Effects of Intermittent Fasting or Calorie Restriction on Markers of Lipid Metabolism in Human Skeletal Muscle

Abstract

Context

Impaired lipid metabolism is linked with obesity-associated insulin resistance, which may be reversed by caloric restriction (CR).

Objective

In a secondary analysis of a randomized controlled trial, we compared the effects of intermittent fasting (IF) and CR on markers of lipid metabolism in muscle.

Design

Seventy-six women (body mass index, 25-40 kg/m²) were randomly assigned to 1 of 3 diets for 8 weeks and provided foods at 70% (CR70 and IF70) or 100% (IF100) of energy requirements. IF groups ate breakfast prior to a 24-hour fast on 3 nonconsecutive days per week. On nonfasting days, IF70 ate at 100% and IF100 ate at 145% of energy requirements to achieve the prescribed target. Weight, body composition, insulin sensitivity by clamp, nonesterified fatty acids (NEFAs), β-hydroxybutyrate (BHB), and markers of lipid metabolism and oxidative stress in muscle by quantitative polymerase chain reaction were measured at baseline and week 8 following a 12-hour overnight fast (all groups) and 24-hour fast (IF groups).

Results

IF70 resulted in greater weight and fat loss and reduced NEFAs vs CR70 and IF100 after an overnight fast. IF70 and IF100 induced a greater reduction only in mRNA levels of antioxidant enzymes glutathione peroxidase 1 (GPX1), superoxide dismutase 1, soluble (SOD1), and SOD2 vs CR70. Fasting for 24 hours increased NEFAs and BHB in IF groups, but impaired insulin sensitivity and increased PLIN5 mRNA levels.

Conclusions

In comparison to CR, IF did not increase markers of lipid metabolism in muscle, but reduced expression of antioxidant enzymes. However, fasting-induced insulin resistance was detected, alongside increased PLIN5 expression, potentially reflecting transient lipid storage.

calorie restriction, intermittent fasting, lipid metabolism, mitochondria, muscle, obesity

Impairments in lipid oxidation and mitochondrial metabolism, including decreased mitochondrial volume and oxidative capacity in skeletal muscle, are linked with increased lipid deposition and the development of insulin resistance in humans with obesity (1-9). Weight loss by moderate calorie restriction (CR) improves insulin sensitivity (10, 11), but its impact on lipid metabolism in human skeletal muscle remains controversial. Six months of CR increased the messenger RNA (mRNA) expression of peroxisome proliferator-activated receptor γ coactivator 1-α (PGC1α) and mitochondrial transcription factor A (TFAM), which are key markers involved in mitochondrial biogenesis, as well as mitochondrial DNA content and the activity of superoxide dismutase (SOD), which degrades reactive oxygen species (ROS) to hydrogen peroxide and molecular oxygen in skeletal muscle (12). However, CR did not alter markers of mitochondrial biogenesis, mitochondrial content, and mitochondrial oxidative stress response in skeletal muscle at 12 months (13). Moreover, functional analysis revealed that CR did not increase the maximal adenosine 5′-triphosphate (ATP) synthesis rate or mitochondrial efficiency as indicated by ATP turnover (ATPflux)/O₂ (P/O) (13), although a reduction in extramyocellular lipid content was observed (12, 13).

Intermittent fasting (IF) is an alternative to CR that is characterized by periods of fasting for 20 to 24 hours interspersed by free access to food. Some studies suggest that IF may produce greater improvements in metabolic health vs CR, including greater reductions in blood pressure, blood lipids, nonesterified fatty acids (NEFAs), and homeostatic model assessment of insulin resistance (14). A feature of IF is the near daily metabolic “switch” that occurs between the anabolic fed state, which is characterized by high rates of carbohydrate oxidation, lipid synthesis, and fat storage, to the catabolic fasting state, which is characterized by mobilization of lipid and high rates of lipid oxidation and fatty-acid derived ketogenesis (14-16), and IF may provide greater stimulus than CR to increase lipid turnover. However, acute and prolonged fasts of 24 to 72 hours also increase the number and size of lipid droplets and lipid droplet-coating protein perilipin 5 (PLIN5) in human skeletal muscle (17-21). The comparative effects of IF vs CR on markers of lipid and mitochondrial metabolism in human muscle have not previously been reported.

