Is higher protein intake during weight loss interventions in older adults associated with improved outcomes? A secondary data analysis of three randomised controlled trials

Abstract

Background

Weight loss interventions can exacerbate sarcopenia and sarcopenic obesity in older individuals. Increasing protein intake, alongside resistance training, is suggested to help preserve muscle mass, strength, and function during energy-restricted diets, but the benefits are not consistently supported by all studies in older adults.

Objective

To investigate the relationship between protein intake and body composition, muscle strength, and function in weight loss programs with energy restriction, with or without increased protein intake (through supplementation or dietary counseling) and combined with resistance training in older adults with overweight or obesity.

Methods

A secondary data analysis of three randomized controlled trials was conducted, including participants aged ≥ 55 with a BMI ≥ 25 kg/m². Data on protein intake at baseline and follow-up were analyzed. The follow-up period was 10 weeks in one study and 13 weeks in the other two studies.Outcomes included body weight, body fat, appendicular lean soft tissue mass (ALST), handgrip strength (HGS), gait speed, chair stand test (CST), and short physical performance battery (SPPB). Linear mixed models were used to identify associations between protein intake and outcomes.

Results

A total of 191 older adults (mean age 65.1 years, mean BMI 32.9 kg/m²) were included. Protein intake significantly increased from 0.87 to 1.06 g/kg BW/day (p < 0.001) in participants receiving protein supplementation or dietary advice. Higher protein intake was significantly associated with increased ALST (beta = 1.0, p = 0.047), but no associations were found with body weight, body fat, HGS, gait speed, CST, or SPPB.

Conclusions

Increasing protein intake during weight loss interventions may help preserve ALST in older adults with overweight or obesity, potentially reducing the risk of sarcopenia. These findings suggest that incorporating protein strategies in weight loss programs is beneficial for muscle health in older adults. Further high-quality studies are needed to determine the optimal protein intake for this population.

Trial registration

Trial registration in the Dutch Trial Register: MPS: NL2623, https://www.onderzoekmetmensen.nl/en/trial/24688, registration date 11022011 WelPrex NL4434 https://www.onderzoekmetmensen.nl/en/trial/29590, registration date 01052014 PROBE NL4357, https://www.onderzoekmetmensen.nl/en/trial/21917), registration date 08042014.

Introduction

Obesity prevalence is still increasing in older adults [1, 2]. The World Health Organization (WHO) estimates that 59% of the European citizens are overweight or obese, reporting a peak in middle age from 50 to around 65 years of age [2, 3]. Obesity is associated with multiple diseases such as metabolic syndrome, cardiovascular disease, or cancer and is related to disability, which represents a serious threat to healthy ageing [3].

Evidence-based guidelines on obesity management recommend employing lifestyle interventions, such as nutrition and exercise interventions, as effective strategies to promote a healthy lifestyle and address obesity [4–7]. Experts widely agree that following an energy-restricted diet is an effective way for individuals with overweight and obesity to reduce their weight and improve their overall health [7–11]. However, studies clearly document that approximately 25% of the body mass lost during weight loss interventions consists of fat-free mass [11, 12]. This presents a significant challenge, particularly for older adults who are already experiencing age-related declines in muscle mass, strength, and function [13]. Consequently, weight loss interventions may exacerbate sarcopenia and sarcopenic obesity [14], which are further associated with physical and cognitive impairment, mobility disorders, and a loss of independence [13, 15].

Evidence for the benefits of weight loss interventions in obese older adults is limited, leading to ongoing debates among experts regarding the advisability of intentional weight loss in older age [16, 17]. Nonetheless, maintaining muscle mass, strength, and function during weight loss is crucial for both older and middle-aged persons, as it influences the process of healthy ageing. Therefore, the use of energy-restricted diets must be carefully considered, individualised, and supervised by professionals such as dietitians.

Studies show that the most effective strategy for obese middle-aged to older persons to simultaneously lose weight and maintain muscle mass is to combine an energy-restricted diet with resistance training [8]. In addition to resistance training, increasing the protein intake – above the recommended daily allowance (RDA) of 0.8 g/kg/d – is discussed as an effective strategy to preserve skeletal muscle mass during weight loss interventions. Older individuals need higher amounts of proteins to maintain muscle mass due to their lower postprandial muscle protein synthesis rates [18, 19], and experts broadly recommend increasing protein intake in adults 65 years or older to at least 1.0 g/kg BW/day [17, 20, 21].

