The influence of combined training sequence on adiposity indices in overweight women

Highlights

• •This study assessed how training sequence (strength → endurance vs. endurance → strength) affects novel adiposity indices (AVI, BAI, TyG-WC, TyG-BMI, McAuley) in overweight and normal-weight women, with both protocols producing significant improvements.
• •Only the E+S sequence significantly reduced TyG Index and Lipid Accumulation Product (LAP), while the S+E sequence uniquely improved WHtR² and CRI-II, highlighting sequence-specific metabolic benefits.
• •The findings emphasize that exercise order can modulate cardiometabolic responses differently, making it a key variable when designing personalized training programs for metabolic health enhancement.

Abstract

Objectives

This study examined whether the sequence of combined training (strength followed by endurance vs. endurance followed by strength) differentially affects adiposity- and metabolism-related indices in overweight and normal-weight women.

Methods

Sixty-six sedentary women, including overweight (n = 32; 25 ≤ BMI ≤ 30 kg/m²) and normal-weight (n = 34; 20 ≤ BMI ≤ 25 kg/m²), were randomly allocated to six groups: normal-weight S+E (n = 12), normal-weight E+S (n = 12), overweight S+E (n = 10), overweight E+S (n = 10), normal-weight control (n = 12), and overweight control (n = 10). Endurance training was performed on a cycle ergometer (60%–88% HRmax), and resistance training was conducted at 40%–75% of one-repetition maximum.

Results

MANCOVA adjusted for baseline values revealed a significant main effect of group for adiposity- and metabolism-related indices, including AVI, BAI, TyG-WC, TyG-BMI, WHtR², McAuley index, CRI-I, and CRI-II (p < 0.05), with no effects of weight status or group × weight status interaction (p > 0.05). Post hoc analyses indicated both training groups (S+E and E+S) exhibited significantly improved values compared with the control for AVI, BAI, TyG-derived indices, and McAuley index (p < 0.05), with no differences between sequences. Limited sequence-dependent effects were observed for LAP, WHtR², and CRI-II, whereas no between-group differences were detected for CI, HOMA-IR, AIP, MOAD, VAI, or CRI-I.

Conclusion

Both combined training sequences resulted in significant changes in several adiposity- and metabolic-related indices compared with control, with sequence-specific differences evident for selected markers.

Clinical trial registration number

IRCT2017080319995N8.

Introduction

According to current epidemiological data and economic significance, obesity is considered one of the most critical public health challenges of the 21st century.¹ Central and abdominal obesity frequently correlate with an increased prevalence of metabolic syndrome (MetS), leading to the subsequent development of cardiovascular diseases, diabetes, and an elevated mortality risk. MetS has been demonstrated to increase the risk of atherosclerotic cardiovascular disease and significantly elevate susceptibility to type 2 diabetes mellitus (DM) in individuals without diabetes.² Furthermore, prospective observational studies have elucidated a strong association between MetS and the incidence of DM.³

The presence of MetS significantly increases the risk of severe health consequences, including mortality. Early identification of at-risk individuals is crucial for timely intervention and improved health outcomes. Studies have indicated that indices associated with obesity and lipid profiles are potential predictors of MetS onset. In recent years, numerous studies have suggested that novel anthropometric indices such as the waist-to-height ratio (WHtR), waist-to-hip ratio (WHR), body mass index (BMI), body adiposity index (BAI), lipid accumulation product (LAP), visceral adiposity index (VAI), and triglyceride-glucose index (TyG) can be readily computed and measured using parameters such as triglycerides (TGs), body height (BH), hip circumference (HC), waist circumference (WC), and body weight (BW) .⁴ These indices are effective indicators of metabolic syndrome in non-overweight/obese adults, vegetarians, and Nigerians.⁵ The triglycerides-glucose index (TyG), TyG waist circumference index (TyG-WC), TyG body mass index (TyG-BMI), and Mmetabolic Score for Insulin Resistance (METS-IR) have emerged as the most frequently utilised Insulin Resistance Indices (IRIs) ,⁶ demonstrating a substantial degree of accuracy in assessing Insulin Resistance (IR) .⁷ In an effort to establish an alternative marker for evaluating adipose tissue dysfunction and distribution, the VAI has been proposed as an experimental sex-specific model. Recent studies have demonstrated a significant association between VAI utilisation and peripheral glycaemia, insulin resistance, and DM.⁸

In contrast, the LAP postulates a consistent adipose risk function related to the progression of cardiovascular disease and mortality in the adult population. In an examination conducted on a sample of 1518 Peruvian adults employing a variety of measures to explore obesity, VAI, Waist Circumference (WC), and the WHtR emerged as the most reliable predictors of metabolic syndrome. Notably, VAI exhibits a commendable predictive capacity in illustrating vulnerability to visceral obesity and elevated blood pressure.⁹ It is well recognised that behavioural changes and lifestyle modifications, including physical activity, are essential for preventing and managing obesity. Physical activity has been advocated as a nonpharmacological treatment for reducing the risk of CVDs and metabolic disorder.¹⁰ In many physical activities, the type of exercise, along with energy production systems, plays a decisive role in creating the desired adaptations.¹¹ Based on the principle of specificity (i.e., each exercise contributes to specific adaptations, and as a result, it is of the utmost importance to consider different types of exercise. Both strength and endurance training programmes can have significant effects on health status; therefore, a combination of exercises can be exciting and crucial for improving physical performance and health-related factors.¹¹ Monteiro et al. investigated the effects of concurrent training on body composition and metabolic profiles in adolescents with obesity. The findings confirmed that exercise reduced body fat and metabolic risk profiles in these individuals.¹² Mikkonen et al.’s systematic review synthesized evidence indicating that concurrent strength and endurance training elicits positive adaptations in body composition, maximal strength, and endurance capacity. The authors further highlighted a notable paucity of research specifically investigating these training modalities within female athletic populations.¹³ Research by Schroeder et al. (2019) suggests that concurrent aerobic and resistance training may be a more effective approach for reducing the burden of CVD risk factors in at-risk middle-aged adults, compared to performing either aerobic or resistance training in isolation.¹⁰ In contrast, the study by Julián Flores-Moreno et al. examined the impact of a twelve-week concurrent training intervention on glucose homeostasis, hepatic function, lipid metabolism, and oxidative stress markers in a cohort of moderately active young men. The findings revealed a lack of statistically significant alterations in body weight, BMI, glucose levels, total cholesterol, triglycerides, High-Density Lipoprotein cholesterol (HDL-c), Low-Density Lipoprotein cholesterol (LDL-c), Aspartate Aminotransferase (AST), Alanine Aminotransferase (ALT), Very-Low-Density Lipoprotein (VLDL), and Malondialdehyde (MDA) concentrations after the training protocol.¹⁴