This study aimed to 1) compare the effects of 8 weeks of CR vs IF on mRNA expression of genes involved in lipid, mitochondrial metabolism, and oxidative stress in skeletal muscle in women with overweight or obesity following an overnight fast, and 2) examine the metabolic adaptation in these markers during the fed to 24-hour fasted transition in IF groups. We hypothesized that compared to CR, IF would induce a greater increase in markers of lipid metabolism in human skeletal muscle.

Materials and Methods

Participants and diets

This is a secondary analysis of a previously published trial (22). This study was reviewed and approved by the Royal Adelaide Hospital Research Ethics Committee and registered as a clinical trial with Clinicaltrials.gov (NCT01769976). All participants provided written, informed consent before initiation of the study.

The inclusion and exclusion criteria of the study and research protocol are reported elsewhere (22, 23). Briefly, 88 women age 35 to 70 years, with overweight or obesity (body mass index of 25-42 kg/m²), were randomly assigned to 1 of 4 groups in a 2:2:2:1 ratio (CR70, IF70, IF100, and control) for 8 weeks. CR70 group members were provided 70% of calculated energy requirements daily. IF70 group members were provided foods at approximately 32% of daily energy requirements at breakfast on fasting days and at approximately 100% of energy requirements on fed days so their overall energy deficit was equivalent to the CR70. The IF100 group was provided foods at approximately 37% of requirements at breakfast on fasting days and at approximately 145% of energy requirements on fed days, to maintain overall energy balance. IF groups consumed breakfast before 8 AM prior to initiating a 24-hour fast until 8 AM the following day, every other weekday (3 nonconsecutive fasting days per week). During the 24-hour fasting period, participants were allowed to consume water and energy-free foods, black coffee and/or tea, plus one very low-calorie broth (250 mL, ~20 kcal). To monitor adherence, participants were required to complete daily checklists and to return them to the researchers at weekly individual counseling sessions. Self-reported adherence to continuous energy restriction (CR70) and IF70 diets was excellent, whereas the IF100 group ate 9% less than prescribed on fed days as previously reported (22). This report excludes the control group (N = 12) because muscle biopsies were not performed.

Metabolic testing in participants

Participants were required to attend the research clinic at 7:30 AM, at baseline and after 8 weeks following an overnight 12-hour fast. IF groups underwent an additional visit 3 to 7 days later, following a 24-hour fast. Body weight, height, and waist and hip circumference were measured with the participant dressed in a hospital gown after voiding. Intravenous cannulas were placed, and baseline samples collected. After a supine rest, respiratory quotient (RQ) was determined for 30 minutes by indirect calorimetry (ParvoMedics) as mentioned previously (24). The vastus lateralis muscle samples were obtained in a subset of participants (N = 16 in CR70, N = 16 in IF70, N = 14 in IF100) at baseline and at week 8 after a 12-hour fast (all groups) and a 24-hour fast (IF groups only) as previously described (23). Insulin sensitivity was measured using a 2-hour hyperinsulinemic-euglycemic clamp (60 mU/m² body surface area/minute) (22). Total body composition was assessed by dual-energy X-ray absorptiometry (Lunar Prodigy; GE Healthcare) at baseline and after 8 weeks following an overnight fast for all groups.

Biochemical analysis

Blood glucose was examined by photometric assay in the laboratory of SA Pathology. Plasma β-hydroxybutyrate (BHB) levels were measured using a commercial enzymatic kit (RANBUT D-3 hydroxybutyrate kit, Randox) on a Beckman AU480 clinical analyzer (Beckman Coulter Inc). Serum NEFAs were measured by enzymatic colorimetric assay (NEFA-HR (2), Wako Diagnostics) on a VersaMax enzyme-linked immunosorbent assay (ELISA) Microplate Reader. Samples were run in duplicate and samples from each participant were tested within the same run to reduce instrument variation.