However, studies on the impact of protein intakes higher than the recommended RDA of 0.8 g/kg/d during weight loss interventions with energy restriction and resistance training are rare, especially in middle-aged to older adults [8]. Based on intervention data, Weijs and Wolfe suggested that an RDA of higher than 1.2 g/kg/d was advisable to maintain muscle mass when following an energy-restricted diet and performing resistance training [22, 23]. A systematic review summarised the results of 24 randomised controlled trials (RCTs) on dietary protein and their effects on changes in body composition in adults 50 years and older who followed energy-restricted diets but did not perform resistance training. The reviewers concluded that higher protein diets more effectively enable people to lose body fat mass and maintain lean body mass than standard protein diets [24]. A recently published network meta-analysis compared different weight loss interventions in persons around retirement age (55–70 years). No high-quality studies were identified reporting an additional effect of high protein intake during energy restriction and resistance training on fat mass or muscle mass, as compared to energy restriction with resistance training alone [8]. Thus, the question of whether higher protein intake during weight loss phases with energy restriction and physical training (resistance or mixed exercise) provides additional benefits in terms of weight loss, body fat, muscle mass, strength, or function remains equivocal.

Generally, individuals transitioning into a new life stage, such as retirement, need to restructure their daily routines. This period is often accompanied by a reassessment of personal habits and lifestyle choices, with a heightened focus directed towards health and well-being [25–27]. A common objective during this time is to improve the diet and/or to reduce weight, which is pursued to enhance overall health and maintain mobility in older age [25]. Therefore, it is crucial to identify effective weight loss interventions for this demographic, considering the increased risk of sarcopenia associated with aging. The optimal protein intake during weight loss interventions for this age group remains unresolved.

Hence, the aim of this study was to explore the association between protein intake and outcomes of body composition, strength, and physical performance during weight loss interventions, including an energy-restricted diet combined with resistance training, in persons with overweight or obesity aged 55 years or older. The findings contribute to the existing knowledge about the protein intake required to ensure the maintenance of muscle mass and function during weight loss interventions.

Method

Study design

This study is a secondary data analysis of data from three randomised controlled trials (RCTs). Reporting followed the CONSORT 2025 guidelines for randomized trials [28]. The Muscle Preservation Study (MPS) [29], the Protein and lifestyle intervention to preserve muscle mass in obese older type 2 diabetes patients (PROBE) study [30] and the Weight Loss with Protein and Exercise (WelPrex) study [31]. MPS and PROBE were two-arm RCTs and analysed the effect of the same high whey protein-, leucine-, and vitamin D-enriched supplement in addition to energy restriction and resistance training. The MPS and PROBE studies each had a study duration of 13 weeks. WelPrex was a four-arm RCT, had a study duration of 10 weeks, and tested the effect of a high protein diet (without supplements) in addition to energy restriction and resistance training in older adults with overweight or obesity. In the MPS and PROBE studies, participants in the protein groups received a daily supplement providing approximately 21 g of additional protein, aiming for a total protein intake of around 1.2 g/kg body weight/day. In the WelPrex study, participants in the dietary counselling group were advised to reach a protein intake of about 1.3 g/kg body weight/day through regular foods. Across all three studies, the energy-restricted diet targeted an intake approximately 600 kcal/day below estimated energy requirements to achieve gradual weight loss while preserving muscle mass [29–31].

All three studies were conducted at the Amsterdam University of Applied Sciences in the Amsterdam Nutritional Assessment Center, The Netherlands. A more detailed description of the included studies is available in the respective publications for the studies [29–31] as well as in the Dutch Trial Register https://www.onderzoekmetmensen.nl/en (Clinical trial number MPS: NL2623, registration date 11-02-2011, Clinical trial number Wel-Prex: NL4434, registration date 01-05-2014; Clinical trial number PROBE: NL4357, registration date 08-04-2014).

Sample and variables

The criteria for including data from participants of these three trials in the current secondary data analysis were an age of 55 or older, a BMI ≥ 25 kg/m², available data for protein intake at both baseline and follow-up, as well as participation in a weight loss intervention consisting of an energy-restricted diet combined with resistance training. Outcomes for body composition, physical performance, and strength were body weight kg, percentage body fat, appendicular lean soft tissue mass (ALST) in kg, handgrip strength (HGS) in kg, gait speed in m/s, chair stand test (CST) results, and short physical performance battery (SPPB) results.