Although the effectiveness of a combination of strength and endurance training in lowering fat mass and body composition in individuals with obesity and overweight is well known, the exercise order sequence of concurrent training, i.e., which exercise is performed at the beginning of the routine and the type of exercise that follows in the sequence, is still the subject of debate. The order in which strength and endurance training is performed could potentially impact the adaptations induced by training. However, few studies have investigated whether strength training should precede or follow endurance training during a training session, specifically in relation to improvements in adiposity indices.¹⁵ To the best of our knowledge, no comprehensive research has been conducted on the impact of the order sequence of combined strength and endurance training on new adiposity indices. Therefore, recognising the significance of these adiposity indices in metabolic syndrome and various chronic diseases, the present study aimed to assess the effects of eight weeks of concurrent training on novel body adiposity indices in overweight and normal-weight women.

Materials and methods

Study design and participants

The present study was a clinical trial conducted at Shahrekord University. The protocol was registered in the Iranian Clinical Trial Registry (IRCT2017080319995N8). After the local call to recruit subjects in Shahrekord city, 66 sedentary and overweight (n = 32, 25 ≤ BMI ≤ 30 kg/m²) and normal-weight (n = 34, 20 ≤ BMI ≤ 25 kg/m²) women volunteered to participate in this study. Participants were randomly assigned to one of six groups: a normal-weight strength-then-endurance training group (S+E, n = 12), a normal-weight endurance-then-strength training group (E+S, n = 12), an overweight strength-then-endurance training group (S+E, n = 10), an overweight endurance-then-strength training group (E+S, n = 10), a normal-weight control group (C, n = 12), and an overweight control group (C, n = 10). In the S+E groups, participants performed strength training exercises before endurance training, while in the E+S groups, the order was reversed. This design enabled us to investigate the effects of different exercise orderings on adiposity indices.

All participants were closely monitored to ensure proper adherence to the exercise program, which was assessed through weekly training logs and coach reports. This careful oversight aims to reassure the audience that participant compliance was prioritized, fostering confidence in the study's reliability. Participants were allocated to groups using block randomization, and outcome assessors and data analysts were blinded to group assignment (double-masked) to minimize bias.

Sample size was calculated using G*Power 3.1 for a repeated-measures MANOVA with a within–between interaction, with parameters set as effect size f(V) = 0.35, significance level α = 0.05, and desired power = 0.80. Six groups and two measurements were included, yielding a required sample size of 92 participants and an actual power of 0.803 based on prior effect size estimates and study design considerations.

Before the tests, all participants were briefed on the assessment methods as well as the study objectives, benefits, and potential risks, and all signed informed consent forms.

Inclusion criteria for the study were that participants were physically healthy and not participating in any regular exercise and weight loss programs with special diets for six months before the start of the study. The participants were also required to have no history of regular weight training and freedom from diseases, such as CVDs, Hypertension (HTN), and DM, as well as non-use of medication or performance-enhancing drugs. The menstrual cycle status of participants was assessed using specific questions and logbook completions. All participants selected for the study reported regular menstrual cycles. Before participating in the pre- and post-test evaluations, strenuous physical activity was avoided for 48 h. Height, weight, Body Mass Index (BMI), fat percentage, and maximum rate of oxygen consumption (VO_2max) were measured using appropriate methods prior to commencing training and 24 h following the last training session.

To ensure high levels of participant adherence to the study protocol, several monitoring strategies were implemented. The research team conducted regular follow-ups with participants to address inquiries and assess progress. Furthermore, participants were required to complete daily training logs documenting exercise duration, intensity, and perceived exertion. These logs were systematically reviewed by the research team to verify adherence to the prescribed training regimen. Baseline and post-intervention assessments were conducted to evaluate participant progress and adherence to the study protocol. These assessments encompassed measurements of cardiorespiratory fitness and other relevant parameters. To maintain data quality, participants who failed to attend three or more consecutive training sessions were excluded from the study. Through the implementation of these rigorous monitoring strategies, high levels of participant adherence to the study protocol were achieved, thereby ensuring the reliability of the research findings.

The daily energy requirements of the participants were estimated using the Harris-Benedict equation.

BMR (kcal/day) = 655.0955 + (9.5634 × weight in kg) + (1.8496 × height in cm) – (4.6756 × age in years)

Daily Energy Expenditure (kcal) = Basal Metabolic Rate (kcal) × 1.375.¹⁶

This method was employed regardless of the exercise program to which participants were assigned, to standardize the caloric intake based on their basal metabolic needs and estimated activity levels. Notably, although this approach aimed to approximately homogenize energy intake, it did not entail strict and rigorous control over specific dietary components, such as macronutrient ratios or types of food consumed.

