Effect of Whey Protein Supplementation in Postmenopausal Women: A Systematic Review and Meta-Analysis

(1) Background: Whey protein (WP) in combination with resistance training (RT) is beneficial in improving sarcopenic obesity and its damaging effects in older adults, while the difference between men and women should be considered while interpreting results. This review aims to investigate WP’s efficacy on postmenopausal women with or without RT; (2) Material and Methods: We searched electronic databases including PubMed, EMBASE, and the Cochrane Library from inception to August 2021 for randomized controlled trials that included comparison groups to evaluate WP’s efficacy in women aged 55 years and above. The outcomes included body composition, muscular strength, functional capacity, and dietary intake. Standardized mean differences (SMDs) with 95% confidence intervals (CIs) were used to estimate the effect of WP. We also performed subgroup analysis with or without RT; (3) Results: We included 14 studies in the systematic review and 10 studies in the meta-analysis. Subgroup analyses showed RT was a major confounder for muscle strength, lean mass, and dietary protein intake (PI). In the RT subgroup, WP supplementation had a significant positive effect on biceps curl strength (BC) (SMD: 0.6805, 95% CI: 0.176, 1.185, I2: 0%), and lower limb lean-mass (LLLM) (SMD: 1.103, 95% CI: 0.632, 1.574, I2: 14%). In the subgroup without RT, a significant negative effect on PI (SMD: −0.4225, 95% CI: −0.774, −0.071, I2: 47%) was observed, while no significant effect on muscle strength or lean mass was revealed. WP supplementation did not show a significantly different effect on fat mass or body weight loss in both the subgroups; (4) Conclusions: In postmenopausal women, WP supplementation only in combination with RT enhances BC and LLLM compared to placebo controls. Without RT, WP has no significant benefit on muscle strength or lean mass.


Introduction
Aging is an irreversible process which predisposes individuals to many biological alterations, often associated with a progressively decreased muscle mass and strength (sarcopenia), and an increased fat mass (obesity) [1]. This combination, termed sarcopenic obesity, can cause chronic inflammation thus explaining the increased risk of developing metabolic syndrome and comorbidities, such as type-2 diabetes mellitus, dyslipidemia,

Eligibility Criteria and Study Selection
The eligible studies should meet the following inclusion criteria: (1) A randomized controlled trial that included comparison groups to evaluate the efficacy of diet supplemented with WP, either with or without RT; (2) Studies that used validated tools to evaluate body composition, lean mass (depicted as lean mass, muscle mass, or lean soft tissue mass), muscular strength, functional capacity, and dietary intake outcomes; and (3) Study participants that included women above 55 years of age. We excluded articles that met at least one of the following criteria: (1) Study types other than RCTs; (2) Non-human research; and (3) Insufficient outcome data. Titles and abstracts of potentially eligible studies were screened, their full texts were retrieved, and reviewed based on the eligibility criteria. Two independent authors (H. Y. Chang and Y. Y. Kuo) were involved in the retrieval, screening and selection process.

Data Extraction and Quality Assessment
Two independent authors (H. Y. Chang and Y. Y. Kuo) extracted the data from the included studies. The extracted data included the author's surname, year of publication, country, ethnicity, sample size, participant characteristics, exercise interventions, changes between baseline and endpoint on study outcomes, study duration, and time points of data collection. The methodological quality of each included study was assessed by using the modified Cochrane Risk of Bias tool Version 2.0, which included the following five domains: allocation bias, performance bias, attrition bias, detection bias, and reporting bias. The overall bias of each article was determined to be low, unclear, or high risk. A third reviewer (Y.C. Huang) was consulted for any disagreement to reach a consensus.

Data Synthesis and Statistical Analysis
The primary outcomes were changes in muscular strength and functional capacity, and gain of lean mass; the secondary outcomes were changes in fat mass, loss of body weight, and dietary intake from baseline study outcome to the treatment endpoint. Meta-analyses were performed when at least two studies compared the same outcome in participants receiving or not receiving WP supplementation. A subgroup analysis for participants with or without RT was also performed. Continuous variables were expressed by standardized mean differences (SMDs) with 95% confidence intervals (CIs). A random-effects model was adopted due to possible heterogeneity. Data heterogeneity was assessed using the Cochran Q-test and I 2 test. Both funnel plot and Egger's test were used to assess publication bias if more than 10 studies were available [22]. All of the analyses were performed using Comprehensive Meta-Analysis Version 3 software (Biostat, Inc., Englewood, NJ, USA).

Quality of Evidence
The Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) approach including risk of bias, inconsistency, indirectness, imprecision, and publication bias was used to evaluate the quality of evidence of each outcome in this systematic review and meta-analysis.

