Effectiveness of an intermittent fasting diet versus regular diet on fat loss in overweight and obese middle-aged and elderly people without metabolic disease: a systematic review and meta-analysis of randomized controlled trials

Abstract

Objective

As the number of adults aged over 40 with obesity increases dramatically, intermittent fasting interventions (IF) may help them to lose fat and weight. This systematic review investigated the most recent research on the effects of intermittent fasting and a regular diet on body composition and lipids in adults aged over 40 with obesity without the metabolic disease.

Data sources

Randomized controlled trials (RCTs) on IF on adults aged over 40 with obesity were retrieved from PubMed, Web of Science, EBSCO, China Knowledge Network (CNKI), VIP database, Wanfang database with the experimental group using IF and the control group using a regular diet. Revman was used for meta-analysis. Effect sizes are expressed as weighted mean differences (WMD) and 95% confidence intervals (CI).

Study selection

A total of 9 articles of randomised controlled trials that met the requirements were screened for inclusion. Studies typically lasted 2–6 weeks. The experimental population was aged 42–66 years, with a BMI range of 25.7−35 kg/m².

Synthesis

A total of 9 RCTs were included. meta-analysis showed that body weight (MD: −2.05 kg; 95% CI (−3.84, −0.27); p = 0.02), BMI (MD: −0.73 kg/m²; 95% CI (−1.05, −0.41); p < 0.001), fat mass (MD: −2.14 kg; 95% CI (−3.81, 0.47); p = 0.01), and TG (MD = −0.32 mmol/L, 95% CI (−0.50, −0.15, p < 0.001) were significantly lower in the experimental group than in the control group. No significant reduction in lean body mass (MD: −0.31 kg; 95% CI (−0.96, 0.34); p = 0.35).

Conclusion

IF had a reduction in body weight, BMI, fat mass, and TG in adults aged over 40 with obesity without metabolic disease compared to RD, and IF did not cause a significant decrease in lean body mass, which suggests healthy and effective fat loss. However, more long-term and high-quality trials are needed to reach definitive conclusions.

1. Introduction

With the increasing number of overweight and obese people worldwide, the healthcare burden has increased significantly [1,2]. The problem of overweight and obesity has caused many problems in people's lives, and at the same time, population aging is an important issue for human society in the 21 st century, and the adults aged over 40 with obesity is expanding. There were 20.9 million obese people aged 60 or over in the US in 2010 and 32 million obese older people in the EU in 2012 [3]. The middle-aged and older population is defined as those over 40 years of age, the World Health Organization (WHO) refers to 60–74 years of age as the young older adults, for those over 75 years of age as the older adults, and for those over 90 years of age as the long-lived older adults [[4], [5], [6]]. Overweight and obesity are major factors affecting the health of middle-aged and elderly people without metabolic diseases. Overweight and obesity increase the risk of type 2 diabetes, hypertension, cardiovascular disease, stroke and various cancers in middle-aged and older adults [2]. The current interventions for obesity are mainly through diet and exercise and medication. Drugs have some negative effects, and adults aged over 40 with obesity are less able to exercise, are at greater risk of exercise injury, and tend to give up in the middle of the process. Therefore, a dietary intervention model may be more effective in adults aged over 40 with obesity without the metabolic disease. Intermittent fasting (IF) as a new non-drug weight loss therapy and the high feasibility of diet weight loss mode has also attracted extensive research by scientists [[7], [8], [9], [10]].

IF is mainly divided into alternate-day fasting (ADF), 5:2 fasting (2DW), time-restricted eating (TRF), and Ramadan fasting (FCR) [7,11,12]. Alternate-day fasting (ADF) is one of the most common patterns of IF in current research. ADF refers to a dietary approach that alternates between 1 day of fasting (or a low-energy diet with intake of 20%–25% of normal energy) and 1 day of unrestricted intake (or intake of 125%–150% of normal energy) [7]. 5:2 fasting (2DW) generally follow a 5 + 2 pattern. Two non-consecutive days in a week are usually chosen as fasting days, and a very low-calorie diet is adopted on the fasting days, with a typical intake of 1/4 of a normal day, typically 600 kcal/d for men and 500 kcal/d for women, and a normal intake on the remaining 5 d of non-fasting days. During the fasting days, the food is mainly fresh fruits and vegetables, whole grains, soy products and skim milk [8]. Time-restricted eating (TRF) is defined as restricting energy intake or fasting (without restricting water intake) for only a specific period of time (usually within 8−12 h) on 1 day, with free intake for the rest of the day [12]. Ramadan fasting (FCR) is a special form of intermittent fasting, Ramadan is the ninth month of the Islamic lunar calendar, during which Muslims are not allowed to eat or drink from sunrise to sunset every year. Before sunrise and after sunset, they can eat, during which time they reduce their caloric intake. This dietary pattern lasts for one month [11].

Up to now, numerous epidemiological and clinical trials on IF have been conducted in animal and human subjects, showing improvements in body weight, total cholesterol (TC) [13], and body composition [14,15]. However, previous meta-analyses have shown the effects of an intermittent fasting diet versus continuous energy restriction on anthropometric measurements, body composition, and lipid profile in overweight and obese young people, people with type 2 diabetes, and other metabolic diseases [14], with less attention paid to middle-aged and elderly populations, and the effects of intervention-induced lipids, body composition, and body morphology are not well understood. Some meta-analyses had no control group and only compared body parameters before and after the intervention [16], the lack of use of the regular diet (RD) as a control group for comparison may affect the validity of the results. and the larger amount of literature included in the meta-analysis on IF was not all RCTs, and the quality of the literature was not evaluated [13,17]. A regular diet is one that does not involve dieting, has no calorie restriction, and maintains the original eating patterns. IF results in weight loss compared to RD [12,13], however whether the weight loss is due to body fat loss or lean body mass loss or both or just one of them is unclear. Intermittent fasting results in different body compositions also due to gender differences, women aged over 40 with obesity may be somewhat in menopause compared to men, they are at high risk of developing sarcopenia, and hormonal changes may result in loss of lean weight loss [[18], [19], [20]]. Also, the above meta-analysis is more often without subgroup analysis of IF type, intervention area, obesity level, etc. This meta-analysis was also planned by subgroup analysis according to WHO severity of obesity (overweight [25 < BMI < 29.9], mild obesity [30 < BMI < 34.9], moderate obesity [35 < BMI < 39.5], severe obesity [40 < BMI]) [4,21]; study duration. IF type (ADF, TRF, 5:2 light fasting, fasting); and intervention area factors for some subgroup analysis of the primary outcome.

