinuation rate of about 30% (39), and a hypocaloric diet in 28 overweight/obese women resulted in a dropout rate of 40% after 6 months (40). In a trial assessing the effect of intermittent energy/carbohydrate restriction and daily energy restriction on weight loss and metabolic disease risk markers in overweight women, Harvie et al. reported a 23% dropout rate (41). Nonetheless, there are limitations of our trial that should be considered: (i) A relatively small number of subjects in the randomized comparison; (ii) despite providing nourishment and calories for the duration of the FMD, we experienced a higher dropout rate during the FMD intervention than in the control arm; (iii) the findings that the FMD reduced metabolic markers more effectively in at-risk subjects are based on a non-randomized post hoc analysis of the individual factors in generally healthy participants, and thus, further evaluation in subjects with diagnosed disease is needed.
Other less restrictive diets such as those requiring a very low calorie intake twice a week would impose 8 days per month of a severe restriction compared to the 5 days per month or per several months of a less restrictive intervention tested here (41). However, an advantage of these diets is that they may not require as much medical supervision as the longer FMD. FMDs or any type of prolonged fasting interventions lasting more than 12 hours, particularly those lasting several days, require supervision, preferably from a health care professional familiar with prolonged fasting. Although our results suggest that cycles of the plant-based FMD might be safe for elderly individuals, additional studies are necessary to determine its safety for subjects who are 70 years and older.
In summary, and with the limitations outlined above, these results indicate that the periodic FMD cycles are effective in improving the levels of an array of metabolic markers/risk factors associated with poor health and aging and with multiple age-related diseases. As suggested by preclinical studies, interventions that promote longevity should also extend healthspan. Further investigations in larger clinical trials focused on subjects with diagnosed metabolic syndrome, diabetes, and CVDs as well as subjects at high risk for developing cancer and other age-related diseases are needed.
MATERIALS AND METHODS
One-hundred participants without a diagnosed medical condition in the previous 6 months were enrolled (ClinicalTrials.gov; NCT02158897). All participants provided written informed consent, and the University of Southern California (USC) Institutional Review Board (IRB) approved the protocol. Recruitment of subjects was based on fliers, the ClinicalTrials.gov and usc.com websites, and/or word of mouth. Because this was a dietary intervention study, it was not possible for participants or all study personnel to be blinded to group assignment. However, study personnel involved in data collection and specimen analysis were blinded to group assignments.
Flow of participant enrollment and participation was prepared following the CONSORT standards for randomized clinical trials with crossover design. All data were collected at the USC Diabetes and Obesity Research Institute. Subjects were recruited from April 2013 to July 2015 under protocols approved by the USC IRB (HS-12-00391) based on established inclusion (generally healthy adult volunteers and 18 to 70 years of age; BMI, 18.5 and up) and exclusion [any major medical condition or chronic diseases, mental illness, drug dependency, hormone replacement therapy (dehydroepiandrosterone, estrogen, thyroid, and testosterone), pregnant or nursing female, special dietary requirements or food allergies, alcohol dependency, and medications known to affect body weight] criteria. Intention to treat analysis was performed by including all available observations. Eligible participants were randomly assigned using a random-number generator to either arm 1 or arm 2 of the study. All participants completed a health habits questionnaire. Prespecified outcome measures included safety and feasibility, and evaluation of changes in metabolic risk factors for diabetes and CVD and metabolic markers associated with age-related diseases and mortality; these outcomes were measured at baseline during and after completion of the intervention. Laboratory examinations included height, weight, body composition (including total and trunk body fat, soft lean tissue, and bone mineral content) measured by dual-energy x-ray absorptiometry (DEXA), oscillometric blood pressure measurements, and overnight fasting blood draw through venipuncture.
Arm 1 (control).
Participants completed anthropometric measurements and blood collection at enrollment and after 3 months to provide an estimate of non– diet-related changes (Fig. 1). Participants were instructed to maintain their regular eating habits. After 3 months, subjects were crossed over to the experimental FMD group (Fig. 1).
Arm 2 (FMD).
Participants were instructed to consume the FMD, which was provided in a box, for 5 continuous days, and to return to their normal diet after completion until the next cycle that was initiated about 25 days later. Participants completed three cycles of this 5-day FMD (Fig. 1). Participants completed baseline and follow-up examinations at the end of the first FMD (before resuming normal diet to measure the acute FMD effects) and after a washout period of 5 to 7 days of normal caloric intake after the third FMD cycle. An optional follow-up assessment 3 months after the third FMD cycle was offered.