Search Results and Trial Characteristics
The primary search terms initially identified 330 records potentially eligible for inclusion. After excluding 109 duplicated records along with 25 records that were reviewed, guidelines, or conference reports, a total of 196 records entered the screening stage. After assessing the title or abstract, 140 records were excluded as they were irrelevant to our research, resulting in a total of 56 reports that were sought for retrieval and assessed for eligibility. Forty-two reports were excluded as they met the exclusion criteria or did not meeting the inclusion criteria. Fourteen reports were included in the review (Figure 1).

Search Results and Trial Characteristics
The primary search terms initially identified 330 records potentially eligible for inclusion. After excluding 109 duplicated records along with 25 records that were reviewed, guidelines, or conference reports, a total of 196 records entered the screening stage. After assessing the title or abstract, 140 records were excluded as they were irrelevant to our research, resulting in a total of 56 reports that were sought for retrieval and assessed for eligibility. Forty-two reports were excluded as they met the exclusion criteria or did not meeting the inclusion criteria. Fourteen reports were included in the review (Figure 1). All 14 included studies were RCT with a control group. Among these, 2 were single blinded [23,24], and 12 were double blinded [25][26][27][28][29][30][31][32][33][34][35][36]. These studies were implemented in six different countries with the majority having an intervention duration of 12 weeks (ranged from 10 weeks to 72 weeks). Detailed study characteristics are shown in Tables 1  and 2. Among these 14 studies, 11 studies included the data of WP with RT [23][24][25][26][27][28][29][30][31][32]35], and 4 studies included the data of WP without RT [23,33,34,36]. Furthermore, seven studies were conducted by the same team and the participants were from the same group [26][27][28][29][30][31][32]. To address the concern of bias due to overlapped participants, we only selected one study with the largest number of participants among these seven studies for each All 14 included studies were RCT with a control group. Among these, 2 were single blinded [23,24], and 12 were double blinded [25][26][27][28][29][30][31][32][33][34][35][36]. These studies were implemented in six different countries with the majority having an intervention duration of 12 weeks (ranged from 10 weeks to 72 weeks). Detailed study characteristics are shown in Tables 1 and 2. Among these 14 studies, 11 studies included the data of WP with RT [23][24][25][26][27][28][29][30][31][32]35], and 4 studies included the data of WP without RT [23,33,34,36]. Furthermore, seven studies were conducted by the same team and the participants were from the same group [26][27][28][29][30][31][32]. To address the concern of bias due to overlapped participants, we only selected one study with the largest number of participants among these seven studies for each outcome. In addition, another two studies were conducted by the same team with the same participants [34,36]. However, they did not present the same outcomes and we included both of them in our study. We reported each study rather than collapsing to one to clarify the difference between these studies. Finally, four studies were excluded in the meta-analysis for no outcome selected [26,28,29,32], while 10 studies were included in the meta-analysis for quantitative analysis [23][24][25]27,30,31,[33][34][35][36].

Risk of Bias Assessment
The risk of bias assessment of the studies included in this systematic review is presented in Figure 2a,b. High risk of bias was only seen in the domain of attrition bias in one study for incomplete outcome data [35]. In that study, 12 participants completed the study while the outcome data of biceps curl strength (BC), knee extension strength, and knee flexion strength were available from only 10 of them. There was no evidence that the result was unbiased, and the missing data likely depended on true value. Therefore, attrition bias in this study was rated as high.

Participants' Characteristics
The total participants across the included studies were 776, with average age ranging from 57 to 74 years and average BMI ranging from 22 to 28 kg/m 2 . Participants were declared to be "healthy" in four studies [23][24][25]35], and to be "healthy or sarcopenic" in three studies [27,30,31]. Participants were classified to be PI deficient in two studies, reporting a PI at baseline lower than the current adult recommended dietary-allowances [30,32].

RT Characteristics
Among the 10 included studies, 7 studies included the data of WP with RT. The RT interventions duration ranged from 10 to 24 weeks undergoing RT for 2 to 3 days per week, with 4-8 exercises per session, with 0-6 sets per exercise, and with 8-12 repetitions per set (or to fatigue). All the studies involved upper and lower body RT. Each outcome of muscle strength was trained by specific RT programs in all the studies.

Results of Meta-Analysis
The detailed data of each included study are shown in Tables S2-S5. The results of the meta-analysis are shown in Table 3.

Effect of WP Supplementation on Muscle Strength and Functional Capacity
We

Effect of WP Supplementation on Muscle Mass Gains
We included upper-limb lean mass (ULLM), lower-limb lean mass (LLLM), and skeletal muscle mass (SMM) in the muscle mass gain-assessment. The results showed that WP supplementation may not enhance the outcomes related to muscle mass.  (Figures S8-S10).
Meta-analyses of ULLM, LLLM and SMM were not performed in the WP group without RT due to the lack of data availability.