This meta-analysis systematically evaluated the effects of IF on lipids, body composition, and body morphology in adults aged over 40 with obesity without metabolic disorders. Updated and compared the efficacy of different IF modalities on lipid and weight loss in a middle-aged and elderly population [22]. The included literature was all RCTs and the quality of the literature was evaluated. Subgroup analysis (type of intervention, duration, region, severity of obesity) was performed for the primary outcome. This could help the adults aged over 40 with obesity to view the impact of IF on their fat and weight loss outcomes and alleviate obesity in the middle-aged and elderly population. Since intermittent fasting has certain risks for people with metabolic diseases, such as diabetes mellitus, eating disorders, and anorexia nervosa. Therefore, this meta-analysis is instructive for the diet of the adults aged over 40 with obesity without metabolic disease and provides a reference for subsequent experimental studies and practice.

2. Data sources

2.1. Data sources and search strategies

Articles were searched on 5 July 2022. The databases used in this meta-analysis were PubMed, EBSCO, Web of Science, CNKI, Wanfang, and VIP. Different databases are searched for using MeSH subject terms and the AND and OR Boolean operators (Table 1A). A complete search rate is ensured by hand-searching references of relevant studies.

Table 1A.

Keywords used in the article search.

Database	Search formula
PubMed	(“intermittent fasting”OR“intermittent energy restriction”OR“intermittent calorie restriction”OR“alternative day fasting”OR“periodic fasting”OR“fast diet”OR“time restricted feeding”OR“fasting, intermittent”OR“hunger strike”OR“hunger strikes”OR“strikes,hunger”OR“time restricted feeding”OR“feeding, time restricted”OR“time restricted feedings”)AND(“weight loss”OR“obesity”OR“overweight”OR“excess weight”)AND(“body shape”OR“physique”OR“body form”OR“body composition”)
Web of Science	(“intermittent fasting”OR“intermittent energy restriction”OR“intermittent calorie restriction”OR“alternative day fasting”OR“periodic fasting”OR“fast diet”OR“time restricted feeding”OR“fasting, intermittent”OR“hunger strike”OR“hunger strikes”OR“strikes,hunger”OR“time restricted feeding”OR“feeding, time restricted”OR“time restricted feedings”)AND(“weight loss”OR“obesity”OR“overweight”OR“excess weight”)AND(“body shape”OR“physique”OR“body form”OR“body composition”)
EBSOC	(“intermittent fasting”OR“intermittent energy restriction”OR“intermittent calorie restriction”OR“alternative day fasting”OR“periodic fasting”OR“fast diet”OR“time restricted feeding”OR“fasting, intermittent”OR“hunger strike”OR“hunger strikes”OR“strikes,hunger”OR“time restricted feeding”OR“feeding, time restricted”OR“time restricted feedings”)AND(“weight loss”OR“obesity”OR“overweight”OR“excess weight”)AND(“body shape”OR“physique”OR“body form”OR“body composition”)
CNKI	("intermittent fasting" OR "light fasting" OR "intermittent fasting" OR "TRF fasting" OR "TRF fasting" OR "cycle fasting" OR "cycle fasting") AND ("obese" OR "overweight")
VIP database	(Intermittent fasting OR Light fasting OR Intermittent fasting OR TRF fasting OR TRF fasting OR Periodic fasting OR Periodic fasting with obesity OR Overweight)
Wanfang database	("intermittent fasting" OR "light fasting" OR "intermittent fasting" OR "TRF fasting" OR "TRF fasting" OR " periodical fasting" OR "periodical fasting") AND ("obese" OR "overweight")

The reference lists from eligible studies were screened to identify additional relevant research. Screening and study selection were conducted by two authors independently.

2.2. Selection of studies

Studies were included in the systematic review if they met the following criteria:

• (1)A randomized controlled trial in humans, the full text of which is available and written in English.
• (2)Participants >40 years of age (middle-aged people age 40–60, the young older adults age 60–74, the older adults age 75 and above, and the long-lived older adults age 90 and above).
• (3)The experimental group underwent IF intervention alone (time-restricted fasting (TRF), alternate-day fasting (ADF), and religion-related fasting (FCR), 5:2 design (2DW) fasting pattern);
• (4)The control group did not receive dietary intervention and maintained their original dietary habits.
• (5)Overweight and obese people with BMI > 24 were included and classified into obesity severity (overweight [24 < BMI < 28], mild obesity [28 < BMI < 32.5], moderate obesity [32.5 < BMI < 37.5], severe obesity [37.5 < BMI]).
• (6)The study contains outcome indicators including body morphology indicators (body weight, body mass index [BMI]); body composition indicators (body fat percentage [BFP], fat mass, lean body mass); blood lipids level indicators (total cholesterol [TC], triglycerides [TG], low-density lipoprotein [LDL], high-density lipoprotein [HDL]).
• (7)The study populations of the literature included in the analysis were without metabolic disorders, such as the absence of hyperglycemia, hypertension, or hyperlipidemia.

Studies were excluded from the systematic review if they met the following criteria:

• (1)There is a lack of a control group in the literature, and the full text and abstract of the study are not available, so the study is not relevant.
• (2)Study involved patients with Obstructive Sleep Apnea (OSA) and Fatty liver disease.
• (3)The study was not an RCT, such as meta-analyses, conference abstracts, reviews, case reports, and other secondary analyses.
• (4)The study population had chronic diseases and the study population was not overweight or obese middle-aged adults.
• (5)The study population was bariatric surgery patients, critically ill and hospitalized patients, and patients with eating disorders.