Data collection

For this study, we included variables for general participant characteristics which were self-reported at baseline (sex, age, ethnicity, and smoking status). Body weight was measured on a calibrated scale as part of the BODPOD system (Cosmed). Body height was measured to the nearest 0.1 cm with a stadiometer that was mounted on a wall (Seca 222; Seca, Hamburg, Germany and De Grood DGI 250D; De Grood metaaltechniek). Waist circumference was measured with the participant in a standing position at a point halfway between the anterior superior iliac spine and the lower rib after a normal expiration (Seca 201; Seca). Appendicular lean soft tissue mass (ALST) was calculated using a whole-body scan with dual-energy X-ray absorptiometry (DXA, Discovery A, Hologic). Hand grip strength was measured using an isometric JAMAR hand grip dynamometer by the participant in a sitting position, who performed three repetitions with both hands. Results were documented to the nearest 0.1 kg and the mean for each hand was calculated. A 4-m gait speed test was conducted to evaluate the participants’ physical performance [32].

The baseline dietary intake was assessed using a three-day food record, including two weekdays and one weekend day. The records were structured by eating occasions: breakfast, between breakfast and lunch, lunch, between lunch and dinner, dinner, and evening. Participants reported their intake as accurately as possible using standard household measures or a kitchen scale. Food records were reviewed for completeness by trained students, supervised by the dietician, during the study visits with clarifications obtained as needed. Data were verified by following standard operating procedures, coded, and compared to the Dutch Food Composition Table to calculate the energy and macronutrient intake. Additional checks were conducted by a dietician or the investigator after the data were coded.

Statistical analysis

The data sets for the individual studies were harmonised and merged into one data file. Patient characteristics were reported as absolute and relative frequencies for categorical data and as medians and interquartile range (q1, q3) or mean and standard deviations for numerical data, as appropriate. Comparisons between groups were made using the one-way ANOVA, Kruskal-Wallis test, or Pearson’s chi-square tests as appropriate. In a first step, univariable analyses were performed. Furthermore, linear mixed models were applied to identify associations between total protein intake and changes in several outcome variables during a weight loss intervention, i.e. body weight, percentage fat mass, HGS, gait speed, ALST, and CST. A linear model without random effects was calculated to arrive at the outcome SPPB, because data on SPPB were only available from the PROBE study. The number of datapoints available for the participants varied according to the different outcomes. The respective study (MPS, PROBE, and WelPrex) was included as a random effect, and the models were adjusted using the baseline parameter of the corresponding outcome, age, sex, energy intake, and training adherence. Training adherence was defined as the percentage of training sessions attended. Beta coefficients are presented along with their 95% confidence interval (CI) and the p-value. A p-value of 0.05 or less was considered statistically significant. To avoid multicollinearity, the VIF (variance inflation factor) was calculated.

Sensitivity analyses were then performed, namely by repeating the analysis and employing alternative options to calculate the protein intake. Instead of reporting the total protein intake in g/kg BW/day, we also reported the total protein intake in g per day, the protein intake per day in g/kg ALST, and the energy% based on the protein/total energy intake. To perform certain analyses, we categorised protein intake into three groups: low protein intake (< 0.8 g/kg BW/day), moderate protein intake (0.8–1.2 g/kg BW/day), and high protein intake (> 1.2 g/kg BW/day). All statistical analyses were conducted using R version 4.4.1.

Ethical considerations

All studies providing data included in this secondary data analysis were approved by the responsible ethics committees and a written informed consent was obtained from all subjects. The PROBE study was approved by the Medical Ethics Committee (NL46790.056.14) Assen June 2014, The Netherlands. The MPS study was approved by the Medical Ethics Committee of the VU University Medical Center Amsterdam (2010/280). The Welprex study was approved by the Medical Ethics Committee Independent Review Board Nijmegen, Netherlands (NL43226.072.14).

The studies were conducted in line with the ethical principles described in the Declaration of Helsinki.