Messenger RNA quantification in muscle

Total RNA extraction from muscle samples, complementary DNA synthesis, and quantitative real-time polymerase chain reaction (PCR) were performed as previously reported (23) using TaqMan primers (Table 1) and Fast Universal PCR Master Mix (Applied Biosystems). Relative gene expression was analyzed using the 2^−ΔΔCT method and normalized for the mean of actin-β and hypoxanthine guanine phosphoribosyl transferase, which were not different at baseline, or following the intervention.

Table 1.

TaqMan primers used for messenger RNA expression analysis

Gene symbol	Gene Name	Assay ID
Housekeeper
ACTB	Actin β	Hs01060665_g1
HPRT	Hypoxanthine phosphoribosyltransferase 1	Hs02800695_m1
Lipolysis
LIPE	Lipase E, hormone sensitive type	Hs00193510_m1
Lipid synthesis and storage
ACACA	Acetyl-CoA carboxylase α	Hs01046047_m1
FASN	Fatty acid synthase	Hs01005622_m1
DGAT1	Diacylglycerol O-acyltransferase 1	Hs01017541_m1
PLIN5	Perilipin 5	Hs00965990_m1
Lipid oxidation
PPARα	Peroxisome proliferator-activated receptor α	Hs00947536_m1
CD36	CD36 molecule	Hs00354519_m1
CPT1B	Carnitine palmitoyltransferase 1B	Hs03046298_s1
PDK4	Pyruvate dehydrogenase kinase 4	Hs01037712_m1
Ketone oxidation
ACAT1	Acetyl-CoA acetyltransferase 1	Hs00608002_m1
MCT1	Monocarboxylate transporter 1	Hs01560299_m1
MCT4	Monocarboxylate transporter 4	Hs01006127_m1
OXCT1	3-oxoacid CoA-transferase 1	Hs01036203_m1
Mitochondrial function
PGC1α	Peroxisome proliferator-activated receptor γ coactivator 1-α	Hs01016719_m1
SIRT3	Sirtuin 3	Hs00953477_m1
MFN2	Mitofusin 2	Hs00208382_m1
TFAM	Transcription factor A, mitochondrial	Hs00273372_s1
Oxidative stress
SOD1	Superoxide dismutase 1, soluble	Hs00533490_m1
SOD2	Superoxide dismutase 2, soluble	Hs00167309_m1
GPX1	Glutathione peroxidase 1	Hs00829989_gH

Statistical analysis

Data are expressed as mean ± SEM. Individuals who withdrew from the study were not included in the analysis. Participants who completed baseline and week 8 (12-hour fast) biopsies were included for gene expression analyses (CR70, N = 12; IF70, N = 16; IF100, N = 14). All statistical analyses were performed using IBM SPSS Statistics 26. A maximum likelihood mixed-effects model was employed to examine the group differences of 8-week intervention following an overnight 12-hour fast, as well as the time effects of a 24-hour fast within each IF group. The model included fixed effects for intervention, visit and the intervention by visit interaction, and a random effect for subject with an unstructured covariance matrix to account for the repeated visits. The effect of intervention was assessed with planned contrasts between groups in the change from baseline to week 8 following the 12-hour overnight fast. Bonferroni-adjusted pairwise comparisons were also conducted within each IF group to assess differences from a 12-hour to a 24-hour fast after 8 weeks of intervention. All analyses were performed with 2-tailed tests. Data were log-transformed for analysis if skewness in the residuals was observed. Correlations were calculated using Pearson correlation coefficients. Significance was accepted as P less than .05.

Results

The baseline characteristics of the participants are summarized in Table 2 and appeared evenly distributed across groups. As previously reported (22, 23), greater reductions in body weight and fat mass were observed in IF70 vs CR70 and IF100 (all P < .05, Fig. 1A and 1B). Greater reductions in NEFA levels were observed in IF70 vs CR70 on fed days (P = .02, Fig. 1D) and in RQ in CR70 and IF70 vs IF100 (both P < .05, Fig 1F), whereas the change in insulin sensitivity by clamp was not statistically different between groups (Fig. 1C). During the fed to 24-hour fasted transition, NEFA and BHB levels were increased and insulin sensitivity was reduced in both IF groups (all P < .05, Fig. 1C-1E); RQ was significantly reduced in the IF100 group only (P < .01, Fig. 1F).