Blood sampling

Blood samples (10 mL) from the antecubital vein in a sitting position were collected 72 h before the exercise protocol and 72 h after the last session of the training program in 12 h of fasting state. Next, the samples were centrifuged at 3500 rpm for 10 min at 4 °C to isolate the serum, followed by extraction of the serum in special microtubes and storage at −70 °C until further measurement. Fasting Blood Glucose (FBG) concentration was determined using the ELISA kit, Pars Azmun Co., Tehran, Iran (CV% = 1.28). Insulin was measured using an ELISA kit (Monobind Inc., USA) (CV% = 4.9). Serum TG levels were further determined using an ELISA kit (Pars Azmun Co., Tehran, Iran (CV% = 1.53). Serum Total Cholesterol (TC) levels were also measured using an ELISA kitPars Azmun Co., Tehran, Iran (CV% = 1.62). The serum HDL-C level was subsequently calculated using an ELISA kitPars Azmun Co., Tehran, Iran (CV% = 3.7). Low-Density Lipoprotein (LDL)-C was calculated using an ELISA kitPars Azmun Co., Tehran, Iran (CV% = 1.29).

Anthropometric measurements

The participants' weight was correspondingly measured by a calibrated digital scale, with a minimum accuracy of 0.1 kg (Model WS 80, Switzerland), and their height was determined using a metric stadiometer in a standing position next to the wall and without shoes, with an accuracy of 0.1 cm. The formula for BMI was thus the weight in kg divided by height in m². WC was further measured in the narrowest area, parallel to the umbilicus, when the person was at the end of a normal exhalation, with an accuracy of 0.1 cm. HC was then obtained by transversely measuring the largest diameter in the hip area without skin compression. Body fat percentage was determined by measuring subcutaneous fat at three points on the body (triceps, abdomen, and thighs) with a caliper (accuracy of 1 mm, Harpenden, UK), and body density was estimated using the Jackson-Pollack formula. Notably, WC was measured in the narrowest part of the trunk between the last rib and iliac crest, and pelvic circumference in the widest part was measured with a tape measure. Finally, the WHR was obtained by dividing the waist-to-pelvic area.

Body adiposity formula

The calculations of novel body adiposity indices were performed with respect to the following equations:

AVI: [2 × (waist (cm))² + 0.7 cm × (waist (cm)-hip (cm))²]÷1000¹⁷

BAI: [hip circumference (cm)÷height (m)^1.5]−18¹⁷

CI: waist circumference (m)÷0.109√(weight (kg)/height (m))¹⁷

TyG index: Ln [TG (mg/dL) × FPG (mg/dL)/2]²²

TyG-BMI: TyGindex × BMI¹⁸

TyG-WC: TyGindex × WC (cm)¹⁸

WHtR: waist circumference (cm)÷height (cm)

McAuley: exp (2.63–0.28 ln (insulin in mU/L)-0.31 ln(triglycerides in mmoL/L)¹⁹

AIP=log TG/HDL-C²⁰

CRI-I: TC/HDL-C²⁰

CRI-II: LDL-C/HDL-C²⁰

MOAD: [WC/(36.58+1.89 × BMI)]²¹

VAI: [WC/(36.58+1.89 × BMI)] × (TG/0.81) × (1.52/HDL-C) where both TG and HDL-C levels are expressed in mmoL/L.³

LAP: (waist circumference [cm]−58) × (triglyceride concentration [mM])²²

Exercise training program

All exercises were performed using weights and standard body-building equipment. For the VO_2max measurement, the participants performed the Cooper 12-min walk-run test, in which they ran around the basketball court for 12-min, the distance travelled by each one was calculated, and then their VO_2max was obtained.²³

The training programs implemented in this study were conducted for eight weeks, three days per week. The exercises were performed from simple to difficult and from low- to high-intensity, taking into account the principle of overload and increasing the intensity of the exercise. Each exercise further included a warm-up, basic workouts, and a cool-down that lasted approximately 60‒70 min.

The strength training program was conducted three times per week (every other day) for eight weeks. Each training session included a general warm-up (Stretching and dynamic movements) and a main training program. The main training program consisted of a variety of resistance exercises, including leg press, knee extension, chest press, lat pulldown, calf raise, bilateral triceps pushdown, and bilateral biceps curl. The specific number of sets, repetitions, and rest periods for each exercise was based on a progressive overload protocol and can be found in Table 1. To ensure progressive overload, a one-Repetition Maximum (1RM) test was performed for each exercise every eight sessions, and the weight was adjusted accordingly. During the two sessions of familiarity with exercises, the values of one Repeat Maximum (1RM) used in the experimental group were determined using the following formula:^[24 Equations to determine 1RM=weight ÷ (1.0278 - (0.0278 × number of repetitions)).

Table 1.

Strength and endurance training protocol.

Sessions	Endurance training		Strength training
	Volume	Intensity	Intensity	Repetitions	Sets	Rest interval between sets
1‒8	25 min	60%‒65% HRmax	50%‒60% 1RM	10‒15	3‒5	2 min
9‒16	35 min	65%‒75% HRmax	60%‒70% 1RM	10‒12	3‒5	2 min
17‒24	40 min	75%‒85% HRmax	70%‒80% 1RM	8‒12	3‒5	2 min

HRmax, Maximum Heart Rate; 1RM, One Repeation Maximum.