Results

General participant characteristics

In total, data from 191 adults aged 55 and older with overweight or obesity who were taking part in weight loss programmes and following an energy-restricted diet combined with resistance training were included in this secondary data analysis (see Fig. 1). Of these participants, 95 (49.7%) additionally received an intervention to increase their protein intake, consisting either of a high protein nutritional supplement (PROBE, MPS) or advice from a dietitian on how to increase dietary protein intake (WelPrex).

Fig. 1.

Flowchart illustrating the inclusion of data from participants in the data analysis. * Participants could be excluded for one or more than one reason

Less than half of the participants were female (45.0%), and the mean age of the participants was 65.1 (± 6.1) years. The mean BMI at baseline was 32.9 kg/m² (± 4.2). Table 1 displays the participants’ characteristics in detail.

Table 1.

Participants’ characteristics and protein/energy intake at baseline

	Total study population (n = 191)	MPS (n = 59)	PROBE (n = 97)	WelPrex (n = 35)	p-value*
Sex female, % (n)	45.0 (86)	55.9 (33)	36.1 (35)	51.4 (18)	0.038
Age in years, mean (SD)	65.1 (6.1)	63.2 (5.8)	66.0 (6.0)	63.8 (5.6)	< 0.001
Ethnicity (caucasian), % (n)	86.4 (165)	88.1 (52)	84.5 (82)	88.6 (31)	0.815
Current smoker yes, % (n)	7.4 (14)	5.2 (3)	8.3 (8)	8.6 (3)	0.756
BMI in kg/m2, mean (SD)	32.9 (4.2)	32.8 (3.8)	33.4 (4.6)	31.4 (3.4)	0.048
Waist circumference in cm, median (IQR)	112 (105–120)	109 (102–117)	115 (109–122)	108 (100–115)	< 0.001
Appendicular lean soft tissue mass (ALST) in kg, median (IQR)	25.3 (21–30)	22.3 (18–27)	26.6 (23–30)	NA	< 0.001
Handgrip strength in kg, median (IQR)	42 (29–53)	27 (22–38)	53 (41–64)	36 (27–48)	< 0.001
Gait speed in m/s, median (IQR)	1.2 (1.0–1.3)	1.1 (1.0-1.2)	1.1 (1.0–1.3)	1.3 (1.2–1.5)	< 0.001
Protein intake in g/kg/day, mean (SD)	0.88 (0.26)	0.86 (0.26)	0.86 (0.25)	0.98 (0.26)	< 0.038
Energy intake in kcal/day, mean (SD)	1876 (520)	1921 (594)	1808 (476)	1989 (492)	< 0.135

* p-value for differences between the three included studies. For nominal variables, the Pearson’s chi-square test was used; for continuous variables, the one-way ANOVA was used; for non-parametric variables, the Kruskal-Wallis rank sum test was used

Protein intake

Mean protein intake at baseline was 0.88 g/kg/BW (± 0.26). After categorising protein intake as low (< 0.8 g/kg BW/day), medium (0.8–1.2 g/kg BW/day), or high (> 1.2 g/kg BW/day), we observed that 40.3% of the participants had a low protein intake at the baseline in the intervention studies and 31.4% had a low protein intake after the respective intervention. A medium protein intake was identified in 49.7% before and 47.6% after the intervention, and a high protein intake was reported by 9.9% of the participants before and 20.9% after the intervention.

The mean total protein intake increased significantly from baseline up to the point final measurements were taken in participants who received either a high protein supplement or advice on how to increase their dietary protein intake (from a mean intake of 0.89 g/kg BW/day (± 0.24) to 1.12 (± 0.29) g/kg BW/day (p < 0.001)). In participants who did not receive a protein intervention in addition to following a energy-restricted diet and performing resistance training, we identified a trend of decreasing protein intake (i.e. from 0.88 g/kg BW/day (0.28) to 0.83 g/kg BW/day (± 0.28), p = 0.147). Both participants with and without a protein intervention decreased their mean total energy intake significantly from 1875 (± 520) kcal/day to 1730 (± 436) kcal over the study duration (see Fig. 2).

Fig. 2.