2.3. Data extraction

Two reviewers separately retrieved the following traits for the articles that were included: name of the first author, year of publication, study origin (country), different types of fasting methods, sample size, sex (male/female), mean age, mean BMI, study length (weeks).Relevant ending indicators (primary outcomes: body mass index (BMI), body weight, body fat percentage (BFP), fat-free mass (FFM), fat mass; secondary outcomes: low-density lipoprotein cholesterol(LDL-C), high-density lipoprotein cholesterol (HDL-C), triacylglycerols (TG), total cholesterol (TC). If data were missing, the corresponding author was contacted to ask for the original data and if this was unsuccessful, the Cochrane manual formula was used to estimate the mean and standard deviation.

2.4. Risk of bias assessment

The risk of bias was assessed by two authors using the Risk-of-bias tool (ROB2) to evaluate the quality of the literature for bias where five domains were present. These were: (1) randomization process, (2) biased established interventions, (3) missing outcome data, (4) outcome measurement, and (5) selective reporting of outcomes. All RCTs were assessed as having three levels of risk of bias: high, moderate, and low.

2.5. Data analysis

Statistical analysis was performed using Review Manager 5.4. Data synthesis was performed for outcomes that were reported in at least three studies. The pooled results were expressed as the weighted mean differences (WMD) and 95% confidence intervals (CI) of mean outcome values measured at the end of follow-up between the intermittent fasting (IF) and control (RD) arms in each randomized controlled trial (RCT) study. Fixed-effects model analysis was used P < 0.1and I² < 50% as reported in the literature. We used random effects analysis when the literature showed P > 0.1, I² > 50% or I² = 50%. We performed subgroup and sensitivity analyses to explore I² values greater than 50%. We intended to conduct subgroup analyses of the primary outcomes based on the following factors: namely, severity of obesity (overweight [25 < BMI < 29.9], mild obesity [30 < BMI < 34.9], moderate obesity [35 < BMI < 39.5], severe obesity [40 < BMI]) [4,21], study duration (6 or 12 weeks); and types of intermittent fasting (ADF, TRF, FCR, and 2DW).

3. Synthesis

3.1. Search results

The search was conducted in six database engines, and review papers on similar topics were also searched to ensure completeness. A total of 2300 were selected for screening based on the search strategy. The titles and abstracts (occasionally including a methods section) of 1235 studies were reviewed after removing 1065 duplicate papers. Papers from 1193 were excluded based on the PICO not being relevant to this meta or not meeting the inclusion and exclusion criteria. After assessing full-text eligibility, the search was narrowed to 16 studies, but after careful and critical reading, only 9 studies were ultimately included in meta-analysis [[23], [24], [25], [26], [27], [28], [29], [30], [31]] (Fig. 1).

Fig. 1.

Flow of the study.

3.2. Characteristics of the included studies

As shown in Table 1B, a total of 9 articles [22] were included in this meta-analysis. There were 13 studies involved in those 9 articles. The experimental group used IF, IF interventions were primarily FCR, followed by TRF and ADF. The control group used RD, RD means that there is no change in diet, no dietary intervention, and the original diet is maintained. Exercise interventions were performed in two of the groups, but the experimental and control groups exercised in exactly the same way and with the same amount of exercise, which had less of an effect on the observation of fasting. Three RCTs were conducted in the US [24,27,30], three in Malaysia [28,29,31], and one each in Poland [25], Germany [26], and China [23]. The study length ranged from 6 to 12 weeks; the mean age of the participants ranged from 42 to 66 years, and their BMI was from 25.7 to 35 kg/m². One RCT [1] (102 subjects in total) compared 5:2 design (2DW) with RD; three RCTs [11] (129 subjects in total) compared TRF with RD; two RCTs [7] (113 subjects in total) compared ADF with RD, and three RCTs [8] (112 subjects in total) compared FCR with RD. In the RCT study in Malaysia, since the life expectancy of men is shorter than that of women. At the same time the health status of males is poor. Therefore the study of whether FCR can improve metabolic parameters was conducted only in men. In the China study, menopausal women often experience deterioration in vascular elasticity and elevated blood pressure due to hormonal changes. Therefore, this study examines the effects of TRF on body composition and cardiometabolic risk factors in pre and postmenopausal women.

Table 1B.

Characteristics of studies included in the meta-analysis.