Participants completed endurance cycling training using a Proteus PEC-3320 bicycle ergometer manufactured in China. Training sessions consisted of continuous cycling at prescribed intensities (Table 1). The heart rate was used to monitor and adjust exercise intensity. To measure heart rate, all participants used Polar and integrated heart rate monitors on their bicycles. Each training session consisted of a 10-min warm-up (including stretching and dynamic movements), a primary endurance training phase, and a 5-min cool-down (static stretching). The specific intensity and duration of each training session are presented in Table 1.^[25

Statistical analysis

Data are presented as mean ± Standard Deviation (SD). Normality was assessed using the Shapiro-Wilk test, and homogeneity of variances was examined using Levene’s test. All statistical assumptions were adequately satisfied prior to analysis. Between-group differences were evaluated using Multivariate Analysis of Covariance (MANCOVA), with baseline values entered as covariates to control for initial intergroup variability. When significant multivariate effects were detected, Bonferroni-adjusted post-hoc tests were performed to identify pairwise differences. Effect sizes were reported as Cohen’s d with corresponding 95% Confidence Intervals to facilitate interpretation of the magnitude of observed effects. To enhance model robustness and reduce the risk of overfitting, a parsimonious multivariable modelling approach was additionally employed. Predictor selection was guided by theoretical relevance, consistency with prior empirical evidence, and support from univariate analyses, with priority given to variables most consistently associated with cardiometabolic outcomes in the literature. This approach ensured a stable and interpretable model structure while minimizing spurious inference due to overfitting and enhancing overall model stability.

Results

Table 2, Table 3 present baseline and post-intervention descriptive statistics (means ± standard deviations) for anthropometric, metabolic, and derived cardiometabolic indices across the study groups. Baseline values were generally comparable between groups. Following the intervention, both training conditions (S+E and E+S) demonstrated favorable changes in several anthropometric and metabolic variables relative to the control group, whereas changes in the control group were small or inconsistent. These descriptive patterns were further examined using inferential analyses (Table 4, Table 5).

Table 2.

Descriptive statistics (Mean ± SD) of metabolic and anthropometric indices across groups.

Variable	Time	Over weight			Normal weight
		S+E	E+S	C	S+E	E+S	C
Weight (kg)	Pre-test	71.67±4.77	73.86±4.98	70.62±4.81	60.00±4.67	57.00±4.01	61.22±6.85
	Post-test	68.94±5.23	70.64±4.40	71.00±5.09	57.50±5.14	54.70±3.83	61.51±7.05
BMI (kg/m2)	Pre-test	27.51±1.72	27.70±2.28	27.47±1.90	22.29±1.85	21.73±1.94	21.10±1.43
	Post-test	26.41±1.71	26.47±2.15	27.76±1.93	21.47±2.09	20.86±1.95	22.34±2.77
WC (cm)	Pre-test	86.38±3.81	81.43±9.41	85.12±9.66	71.63±9.53	68.25±6.22	72.44±13.26
	Post-test	83.00±3.85	79.14±9.74	85.31±9.59	69.58±9.97	66.90±5.84	72.44±13.26
HC (cm)	Pre-test	106.33±4.77	106.86±6.82	106.50±6.65	98.58±5.72	93.30±6.79	99.56±7.19
	Post-test	102.22±4.82	103.43±6.60	107.00±6.59	95.88±5.11	90.05±6.74	100.67±7.62
WHR	Pre-test	0.81±0.05	0.76±0.06	0.79±0.07	0.73±0.09	0.73±0.07	0.73±0.10
	Post-test	0.81±0.05	0.76±0.06	0.80±0.07	0.73±0.09	0.74±0.07	0.71±0.10
TG (mg/dL)	Pre-test	120.44±59.31	99.86±50.94	114.5 ± 42.86	92.00±24.53	99.20±39.87	69.33±14.24
	Post-test	95.89±27.01	77.86±27.56	101.38±31.48	75.58±10.53	78.10±23.19	71.56±16.98
FBG (mg/dL)	Pre-test	84.33±11.43	84.19±4.77	80.00±6.78	82.67±5.58	86.50±5.76	79.56±9.76
	Post-test	79.89±9.16	78.43±5.50	81.50±3.82	78.17±4.84	81.50±5.36	84.89±9.33
Insulin (µIU/mL)	Pre-test	13.5 ± 3.35	8.16±3.51	8.52±4.09	5.61±2.89	6.89±4.12	6.79±3.16
	Post-test	7.26±3.25	6.72±4.02	11.20±8.23	4.06±2.33	6.19±4.36	8.67±3.94
HOMA-IR	Pre-test	2.81±0.70	1.70±0.50	1.68±0.55	1.14±0.30	1.47±0.40	1.33±0.45
	Post-test	1.43±0.45	1.30±0.40	2.25±0.80	0.78±0.25	1.24±0.38	1.82±0.60
HDL (mg/dL)	Pre-test	42.78±14.22	42.68±6.57	40.75±6.79	43.67±10.40	46.80±5.69	41.89±6.09
	Post-test	38.44±9.38	47.29±8.96	47.38±7.69	45.17±8.39	49.50±7.50	42.11±6.72
LDL (mg/dL)	Pre-test	88.76±14.77	85.54±13.75	94.55±24.96	97.77±23.41	95.32±16.98	76.11±20.79
	Post-test	74.76±14.40	79.06±16.74	106.35±34.65	79.11±24.00	83.72±14.51	85.67±14.88

The table above presents the mean ± standard deviation of various variables in two groups, individuals with overweight and individuals with normal weight, both pre- and post-intervention. The intervention groups consisted of S+E (strength and endurance training), E+S (endurance and strength training with a different sequence), and C (control group). The measured variables included body mass, Body Mass Index (BMI), Waist Circumference (WC), Hip Circumference (HC), Waist-to-Hip Ratio (WHR), Triglycerides (TG), Fasting Blood Glucose (FBG), insulin, Homeostatic Model Assessment for Insulin Resistance (HOMA-IR), High-Density Lipoprotein cholesterol (HDL), and Low-Density Lipoprotein cholesterol (LDL).