Protein and energy intake at baseline and after the intervention phase, in (a) all participants, (b) participants with protein intervention, (c) participants without protein intervention^* Protein intervention = either high protein supplements or advice on how to increase dietary protein intake

Linear mixed models analysis

The mean total protein intake within a weight loss intervention, which included an energy-restricted diet and resistance training, was significantly associated with increased ALST after the intervention (beta 1.03 (0.01, 2.1), p = 0.047) after adjusting for ALST baseline, age, sex, energy intake, and training adherence. This means that an increase in protein intake of 0.1 g/kg BW/day increased ALST during the intervention by 0.1 kg. The total protein intake was not associated with other outcomes related to obesity and sarcopenia, such as body weight, percentage body fat, HGS, gait speed, CST, or SPPB. The sensitivity analyses did not reveal results that differed from those of the main analysis. Detailed results for associations between protein intake and body weight, percentage body fat, ALST, HGS, and gait speed are shown in Fig. 3.

Fig. 3.

Linear mixed models for the association of protein intake (in g/kg BW/day), relating outcomes to obesity and sarcopenia. HR = Hazard Ratio, CI = Confidence Interval, HGS = hand grip strength, ALST = appendicular skeletal muscle mass, CST = chair stand test, SPPB = short physical performance battery; analyses were adjusted for age, sex, energy intake, training adherence, and the respective outcome variable at baseline; * SPBB was only measured in PROBE; therefore, there is no random effect for the study (linear model)

Discussion

The aim of this study was to evaluate associations between protein intake during weight loss interventions – combining a low-calorie diet and resistance training – and outcomes related to obesity and sarcopenia in older adults with overweight or obesity. The total protein intake in g/kg BW/day was significantly associated with increased appendicular lean soft tissue mass post-intervention. No added value of higher protein intake could be shown for other outcomes related to obesity and sarcopenia, namely body weight, percentage body fat, HGS, gait speed, CST, and SPPB.

Our study findings suggest that a higher protein intake plays a beneficial role in muscle mass maintenance during a weight loss intervention. These findings are supported by those of a systematic review in adults, which found that high-protein, low-calorie diets were more effective than standard-protein, low-calorie diets for preserving fat-free mass and resting energy expenditure. This study also noted that the former diets helped preventing weight regain after the diet ended [33]. Another systematic review reported a positive association between a higher retention of lean body mass and higher protein intake (≥ 1 g/kg/d) during an energy restriction intervention in middle-aged to older adults (≥ 50 years), whereby the participants also experienced greater weight loss when following the high-protein diet [24]. However, both of these reviews excluded studies incorporating structured training programmes; thus, whether increasing protein intake while following an energy-restricted diet and performing exercise training offers additional benefits remains unclear. This knowledge gap is highlighted by the results of a network meta-analysis that compared various nutrition and exercise interventions in individuals around retirement age. The results indicate that combining energy restriction and high protein content diet interventions seems to be slightly better than applying an energy restriction intervention alone regarding the retention of lean mass, but the study participants still tended to lose lean body mass if they did not take part in training programmes [8]. Since physical training not only improves obesity and sarcopenia parameters but also offers a variety of other physiologic benefits (e.g. improved cardiometabolic outcomes), we recommend its use in conjunction with weight reduction by following an energy-reduced and protein-adapted diet.

Our study found that higher protein intake was associated with higher ALST, but no association was found with improvements in other outcomes such as handgrip strength and gait speed. This could be explained by the fact that, while protein intake supports muscle protein synthesis and muscle preservation, it does not necessarily lead to significant changes in strength and function. Improvements in functional outcomes such as handgrip strength and gait speed may be most strongly affected by resistance training, which was part of the intervention for all of our study participants. Increasing protein may not provide additional benefits during the period of the intervention; however, this can only be assumed, as trials addressing this question are rare. An RCT included in this secondary analysis (WelPrex) demonstrated that overweight or obese older adults who following an energy-restricted diet, performed resistance training, and received dietary advice for increasing their protein intake showed greater improvements in their fat-free mass, handgrip strength, 4-m gait speed, and body fat reduction compared to those who following an energy-restricted diet and performed resistance training alone [31]. In the current mixed models analysis, however, we found no significant correlation between the level of protein intake and handgrip strength or gait speed.