		Age (year)	Sample size (sex)	BMI (kg/m2)	Weight (kg)	Interventions
Author(year)	Country	EG/CG	EG(M/F)	CG(M/F)	EG/CG	EG/CG	EG	CG	Duration	Outcome	Any significant differences in baseline characteristics
Domaszewski (2020)	Poland	65/66	0/25	0/20	28.99/27.61	69.93/69.66	TRF Complete fasting for 16 h a day from 20:00 pm until 12:00 am the following day	RD Maintain current lifestyle	6wk	①②③④	NA
Teng a (2013)	Malaysia	59.6/59.1	28/0	28/0	26.8/26.7	73.1/71.2	FCR FCR consisted of two days of Muslim Sunnah fasting, During fasting day, subjects needed to take a light meal before sunrise (Sahur), no food and drink on the day was allowed (approximately for 13 h) and a complete meal after sunset (Iftar).	RD Maintain current lifestyle	6wk	①②③④⑤⑥⑦⑧⑨	NA
Teng b (2013)	Malaysia	59.6/59.1	28/0	28/0	26.8/26.7	73.1/71.2	FCR FCR consisted of two days of Muslim Sunnah fasting, During fasting day, subjects needed to take a light meal before sunrise (Sahur), no food and drink on the day was Allowed (approximately for 13 h) and a complete meal after sunset (Iftar).	RD Maintain current lifestyle	12wk	①②③④⑤⑥⑦⑧⑨	NA
Kotarsky (2021)	USA	45/44	2/9	1/9	29.8 /29.4	166/168	TRF Fasting daily from 12:00 p.m. to 8:00 p.m. EG and CG complete three sets of 12 resistance training repetitions per week and aerobic training three or more days per week based on physical activity goals	RD Maintain current lifestyle	8wk	①②③⑦	NA
Yan (2022)	China	50.1/54.2	0/30	0/33	25.9/25.7	65.9/65.8	TRF Food is allowed only for 8 h and fasting for the remaining 16 h.	RD Maintain current lifestyle	8wk	①②⑤⑥⑦⑨	NA
Varady (2013)	USA	47/ 48	5/10	3/12	26/26	77/77	ADF Meals are provided on each fasting day (range 400−600 kcal) and are eaten freely at home on non-fasting days.	RD Maintain current lifestyle	12wk	⑥⑦⑧⑨	NA
Bhutani a (2013)	USA	45/42	0/18	1/23	35/35	91/93	ADF Consume 25% of baseline energy requirements on each fasting day and eat ad libitum on non-fasting days. EG and CG perform 60% of your maximum heart rate during 25 min of exercise three times a week.	RD Maintain current lifestyle.	12wk	①②④⑤⑥⑦⑧⑨	NA
Bhutani b (2013)	USA	42/ 49	1/24	1/15	35/35	94/93	ADF Consume 25% of baseline energy requirements on each fasting day and eat freely on non-fasting days.	RD Maintain current lifestyle	12wk	①②④⑤⑥⑦⑧⑨	NA
N.M. HUSSIN a (2012)	Malaysia	59.7/59.7	16/0	15/0	26.7/26.8	74.2/69.9	FCR Reduction in baseline energy intake from participants by 300–500 kcal/day, combined with two days per week of Muslim Sunni fasting.	RD Maintain current lifestyle	6wk	①②⑤	NA
N.M. HUSSIN b (2012)	Malaysia	59.7/59.7	16/0	15/0	26.7/26.8	74.2/69.9	FCR Reduction in baseline energy intake from participants by 300–500 kcal/day, combined with two days per week of Muslim Sunni fasting.	RD Maintain current lifestyle	12wk	①②⑤	NA
Schübel (2018)	Germany	49.4/50.7	25/24	25/27	32.0/31.1	96.4/93.3	2DW No energy restriction 5 days a week, 75% energy deficit 2 days a week, net energy deficit ≈ 20% per week	RD Maintain current lifestyle	12wk	⑥⑦⑧⑨	NA
Teng a (2011)	Malaysia	59.3/58.3	12/0	13/0	27.0/26.5	71.6/72.9	FCR Reduce your daily energy intake by 300–500 kcal/day and fast 2 days a week.	RD Maintain current lifestyle	6wk	①②④⑤	NA
Teng b (2011)	Malaysia	59.3/58.3	12/0	13/0	27.0/26.5	71.6/72.9	FCR Reduce your daily energy intake by 300–500 kcal/day and fast 2 days a week.	RD Maintain current lifestyle	12wk	①②④⑤	NA

① is Body Weight (Weight); ② is Body Mass Index (BMI); ③ is Fat Mass (FM); ④ is Fat Free Weight (FFM) or Lean Body Mass (FFM); ⑤ is Body Fat Percentage (BFP); ⑥ is Total Cholesterol (TC); ⑦ is Triglycerides (TG); ⑧ is Low Density Lipoprotein (LDL); ⑨ is High Density Lipoprotein (HDL).

EG is for the experimental group, CG is for the control group, M is for male, F is for female, and wk is for weekly.

None of the baseline characteristics were found to have any significant between-group differences.

As shown in Table 2, the outcome measures and p-values for all studies included in this meta-analysis.

Table 2.

Outcome measures and p-values for all studies.