Table 3.

Descriptive statistics, mean, and Standard Deviation of new metabolic and anthropometric indices across groups.

Variable	Time	Over weight			Normal weight
		S+E	E+S	C	S+E	E+S	C
AVI	Pre-test	15.56±1.09	14.36±2.99	15.37±3.15	11.55±2.33	10.33±1.49	11.92±2.82
	Post-test	14.37±1.14	13.56±3.00	15.45±3.13	10.95±2.17	9.82±1.39	11.99±2.84
BAI	Pre-test	33.85±2.37	33.23±3.28	34.58±4.44	29.14±3.00	27.15±4.02	30.41±3.31
	Post-test	31.84±2.19	31.58±3.12	34.83±4.47	27.84±2.63	25.58±3.93	30.95±3.54
CI	Pre-test	1.19±0.06	1.11±0.12	1.18±0.12	1.09±0.13	1.06±0.09	1.08±0.16
	Post-test	1.17±0.07	1.10±0.13	1.18±0.12	1.08±0.15	1.06±0.09	1.07±0.16
TyG.index	Pre-test	8.45±0.43	8.24±0.45	8.37±0.39	8.21±0.25	8.29±0.42	7.89±0.25
	Post-test	8.22±0.28	7.97±0.36	8.28±0.38	7.98±0.16	8.03±0.33	7.99±0.20
TyG.BMI	Pre-test	232.27±15.65	228.66±26.5	230.09±21.84	183.04±17.3	180.53±19.9	166.56±11.1
	Post-test	216.87±14.11	211.16±22.6	229.69±17.41	171.39±17.8	167.51±17.6	178.52±22.5
TyG.WC	Pre-test	729.3 ± 34.78	672.7 ± 103.8	713.7 ± 103.07	587.9 ± 82.00	566.9 ± 65.58	571.8 ± 107.1
	Post-test	681.8 ± 36.93	631.4 ± 92.93	705.7 ± 82.11	555.1 ± 79.59	537.4 ± 57.63	578.9 ± 108.1
WHtR2	Pre-test	0.53±0.03	0.49±0.05	0.53±0.07	0.44±0.06	0.42±0.04	0.45±0.08
	Post-test	0.51±0.03	0.48±0.06	0.53±0.07	0.43±0.06	0.41±0.04	0.45±0.08
Mcauley	Pre-test	3.62±0.53	4.24±0.96	4.10±0.59	4.64±0.63	4.92±2.22	4.46±0.63
	Post-test	4.29±0.59	4.72±1.39	3.92±0.66	5.25±1.07	5.13±2.11	4.10±0.51
AIP	Pre-test	0.44±0.28	0.33±0.23	0.43±0.15	0.32±0.18	0.30±0.15	0.22±0.12
	Post-test	0.39±0.22	0.2 ± 0.19	0.32±0.16	0.23±0.09	0.19±0.12	0.23±0.13
CRI.I	Pre-test	3.77±1.06	3.56±0.73	4.08±1.09	3.78±0.88	3.57±0.73	3.19±0.47
	Post-test	3.82±1.08	3.15±0.65	3.66±1.08	3.14±0.56	3.15±0.75	3.45±0.68
CRI.II	Pre-test	2.25±0.67	2.06±0.51	2.38±0.74	2.23±0.72	2.06±0.44	1.86±0.69
	Post-test	2.03±0.50	1.74±0.52	2.28±0.75	1.78±0.54	1.73±0.49	2.08±0.49
MOAD	Pre-test	0.98±0.04	0.92±0.11	0.96±0.08	0.91±0.10	0.88±0.09	0.95±0.16
	Post-test	0.96±0.05	0.91±0.11	0.96±0.09	0.90±0.12	0.88±0.09	0.92±0.16
VAI	Pre-test	6.18±5.03	4.29±3.02	5.10±1.80	3.81±1.63	3.50±1.28	3.08±1.37
	Post-test	5.19±3.52	3.02±1.71	3.87±1.17	2.87±0.55	2.65±0.89	3.02±1.20
LAP	Pre-test	38.60±10.85	26.51±9.87	35.80±13.55	14.16±5.31	11.48±5.71	11.30±4.95
	Post-test	27.80±12.55	18.79±8.24	31.55±12.39	9.88±4.29	7.89±3.58	11.64±5.98

AVI, Abdominal Volume Index; BAI, Body Adiposity Index; CI, Conicity Index; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance; TyG index, Triglyceride and Glucose Index; TyG-BMI, TyG index multiplied by Body Mass Index; TyG-WC, TyG index multiplied by Waist Circumference; WHtR², Waist-to-Height Ratio squared; McAuley, McAuley Index of insulin sensitivity; AIP, Atherogenic Index of Plasma; CRII, Castelli Risk Index II; CRIII, Castelli Risk Index III; MOAD, Metabolic Overweight and Dyslipidemia Index; VAI, Visceral Adiposity Index; LAP, Lipid Accumulation Product; S+E, Strength followed by Endurance Training; E+S, Endurance followed by Strength Training; C, Control Group.

Table 4.

Multivariate analysis of covariance (MANCOVA) examining the main and interaction effects of exercise type and weight status on composite metabolic variables.