The FMD is a plant-based diet designed to attain fasting-like effects on the serum levels of IGF-1, IGFBP-1, glucose, and ketone bodies while providing both macro- and micronutrients to minimize the burden of fasting and adverse effects (5). Day 1 of the FMD supplies ~4600 kJ (11% protein, 46% fat, and 43% carbohydrate), whereas days 2 to 5 provide ~3000 kJ (9% protein, 44% fat, and 47% carbohydrate) per day. The FMD comprises proprietary formulations belonging to USC and L-Nutra (www.prolonfmd.com) of vegetable-based soups, energy bars, energy drinks, chip snacks, tea, and a supplement providing high levels of minerals, vitamins, and essential fatty acids (fig. S4). All items to be consumed per day were individually boxed to allow the subjects to choose when to eat while avoiding accidentally consuming components of the following day.
Common Terminology Criteria for Adverse Events
Study participants were asked about adverse events at each study visit; events were graded according to the general CTCAE guidelines (see the Supplementary Materials for details).
Blood tests and serum markers
Complete metabolic and lipid panels (overnight fasting) were completed at the Clinical Laboratories at the Keck Medical Center of USC and analyzed immediately after the blood draw of each visit (see the Supplementary Materials for details).
The primary comparisons of randomized groups involved changes in outcomes observed in the control period of arm 1 versus the changes observed in the FMD group (arm 2) after completion of three FMD cycles. Secondary observational analyses involved (i) comparing the FMD effects in arm 2 (randomized to FMD) versus arm 1 (receiving FMD after completion of the randomized control period) and (ii) summarizing the changes for arms 1 and 2 combined after completion of the first and third FMD cycles. Changes from baseline were normally distributed. Comparison of changes from baseline within the treatment arms was performed using paired two-tailed Student’s t tests, and P values <0.05 were considered significant. The betweenarm comparison of treatment changes from baseline was performed using two-tailed two-sample equal variance t tests, and P values <0.05 were considered significant. To control for multiple testing, we used the Benjamini-Hochberg FDR method. All reported P values are nominal two-sided P values; those that met the FDR criteria and remained “significant” at P < 0.05 are indicated with an asterisk.M.W. generated the random allocation sequence and enrolled and assigned participants to interventions. M.W. was not involved in outcome assessments. For this initial randomized trial, the sample size of 100 total subjects was based on detection of a 25% reduction in mean IGF-1, with a two-sided a of 0.05 and 70% power. The estimated control group mean (SD) IGF-1 of 194 (97) used published data on males and females aged 26 to 40 years (42). Statistical analyses were performed on deidentified data. Baseline information and changes from baseline were summarized using means ± SDs for subjects randomized to the control (arm 1, n = 48) and the diet group (arm 2, n = 52). All subjects are included in the arm assigned regardless of treatment adherence (intention to treat); no attempt was made to impute missing values (primarily because if data after completion of the third FMD cycle were not available, then other measurement time points were usually unavailable).In post hoc subgroup analyses, we compared FMD-control group differences over the randomized trial period (three FMD cycles versus control) within high/lower-risk subgroups and tested whether those treatment effects differed in the higher-risk versus lower-risk groups. This subgroup analysis was completed using analysis of variance, with main effects of treatment (FMD and control) and risk group (high and low); the interaction of treatment-by-risk group tested whether the randomized FMD effect differed in high-risk versus low-risk groups. In observational analyses of the pre-post FMD changes combining the two treatment arms, pre-post changes in markers within risk subgroups were tested using paired t test; pre-post changes over risk subgroups were compared using two-sample t test or analysis of variance.
Materials and Methods
Table S1. Complete metabolic panel.
Table S2. Arm-specific markers of adherence and changes in risk factors, including arm 1 after crossover to FMD, and summary of FMD arms 1 and 2.
Table S3. Changes in risk factors and metabolic markers of adherence after the first FMD.
Table S4. Changes in risk factors and metabolic markers of adherence 3 months after intervention.
Fig. S1. Subject self-reported adverse effects based on CTCAE.
Fig. S2. Comparison of participants who completed the trial versus dropouts. Fig. S3. Baseline to 3 months before/after comparison of individual subjects in the control cohort and all subjects who completed the FMD.