		Weight (kg)	BMI (kg/m2)	BFP (%)	LBM (kg)	FM (kg)	TC (mmol/L)	TG (mmol/L)	HDL-c(mmol/L)	LDL-c (mmol/L)
Domaszewski (2020)	EG	68.6 ± 12.4*	27.7 ± 5.1*	NA	39.9 ± 4.8	28.7 ± 8.6*	NA	NA	NA	NA
	CG	70.2 ± 13.1*	27.6 ± 4.7*	NA	39.8 ± 6.1	30.4 ± 8.0*	NA	NA	NA	NA
Teng a (2013)	EG	71.1 ± 6.9+	26.3 ± 2.1*	26.3 ± 3.2+	53.2 ± 4.3	18.8 ± 3.3*	5.7 ± 0.9+	1.4 ± 0.7	1.1 ± 0.2	3.9 ± 0.7*
	CG	71.2 ± 8.2+	26.7 ± 2.3*	26.2 ± 3.0+	52.4 ± 5.3	19.1 ± 3.9*	5.6 ± 0.9+	1.9 ± 0.9	1.1 ± 0.2	3.6 ± 0.9*
Teng b (2013)	EG	70.6 ± 6.7+	25.9 ± 1.9*	25.1 ± 3.1+	54.0 ± 4.1	17.9 ± 3.3*	5.5 ± 0.9+	1.4 ± 0.4	1.1 ± 0.2	3.7 ± 0.8
	CG	71.2 ± 8.2+	26.7 ± 2.2	26.2 ± 3.1+	51.6 ± 8.8	19.1 ± 4.7*	5.6 ± 1.0+	1.8 ± 1.0	1.1 ± 0.2	3.7 ± 0.9
Kotarsky (2021)	EG	79.0 ± 3.0*	28.8 ± 0.8*	NA	47.0 ± 2.0*	30.0 ± 2.0*	5.3 ± 0.3	NA	1.4 ± 0.8	NA
	CG	83.0 ± 3.0*	29.3 ± 0.9*	NA	48.0 ± 2.0*	32.0 ± 2.0*	5.1 ± 0.3	NA	1.4 ± 0.1	NA
Yan (2022)	EG	63.2 ± 9.8*	24.5 ± 3.3*	31.3 ± 3.9	39.6 ± 4.5	NA	4.9 ± 0.8	NA	1.7 ± 0.4	2.8 ± 0.7
	CG	64.2 ± 8.5*	25.2 ± 3.6*	32.4 ± 3.9	39.5 ± 4.3	NA	4.9 ± 0.8	NA	1.6 ± 0.3	2.9 ± 0.8
Varady (2013)	EG	NA	NA	NA	NA	NA	4.5 ± 0.3	1.0 ± 0.1*	1.4 ± 0.1	2.6 ± 0.2
	CG	NA	NA	NA	NA	NA	5.2 ± 0.2	1.3 ± 0.2*	1.5 ± 0.1	3.1 ± 0.2
Bhutani a (2013)	EG	85.0 ± 6.0	33.0 ± 1.0	NA	46.0 ± 2.0	40.0 ± 2.0	4.8 ± 0.3	1.0 ± 0.1	1.5 ± 0.1	2.8 ± 0.3
	CG	92.0 ± 2.0	34.0 ± 1.0	NA	47.0 ± 1.0	45.0 ± 2.0	4.7 ± 0.2	0.9 ± 0.1	2.1 ± 0.1	2.9 ± 0.2
Bhutani b (2013)	EG	91.0 ± 3.0	34.0 ± 1.0	NA	50.0 ± 2.0	41.0 ± 2.0	4.7 ± 0.3	1.0 ± 0.1	1.3 ± 0.1	2.9 ± 0.2
	CG	93.0 ± 5.0	35.0 ± 1.0	NA	49.0 ± 2.0	43.0 ± 4.0	4.8 ± 0.3	1.2 ± 0.1	1.5 ± 0.1	3.2 ± 0.2
N.M. HUSSIN a (2012)	EG	72.1 ± 7.4+	26.2 ± 2.3*	25.9 ± 2.5*	NA	NA	NA	NA	NA	NA
	CG	69.4 ± 7.6+	26.7 ± 2.6*	26.9 ± 3.1*	NA	NA	NA	NA	NA	NA
N.M. HUSSIN b (2012)	EG	71.4 ± 7.2+	25.7 ± 1.8*	24.9 ± 2.5*	NA	NA	NA	NA	NA	NA
	CG	69.0 ± 7.6+	26.5 ± 2.5*	26.8 ± 3.2*	NA	NA	NA	NA	NA	NA
Schübel (2018)	EG	NA	NA	NA	NA	NA	4.8 ± 1.0	1.1 ± 0.5	1.3 ± 0.4	3.0 ± 0.7
	CG	NA	NA	NA	NA	NA	4.9 ± 0.8	1.4 ± 0.7	1.2 ± 0.3	3.1 ± 0.6
Teng a (2011)	EG	69.8 ± 6.3+	26.5 ± 1.9+	26.7 ± 4.1*	51.4 ± 4.3*	NA	NA	NA	NA	NA
	CG	73.3 ± 8.7+	26.7 ± 1.9+	25.4 ± 2.9*	54.5 ± 5.2*	NA	NA	NA	NA	NA
Teng b (2011)	EG	69.3 ± 6.0+	26.3 ± 1.9+	25.3 ± 3.8*	52.2 ± 4.1*	NA	NA	NA	NA	NA
	CG	73.7 ± 8.4+	26.8 ± 1.9+	25.5 ± 2.9*	54.8 ± 5.2*	NA	NA	NA	NA	NA

*: indicates P < 0.05, +: indicates P < 0.01. Body Mass Index (BMI); Fat Mass (FM); Lean Body Mass (FFM); Body Fat Percentage (BFP); Total Cholesterol (TC); Triglycerides (TG); Low Density Lipoprotein (LDL); High Density Lipoprotein (HDL).

3.3. Risk of bias

The results of the Rob assessment are provided in supplementary material. All of the studies [[24], [25], [26], [27], [28], [29], [30], [31]] used a randomization method, All trials did not meet the requirement of blinding the subjects. However, given the dietary interventions, it does not appear feasible to use a blinded approach. All trials did not appear to have selective outcomes reported and miss outcome data. Overall, seven RCTs [[24], [25], [26],[30], [31], [32]] were rated at high risk of bias due to the inability to perform blinding, blinding of outcome assessments.

3.4. Effectiveness of IF on anthropometric parameters (body weight, BMI)

3.4.1. Effectiveness of IF on body weight

Seven RCTs (11 data points) reported changes in body weight and meta-analysis showed I² = 50%, p = 0.02, with moderate heterogeneity between studies, and using random effects model analysis IF have a significant effect in reducing body weight compared to RD (MD: −2.05 kg; 95% CI (−3.84, −0.27); p = 0.02) (Fig. 2A). Subgroup analysis showed that body weight was effectively reduced in overweight or mild obesity people with a BMI less than 32.5 (MD = −1.63 kg, 95% CI: −3.08 to −0.18, p = 0.03). Subgroup analysis showed effective weight loss with the TRF intervention (MD = −3.14 kg, 95% CI: −5.29 to −1.00, p = 0.004). The results of the subgroup analysis are shown in Table 3.

Fig. 2A.

Effectiveness of IF on body shape and body composition. (a) Effectiveness of IF on body weight, (b) Effectiveness of IF on BMI, (c) Effectiveness of IF on body fat percentage, (d) Effectiveness of IF on lean body mass, (e) Effectiveness of IF on fat mass.

Table 3.

Subgroup analysis on body weight, fat mass, triglycerides, HDL, LDL of overweight and obese middle-aged and elderly people without metabolic disease.