Table 2.

Anthropometric and metabolic measures of participants at baseline

Variable	DR70	IF70	IF100
No.	26	25	25
Ethnicity	1 AS/22 W/3 NA	1 AA/18 W/6 NA	1 AA/20 W/4 NA
Age, y	51 ± 2	50 ± 2	51 ± 2
Premenopause/postmenopause	12/14	13/12	10/15
Weight, kg	88.4 ± 2.8	89.4 ± 2.8	84.1 ± 2.8
Height, cm	164.7 ± 1.0	166.0 ± 1.7	162.2 ± 1.5
Waist, cm	99.0 ± 1.8	100.5 ± 2.2	98.8 ± 2.6
Hip, cm	115.7 ± 2.4	115.1 ± 2.1	112.1 ± 1.9
Waist to hip ratio	0.86 ± 0.01	0.87 ± 0.02	0.88 ± 0.02
BMI, kg/m2	32.6 ± 1.0	32.4 ± 0.8	31.2 ± 0.9
Triglyceride, mmol/L	1.3 ± 0.1	1.2 ± 0.1	1.5 ± 0.1
Total cholesterol, mmol/L	4.9 ± 0.1	4.8 ± 0.1	5.0 ± 0.2
HDL cholesterol, mmol/L	1.4 ± 0.1	1.4 ± 0.1	1.4 ± 0.2
LDL cholesterol, mmol/L	3.0 ± 0.1	2.9 ± 0.1	3.0 ± 0.2
Glucose, mmol/L	4.9 ± 0.1	4.9 ± 0.1	4.9 ± 0.1
hs-CRP, mg/L	2.7 ± 0.5	2.9 ± 0.5	2.8 ± 0.5

Data are presented as mean ± SEM.

Abbreviations: AS, Asian; AA, Aboriginal Australian; BMI, body mass index; W, White; NA, not available because participants did not provide this information; DR70, continuous energy restriction at 70% baseline energy requirements; HDL, high-density lipoprotein; hs-CRP, high-sensitivity C-reactive protein; IF70, intermittent fasting diet at 70% baseline energy requirements; IF100, intermittent fasting diet at 100% baseline energy requirements; LDL, low-density lipoprotein.

Figure 1.

Eight weeks of intermittent fasting (IF) differentially altered body weight and lipid metabolism. The change in A, body weight; B, fat mass; C, insulin sensitivity assessed by hyperinsulinemic-euglycemic clamp; D, nonesterified fatty acids (NEFA); E, β-hydroxybutyrate (BHB); and F, respiratory quotient (RQ) from baseline to week 8 following an overnight 12-hour fast (open bar) and to week 8 following a 24-hour fast (solid bar). AU, arbitrary unit; CR70, daily calorie restriction at 70% baseline energy requirements; FFM, fat-free mass; IF70, intermittent fasting diet at 70% baseline energy requirements; IF100, intermittent fasting diet at 100% baseline energy requirements. ^P less than .05 vs CR70, ^‡P less than .05 vs IF100, and ^#P less than .05 vs IF70 in the change from baseline to week 8 following 12-hour fast; *P less than .05; ^**P less than .01, and ^***P less than .001.

There was no statistical significance between groups in the change in mRNA levels of any lipolysis or lipogenesis markers assessed in skeletal muscle after the fed day (Fig. 2A-2E). The fed to 24-hour fasted transition increased mRNA levels of PLIN5 (both P < .05, Fig. 2D), and tended to increased lipase E, hormone sensitive type (LIPE1) and diacylglycerol O-acyltransferase 1 (DGAT1) mRNA expression in both IF groups (Fig. 2A and 2E). Within the combined IF groups, the change in PLIN5 mRNA following a 24-hour fast was positively correlated with the change in serum NEFA levels (r = 0.415, P = .044, Fig. 2F).