Effect of WP Supplementation on Daily Dietary Nutrients Intake
We regarded daily dietary intake as an outcome in our study to investigate how the introduction of a daily supplement impacted on daily macronutrient and energy intakes. We included total energy intake (TEI), carbohydrate intake (CHI), protein intake (PI), and fat intake (FI) in daily dietary nutrient intake assessment. We used the dietary nutrient intake with deduction of the WP and placebo supplementation to render the results more objective. Two studies presented total nutrient intake including WP and placebo supplementation [30,36], while they both provided the nutrient composition of the supplements. Therefore, the dietary nutrient intake with WP and the placebo supplementation deduction were calculated according to the supplementation protocol.
The participants with WP supplementation showed that WP supplementation may reduce PI (

Quality of Evidence Assessment by GRADE
The results of the GRADE evaluation are presented in Table 3. We evaluated all of the outcomes with more than one paper providing information. Therefore, all 17 outcomes were assessed in the analysis of combining all included studies, while only 11 and 10 outcomes were assessed in the subgroup analysis of WP supplementation with and without RT, respectively. As the design of the inclusion criteria was rigid, there was no obvious intransitivity. In the analysis of combining all of the included studies, the confidence of the evidence of FI and TEI were high for no downgrading, while that of FM was very low due to the inconsistency, imprecision, and some concerns in relation to the risk of bias. In the subgroup assessment, the extent of inconsistency decreased, while that of imprecision increased. For the subgroup of WP supplementation with RT, the confidence of the evidence of KE and KF was very low due to the imprecision and high risk in relation to the risk of bias. For the subgroup analysis of WP supplementation without RT, the confidence of the evidence of FM was very low due to the inconsistency, imprecision, and some concerns in relation to the risk of bias.

Discussion
Based on evidence, this review summarized the effect of WP supplementation on postmenopausal women. The main findings of this meta-analysis are: (1) WP with RT significantly enhances LLLM gain and BC; (2) WP without RT significantly reduces PI.

Effect of WP Supplementation on Muscle Strength and Functional Capacity
Subgroup analysis showed that RT augments the benefits of whey protein for muscle strength. WP supplementation in the group with RT demonstrated a significant enhance-ment of BC, while that without RT showed no significant enhancement of all kinds of muscle strength and functional capacity. The analysis results partially agree with previous meta-analyses by Finger et al., Morton et al.,and Liao et al. [19,20,37], which investigated the effect of protein supplementation together with RT, and a previous meta-analysis by Richard et al., suggesting that protein interventions augment the effect of RT on muscle strength in older adults. Therefore, despite the difference in metabolism between men and women, for postmenopausal women consuming sufficient quantities of protein, WP supplementation still enhances muscle strength, but only when combined with RT. A possible explanation is that RT increases fasted-state protein losses and the need for protein. In this condition, WP supplementation promotes maintaining whole body protein-balance [38].
Another study concluded that the improvement in strength was related to better physical and social function [39]. Nevertheless, in our study, no significant effect on functional capacity was revealed. More studies with uniformity in outcome measures would be needed to clarify this conclusion.

Effect of WP Supplementation on Muscle Mass
RT played an important role for muscle mass in the subgroup analysis. WP supplementation in the RT group showed a significant enhancement in LLLM. However, the group without RT demonstrated a significant decrease in ULLM and LLLM. The results of our analysis partially agree with previous meta-analyses [19][20][21]37], suggesting that protein interventions augment the effect of RT on appendicular lean mass. For the significant decrease in ULLM and LLLM, the results were constructed by only one study and the intervention time of this study was one year [34], while that of the studies presenting the result of WP with RT was 10-24 weeks, and the effect of age-related decline was taken into consideration. Moreover, this study used an isocaloric carbohydrate for supplementation in control group. Carbohydrate stimulates pancreatic insulin secretion, which can inhibit muscle protein breakdown [40,41]. Elderly people of normal status have lower insulin secretion [42]. This might explain the decrease in ULLM and LLLM.
The results of WP supplementation in the RT group did not show a significant effect on lower limb muscle strength, but demonstrated a significant enhancement of LLLM. A previous study presented the same outcome [43], showing that the relationship between muscle strength and muscle mass differed according to sex and age. For women between 65 and 74 years, no significant relationship between muscle strength and muscle mass was demonstrated. Both "neural" and "muscular" factors are required for muscle strength; meanwhile, this neural activity decreases with aging [44]. In addition, strength tests are performance tests and affected by technique and motivation. In novice individuals, there is more testing variability. These might explain the lack of a relationship between lower limb muscle strength and LLLM.