Subgroup	Article	Data points	MD [95% CI]	P	I2 (%)	Pheterogeneity
Body weight
Severity of obesity
25 < BMI < 34.9	Domaszewski (2020) Teng a (2013) Teng b (2013) Kotarsky (2021) Yan (2022) N.M. HUSSIN a (2012) Teng a (2011) Teng b (2011) N.M. HUSSIN b (2012)	9	−1.63 [−3.08, −0.18]	0.03	23%	0.24
35 < BMI	Bhutani a (2013) Bhutani b (2013)	2	−4.46 [−9.36, 0.44]	0.07	82%	0.02
Intervention
TRF	Domaszewski (2020) Kotarsky (2021) Yan (2022)	3	−3.14 [−5.29, −1.00]	0.004	0%	0.49
ADF	Bhutani a (2013) Bhutani b (2013)	2	−4.46 [−9.36, 0.44]	0.07	82%	0.02
FCR	Teng a (2013) Teng b (2013) Teng a (2011) Teng b (2011) N.M. HUSSIN a (2012) N.M. HUSSIN b (2012)	6	−0.37 [−2.33, 1.59]	0.71	8%	0.37
Fat mass
Severity of obesity
25 < BMI < 34.9	Domaszewski (2020) Kotarsky (2021) Teng a (2013) Teng b (2013)	4	−1.25 [−2.31, −0.18]	0.02	0%	0.63
35 < BMI	Bhutani a (2013) Bhutani b (2013)	2	−3.62 [−6.55, −0.70]	0.02	81%	0.02
Triglycerides
Severity of obesity
25 < BMI < 34.9	Teng a (2013) Teng b (2013) Varady (2013) Schübel (2018)	4	−0.32 [−0.50, −0.15]	0.0002	0%	0.57
35 < BMI	Bhutani a (2013) Bhutani b (2013)	2	−0.04 [−0.31, 0.22]	0.75	97%	<0.001
High density lipoprotein
Intervention
ADF	Bhutani a (2013) Bhutani b (2013) Varady (2013)	3	−0.27 [−0.51, −0.02]	0.03	98%	<0.001
TRF	Kotarsky (2021) Yan (2022)	2	−0.00 [−0.07, 0.07]	0.92	0%	0.36
FCR	Teng a (2013) Teng b (2013)	2	−0.00 [−0.07, 0.08]	0.89	0%	0.69
Low density lipoprotein
Intervention
ADF	Bhutani a (2013) Bhutani b (2013) Varady (2013)	3	−0.31 [0.53, 0.08]	0.008	85%	0.001
FCR	Teng a (2013) Teng b (2013)	2	0.17 [−0.14, 0.47]	0.29	0%	0.48

Body weight, fat mass, and triglycerides were analyzed in subgroups according to severity of obesity. Body weight, HDL, LDL were subgrouped according to intervention modality.

Severity of obesity (overweight [25 < BMI < 29.9], mild obesity [30 < BMI < 34.9], moderate obesity [35 < BMI < 39.5], severe obesity [40 < BMI]).

The source of heterogeneity was searched for using an article-by-article exclusion, it found that the exclusion of the Bhutani literature reduced the heterogeneity to 23%, indicating that Bhutani was the main factor influencing the heterogeneity of the literature.

3.4.2. Effectiveness of IF on body BMI

Seven RCTs (11 data points) reported changes in BMI and meta-analysis showed I² = 0%, p < 0.001, with no significant heterogeneity between studies, and using fixed effects model analysis. IF has a significant effect in reducing BMI compared to RD (MD: −0.73 kg/m²; 95% CI (−1.05, −0.41); p < 0.001) (Fig. 2A).

3.5. Effectiveness of IF on body composition (body fat percentage, lean body mass, and fat mass)

3.5.1. Effectiveness of IF on body fat percentage

Four RCTs (7 data points) reported changes in body fat percentage and meta-analysis showed I² = 0%, p = 0.06, with no significant heterogeneity between studies, and using fixed effects model analysis. IF did not have a significant effect in reducing body fat percentage compared to RD (MD: −0.72%; 95% CI (−1.47, 0.03); p = 0.06) (Fig. 2A).

3.5.2. Effectiveness of IF on lean body mass

Six RCTs (9 data points) reported changes in lean body mass and meta-analysis showed I² = 35%, p = 0.35, with moderate heterogeneity between studies, and using fixed effects model analysis. IF did not have a significant effect in reducing lean body mass compared to RD (MD: −0.31 kg; 95% CI (−0.96, 0.34); p = 0.35) (Fig. 2A).

3.5.3. Effectiveness of IF on fat mass

RCTs (6 data points) reported changes in fat mass and meta-analysis showed I² = 75%, p = 0.01, with high heterogeneity between studies, and using random effects model analysis. IF has a significant effect in reducing fat mass compared to RD (MD: −2.14 kg; 95% CI [−3.81, 0.47]; p = 0.01) (Fig. 2A). Subgroup analysis showed that fat mass was effectively reduced in overweight or mild obesity people with a BMI less than 32.5 (MD = −1.25 kg, 95% CI: −2.31 to −0.18, p = 0.02). The results of the subgroup analysis are shown in Table 3.

The source of heterogeneity was searched for using an article-by-article exclusion, it found that the exclusion of the Bhutani literature significantly reduced the heterogeneity to 0%, indicating that Bhutani was the main factor influencing the heterogeneity of the literature.

3.6. Effectiveness of IF on the lipid profile (TC, HDL-c, LDL-c, and plasma TG)

3.6.1. Effectiveness of IF on TC

6 RCTs (8 data points) reported changes in TC and meta-analysis showed I² = 85%, p = 0.43, with high heterogeneity between studies, and using random effects model analysis. IF did not have a significant effect in reducing TC compared to RD (MD: −0.10 mmol/L; 95% CI [−0.35, 0.15]; p = 0.43) (Fig. 2B). Subgroup analysis showed no significant effect of IF intervention on TC in each subgroup.

Fig. 2B.

Effectiveness of IF on lipid composition. (f) Effectiveness of IF on TC, (g) Effectiveness of IF on TG, (h) Effectiveness of IF on HDL-c, (i) Effectiveness of IF on LDL-c.

The source of heterogeneity was searched for using an article-by-article exclusion, it found that the exclusion of the Varady literature significantly reduced the heterogeneity to 0%, indicating that Varady was the main factor influencing the heterogeneity of the literature.

3.6.2. Effectiveness of IF on TG

RCTs (6 data points) reported changes in TG and meta-analysis showed I² = 90%, p = 0.04, with high heterogeneity between studies, and using random effects model analysis. IF have a significant effect in reducing TG compared to RD (MD: −0.20 mmol/L; 95% CI (−0.40, −0.01); p = 0.04) (Fig. 2B). Subgroup analysis showed that TG was effectively reduced in overweight or mild obesity people with a BMI less than 32.5 (MD = −0.32 mmol/L, 95% CI: −0.50 to −0.15, p = 0.0002). The results of the subgroup analysis are shown in Table 3.