Figure 2.

Intermittent fasting (IF) and caloric restriction (CR) did not differentially affect markers of lipid synthesis or breakdown in muscle, but markers of lipid storage were stimulated by 24 hours of fasting. The change in messenger RNA (mRNA) levels of A, LIPE; B, ACACA; C, FASN; D, PLIN5; and E, DGAT1 in skeletal muscle from baseline to week 8 following an overnight 12-hour fast (open bar) and to week 8 following a 24-hour fast (solid bar). F, Correlation between changes in nonesterified fatty acid (NEFA) levels and PLIN5 mRNA expression in skeletal muscle after a 24-hour fast. AU: arbitrary unit; CR70, daily calorie restriction at 70% baseline energy requirements; IF70: intermittent fasting diet at 70% baseline energy requirements; IF100: intermittent fasting diet at 100% baseline energy requirements. LIPE, lipase E, hormone sensitive type; ACACA, acetyl-CoA carboxylase α; FASN, fatty acid synthase; DGAT1, diacylglycerol O-acyltransferase 1; open dots, IF100; PLIN5, perilipin 5; solid dots: IF70. *P less than .05.

There was a greater reduction in the mRNA levels of CD36 molecule (CD36) in IF100 vs CR70 only (P < .05, Fig. 3A) and acetyl-CoA acetyltransferase 1 (ACAT1) in IF70 and IF100 vs CR70 (both P < .05, Fig. 3E). There was no statistical significance between groups in the change of other markers of lipid oxidation (peroxisome proliferator-activated receptor α [PPARα], carnitine palmitoyltransferase 1 [CPT1], and pyruvate dehydrogenase kinase 4 [PDK4], Fig. 3B-3D) or ketone metabolism (monocarboxylate transporter 1 [MCT1], MCT4, and 3-oxoacid CoA-transferase 1 [OXCT1], Fig. 3F-3H). The fed to 24-hour fasted transition decreased the mRNA levels of PPARα, CPT1, and MCT1 within the IF100 group only (all P < .05, Fig. 3B, 3C, and 3F). Within the combined IF groups, the change in BHB levels following a 24-hour fast was positively correlated with the change in the mRNA levels of MCT1 (r = 0.681, P < .001, Fig. 3I), ACAT1 (r = 0.596, P < .001, figure not shown), and OXCT1 (r = 0.440, P = .041, figure not shown).

Figure 3.

Intermittent fasting (IF) and caloric restriction (CR) did not increase markers involved in lipid or ketone oxidation in skeletal muscle. The change in messenger RNA (mRNA) levels of A, PPARα; B, CD36; C, CPT1; D, PDK4; E, MCT1; F, MCT4; G, OXCT1; and H, ACAT1 in skeletal muscle from baseline to week 8 following an overnight 12-hour fast (open bar) and to week 8 following a 24-hour fast (solid bar). I, Correlation between changes in β-hydroxybutyrate (BHB) levels and MCT1 mRNA expression in skeletal muscle after a 24-hour fast. ACAT1, acetyl-CoA acetyltransferase 1; AU, arbitrary unit; CR70, daily calorie restriction at 70% baseline energy requirements; IF70, intermittent fasting diet at 70% baseline energy requirements; IF100, intermittent fasting diet at 100% baseline energy requirements; open dots, IF100; PPARα, peroxisome proliferator-activated receptor α; CD36, CD36 molecule; CPT1B, carnitine palmitoyltransferase 1B; PDK4, pyruvate dehydrogenase kinase 4; MCT1, monocarboxylate transporter 1; MCT4, monocarboxylate transporter 4; OXCT1, 3-oxoacid CoA-transferase 1; solid dots, IF70. ^P less than 0.05 vs CR70 in the change from baseline to week 8 following a 12-hour fast; *P less than .05; and ^**P less than .01.