Fig. S4. Nutritional information of the FMD.
CONSORT checklist Trial protocols
REFERENCES AND NOTES
- K. G. Alberti, R. H. Eckel, S. M. Grundy, P. Z. Zimmet, J. I. Cleeman, K. A. Donato, J.-C. Fruchart, W. P. T. James, C. M. Loria, S. C. Smith Jr.; International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; International Association for the Study of Obesity, Harmonizing the metabolic syndrome: A joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation 120, 1640–1645 (2009).
- E. S. Ford, W. H. Giles, W. H. Dietz, Prevalence of the metabolic syndrome among US adults: Findings from the third National Health and Nutrition Examination Survey. JAMA 287, 356–359 (2002).
- A. S. Gami, B. J. Witt, D. E. Howard, P. J. Erwin, L. A. Gami, V. K. Somers, V. M. Montori, Metabolic syndrome and risk of incident cardiovascular events and death: A systematic review and meta-analysis of longitudinal studies. J. Am. Coll. Cardiol. 49, 403–414 (2007).
- I. Y. Choi, L. Piccio, P. Childress, B. Bollman, A. Ghosh, S. Brandhorst, J. Suarez, A. Michalsen, A. H. Cross, T. E. Morgan, M. Wei, F. Paul, M. Bock, V. D. Longo, A diet mimicking fasting promotes regeneration and reduces autoimmunity and multiple sclerosis symptoms. Cell Rep. 15, 2136–2146 (2016).
- S. Brandhorst, I. Y. Choi, M. Wei, C. W. Cheng, S. Sedrakyan, G. Navarrete, L. Dubeau, P. Yap, R. Park, M. Vinciguerra, S. D. Biase, H. Mirzaei, M. G. Mirisola, P. Childress, L. Ji, S. Groshen, F. Penna, P. Odetti, L. Perin, P. S. Conti, Y. Ikeno, B. K. Kennedy, P. Cohen, T. E. Morgan, T. B. Dorff, V. D. Longo, A periodic diet that mimics fasting promotes multi-system regeneration, enhanced cognitive performance, and healthspan. Cell Metab. 22, 86–99 (2015).
- V. D. Longo, S. Panda, Fasting, circadian rhythms, and time-restricted feeding in healthy lifespan. Cell Metab. 23, 1048–1059 (2016).
- A. J. Bruce-Keller, G. Umberger, R. McFall, M. P. Mattson, Food restriction reduces brain damage and improves behavioral outcome following excitotoxic and metabolic insults. Ann. Neurol. 45, 8–15 (1999).
- A. L. Hartman, J. E. Rubenstein, E. H. Kossoff, Intermittent fasting: A “new” historical strategy for controlling seizures? Epilepsy Res. 104, 275–279 (2013).
- H. Müller, F. W. de Toledo, K. L. Resch, Fasting followed by vegetarian diet in patients with rheumatoid arthritis: A systematic review. Scand. J. Rheumatol. 30, 1–10 (2001).
- V. D. Longo, M. P. Mattson, Fasting: Molecular mechanisms and clinical applications. Cell Metab. 19, 181–192 (2014).
- C-W. Cheng, G. B. Adams, L. Perin, M. Wei, X. Zhou, B. S. Lam, S. D. Sacco, M. Mirisola, D. I. Quinn, T. B. Dorff, J. J. Kopchick, V. D. Longo, Prolonged fasting reduces IGF-1/PKA to promote hematopoietic-stem-cell-based regeneration and reverse immunosuppression. Cell Stem Cell 14, 810–823 (2014).
- W. E. Sonntag, D. Lynch, W. T. Cefalu, R. L. Ingram, S. A. Bennett, P. L. Thornton, A. S. Khan, Pleiotropic effects of growth hormone and insulin-like growth factor (IGF)-1 on biological aging: Inferences from moderate caloric-restricted animals. J. Gerontol. A Biol. Sci. Med. Sci. 54, B521–B538 (1999).
- Y. Ikeno, R. T. Bronson, G. B. Hubbard, S. Lee, A. Bartke, Delayed occurrence of fatal neoplastic diseases in Ames dwarf mice: Correlation to extended longevity. J. Gerontol. Sci. Med. Sci. 58, 291–296 (2003).