In general, we noted that a considerable number of our study participants did not manage to consume the protein RDA of 0.8 g/kg/d. Similar results were demonstrated in the Health ABC study, where 43% of the participating older adults (70–79 years of age) reported protein intakes below the RDA of 0.8 g/kg/d [34]. However, the sample in the current study was considerably younger (i.e. median age of 65 years), indicating that many “younger” older persons have protein intakes below the RDA. In our sample, the BMI was much higher (mean of 32.9) than, for example, the BMI reported in the Health ABC study (mean of 26.8). The latter leads to high protein requirements and consequently a need to consume a high amount of protein-rich foods [23]. This finding also raises the question of whether there should be an upper limit for total protein intake, since some studies suggest that high protein intake can have adverse effects, e.g. increasing bone, kidney, or cancer risks [35]. To tackle this question and evaluate the optimal protein intake level in older adults with overweight and obesity during weight loss periods, an exploratory study using data from the MPS study identified a required protein intake of at least 1.2 g/kg/d as the optimal daily protein intake in people following energy-restricted diets and performing resistance training, based on their muscle mass accretion [36]. These results imply that a protein intake higher than 1.2 g/kg/d is not only important for older adults in general, as has been recognised by expert groups [21, 37], but is particularly important in older adults with overweight and obesity undergoing weight loss interventions to retain muscle mass. In addition to its potential beneficial effects on muscle mass, a high protein intake may also facilitate lower calorie consumption by promoting greater satiety. In persons older than 80 years or in vulnerable, frail individuals, weight loss interventions should not be conducted due to the risk of worsened health outcomes [16].

This study adds new insights, adding to our limited knowledge of protein intake in older persons who undergo weight loss interventions. Nevertheless, this study has some limitations. First, we had to deal with missing data. For example, when applying the linear mixed model with handgrip strength data, data were available from only 72 participants; these data gaps potentially bias inferences made from the results. However, linear mixed model estimates are valid with missing data on the observed outcome and the baseline variables in the analysis model. Second, nutritional intake in the three included RCTs was self-reported via dietary food records. We found a low protein intake, i.e. below the RDA, in more than 40% of the participants at baseline, which may be caused by the relatively low self-reported energy intake. Furthermore, participants may have underestimated their dietary intake, which occurs frequently in overweight persons [38]. Nevertheless, the protein intake based on the food records correlates significantly with the ratio of urinary urea to creatinine excretion over a period of 24 h, providing reliable data for evaluating the current dietary intake [39]. Third, we chose the included RCTs conveniently according to their availability. The PROBE, MPS, and WelPrex studies were all conducted in the Amsterdam Nutritional Assessment Center at the Amsterdam University of Applied Sciences, Amsterdam, The Netherlands, which was a partner in the SO-NUTS project.

Conclusion

Increasing the protein intake in older persons who undergo a weight loss intervention by following an energy-restricted diet and performing resistance training may help them to maintain muscle mass. Since the baseline mean protein intake level was already above the recommended RDA of 0.8 g/kg/d, protein requirements under these conditions are believed to be higher than this RDA [23]. To answer the question of exactly how much protein is needed by older persons with overweight and obesity who follow an energy-restricted diet and combine this with exercise, more high-quality studies are necessary.

Authors’ contributions

Doris Eglseer: Conceptualisation, Methodology, Validation, Writing – Original Draft, Writing – Review & Editing, Project Administration. Lea Reiter: Writing – Original Draft. Josje D Schoufour: Conceptualisation, Methodology, Validation, Writing – Original Draft. Tereza Vágnerová: Writing – Review & Editing. Robert G Memelink: Writing – Review & Editing, Investigation. Amely M Verreijen: Writing – Review & Editing, Investigation. Andrea Borenich: Methodology, Formal analysis. Silvia Bauer: Conceptualisation, Methodology, Writing – Review & Editing Peter JM Weijs: Conceptualisation, Methodology, Validation, Investigation, Writing – Review & Editing.

Funding

The SO-NUTS project is funded by JPI HDHL, and the funding agencies supporting this work are as follows: the Netherlands Organisation for Health Research and Development (ZonMw), French National Research Agency (ANR), Federal Ministry of Education, Science and Research represented by the Austrian Research Promotion Agency (BMBWF represented by FFG), Spanish State Research Agency (AEI: PCI2020-120683-2), and the Ministry of Education, Youth and Sports Department of Research and Development (MSMT). This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the ERA-NET Cofund action No. 727565.

Data availability

Upon request, summary data may be made available for reasonable academic or research purposes, subject to institutional review and approval. Researchers interested in accessing summary data or further details may contact the corresponding author.

Declarations

Competing interests

The authors declare no competing interests.