Variable	Exercise type (df = 2)			Weight status (df = 1)			Interaction (df = 2)
	F	p-value	Partial η²	F	p-value	Partial η²	F	p-value	Partial η²
AVI	11.8	<0.001	0.39	0.31	0.58	0.009	1.84	0.17	0.098
BAI	18.43	<0.001	0.52	0.53	0.46	0.016	1.34	0.27	0.073
CI	1.2	0.314	0.06	0.23	0.62	0.007	0.67	0.51	0.038
HOMA-IR	9.14	0.001	0.14	0.04	0.84	0.00	1.10	0.34	0.060
TYG-Index	3.11	0.084	0.13	0.44	0.507	0.013	0.451	0.64	0.026
TYG-WC	7.73	0.002	0.31	0.01	0.90	0.001	0.960	0.39	0.053
TYG-BMI	10.24	<0.001	0.37	0.27	0.60	0.008	0.409	0.66	0.023
WHtR²	6.77	0.003	0.28	0.42	0.51	0.01	1.11	0.34	0.061
McAuley	7.54	0.002	0.30	0.95	0.33	0.02	0.28	0.75	0.017
AIP	3.66	0.036	0.17	0.14	0.70	0.004	1.05	0.35	0.058
CRI-I	2.06	0.143	0.10	0.30	0.58	0.009	0.73	0.48	0.042
CRI-II	5.54	0.008	0.24	0.03	0.84	0.001	1.03	0.36	0.058
MOAD	0.56	0.573	0.03	0.10	0.74	0.003	1.04	0.36	0.058
VAI	3.07	0.059	0.15	2.17	0.14	0.06	1.49	0.23	0.081
LAP	8.95	0.001	0.34	2.10	0.15	0.06	2.12	0.03	0.145

Note: weight status, and their interaction on composite variables. η², Partial eta squared; * p < 0.05, ** p < 0.01. AVI, Abdominal Volume Index; BAI, Body Adiposity Index; CI, Conicity Index; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance; TyG index, Triglyceride and Glucose Index; TyG-BMI, TyG index multiplied by Body Mass Index; TyG-WC, TyG index multiplied by Waist Circumference; WHtR², Waist-to-Height Ratio squared; McAuley, McAuley Index of insulin sensitivity; AIP, Atherogenic Index of Plasma; CRII, Castelli Risk Index II; CRIII, Castelli Risk Index III; MOAD, Metabolic Overweight and Dyslipidemia Index; VAI, Visceral Adiposity Index; LAP, Lipid Accumulation Product.

Table 5.

Pairwise comparisons of the residual means of Adiposity Indicators scores based on the type of combined training received.

Variable	Groups	Means	Comparison	Mean Difference	p-value	Cohen’s d
AVI	S+E	12.116	S+E vs. E+S	−0.21	0.666	−0.06
	E+S	12.326	S+E vs. C	−0.856	<0.001	−0.23
	C	12.972	E+S vs. C	−0.646	0.006	−0.17
BAI	S+E	29.495	S+E vs. E+S	0.104	1	0.02
	E+S	29.391	S+E vs. C	−1.908	<0.001	−0.35
	C	31.403	E+S vs. C	−2.012	<0.001	−0.37
CI	S+E	1.097	S+E vs. E+S	−0.015	0.471	−0.08
	E+S	1.112	S+E vs. C	−0.012	0.853	−0.06
	C	1.109	E+S vs. C	0.003	1	0.02
HOMA-IR	S+E	1.30	S+E vs. E+S	−0.12	0.63	−0.15
	E+S	1.42	S+E vs. C	−0.88	0.002	−0.70
	C	2.18	E+S vs. C	−0. 76	0.004	−0.62
TYG-Index	S+E	8.076	S+E vs. E+S	0.098	0.577	0.26
	E+S	7.977	S+E vs. C	−0.096	0.728	−0.25
	C	8.171	E+S vs. C	−0.194	0.084	−0.52
TYG-BMI	S+E	188.885	S+E vs. E+S	2.012	1	0.05
	E+S	186.874	S+E vs. C	−14.172	0.001	−0.37
	C	203.058	E+S vs. C	−16.184	<0.001	−0.42
TYG-WC	S+E	599.192	S+E vs. E+S	−0.817	1	−0.01
	E+S	600.009	S+E vs. C	−27.568	0.003	−0.22
	C	626.760	E+S vs. C	−26.751	0.006	−0.22
WHtR2	S+E	0.457	S+E vs. E+S	−0.006	0.349	−0.01
	E+S	0.464	S+E vs. C	−0.016	0.002	−0.02
	C	0.473	E+S vs. C	−0.009	0.131	−0.01
McAuley	S+E	4.961	S+E vs. E+S	0.131	1	0.08
	E+S	4.830	S+E vs. C	0.894	0.002	0.56
	C	4.067	E+S vs. C	0.763	0.013	0.48
AIP	S+E	0.290	S+E vs. E+S	0.088	0.04	0.43
	E+S	0.202	S+E vs. C	0.015	1	0.07
	C	0.275	E+S vs. C	−0.073	0.2	−0.35
CRI-I	S+E	3.274	S+E vs. E+S	0.044	1	0.04
	E+S	3.229	S+E vs. C	−0.371	0.266	−0.36
	C	3.645	E+S vs. C	−0.415	0.21	−0.4
CRI-II	S+E	1.769	S+E vs. E+S	−0.107	1	−0.15
	E+S	1.876	S+E vs. C	−0.427	0.007	−0.61
	C	2.196	E+S vs. C	−0.320	0.073	−0.56
MOAD	S+E	0.918	S+E vs. E+S	−0.010	1	−0.07
	E+S	0.928	S+E vs. C	0.002	1	0.01
	C	0.916	E+S vs. C	0.011	1	0.08
VAI	S+E	3.646	S+E vs. E+S	0.57	0.063	0.24
	E+S	3.076	S+E vs. C	−0.127	1	0.05
	C	3.519	E+S vs. C	−0.443	0.327	−0.19
LAP	S+E	32.20	S+E vs. E+S	9.55	0.041	0.44
	E+S	22.65	S+E vs. C	−4.60	0.018	−0.29
	C	36.80	E+S vs. C	−14.15	0.001	−0.68

The data are reported as a mean. S+E: Strenght+Endurance; E+S: Endurance+ Strenght; C, Control; AVI, Abdominal Volume Index; BAI, Body Adiposity Index; CI, Conicity Index; HOMA-IR, Homeostasis Model Assessment of Insulin Resistance; TyG index, Triglyceride and Glucose Index; TyG-BMI, TyG index multiplied by Body Mass Index; TyG-WC, TyG index multiplied by Waist Circumference; WHtR², Waist-to-Height Ratio squared; McAuley, McAuley Index of insulin sensitivity; AIP, Atherogenic Index of Plasma; CRII, Castelli Risk Index II; CRIII, Castelli Risk Index III; MOAD, Metabolic Overweight and Dyslipidemia Index; VAI, Visceral Adiposity Index; LAP, Lipid Accumulation Product.