- K. Flurkey, J. Papaconstantinou, R. A. Miller, D. E. Harrison, Lifespan extension and delayed immune and collagen aging in mutant mice with defects in growth hormone production. Proc. Natl. Acad. Sci. U.S.A. 98, 6736–6741 (2001).
- S. E. Dunn, F. W. Kari, J. French, J. R. Leininger, G. Travlos, R. Wilson, J. C. Barrett, Dietary restriction reduces insulin-like growth factor I levels, which modulates apoptosis, cell proliferation, and tumor progression in p53-deficient mice. Cancer Res. 57, 4667–4672 (1997).
- M. S. Bonkowski, F. P. Dominici, O. Arum, J. S. Rocha, K. A. Al Regaiey, R. Westbrook, A. Spong, J. Panici, M. M. Masternak, J. J. Kopchick, A. Bartke, Disruption of growth hormone receptor prevents calorie restriction from improving insulin action and longevity. PLOS ONE 4, e4567 (2009).
- M. E. Levine, J. A. Suarez, S. Brandhorst, P. Balasubramanian, C.-W. Cheng, F. Madia, L. Fontana, M. G. Mirisola, J. Guevara-Aguirre, J. Wan, G. Passarino, B. K. Kennedy, M. Wei, P. Cohen, E. M. Crimmins, V. D. Longo, Low protein intake is associated with a major reduction in IGF-1, cancer, and overall mortality in the 65 and younger but not older population. Cell Metab. 19, 407–417 (2014).
- J. Guevara-Aguirre, P. Balasubramanian, M. Guevara-Aguirre, M. Wei, F. Madia, C.-W. Cheng, D. Hwang, A. Martin-Montalvo, J. Saavedra, S. Ingles, R. de Cabo, P. Cohen, V. D. Longo, Growth hormone receptor deficiency is associated with a major reduction in pro-aging signaling, cancer, and diabetes in humans. Sci. Transl. Med. 3, 70ra13 (2011).
- M. W. Gillman, Developmental origins of health and disease. N. Engl. J. Med. 353, 1848–1850 (2005).
- L. Fontana, B. K. Kennedy, V. D. Longo, D. Seals, S. Melov, Medical research: Treat ageing. Nature 511, 405–407 (2014).
- M. Nayor, R. S. Vasan, Recent update to the US cholesterol treatment guidelines: comparison with international guidelines. Circulation 133, 1795–1806 (2016).
- American Diabetes Association, Diagnosis and classification of diabetes mellitus.
Diabetes Care 37 (suppl. 1), S81–S90 (2014).
- M .Miller, N. J. Stone, C. Ballantyne, V. Bittner, M. H. Criqui, H. N. Ginsberg, A. C. Goldberg, J. Howard, M. S. Jacobson, P. M. Kris-Etherton, T. A. Lennie, M. Levi, T. Mazzone, S. Pennathur, Triglycerides and cardiovascular disease: A scientific statement from the American Heart Association. Circulation 123, 2292–2333 (2011).
- T. A. Pearson, G. A. Mensah, R. W. Alexander, J. L. Anderson, R. O. Cannon, M. Criqui, Y. Y. Fadl, S. P. Fortmann, Y. Hong, G. L. Myers, N. Rifai, S. C. Smith, K. Taubert, R. P. Tracy, Vinicor; Centers for Disease Control and Prevention, American Heart Association, Markers of inflammation and cardiovascular disease: Application to clinical and public health practice: A statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation 107, 499–511 (2003).
- M. N. Pollack, Insulin, insulin-like growth factors, insulin resistance, and neoplasia. Am. J. Clin. Nutr. 86, s820–s822 (2007).
- M. E. Levine, Modeling the rate of senescence: Can estimated biological age predict mortality more accurately than chronological age? J. Gerontol. A Biol. Sci. Med. Sci. 68, 667–674 (2013).
- L. Fontana, L. Partridge, V. D. Longo, Extending healthy life span—From yeast to humans. Science 328, 321–326 (2010).
- S. Brandhorst, M. Wei, S. Hwang, T. E. Morgan, V. D. Longo, Short-term calorie and protein restriction provide partial protection from chemotoxicity but do not delay glioma progression. Exp. Gerontol. 48, 1120–1128 (2013).
- A. G. Renehan, M. Zwahlen, C. Minder, S. T. O’Dwyer, S. M. Shalet, M. Egger, Insulin-like growth factor (IGF)-I, IGF binding protein-3, and cancer risk: Systematic review and meta-regression analysis. Lancet 363, 1346–1353 (2004).