The change in the mRNA levels of markers involved in mitochondrial metabolism (PGC1α, sirtuin 3 [SIRT3], and mitofusin 2 [MFN2], Fig. 4A-4C) were not statistically significant between groups, except for a greater reduction in TFAM mRNA levels in the IF100 vs CR70 group (P < .05, Fig. 4D). Greater reductions in glutathione peroxidase 1 (GPX1), SOD1, and SOD2 mRNA levels were observed in both IF groups vs CR70 (all P < .05, Fig. 4E and 4F) but the change in serum protein carbonyl levels was not statistically significant between groups (Fig. 4G). The fed to fasted transition did not alter any mitochondrial or oxidative stress markers in either IF group (Fig. 4A-4H). There was no association between change in mRNA levels of markers of mitochondrial metabolism, redox homeostasis, insulin sensitivity, or weight loss.

Figure 4.

Intermittent fasting (IF) decreased markers involved in mitochondrial metabolism and redox homeostasis in skeletal muscle without altering systemic oxidative stress. The change in messenger RNA levels of A, PGC1α; B, SIRT3; C, MFN2; D, TFAM; E, GPX1; F, SOD1; and G, SOD2 in skeletal muscle, and H, serum protein carbonyl from baseline to week 8 following an overnight 12-h fast (open bar) and to week 8 following a 24-hour fast (solid bar). AU: arbitrary unit; CR70, daily calorie restriction at 70% baseline energy requirements; GPX1, glutathione peroxidase 1; IF70, intermittent fasting diet at 70% baseline energy requirements; IF100, intermittent fasting diet at 100% baseline energy requirements; MFN2, mitofusin 2; PGC1α, peroxisome proliferator-activated receptor γ coactivator 1-α; SIRT3, sirtuin 3; SOD1, superoxide dismutase 1, soluble; SOD2, superoxide dismutase 2, soluble; TFAM, transcription factor A, mitochondrial. ^P less than .05 and ^^P less than .01 vs CR70 in the change from baseline to week 8 following a 12-hour fast.

Discussion

The present study did not detect significance in the ability of daily CR or IF to alter mRNA levels of markers of lipid or mitochondrial metabolism when assessed following an overnight fast. However, the mRNA levels of antioxidant enzymes were reduced in IF vs CR, despite no change in serum protein carbonyls, suggesting either reduced production of ROS or that alternative antioxidant forces are upregulated by IF to maintain balance in systemic redox homeostasis. The intermittent 24-hour fast transiently increased circulating NEFAs and PLIN5 mRNA levels in muscle, suggesting increased lipid droplet formation. Although PLIN5 is generally recognized to protect against lipotoxicity and stimulate lipid oxidation (20), this was insufficient to prevent transient fasting-induced insulin resistance in women with obesity.

The mitochondria are the main sites oxidizing lipids and generating ROS, leading to oxidative damage (25). To counterbalance this, an antioxidant defense network exists. GPX and SOD are antioxidants that dismutate superoxide radicals and degrade hydrogen peroxides to harmless molecules (26). A key finding from this study is that compared to CR, IF decreased GPX1, SOD1, and SOD2 mRNA levels in skeletal muscle. This could be interpreted that IF was associated with a reduction in ROS production, meaning that lower levels of antioxidant enzymes were required to maintain systemic redox homeostasis, as assessed by serum protein carbonyls. Alternatively, the increased generation of ketones by IF may have provided an alternative antioxidant service to scavenge ROS directly. In vitro, ketones scavenge hydroxyl radical (^•OH) and acetoacetate, capable of neutralizing singlet oxygen (¹O₂), hypochlorous acid (HOCl), and peroxynitrite (ONOO⁻) (27), which may lower requirements for other antioxidant enzymes, and decreased the mRNA expression of catalase, GPX, and SOD in cultured hepatocytes (28). However, ketone levels were low after fed days in all groups, when the reduction in antioxidant enzyme expression was detected.