Effect of WP Supplementation on FM and BW Loss
Subgroup analysis showed that RT was not a key factor in FM and BW loss. Both WP groups, with and without RT, showed a positive effect on FM loss, corresponding to most studies showing that increasing dietary protein increases diet-induced thermogenesis and promotes greater fat loss. Compared to the group of WP without RT, WP supplementation with RT showed less effect on FM loss. A possible explanation is that the intervention time of this study was 18 months [33], while that of the other three studies which provided the results of FM loss were 12-16 weeks [23,24,27]. With a longer intervention time, the effect on FM loss might be more significant.
Our results demonstrated that WP supplementation might not be necessary to be combined with RT for FM loss in postmenopausal women.

Effect of WP Supplementation on Daily Dietary Nutrients Intake
WP supplementation in the group without RT demonstrated a negative effect on PI without significant difference, while that with RT showed a significant positive ef- fect on PI. The results of our analysis partially agree with a previous meta-analysis by Colonetti et al. [45], which investigated the effect of dietary protein supplementation together with RT. For the negative effect on PI in the group without RT, it might be explained by the use of the dietary nutrient intake with deduction of WP and placebo supplementation in our study. The Study Group on meeting protein needs of older people (PROT-AGE) and the European Society for Clinical Nutrition and Metabolism (ESPEN) study showed that 1.0-1.2 g/kg BM −1 day −1 of protein for well-nourished active older adults is sufficient [46,47], while RT can increase fasted-state protein losses and the need of protein. Total protein intake remained unchanged in the group of WP without RT. With extra supplementation of protein, a new balance of dietary intake was achieved, thus contributing to a decline in PI. This might indirectly conclude that WP supplementation was not very helpful for postmenopausal women without RT, especially for those who already consumed sufficient quantities of protein at baseline. On the contrary, for those who underwent RT, WP supplementation might augment the effect of RT and increase the need for protein. Therefore, WP supplementation should be combined with RT for increasing daily dietary PI in postmenopausal women.

Limitation
First, due to restricted number of included studies, subgroup analysis other than RT/non-RT is not constructed in our study. We did not divide the group into trained/untrained individuals, healthy/sarcopenia individuals, sufficient/deficient PI individuals, or build a subgroup analysis of protein supplementation dosage. We also did not discriminate the role of WP from being a supplement or a correction of a deficient diet. Furthermore, some studies suggested that resistance-trained individuals might need a higher dietary intake and protein supplementation [20,41]. Second, the intervention time in non-RT studies is longer than RT studies. The adherence to RT, WP, and PLA supplementation, and the effect of aging on muscle mass, muscle strength, and FM should be considered. Third, we included lean mass, muscle mass, and lean soft tissue mass in the analysis of lean mass gain, and the variation in bone and water should be also considered. Fourth, participants in Sugihara et al. had a basal PI lower than the current adult-recommended dietary allowances, and the results of TEI, CHI, PI, and FI should be interpreted with caution. Last, the restricted number of included studies might lead to some results not showing a significant effect.

Conclusions
Compared to placebo control, WP supplementation causes an improvement in BC and LLLM in postmenopausal women, only when combined with RT. However, the quality of evidence of BC and LLLM in the group of WP with RT was low and moderate, respectively. With both of them presenting some risk of bias, the results should be interpreted with caution. More large scale RCTs are required for a better understanding of the effects of WP supplementation combined with RT.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/nu14194210/s1, Method S1: The full search strings: example of PubMed; Table S1: PRISMA 2020 checklist; Table S2: Differences in muscle strength between whey protein supplements and control supplements considering the subgroup analysis of RT/non-RT; Table S3: Differences in muscle mass between whey protein supplements and control supplements considering the subgroup analysis of RT/non-RT; Table S4: Differences in fat mass and body weight between whey protein supplements and control supplements considering the subgroup analysis of RT/non-RT; Table S5: Differences in daily nutrients intake between whey protein supplements and control supplements considering the subgroup analysis of RT/non-RT; Figure S1: Forest plot of the grip strength; Figure S2: Forest plot of the biceps curl strength; Figure S3: Forest plot of the knee extension strength; Figure S4: Forest plot of the knee flexion strength; Figure S5: Forest plot of the gait speed test; Figure Figure S12: Forest plot of the body weight; Figure S13: Forest plot of the total energy intake; Figure S14: Forest plot of the carbohydrate intake; Figure S15: Forest plot of the fat intake; Figure S16: Forest plot of the protein intake. Institutional Review Board Statement: Ethical review and approval were waived due to this study being a systematic review and meta-analysis.
Informed Consent Statement: Patient consent was waived due to this study being a systematic review and meta-analysis.
Data Availability Statement: Data were available within the article and its supplementary materials.

Conflicts of Interest:
The authors declare no conflict of interest.