- J. M. Chan, M. J. Stampfer, J. Ma, P. Gann, J. M. Gaziano, M. Pollak, E. Giovannucci, Insulin-like growth factor-I (IGF-I) and IGF binding protein-3 as predictors of advanced-stage prostate cancer. J. Natl. Cancer Inst. 94, 1099–1106 (2002).
- N. E. Allen, A. W. Roddam, D. S. Allen, I. S. Fentiman, I. dos Santos Silva, J. Peto, J. M. P. Holly, T. J. Key, A prospective study of serum insulin-like growth factor-I
(IGF-I), IGF-II, IGF-binding protein-3 and breast cancer risk. Br. J. Cancer 92, 1283–1287 (2005).
- O. Fletcher, L. Gibson, N. Johnson, D. R. Altmann, J. M. P. Holly, A. Ashworth, J. Peto, I. dos Santos Silva, Polymorphisms and circulating levels in the insulin-like growth factor system and risk of breast cancer: A systematic review. Cancer Epidemiol. Biomarkers Prev.
14, 2–19 (2005).
- K. Rapp, J. Schroeder, J. Klenk, H. Ulmer, H. Concin, G. Diem, W. Oberaigner, S. K. Weiland, Fasting blood glucose and cancer risk in a cohort of more than 140,000 adults in Austria. Diabetologia 49, 945–952 (2006).
- T. Stocks, K. Rapp, T. Bjørge, J. Manjer, H. Ulmer, R. Selmer, A. Lukanova, D. Johansen, H. Concin, S. Tretli, G. Hallmans, H. Jonsson, P. Stattin, Blood glucose and risk of incident and fatal cancer in the metabolic syndrome and cancer project (Me-Can): Analysis of six prospective cohorts. PLOS Med. 6, e1000201 (2009).
- M. N. Harvie, M. Pegington, M. P. Mattson, J. Frystyk, B. Dillon, G. Evans, J. Cuzick, S. A. Jebb, B. Martin, R. G. Cutler, T. G. Son, S. Maudsley, O. D. Carlson, J. M. Egan, Flyvbjerg, A. Howell, The effects of intermittent or continuous energy restriction onweight loss and metabolic disease risk markers: A randomized trial in young overweight women. Int. J. Obes. 35, 714–727 (2011).
- L. Fontana, E. P. Weiss, D. T. Villareal, S. Klein, J. O. Holloszy, Long-term effects of calorie or protein restriction on serum IGF-1 and IGFBP-3 concentration in humans. Aging Cell 7, 681–687 (2008).
- R. D. Hays, R. L. Kravitz, R. M. Mazel, C. D. Sherbourne, M. R. DiMatteo, W. H. Rogers, S. Greenfield, The impact of patient adherence on health outcomes for patients with chronic disease in the medical outcomes study. J. Behav. Med. 17, 347–360 (1994).
- T. P. Wycherley, M. Noakes, P. M. Clifton, X. Cleanthous, J. B. Keogh, G. D. Brinkworth, A high-protein diet with resistance exercise training improves weight loss and body composition in overweight and obese patients with type 2 diabetes. Diabetes Care 33, 969–976 (2010).
- D. Florakis, E. Diamanti-Kandarakis, I. Katsikis, G. P. Nassis, A. Karkanaki, N. Georgopoulos, D. Panidis, Effect of hypocaloric diet plus sibutramine treatment on hormonal and metabolic features in overweight and obese women with polycystic ovary syndrome: randomized, 24-week study. Int. J. Obes. 32, 692–699 (2008).
- M. Harvie, C. Wright, M. Pegington, D. McMullan, E. Mitchell, B. Martin, R. G. Cutler, Evans, S. Whiteside, S. Maudsley, S. Camandola, R. Wang, O. D. Carlson, J. M. Egan, M. P. Mattson, A. Howell, The effect of intermittent energy and carbohydrate restriction v. daily energy restriction on weight loss and metabolic disease risk markers in overweight women. Br. J. Nutr. 110, 1534–1547 (2013).
- G. Brabant, A. von zur Mühlen, C. Wüster, M. B. Ranke, J. Kratzsch, W. Kiess, J.-M. Ketelslegers, L. Wilhelmsen, L. Hulthén, B. Saller, A. Mattsson, J. Wilde, R. Schemer, P. Kann, Serum insulin-like growth factor I reference values for an automated chemiluminescence immunoassay system: Results from a multicenter study. Horm. Res. 60, 53–60 (2003).