The source of heterogeneity was searched for using an article-by-article exclusion, it found that the exclusion of the Bhutani literature significantly reduced the heterogeneity to 0%, indicating that Bhutani was the main factor influencing the heterogeneity of the literature.

3.6.3. Effectiveness of IF on HDL-c

6 RCTs (8 data points) reported changes in HDL-c and meta-analysis showed I² = 96%, p = 0.24, with high heterogeneity between studies, and using random effects model analysis, IF did not have a significant effect in reducing HDL-c compared to RD (MD: -0.09 mmol/L; 95% CI (−0.24, 0.06); p = 0.24) (Fig. 2B). Subgroup analysis showed that the intervention of ADF resulted in a significant reduction in HDL-c (MD = −0.27 mmol/L, 95% CI: −0.51 to −0.02, p = 0.03). The results of the subgroup analysis are shown in Table 3.

The source of heterogeneity was searched for using an article-by-article exclusion, it found that the exclusion of the Bhutani literature significantly reduced the heterogeneity to 37%, indicating that Bhutani was the main factor influencing the heterogeneity of the literature.

3.6.4. Effectiveness of IF on LDL-c

5 RCTs (7 data points) reported changes in LDL-c and meta-analysis showed I² = 76%, p = 0.06, with high heterogeneity between studies, and using random effects model analysis. IF did not have a significant effect in reducing LDL-c compared to RD (MD: −0.17 mmol/L; 95% CI (−0.35, 0.01); p = 0.06) (Fig. 2B). Subgroup analysis showed that the intervention of ADF resulted in a significant reduction in LDL-c (MD = −0.31 mmol/L, 95% CI: −0.53 to −0.08, p = 0.008). The results of the subgroup analysis are shown in Table 3.

The source of heterogeneity was searched for using an article-by-article exclusion, it found that the exclusion of the Varady literature significantly reduced the heterogeneity to 38%, indicating that Varady was the main factor influencing the heterogeneity of the literature.

4. Discussion

4.1. Effects of IF on body morphology (body weight, BMI)

In this systematic review, our study showed that IF was effective in significantly reducing body weight and BMI in overweight obese middle-aged people without metabolic disease compared to the regular diet. Similar to the findings of most previous studies [15,33,34].In a recent meta-analysis of RCTs, IF was more effective than a usual diet to induce weight loss [35]. Body weight loss (4.83 kg) observed in one study was also comparable to the intervention effect of dietary replacement therapy (4 kg) after IF compared to the conventional diet [15]. By comparing it with the conventional diet, it can be more objectively indicated that adults aged over 40 with obesity without metabolic diseases can be helped to lose weight by intermittent fasting, which is an effective dietary strategy. But there was inconsistency with the findings of a small number of previous studies [13,36]. This may be due to the length of the intervention or the way it was measured. In this meta-analysis, we found that: the moderate heterogeneity in body weight may reflect different dietary and lifestyle behaviors in IF, as well as variations in fasting duration and climatic and geographic conditions in fasting populations in different countries, which can influence the effect of IF protocols on body weight, an inference similar to the problem found in previous studies [[37], [38], [39], [40], [41], [42]].

Our results on body weight subgroup analyses suggest that: Heterogeneity of TRF fasting patterns was low, with a p-value of 0.004 indicating a better intervention. The reason is that during TRF, overweight or obese middle-aged and older people consume a large amount of liver glycogen, etc. Since glycogen has a higher water content than fat, consuming a large amount of glycogen may cause water loss, resulting in a certain amount of dehydration, which will cause a significant effect on weight loss [11,[43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [53]]. Besides, TRF can improve body shape and reduce diastolic blood pressure [23]. Therefore, TRF is [54] a viable strategy for Overweight or obese middle-aged and elderly people without the metabolic disease. However, some studies have found that Ramadan fasting causes most of the weight changes to rebound after Ramadan, and some studies have found ramadan fasting interventions in healthy adults produce transient weight loss [[55], [56], [57], [58]].So the necessary follow-up studies are needed to clarify the weight change after ramadan fasting intervention.

In summary, through this meta-analysis, we found that IF significantly affects body weight and BMI compared to RD. This is similar to the vast majority of studies. Subgroup analysis showed that: The TRF model had the best effect among the IF models. In this regard, we suggest that overweight or obese middle-aged and elderly people without metabolic disease are supposed to try TRF for fat and weight loss [59].

4.2. Effects of IF on body composition (body fat percentage, lean body mass, fat mass)

Causes of weight loss can include loss of fat mass and muscle mass [55], our study shows that there is some improvement in body composition compared to RD, which has a significant reduction in FM. Our meta results for body weight showed that subjects in the experimental group (IF) experienced significant weight loss. The skeletal muscle mass did not change significantly. This implies that most of the weight loss is due to a reduction in fat mass, in line with most previous meta-analyses and systematic reviews that fat loss caused by IF is significant [14,15,44,49,[60], [61], [62], [63], [64]].

We found that: there was no significant difference in the effects caused by IF intervention on FFM compared to RD, which is consistent with some of the previous studies [65,66]. The possible reason for the inconsistency is that the diversity of measurement methods may affect the experimental results, with Kotarsky, Varady using DXA for body composition, Bhutani, N.M. HUSSIN uses a bioelectrical impedance analyzer, Yan uses X-ScanPlus II for BFP, Domaszewski uses the SECA mBCA 515 analyzers, and Teng uses TANITA BC-418 body composition analyzers. The DXA is considered to be the gold standard for measuring body composition, but it is non-portable and expensive, has a certain amount of radiation. The bioelectric impedance is more convenient, simple, and fast, but not very accurate and it may have errors when the body has moisture fluctuations. Since our study found that: IF did not have a significant effect on FFM but had a significantly decreasing effect on FM, this suggests that IF can be used to treat overweight or obese middle-aged and elderly people without the metabolic disease, because IF can selectively reduce body fat, thus ensuring that the muscle component that maintains a healthy state without running off [67,68]. Two other studies in premenopausal and postmenopausal women [23,69] have shown that lean weight loss is greater in premenopausal and postmenopausal women after IF, possibly because of the high risk of sarcopenia and lower FFM caused by hormonal changes [18,19]. Therefore, postmenopausal women need to use IF cautiously with safety preconditions.