Both CR and IF70 stimulate whole-body lipid oxidation as evidenced by a reduction in RQ (29, 30), although some studies suggest that IF provides greater stimulus for lipid oxidation (16), and improved glucose metabolism (31). In this study, greater reductions in body weight, fat mass, NEFAs, and cholesterol levels were observed in IF70 vs CR70 (22), but there was no statistical significance between groups in the change in RQ or the mRNA levels of markers of lipid or mitochondrial metabolism. Our results agree with a previous study that showed 12 months of CR do not alter mitochondrial function or the levels of genes or proteins involved in mitochondrial function and lipid oxidation in human skeletal muscle, although reduced intramyocellular lipid was observed (13).

In response to an acute 24-hour fast, muscle rapidly switches from carbohydrate to lipid oxidation (30). In this study, marked elevations in NEFAs and BHB, and a reduction in RQ were observed during the fed to fasted transition, which indicates a switch toward lipid metabolism and high adherence to the fasting protocol. However, no changes were detected at the mRNA level in muscle of markers of lipid or ketone oxidation in the IF70 group. Few studies have reported the effects of cycles of repeated fasting and refeeding on markers of lipid or ketone oxidation. Three weeks of alternate-day fasting did not alter the mRNA levels of markers of lipid metabolism in lean individuals (32). One acute 24-hour fast did not alter the mRNA level of acetyl-CoA carboxylases, acyl-CoA dehydrogenases for short-, branched-, straight-, and very long-chain fatty acids as well as acyl-CoA oxidase, although reductions in the mRNA levels of PPARα and CD36 were observed in 14 individuals of normal to overweight (33). In 12 young and healthy men and women, 15-hour and 40-hour fasts also did not alter the mRNA expression of genes responsible for lipid uptake and oxidation, including fatty acid-binding protein, CD36, CPT1, and long-chain acyl-CoA dehydrogenase (34). Similarly, an acute 48-hour fast did not alter the expression of genes responsible for ketone oxidation in muscle in young adults (33), but 72-hour fasting upregulated these markers (35, 36). Tsintzas et al also demonstrated that increased utilization of fatty acids did not require an adaptive response at the transcriptional level in CD36 or CPT1, but showed inhibition of glucose oxidation via upregulating pyruvate dehydrogenase kinase 4 (PDK4) (37). Although PDK4 mRNA levels tended to increase during the fed to fasted transition, this did not reach statistical significance in this study.

Unexpectedly, we observed reduced mRNA expression of PPARα, CPT1, MCT1, and TFAM during the fed to fasted state in the IF100 group. That group was included to determine whether weight loss is required for the metabolic health benefits of IF. As we reported previously (22), this group experienced poorer clinical outcomes, including transient increases in homeostatic model assessment of insulin resistance and smaller reductions in total and low-density lipoprotein cholesterol than IF70, which led us to the conclusion that weight loss is required for the beneficial health effects of IF (22). To maintain overall energy balance, the IF100 group was prescribed intermittent overfeeding at approximately 145% of energy requirements during the fed days. We speculate the transient reduction in genes involved in lipid oxidation could have been in response to the overfed state; however, we have previously shown that 3 days of continuous overfeeding transiently increased markers of lipid oxidation and mitochondrial function in muscle (38).