Acknowledgments: We thank Y. Guan for help in data curation and acknowledge the technical expertise of R. Buono (USC Leonard Davis School of Gerontology). Funding: Funding was provided by the USC Edna Jones chair fund to V.D.L. W.J.M.’s contributions were provided through the Southern California Clinical and Translational Science Institute supported by NIH UL1TR001855. Author contributions: V.D.L., M.W., T.E.M., and T.D. designed the clinical trial. M.W., K.H., and T.D. supervised the clinical trial. J.B. and M.S. contributed to patient recruitment and supervision. S.B., H.M., C.W.C., E.G., S.G., and W.J.M. performed data analysis. S.B. and V.D.L. wrote the paper. All authors discussed the results and commented on the manuscript. Competing interests: The experimental FMD was provided by L-Nutra Inc. The funding sources had no involvement in study design; collection, analysis, and interpretation of data; writing of the report; or decision to submit the article for publication. The USC has licensed intellectual property to L-Nutra that is under study in this research. As part of this license agreement, the University has the potential to receive royalty payments from L-Nutra. V.D.L. and T.E.M., who have equity interest in L-Nutra, did not participate in the collection and analysis of the data. One-hundred percent of V.D.L.’s equity will be assigned to the nonprofit foundation Create Cures. U.S. patents 20140227373A1 (“Methods and diets to protect against chemotoxicity and age related illnesses”) and 20140112909A1 (“Methods and formulations promoting tissue regeneration, longevity and healthspan”) related to the work described here have been filed. Data and materials availability:
Use of the FMD may require a material transfer agreement.
Submitted 27 April 2016
Resubmitted 23 August 2016 Accepted 20 December 2016
Published 15 February 2017
Citation: M. Wei, S. Brandhorst, M. Shelehchi, H. Mirzaei, C. W. Cheng, J. Budniak, S. Groshen,
W. J. Mack, E. Guen, S. Di Biase, P. Cohen, T. E. Morgan, T. Dorff, K. Hong, A. Michalsen, A. Laviano, V. D. Longo, Fasting-mimicking diet and markers/risk factors for aging, diabetes, cancer, and cardiovascular disease. Sci. Transl. Med. 9, eaai8700 (2017).
Fasting-mimicking diet and markers/risk factors for aging, diabetes, cancer, and cardiovascular disease
Min Wei, Sebastian Brandhorst, Mahshid Shelehchi, Hamed
Mirzaei, Chia Wei Cheng, Julia Budniak, Susan Groshen, Wendy J.
Mack, Esra Guen, Stefano Di Biase, Pinchas Cohen, Todd E.
Morgan, Tanya Dorff, Kurt Hong, Andreas Michalsen, Alessandro Laviano and Valter D. Longo (February 15, 2017) Science Translational Medicine 9 (377), . [doi:10.1126/scitranslmed.aai8700]
Fasting: More than a fad
Mice that fast periodically are healthier, metabolically speaking. To explore whether fasting can help people as well, Wei et al. studied 71 people who either consumed a fasting-mimicking diet for 5 days each month for 3 months or maintained their normal diet for 3 months and then switched to the fasting schedule. The fasting-like diet reduced body weight and body fat, lowered blood pressure, and decreased the hormone IGF-1, which has been implicated in aging and disease. A post hoc analysis replicated these results and also showed that fasting decreased BMI, glucose, triglycerides, cholesterol, and C-reactive protein (a marker for inflammation). These effects were generally larger in the subjects who were at greater risk of disease at the start of the study. A larger study is needed to replicate these results, but they raise the possibility that fasting may be a practical road to a healthy metabolic system.
The following resources related to this article are available online at http://stm.sciencemag.org. This information is current as of February 15, 2017.
|Article Tools||Visit the online version of this article to access the personalization and article tools:|
|Supplemental Materials||“Supplementary Materials”|
|Permissions||Obtain information about reproducing this article:|
Science Translational Medicine (print ISSN 1946-6234; online ISSN 1946-6242) is published weekly, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue, NW, Washington, DC 20005. Copyright 2017 by the American Association for the Advancement of Science; all rights reserved. The title Science Translational Medicine is a registered trademark of AAAS.