After adjustment for baseline values, MANCOVA results (Table 4) revealed a significant main effect of exercise type on AVI (F = 11.8, p < 0.001, η² = 0.39), BAI (F = 18.43, p < 0.001, η² = 0.52), HOMA-IR (F = 9.14, p = 0.001, η² = 0.14), TyG-WC (F = 7.73, p = 0.002, η² = 0.31), TyG-BMI (F = 10.24, p < 0.001, η² = 0.37), WHtR² (F = 6.77, p = 0.003, η² = 0.28), McAuley index (F = 7.54, p = 0.002, η² = 0.30), AIP (F = 3.66, p = 0.036, η² = 0.17), CRI-II (F = 5.54, p = 0.008, η² = 0.24), and LAP (F = 8.95, p = 0.001, η² = 0.34). No significant main effects were observed for CI, TyG-index, CRI-I, MOAD, or VAI (all p > 0.05).

Neither weight status nor the interaction between exercise type and weight status reached statistical significance for any of the examined variables (Table 4).

Pairwise comparisons (Table 5) indicated that both S+E and E+S differed significantly from the control group for AVI, BAI, HOMA-IR, TyG-WC, TyG-BMI, McAuley index, and LAP (all p < 0.05), while no significant differences were observed between the two training sequences for these variables.

For WHtR² and CRI-II, significant reductions were observed in S+E compared with the control group (p = 0.002 and p = 0.007, respectively), whereas the corresponding comparisons for E+S were not statistically significant. A significant difference between S+E and E+S was detected for AIP (p = 0.040), with no significant differences between either intervention group and the control group for this variable.

For LAP, all pairwise comparisons were statistically significant (S+E vs. E+S, p = 0.041; S+E vs. control, p = 0.018; E+S vs. control, p = 0.001), indicating differential responses across all groups. No significant pairwise differences were found for CI, TyG-index, CRI-I, MOAD, or VAI (all p > 0.05).

Overall, the pattern of findings across Table 2, Table 3, Table 4, Table 5 indicates that the response to the intervention was variable across indices, with more consistent improvements observed in composite adiposity and insulin-related markers compared with other cardiometabolic indices.

The table above presents the mean ± standard deviation of various variables in two groups, individuals with overweight and individuals with normal weight, both pre- and post-intervention. The intervention groups consisted of S+E (strength and endurance training), E+S (endurance and strength training with a different sequence), and C (control group). The measured variables included body mass, Body Mass Index (BMI), Waist Circumference (WC), hip Circumference (HC), Waist-to-Hip Ratio (WHR), triglycerides (TG), Fasting Blood Glucose (FBG), insulin, Homeostatic Model Assessment for Insulin Resistance (HOMA-IR), High-Density Lipoprotein cholesterol (HDL), and Low-Density Lipoprotein cholesterol (LDL).

Discussion

The purpose of the present study was to examine the effects of two combined training sequences (strength-endurance vs. endurance-strength) on a range of adiposity and metabolism-related indices in overweight and normal-weight women. The interventions were associated with significant improvements in several composite markers of adiposity and metabolism, particularly those representing central fat distribution and lipid-glucose interactions, including AVI, BAI, TyG-derived indices (TyG-WC and TyG-BMI), HOMA-IR, McAuley index, and LAP. By contrast, the changes in some selected cardiometabolic risk indices, such as CRI-II and AIP, were more variable and less consistently expressed across comparisons.

In conclusion, the modifications observed in AVI and BAI, which are indicators of central and general adiposity, along with the favorable changes in TyG-based indices, suggest beneficial adaptations in body composition and metabolic regulation in response to both training protocols. These indices have been well-validated as surrogate markers of cardiometabolic risk, particularly in the context of visceral fat accumulation and insulin resistance.^17,26 Overall, HOMA-IR showed significant reductions compared with the control group, but this pattern was not completely consistent across all pairwise comparisons. Conversely, the McAuley index indicated more consistent improvements relative to the control condition; however, differences between the two training sequences were not statistically significant. Overall, these findings suggest that not all surrogate measures of insulin resistance may respond equally to exercise training, with composite lipid-glucose-based measures. Taken together, these results suggest that not all surrogate measures of insulin resistance may respond similarly to exercise training. Composite lipid-glucose indices appear to be more responsive than single fasting insulin measures. In this context, static fasting markers such as HOMA-IR may be more variable in detecting subtle metabolic adaptations over relatively short intervention periods, underscoring the importance of using a combination of metabolic markers to gain a more complete picture of insulin-related changes.^27,28

The observed improvements in adiposity-related parameters may be explained in part by the increased energy expenditure associated with both training modalities, which may promote a sustained negative energy balance and subsequently reduce fat accumulation.²⁹ It has previously been shown that combined resistance and endurance training can induce favourable metabolic adaptations, such as improved insulin sensitivity and glucose-lipid regulation,³⁰ frequently accompanied by positive changes in body composition and cardiometabolic health profiles.³¹ However, these mechanisms are indirect and must be interpreted with caution in the absence of direct hormonal or molecular measurements.