In addition, our study found that: IF intervention did not have a significant effect on BFP compared to RD, which may be related to the shorter duration of the intervention. Since there were no follow-up studies beyond one year in the literature included in this meta-analysis, there is no way to know the long-term results after the intervention, which can be studied in future studies for long-term observation.

In summary, with this meta-analysis, we found that IF resulted in significant reductions in FM compared to RD, with no significant effect on lean body mass or BFP. Previously, the studies conducted by IF suggested that during fat loss, there may also be some FFM loss. However, through a comprehensive analysis of nice RCTs, we found that compared with RD, IF caused weight loss mainly due to FM reduction rather than a significant loss of FFM and then maintaining body function by minimizing body fat and maximizing FFM retention [17]. Thus, confirming the objective effectiveness of IF for healthy fat loss. At the same time, compared with some studies on CER, the significant reduction in body weight also caused a significant reduction in FFM [66,70,71], so IF may be a more worthwhile way to consider for fat loss than this. IF will be suitable for all adults aged over 40 with obesity.

4.3. Effects of IF on blood lipids (TC, HDL-c, LDL-c, and plasma TG)

In this systematic review, our study showed that: IF failed to have a significant effect on TC but produced a significant reduction in TG, especially in overweight and mildly obese people. This is consistent with most previous meta-analyses and systematic reviews [7,60,72]. However, several researchers found a certain increase in lipids, probably due to the blood collection and measurement right after the end of the fast, probably due to lack of sugar in the body, mental stress and lipolysis accelerated, which was accompanied by a dramatic increase in TG, plasma free fatty acids and glycerol, making an acute increase in lipid levels in the subjects [53,73]. As a result, IF produced biochemical transformation of lipids, weight loss, and effects on lipid component concentrations [74].

Our study showed that: IF was not effective on LDL and HDL, which is consistent with the study of Yang Bo et al. [75]. However, subgroup analysis showed that ADF interventions can effectively reduce LDL, especially in the obese middle-aged and elderly population. Some studies showed that obese middle-aged and older adults have significantly lower LDL after IF intervention [8]. As the longest intervention in this included RCT was 12 weeks, the inconsistent results may be due to the short duration of the intervention and the lack of significant effect of IF on LDL, and HDL. Follow-up long-term experimental studies are needed to verify.

In summary, the IF diet prevents the occurrence of dyslipidemia by limiting many of the risk factors. This meta-analysis shows that: IF can cause a significant decrease in TG compared to RD, and ADF interventions can be effective in reducing LDL, especially in the obese middle-aged, which can have a better effect on cardiovascular disease. Decreasing LDL can reduce the increasing accumulation on the arterial wall by the cholesterol it carries, and increasing HDL can transfer more excess cholesterol in the blood to the liver, and then be broken down into bile salts and enhance lipid metabolism, to significantly reduce myocardial infarction and atherosclerosis [1,22,76]. In this regard, we suggest that adults aged over 40 with obesity without metabolic diseases are supposed to try IF to reduce caloric intake while mobilizing fat and promoting triglycerides in the blood for energy supply, which can improve body composition and morphology while improving cardiovascular disease.

4.4. Strengths and limitations

This study is the first meta-analysis that focuses on an intermittent fasting diet versus a regular diet on body shape and composition and blood lipids in adults aged over 40 with obesity without the metabolic disease. Our systematic review has several strengths and limitations to consider. Strengths include: Firstly, this review searched various databases, including three English and three Chinese databases for relevant literature, and the included meta-analyses were all RCT articles with a high strength of evidence. Secondly, two authors independently searched and selected the included studies, extracted the data, and used the recommended protocols to assess bias for each indicator. Thirdly, this study included metabolically healthy obese (MHO) individual who are at some risk of exercise injury in exercise weight loss due to obesity. Intermittent fasting, as a new dietary therapy, is safer and more practicable for metabolically healthy obese (MHO) individuals than exercise weight loss [77].

Limitations include: Limitations are the lack of long-term IF trials (>12 weeks). It is unclear what body composition and morphologic changes occur in long-term interventions with IF interventions. Moreover, due to the small number of randomized controlled trials studying a single subject (adults aged over 40 with obesity without metabolic disease), it was not possible to investigate propagation bias for most outcomes through funnel plots. Importantly, there was significant heterogeneity in the literature in the studies we included in the Meta-analysis, and this heterogeneity may have skewed the results of the study. At last, the certainty of the evidence for most outcomes was mainly rated as low or moderate, the instruments used for indicator measurement were inconsistent with possible measurement bias, and although all trials used randomization methods, none of the studies detailed allocation concealment. None of the studies met participant blinding, although the use of blinding did not appear feasible given the practice.

5. Conclusion

This systematic review and meta-analysis demonstrated that IF may improve body weight, BMI, fat mass, and TG as compared to a regular diet in adults aged over 40 with obesity without the metabolic disease. IF causes weight loss in this middle-aged and elderly population mainly due to a reduction in fat mass, keeping lean body mass intact, making it a healthy and effective weight loss solution.

Author contribution

KY, HS, and YG and designed the systematic review and supervised the entire program. KY and YG reviewed all the studies and extracted the information from the eligible trials. ZTH and QW analyzed the data and prepared the figures and table. KY, QW wrote the paper. KY, TL, TZ, HS, YG, and SNC revised the manuscript. All authors reviewed and approved the manuscript.

Conflict of interest

The authors declare that they have no conflict of interest.

Appendix A. Supplementary data

The following is Supplementary data to this article: