Muscle quality index and phase angle responses to matched-volume resistance training at different frequencies in older women

Abstract

[Purpose]

To compare the effects of two resistance training (RT) frequencies (two vs. three sessions per week) on phase angle (PhA), a bioimpedance-derived marker, and muscle quality index (MQI) in older women under matched weekly training volume.

[Methods]

Forty-five older women (G2x: n = 22; G3x: n = 23; 63.2 ± 4.5 years) completed 12 weeks of supervised RT. Outcomes included fat mass, FFM, resting metabolic rate, total body water, PhA, and MQI, calculated as one-repetition maximum (1RM) divided by the corresponding segment-specific fat-free mass (FFM) measured by DXA for the chest press, leg extension, and preacher curl exercises. This study represents a secondary analysis of data from a previously published trial, focusing on neuromuscular efficiency and PhA.

[Results]

MQI increased significantly over time in both groups (p < 0.001), with no time × group interactions observed across exercises. No group-level changes were detected for PhA, and mean fluctuations remained below the minimal detectable change. Fat mass decreased modestly in the 3×/week group, whereas FFM increased slightly in both groups.

[Conclusion]

When weekly training volume was equated, RT performed two or three times per week resulted in similar improvements in MQI, with no group-level changes observed for PhA in older women.

INTRODUCTION

Aging is accompanied by progressive reductions in skeletal muscle mass, neuromuscular performance, and cellular structural integrity, which collectively compromise mobility and functional independence in older adults. These alterations are often accompanied by unfavorable shifts in hydration and body composition, amplifying the risk of chronic disease and disability [1-3]. In older women, such changes are associated with increased risks of frailty, falls, metabolic disorders, and functional limitations [4,5].

Resistance training (RT) is widely recommended as a non-pharmacological strategy to attenuate these age-related impairments. RT promotes increases in muscle strength and fat-free mass (FFM), reductions in fat mass, and improvements in neuromuscular function and metabolic health [2,6,7]. While the benefits of RT are well established, the role of training frequency, particularly when weekly training volume is matched, remains a topic of debate. In practice, prescribing two versus three sessions per week is particularly relevant for older adults, as these are the most common frequencies used in community and clinical programs [8,9]. Although weekly volume may be the primary driver of adaptations [10,11] frequency may still affect adherence, recovery dynamics, and perceived feasibility. Therefore, clarifying whether similar benefits can be achieved with fewer sessions is essential for optimizing training recommendations in this population.

Overall, some studies suggest that higher frequencies may enhance neuromuscular and morphological adaptations [10,12], and others emphasize that frequency may play a supportive role in stimulating hypertrophy and optimizing recovery dynamics, especially in aging populations [13,14]. In addition to these reviews, some randomized controlled trials have compared training frequencies in older women. Nascimento et al. [15] and Dos Santos et al. [7] reported comparable gains in strength and body composition when weekly volume was equated, while Souza et al. [16] demonstrated improvements in bioimpedance-derived markers such as phase angle (PhA). More recently, Baxter et al. [17] confirmed that different weekly frequencies of eccentric training also produced similar functional outcomes in older adults. These findings are consistent with previous meta-analyses [10,18] showing that, when weekly volume is equated, training frequency has little impact on hypertrophy. Similarly, Souza et al. [16] reported improvements in bioimpedance-derived markers such as PhA, although their longer intervention period may partly explain the divergence with the present null PhA findings.

However, some studies suggest that when training volume is equated, gains in muscle mass and strength are comparable across different frequencies in older adults [19-21], supporting, the hypothesis that weekly volume, rather than frequency alone, is the primary driver of RT-induced adaptations in aging populations [21]. Nonetheless, exploring more specific and sensitive outcome measures may enhance our understanding of how training frequency modulates physiological adaptations in older women. In this context, the relative balance between adiposity and muscle mass has gained increasing attention. The fat-to-muscle ratio (FMR), which integrates fat mass and FFM into a single metric, has been proposed as a sensitive indicator of sarcopenic obesity and metabolic risk in older adults [22-24].

Recently, a shift toward tissue-level and cellular markers has emerged, incorporating assessments of fat-free mass (FFM), fat mass, total body water (TBW), resting metabolic rate (RMR), and PhA. PhA, derived from bioelectrical impedance analysis, reflects impedance-related properties associated with cellular structural integrity and fluid distribution and has been consistently associated with functional capacity, nutritional status, and survival in older adults [25,26]. Moreover, TBW offers an indirect assessment of whole-body hydration, which plays a critical role in metabolic and neuromuscular function, particularly in aging populations [26]. In addition, we specifically focused on the muscle quality index (MQI), operationalized as exercise-specific maximal strength normalized to the corresponding segment-specific fat-free mass (FFM) measured by DXA (i.e., 1RM/segment- specific FFM), which estimates the ability to generate force relative to lean tissue and serves as a proxy for neuromuscular efficiency [27,28]. This outcome highlights the quality of muscle tissue rather than its quantity alone, a particularly relevant aspect in older women, where hypertrophic responses may be modest but functional improvements are clinically meaningful. Together, these outcomes provide unique insights into how training frequency influences adaptations beyond traditional markers, such as muscle mass and absolute muscle strength.

In addition to assessing average group changes, there is growing recognition of the need to examine interindividual variability in responsiveness to exercise interventions. Earlier studies have often referred to the existence of “responders” and “non-responders” in several outcome domains, including muscle mass, PhA, and strength, emphasizing the importance of individualized approaches in exercise prescription for older adults. However, dichotomous classifications based solely on measurement precision (e.g., SEM) may be misleading, as they do not account for population variability. Recent methodological work has highlighted that more conservative definitions based on variability indices, such as standard deviation (SD) or the smallest worthwhile change (SWC), provide a more accurate representation of heterogeneity in training adaptations [27,29,30].

Therefore, this study aimed to compare the effects of two RT frequencies (2 vs. 3 sessions/week) on PhA, a bioimpedance-derived marker, and MQI, assessed by strength-to-FFM ratio in older women, under matched weekly training volume. As an exploratory analysis, we also examined changes in FMR to provide additional insight into qualitative body composition adaptations. As part of this approach, we examined interindividual variability in physiological adaptations through a responder-type analysis. This analysis acknowledges that group averages may mask clinically meaningful differences, and offers complementary insight into the frequency literature, supporting the discussion of personalized exercise prescriptions [31,32]. We hypothesized that both groups would experience positive adaptations when weekly training volume was equated, consistent with evidence that volume is the primary driver of RT outcomes. Nonetheless, we speculated that the higher-frequency group might demonstrate marginally greater improvements in MQI and PhA, given the potential benefits of more frequent stimulation [14,26,33].

METHODS

Experimental Approach to the Problem

This trial employed a parallel group repeated-measure design to investigate how different RT frequencies influence morphological, neuromuscular, and bioelectrical impedance-derived outcomes in older women. The study lasted 16 weeks in total. During the initial 2 weeks (weeks 1-2), participants underwent baseline testing and familiarization with both training protocol and assessment procedures. The subsequent 12 weeks (weeks 3-14) comprised the intervention, in which participants performed supervised RT either two (G2x) or three (G3x) times per week, with total weekly workload carefully equated between groups. The final 2 weeks (weeks 15-16) were reserved for post-intervention testing.

Outcome measures included dual-energy X-ray absorptiometry (DXA) for fat mass and FFM, and bioelectrical impedance analysis (BIA) for TBW, PhA, and RMR. FMR was calculated as the ratio of total fat mass to FFM obtained by DXA, and was included as an exploratory outcome. Maximal strength was assessed via 1RM tests (chest press, leg extension, biceps curl), and MQI was calculated as the ratio of 1RM to FFM for these exercises. All testing was performed under standardized laboratory conditions, with participants instructed to maintain their usual diet and habitual physical activity outside of the training sessions.

This manuscript represents a secondary analysis of data previously collected in the randomized controlled trial published by Pina et al. [32]. Although the cohort, intervention design, and ethical approval were the same, the present analysis focused on different endpoints, namely MQI, PhA, TBW, RMR, FMR and individual responsiveness, which were not the primary outcomes in the original report.

Participants

Recruitment was carried out using a multi-channel approach, including newspaper announcements, local radio broadcasts, and distribution of flyers in central and residential neighborhoods. Women interested in the study first completed health history and physical activity questionnaires as part of the initial screening. Eligibility criteria required participants to be female, 60 years or older, physically independent, not undergoing hormone replacement therapy, and without orthopedic limitations that would hinder the execution of training or testing. In total, 350 women volunteered and attended interviews at the university laboratory, where the study design, procedures, and expectations were explained in detail. Of these, 47 individuals satisfied all criteria and agreed to participate on the next stage.

Clinical screening consisted of medical examination by a cardiologist, including a resting 12-lead electrocardiogram, a structured health interview, and a treadmill stress test when indicated. All selected participants were cleared for exercise participation with no contraindications. Randomization was performed by an investigator blinded to group allocation using the random.org website. Participants were assigned to one of two intervention groups:

• G2x: resistance training two times per week (3 sets per exercise; n = 23; 65.4 ± 4.4 years; 62.5 ± 7.8 kg; 156.2 ± 5.9 cm; 25.7 ± 3.3 kg/m²).

• G3x: resistance training three times per week (2 sets per exercise; n = 24; 64.9 ± 4.6 years; 61.0 ± 8.8 kg; 156.3 ± 5.8 cm; 24.9 ± 3.2 kg/m²).

Sample size determination was based on FFM as the primary outcome. Pilot data indicated a standard deviation of ~0.84 kg for changes in FFM. Assuming α = 0.05 (two-tailed), 80% power, and a minimum detectable between-group difference of 0.80 kg (close to our conservative threshold for meaningful change, 0.77 kg), the required sample size was 18 participants per group. To account for attrition, the recruitment goal was ≥ 22 per group, which was reached (G2x: n = 22; G3x: n = 23). Other endpoints such as PhA and MQI were treated as exploratory for the purposes of power estimation. All volunteers provided written informed consent before participation. The protocol was approved by the local University Ethics Committee (Project 04743, approval number 21750/2006) and followed the Declaration of Helsinki. Participant flow through the study is shown in Figure 1 (CONSORT diagram).

Figure 1. CONSORT flow diagram of participant enrollment, allocation, follow-up, and analysis.

Procedures

Anthropometry

Body mass was assessed to the nearest 0.1 kg using a calibrated digital scale (Filizola, model ID 110, São Paulo, Brazil) with participants dressed in light clothing and barefoot. Height was measured with a wall-mounted stadiometer (E120A - Tonelli) to the nearest 0.1 cm. Body mass index (BMI) was derived as body mass (kg) divided by height squared (m²).

Primary Outcomes

Body composition

Dual-energy x-ray absorptiometry

All assessments of body composition were performed in the morning following an overnight fast. Dual-energy X-ray absorptiometry (DXA; Lunar Prodigy, model NRL 41990, GE Lunar, Madison, WI, USA) was used to quantify relative fat mass (%), absolute fat mass (kg), and whole-body fat-free mass (FFM, kg). FMR was subsequently calculated as the ratio between absolute fat mass (kg) and FFM (kg) derived from DXA data. Prior to scanning, participants removed all metallic items. Each scan was conducted with the subject positioned supine along the longitudinal axis of the table. The feet were gently secured together at the toes to minimize leg movement, and the hands were placed in pronation within the scanning field.

Participants were instructed to remain still for the entire procedure. A trained laboratory technician performed the device calibration and analysis according to the manufacturer’s guidelines. Calibration was executed daily. To evaluate reproducibility, test-retest scans were conducted in a subsample of seven participants 24 hours apart. Reliability was excellent, with an intraclass correlation coefficient (ICC) of 0.98, and coefficients of variation (CV) of 0.86% for FFM and 1.5% for fat mass.

Bioelectrical impedance

Total body water (TBW), resting metabolic rate (RMR), resistance (R), and reactance (Xc) were assessed using a single-frequency bioelectrical impedance analyzer (Biodynamic Body Composition Analyzer, model 310e; Biodynamics Corporation, Seattle, WA, USA). Phase angle (PhA, degrees) was derived from the equation: PhA = arctan (Xc/R) x 180/ π. To minimize hydration-related variability, participants were instructed to empty their bladder immediately before testing, abstain from alcohol, caffeinated drinks, and diuretics for the previous 48 h, and avoid vigorous physical activity for at least 24 h. Measurements were scheduled in the morning after an overnight fast, within approximately one hour of waking.

Prior to testing, participants removed all metallic objects and rested supine for 10 min with arms and legs positioned at ~45°. After the skin was cleaned with alcohol, four electrodes were placed on the dorsal surfaces of the right hand and foot, following established procedures [33]. Reproducibility was verified in a subsample of seven participants who completed duplicate scans 24 hours apart. Reliability was excellent: ICCs were 0.98 for TBW, 0.95 for R, 0.96 for Xc, and 0.96 for PhA. The corresponding CV were 1.3%, 2.2%, 4.1%, and 2.5%, respectively. The validity of single-frequency BIA for estimating body composition and PhA in older adults has been confirmed in previous studies, supporting its use in both clinical and exercise settings [34,35].

Secondary Outcomes

Muscle strength

Maximal dynamic strength was assessed using one-repetition maximum (1RM) tests in the chest press, leg extension, and biceps curl exercises. Testing followed a protocol specifically validated for older women, involving two to three sessions to ensure familiarization and reliability of measurements [36]. Each session began with a warm-up set of 6-10 repetitions at approximately 50% of the estimated 1RM. After a 2-minute rest, participants performed up to three attempts to determine their maximal load, with 3-5 min of rest between attempts and 5 min between exercises. The 1RM was defined as the highest load lifted through the full range of motion with correct technique and no assistance. Loads were adjusted by 3-10% depending on performance in each attempt. All sessions were conducted under direct supervision by trained professionals to ensure safety and proper execution. This method has demonstrated high test-retest reliability in untrained older women, with ICCs exceeding 0.95 for all exercises, and CV ≤ 3.2% [36].

Muscle quality index

Muscle quality index (MQI) was calculated separately for each exercise as the ratio of one-repetition maximum (1RM, kg) to the corresponding segment-specific fat-free mass (FFM, kg) obtained by DXA. Thus, MQI was operationalized as exercise-specific maximal strength normalized to regional lean tissue (1RM/segment-specific FFM), rather than a whole-body strength-to-FFM ratio. The specific calculations were as follows:

• Chest Press MQI = Chest press 1RM (kg) / FFM (kg)

• Leg Extension MQI = Leg extension 1RM (kg) / FFM (kg)

• Preacher Curl MQI = Preacher curl 1RM (kg) / FFM (kg)

Dietary intake

Participants were instructed by a registered dietitian to complete food diaries on three nonconsecutive days, two weekdays and one weekend day, both at baseline and at the end of the intervention. To improve accuracy, they received guidance on estimating portion sizes and recording all foods and beverages consumed. Standardized food models were also used to aid portion reporting. Nutrient intake was analyzed with Avanutri Processor software (version 3.1.4; Avanutri, Rio de Janeiro, Brazil), providing estimates of total energy intake as well as macronutrient distribution (protein, carbohydrate, and fat). Throughout the intervention, participants were encouraged to maintain their habitual diet without intentional modifications.

Resistance training program

The resistance training (RT) intervention lasted 12 weeks and was conducted in the morning at the university’s exercise facility, under full supervision by certified professionals. The training protocol adhered to ACSM guidelines for improving muscle strength and hypertrophy in older populations [8,9]. Participants performed eight exercises targeting major muscle groups using a mix of machines and free weights: chest press, lat pulldown, preacher curl, triceps pushdown, leg extension, leg curl, seated calf raise, and abdominal trunk flexion. Training frequency differed by group while maintaining equated weekly volume. The G2x group trained two days per week, performing three sets per exercise, while the G3x group trained three days per week, performing two sets per exercise. Weekly training volume was equated between groups, since total volume was calculated as sets × repetitions × load (kg), resulting in six sets per exercise per week in both conditions. Most exercises were performed at 10-15 repetition maximum (RM), with exceptions for seated calf raises (15-20 RM) and abdominal exercises (20-30 repetitions, body weight only). Rest intervals were standardized: 1-2 min between sets and 2-3 min between exercises. Participants were instructed to inhale during the eccentric muscle action and exhale during the concentric muscle action while maintaining a constant velocity of movement at a ratio of 1:2 (concentric and eccentric muscle actions, respectively). Progression was planned so that when 15 repetitions were completed for 2 consecutive sessions in an exercise, weight was increased 2-5% for upper limb exercises and 5-10% for lower limb exercises in the next training session [9].

The methodological procedures described in this study reflect standardized protocols routinely implemented in our laboratory. Because multiple trials have been conducted within our research group involving similar populations, training regimens, and testing frameworks, some overlap in methodological descriptions is unavoidable. Previous applications of these procedures in older women have been reported elsewhere [37], and the present trial extends this line of work by focusing specifically on MQI and PhA as primary endpoints.

Statistical Analyses

Descriptive statistics were calculated and presented as mean and standard deviation. The Shapiro-Wilk test was used to assess the normality of continuous variables. For the primary analysis of intervention effects, Generalized Estimating Equations (GEE) were applied with an exchangeable working correlation matrix to account for within-subject repeated measures. GEE models were used to evaluate the main effects of time (pre vs. post), group (G2x vs. G3x), and the interaction between time and group for all outcome variables, including FFM, fat mass (kg and %), PhA, TBW, RMR, MQI, and FMR. To explore interindividual variability, individual pre-to-post changes were examined relative to conservative thresholds. The smallest worthwhile change (SWC) was calculated as 0.2 × baseline SD for each variable (FFM: 0.71 kg; PhA: 0.12°). When combined with the standard error of measurement (SEM; 0.30 kg for FFM and 0.21° for PhA), the thresholds for meaningful individual change were defined as 0.77 kg for FFM and 0.24° for PhA (√SEM² + SWC²), which were operationally defined as the minimal detectable change (MDC) for each outcome. In response to reviewer feedback, a sensitivity analysis was also conducted using a variability-based criterion (0.5 SD) to contextualize population-level heterogeneity. Individual changes exceeding these thresholds were reported descriptively (e.g. waterfall plots). Additionally, percentage changes in MQI were calculated for each exercise (chest press, leg extension, biceps curl), and interaction effects were verified using GEE models. The level of statistical significance was set at p < 0.05. Data were stored and analyzed using JAMOVI software version 2.3.28.

RESULTS

A total of 45 participants were included in the final analysis (G2x = 22; G3x = 23), corresponding to the final sample reported in the original randomized controlled trial [32]. One participant from G2x discontinued participation due to an accident not related to the intervention, and one participant from G3x withdrew due to loss of interest. Baseline characteristics were comparable between groups regarding age, height, body mass, BMI, body composition, and nutritional intake (p > 0.05) (Table 1).

Table 1.

Sample characterization at the beginning of the intervention

Variable	G2x (n = 22)	G3x (n = 23)	p-value
Age (years)	65.43 ± 4.36	64.92 ± 4.60	0.693
Height (cm)	156.20 ± 5.92	156.33 ± 5.91	0.945
Body mass (kg)	62.08 ± 8.06	59.77 ± 9.14	0.369
Body mass index (kg/m²)	25.49 ± 3.48	24.43 ± 3.32	0.294
Body Fat (%)	40.80 ± 6.46	40.50 ± 6.21	0.870
Fat mass (kg)	25.07 ± 6.36	24.28 ± 6.82	0.682
Fat-free mass (kg)	35.47 ± 3.50	34.75 ± 3.61	0.545
Caloric intake (kcal/day)	1395.19 ± 201.75	1392.64 ± 209.30	0.967
Caloric intake (kcal/kg/day)	22.70 ± 4.70	23.63 ± 5.79	0.560
Protein intake (g/kg/day)	0.93 ± 0.30	1.02 ± 0.32	0.522
Carbohydrate Intake (g/kg/day)	3.22 ± 0.82	3.37 ± 0.94	0.943
Fat intake (g/kg/day)	0.66 ± 0.18	0.74 ± 0.27	0.405
Resting metabolic rate (kcal)	1262.70 ± 153.60	1230.22 ± 172.62	0.504

Note. G2x = group that performed resistance training two times/week; G3x = group that performed resistance training three times/week. 1RM = one repetition maximum test. Data are expressed as mean and standard deviation.

Significant main effects of time were observed for all exercises, indicating improvements in MQI following the intervention: chest press (p < 0.001), leg extension (p < 0.001), and preacher curl (p = 0.001). No significant time × group interactions were found for any exercise (all p ≥ 0.16), suggesting that gains in MQI were consistent between the 2×/week and 3×/week groups (Table 2). Descriptively, percentage changes in MQI were similar between groups across exercises, with increases of 10.1% vs. 13.2% (chest press), 17.9% vs. 11.9% (leg extension), and 4.8% vs. 4.9% (preacher curl) in G2x and G3x, respectively (Table 2).

Table 2.

Muscle quality index (strength-to-FFM ratio, MQI) for chest press, leg extension, and preacher curl at pre- and post-training in older women

Exercise	G2x MQI		G3x MQI		p-interaction
	Pre-training	Post-training	Pre-training	Post-Training
Chest Press (1RM / FFM)	0.69 ± 0.09	0.76 ± 0.12*	0.68 ± 0.14	0.77 ± 0.14*	0.947
Leg Extension (1RM / FFM)	0.56 ± 0.07	0.66 ± 0.099*	0.59 ± 0.10	0.66 ± 0.09*	0.937
Preacher Curl (1RM /FFM)	0.42 ± 0.06	0.44 ± 0.06*	0.41 ± 0.05	0.43 ± 0.06*	0.957

Note. MQI = muscle quality index, represented by strength-to-FFM ratio; G2x = group that performed resistance training two times/week; G3x = group that performed resistance training three times/week. 1RM = one repetition maximum test. Data are expressed as mean and standard deviation. FFM = fat-free mass. 1RM = one repetition maximum test.

^*= p < 0.05 vs. pre-training.

Table 3 summarizes body composition and phase angle-related variables. In line with prior analyses of this dataset [32], only the 3×/week group showed reductions in fat mass (≈ −3%), whereas the 2×/week group showed no meaningful change. FFM increased modestly in both groups (≈ +1.5%). FMR did not change from pre- to post-intervention in either group. PhA did not change significantly in either group, and mean fluctuations (~0.2-0.3°) were below the minimal detectable change (MDC = 0.39°), indicating that variations were within measurement error. No significant changes were observed for TBW or RMR in either group.

Table 3.

Changes in body composition and bioimpedance-derived outcomes following the intervention

Variable	G2x (n = 22)		G3x (n = 23)		Interaction P-value
	Pre-training	Post-training	Pre-training	Post-training
PhA (°)	6.28 ± 0.67	6.22 ± 1.52	5.96 ± 0.41	6.07 ± 0.42	0.623
TBW (L)	31.00 ± 3.20	31.56 ± 3.40	30.47 ± 3.39	30.32 ± 3.25	0.067
RMR (kcal)	1262.70 ± 153.60	1279.41 ± 152.49	1230.22 ± 172.62	1232.29 ± 160.47	0.475
Body fat (%)	40.80 ± 6.46	40.60 ± 7.02	40.50 ± 6.21	39.50 ± 6.19*	0.264
Fat Mass (kg)	25.07 ± 6.36	25.35 ± 6.75	24.28 ± 6.82	23.60 ± 6.66*	0.080
FFM (kg)	35.47 ± 3.50	36.11 ± 3.38*	34.75 ± 3.61	35.33 ± 4.07*	0.914
FMR	0.71 ± 0.19	0.70 ± 0.20	0.69 ± 0.18	0.67 ± 0.17	0.151

Note. G2x = group that performed resistance training two times/week; G3x = group that performed resistance training three times/week. 1RM = one repetition maximum test. Data are expressed as mean and standard deviation. RMR = resting metabolic rate; FFM = fat-free mass; PhA = phase angle; TBW = total body water; FMR = fatto-muscle ratio;

^*= p < 0.05 vs. pre-training.

Individual changes were evaluated against conservative thresholds combining measurement precision and population variability (MDC, calculated as √SEM² + SWC: 0.77 kg for FFM; 0.24°for PhA). For PhA, a proportion of participants in both groups exceeded the MDC (G2x: 73%; G3x: 35%; Figure 2). For FFM, similar proportions exceeded the MDC in both groups (G2x: 55%; G3x: 48%; Figure 3). These proportions are reported for descriptive purposes only and should not be interpreted as evidence of frequency-dependent effects. Individual ΔFFM values ranged from approximately −2.0 to +2.5 kg, while ΔPhA values ranged from approximately −0.9° to +1.3°, illustrating substantial inter-individual variability.

Figure 2. Individual changes in phase angle (ΔPha) after traning. Note.

G2x = group that performed resistance traning two times/week; G3x = group that performed resistance traning three times/week.

Figure 3. Individual changes in fat-free mass (ΔFFM) after traning. Note.

G2x = group that performed resistance traning two times/week; G3x = group that performed resistance traning three times/week.

DISCUSSION

This study investigated the effects of RT frequency on functional and bioelectrical impedance-derived adaptations in older women, using a volume-equated randomized controlled design. The current findings are derived from the same trial previously reported by Pina et al. [32], but the present secondary analysis focused on MQI and PhA. At the group level, both 2×/week (G2x) and 3×/week (G3x) programs resulted in significant improvements in MQI, with no significant time × group interactions. In contrast, no significant group-level changes were observed for PhA under the present conditions, although substantial inter-individual variability was evident. Accordingly, the main inferential conclusion is limited to similar MQI improvements under volume-matched conditions, whereas the individual-response analyses are presented as descriptive and hypothesis-generating [12,13,27].

The primary finding was that performing RT two or three times per week, under volume-controlled conditions, led to comparable improvements in MQI, reflecting strength gains relative to active tissue, consistent with earlier analyses of this dataset [32] and corroborating findings from recent meta-analyses demonstrating that weekly training volume, rather than frequency alone, is the predominant factor in driving adaptations in older adults [10,31]. The inclusion of MQI provided a functionally relevant index of strength normalized to FFM, capturing changes beyond absolute FFM accretion or maximal strength alone. Importantly, because MQI was calculated using whole-body FFM, it should be interpreted as a functional index of strength relative to total lean mass, rather than a direct marker of site-specific or tissue-level muscle quality. Therefore, mechanistic inferences regarding neural versus morphological drivers cannot be established from the present data. Moreover, because functional performance outcomes such as power, balance, or gait speed were not assessed, the clinical implications of the observed MQI improvements cannot be directly inferred and should be interpreted with caution.

In response to recent literature emphasizing the relevance of relative indices of body composition, we also examined FMR as an exploratory outcome. FMR integrates information on adiposity and lean tissue into a single metric and has been proposed as a sensitive marker of sarcopenic obesity and metabolic risk in older adults [22-24]. Previous studies have further suggested that the relative balance between fat mass and muscle mass may be more strongly associated with functional and metabolic outcomes than isolated measures of either compartment alone [24]. In the present study, however, FMR did not change following the intervention in either group, despite modest reductions in fat mass and increases in fat-free mass. This finding suggests that, under volume-matched resistance training conditions and over a 12-week period, changes in individual body composition components may not have been sufficient to substantially alter their relative balance, reinforcing the notion that training frequency per se has limited influence on qualitative indices of body composition in older women when weekly volume is equated.

The analysis of individual variability indicated that a considerable proportion of participants exceeded conservative thresholds for meaningful change (0.77 kg for FF and 0.24° for PhA), underscoring the heterogeneous nature of RT adaptations. Importantly, these responder-type analyses are descriptive and exploratory and should not be interpreted as evidence of frequency-dependent effects or superiority of one training condition over another. However, these results should be interpreted with caution, as single-trial designs cannot fully distinguish biological responsiveness from measurement error, regression to the mean, or sampling variability [27,29,30]. Exploratory sensitivity analyses using SD-based riteria produced lower proportions, reinforcing that the interpretation of interindividual differences depends on the thresholds applied. These descriptive findings do not alter the primary inferential conclusion based on group-level analyses. Nonetheless, examining individual variation provides complementary insight, highlighting that group averages may mask clinically relevant changes.

The limited extent to which participants exceeded thresholds for meaningful changes in PhA may be partly explained by the nature of this biomarker. PhA reflects impedance- derived properties related to membrane capacitance and fluid distribution, and meaningful improvements may require longer intervention durations or adjunctive strategies such as nutritional support [25]. Importantly, the absence of significant group-level changes in PhA should be interpreted as a null finding and does not constitute evidence of equivalence between training frequencies. In addition, the modest variation observed in PhA (−0.9° to +1.3°) may lie within the biological variability of single-frequency bioimpedance analysis, particularly in populations with relatively low protein intake, reinforcing the need for cautious interpretation in short-term interventions [34,35].

Despite these considerations, the analysis of individual- level data provides descriptive insight into the heterogeneous nature of training adaptations. It reinforces the importance of multidimensional outcome assessment, beyond strength and muscle mass, particularly in older populations with diverse physiological profiles [15]. Furthermore, identifying variability in individual responses may inform hypothesis generation for future studies, including the examination of adjunctive strategies, such as nutritional interventions or alternative training modalities, to optimize outcomes.

A methodological strength of this study lies in the rigorous control of training volume, the fully supervised intervention, and the standardized assessment procedures, all of which enhance internal validity. Furthermore, both DXA and BIA demonstrated excellent reproducibility in our protocol, with test-retest ICCs ≥ 0.95 for key outcomes such as FFM, PhA, and TBW33,36. Nonetheless, some limitations must be acknowledged. First, although the sample size was larger than that of several comparable trials in this field, it was still relatively small, limiting statistical power to detect subtle group × time interactions and increasing the influence of sampling variance. Second, the intervention lasted 12 weeks, which is consistent with previous RT studies in older adults, but may be insufficient to capture adaptations in outcomes such as PhA or RMR, which may require longer-term interventions or combined strategies to change meaningfully. In addition, PhA represents an indirect, bioimpedance-derived marker related to impedance properties and hydration status, rather than a direct measure of cellular function or integrity; therefore, interpretations regarding cellular adaptations should be made with caution, particularly in light of the absence of group-level changes observed in the present study. Third, dietary intake was not controlled, and average protein consumption was relatively low for this population, which may have attenuated potential gains in muscle mass or PhA. Fourth, participants were healthy, community-dwelling older women without severe comorbidities. While this homogeneity enhances internal validity, it restricts the generalizability of the findings to frailer populations, men, or individuals with chronic diseases. Fifth, although we assessed a range of morphological and cellular outcomes, we did not include neuromuscular performance measures such as power output, balance, or functional mobility, which could provide additional insight into the practical implications of training frequency in older adults. Finally, the exploratory analysis of interindividual variability cannot disentangle true biological differences from measurement error or baseline dependency, even though conservative thresholds (SEM + SWC) and sensitivity analyses (SD-based) were applied.

In conclusion, under volume-matched conditions, RT performed two or three times per week resulted in similar improvements in MQI in older women. No significant group-level changes were observed for PhA over the 12-week intervention. Therefore, when total weekly training volume is equated, the present data do not indicate differential group-level effects of RT frequency (2 vs. 3 sessions/week) on MQI, and no frequency-related differences were detected for PhA. Inter-individual variability was evident, but these analyses were descriptive and should not be used to support comparative inferences regarding frequency superiority. From a practical perspective, training frequency may be selected based on adherence, recovery, and logistical feasibility, as volume is a key indicator for adaptations [38], provided that weekly volume and progression are appropriately prescribed and supervised [2,39].