No statistically significant effects between groups or interactions between conditions were found for the TyG index. In some comparisons, small numerical changes were observed, but these did not reach statistical significance, indicating stability of this index after intervention. The TyG index is a combined marker of insulin resistance,^32,33 but no significant changes were observed after the intervention.

Similar results were found for LAP, which presented statistically significant differences between groups in the overall model and also in the pairwise comparisons, showing that this index is clearly responsive to the training interventions applied. These results indicate that LAP may be a relatively sensitive marker to detect changes in lipid accumulation and cardiometabolic risk induced by exercise in the context of this study. Exercise-induced adaptations in lipid and glucose metabolism are probably mediated through multiple and partially separate physiological pathways.^34,35 However, the degree to which these adaptations are reflected in fasting-based composite indices such as TyG and LAP might still differ depending on intervention duration and the availability of direct mechanistic measures, which were not assessed in the current study.³⁶

The combined strength and aerobic training program did not produce uniform effects across cardiovascular risk markers. CRI-II and WHtR² showed significant reductions only in the S+E group compared with the control condition, whereas no significant differences were observed for E+S versus control or between the two exercise sequences. Importantly, no significant interaction effects were detected, indicating a lack of sequence-dependent adaptation. Overall, these findings suggest that observed changes in CRI-II and WHtR² were driven primarily by a single training condition rather than a consistent effect of combined exercise training. This pattern indicates marker-specific responsiveness to training, consistent with evidence that cardiometabolic risk indices may respond heterogeneously to exercise interventions depending on their underlying physiological determinants.³⁷

The Atherogenic Index of Plasma (AIP) showed a statistically significant overall effect in the multivariate analysis; however, the post hoc pairwise results were less consistent. A significant difference emerged only between the two exercise sequences, while neither intervention group demonstrated a clear or consistent difference when compared with the control group across the remaining comparisons. Overall, these findings suggest that the observed pattern for AIP should be interpreted with caution, as it appears to reflect a somewhat variable response that is not consistently replicated across pairwise analyses.

The sequence order of training (i.e., whether strength or endurance exercise is performed first within a session) has been proposed as a potential determinant of training adaptations; however, current evidence does not support a clear or consistent sequence effect. Only a limited number of studies have directly compared different exercise orders within concurrent training models, and the available findings remain mixed. In the present context, previous work by Taipale et al. suggests that altering the order of endurance and resistance exercise may influence short-term physiological responses;^[38,39 however, such acute responses cannot be directly extrapolated to long-term adaptations. Overall, the literature indicates that training adaptations are shaped by multiple interacting variables, including exercise modality, intensity, volume, frequency, and program design, rather than sequence order alone.

A limitation of the present study is the absence of dietary control, which may have contributed to variability in metabolic outcomes. In addition, the relatively short duration of the intervention limits the ability to conclude longer-term adaptations to combined training.

Methodological considerations should also be acknowledged. The use of skinfold calipers for body composition assessment and the Cooper test for estimating VO_2max, although practical and widely used in field settings, may introduce some degree of measurement error compared with more advanced laboratory techniques. Furthermore, the relatively small subgroup sample sizes may have limited statistical power, particularly for detecting subtle between-group or interaction effects. Collectively, these factors suggest that the findings should be interpreted in the context of these methodological constraints and may not fully reflect long-term or highly individualized training adaptations. Future research should address these limitations by incorporating more controlled nutritional protocols, longer intervention durations, and more precise assessment techniques such as DXA or laboratory-based cardiopulmonary testing. Studies with larger and more diverse samples would also strengthen the generalizability of findings. In addition, combining exercise interventions with structured nutritional strategies and mechanistic biomarkers (e.g., inflammatory or hormonal indices) may provide a more comprehensive understanding of combined training adaptations. Such approaches could ultimately contribute to more individualized and evidence-based exercise prescriptions.

Conclusions

The present study showed that both Strength-Endurance (S+E) and Endurance-Strength (E+S) training sequences were associated with meaningful improvements in several adiposity- and metabolism-related indices, including AVI, BAI, TyG-WC, TyG-BMI, HOMA-IR, the McAuley index, and LAP. Additional favorable changes were also noted in WHtR², CRI-II, and AIP, although these responses were not uniform across all variables and differed in magnitude and consistency depending on the index.

While some variables demonstrated significant changes within intervention groups compared with the control condition, no significant interaction effect was found between training sequences. Overall, these results suggest that combined training is effective in improving selected cardiometabolic risk markers; however, the pattern of response appears to depend more on the specific index than on the order of exercise execution.

In summary, combined resistance and aerobic training may be a practical approach for improving metabolic and cardiovascular risk profiles in both normal-weight and overweight women, although the extent of adaptation may vary depending on the particular outcome measure considered.

Ethical approval

The present study was a clinical trial conducted at Shahrekord University. The protocol was registered in the Iranian Clinical Trial Registry, IRCT2017080319995N8.

Authors’ contributions

Methodology: Mohammad Faramarzi, Zahra Hemati Farsani, Zohreh Shanazari. Project administration: Mohammad Faramarzi. Supervision: Mohammad Faramarzi, Zahra Hemati Farsani. Writing-original draft: Mohammad Faramarzi, Zahra Hemati Farsani, Zohreh Shanazari, Zeinab Gorgin Karaji. Writing-review & editing: Mohammad Faramarzi, Zahra Hemati Farsani, Zohreh Shanazari, Zeinab Gorgin Karaji.

Funding

None.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Conflicts of interest

The authors declare no conflicts of interest.

Edited by: José Maria Soares Junior