Increased lipid deposition in skeletal muscle has been reported in humans with obesity (1-5) but also occurs in response to acute, prolonged fasts, and both are associated with development of insulin resistance (17, 18, 35). However, developing insulin resistance in response to fasting is a protective mechanism to spare glucose that is physiologically different from the insulin resistance that develops in response to obesity, which is coassociated with hyperglycemia and hyperinsulinemia (39). In this study, 24 hours of intermittent fasting induced transient insulin resistance in both groups after 8 weeks. Fasting for 60 hours has previously been shown to increase the number and size of lipid droplets in muscle, as well as the fraction of PLIN5 protein associated with lipid droplets in lean, normoglycemic males. This increase in PLIN5 was partly protective against lipid-induced lipotoxicity, as males with the most prominent increase in PLIN5-associated lipid droplets showed the least reduction in insulin sensitivity (20). Animal studies have shown that PLIN5 depletion induced insulin resistance in skeletal muscle as assessed by clamp (40), although skeletal muscle–specific overexpression of PLIN5 did not alter insulin sensitivity (41). As little as 24 hours of fasting is sufficient to increase lipid deposition within skeletal muscle (21), although the effects of repeated cycles of feasting and fasting are unknown. In the present study, the mRNA levels of PLIN5 were elevated after fasting, and this correlated with an increase in serum NEFAs, suggesting a transient increase in lipid-droplet formation and storage within the myocyte in response to IF. However, the increase in PLIN5 mRNA was not sufficient to prevent, and did not appear to influence the degree of, insulin resistance, as we did not observe a correlation between the change in PLIN5 and the reduction in insulin sensitivity. Although we also did not detect a change in CD36 mRNA levels as a marker of fatty acid uptake, this does not preclude increased uptake of lipid into the myocyte, as the activity of CD36 was not assessed. Our results also suggest that fasting may have increased conversion of existing deleterious lipid intermediates within the myocyte into triglyceride, as a trend toward increased DGAT1 mRNA was observed, which functions to convert diacylglycerol to triacylglycerol. Unfortunately, the histological quantification of lipid deposition in response to IF was not assessed because of a lack of biopsy tissue. The clinical implications of a transient increase in insulin resistance in response to IF are as yet unclear, but require further investigation, particularly once weight loss has slowed, or if compensatory overfeeding is evident.

Unlike other IF studies in which fasting was introduced at lunch or dinner time (42-44), a strength of this study was that participants commenced their 24-hour fast immediately after they had consumed breakfast. This timing was selected to mitigate any potential adverse impact that skipping breakfast may have to induce circadian misalignment (45-47). The limitations of this study include the small sample size, that muscle biopsies were obtained from a subset of individuals in the intervention groups and were not obtained from the control group, and that analysis was limited to the mRNA level. Further, the study was a highly controlled, short-term intervention that was performed solely in women with overweight or obesity and thus, responses in free-living populations, men, or in individuals with normal body weight require further investigation.

This study highlights that IF reduced the mRNA expression of antioxidant enzymes as compared to CR, potentially reflecting reduced ROS production. IF also transiently increased mRNA levels of PLIN5, a marker for lipid-droplet formation, but this was not sufficient to prevent transient insulin resistance in response to a 24-hour fast in women with overweight and obesity.

Abbreviations

Abbreviations
 
• ATP

  adenosine 5′-triphosphate

 
• BHB

  β-hydroxybutyrate

 
• CR

  calorie restriction

 
• CR70

  daily calorie restriction at 70% baseline energy requirements

 
• IF

  intermittent fasting

 
• IF70

  intermittent fasting diet at 70% baseline energy requirements

 
• IF100

  intermittent fasting diet at 100% baseline energy requirements

 
• mRNA

  messenger RNA

 
• NEFAs

  nonesterified fatty acids

 
• PLIN5

  protein perilipin 5

 
• ROS

  reactive oxygen species

 
• RQ

  respiratory quotient

 
• SOD

  superoxide dismutase

 
• TFAM

  mitochondrial transcription factor A

Acknowledgments

The authors thank Briohny Johnston for her assistance in recruiting, screening, and conducting metabolic visits, and all the volunteers who participated in this research study. Parts of this study were presented as an oral presentation at the Australian & New Zealand Obesity Society Annual Scientific Meeting, October 16 to 18, 2019; Sydney, Australia.

Financial Support: This work was supported by the National Health and Medical Research Council (Project Grant APP1023401) and an Australian Research Council Future Fellowship (FT120100027 to L.K.H.).

Author Contributions: B.L. and A.T.H. performed the study, and collected and analyzed the data. G.A.W. and C.H.T. provided clinical support. L.K.H. and G.A.W. designed the study. K.L. performed the statistical analysis. All authors contributed to data interpretation and preparation of the manuscript. L.K.H. is the guarantor of this study.

Additional Information

Disclosure Summary: The authors have nothing to disclose.

Data Availability

The data sets generated during and/or analyzed during the present study are not publicly available but are available from the corresponding author on reasonable request.