Effects of resistance training on body composition and functional capacity among sarcopenic obese residents in long-term care facilities: a preliminary study

A 12-week quasiexperimental study with intervention and comparison groups was conducted to determine the effects of chair resistance training on body composition and physical performance among SO residents of nursing homes. Considering the practical limitations, a nonrandomized study design was applied because we could not quantify the number of individuals in order to randomly assign facilities or clients to groups due to the weak epidemiological evidence for investigating elderly with SO in the nursing homes [5]. Therefore, to ensure high design quality, the statement of transparent reporting of evaluations with nonrandomized designs was used as a guide in this study to improve the reporting standards of nonrandomized evaluations of health interventions [22].

The eligibility criteria for inclusion were residents 60 years or older, those who had no limited exercise recommendation from a physician, and those living continuously in LTC facilities for the past 3 months. The respondents lived a sedentary lifestyle (defined as exercising less than 150 min in a week) and could sit independently for at least an hour on the edge of a bed as well as understand Chinese or Taiwanese and follow instructions. Exclusion criteria included fluctuating weight (> ± 3 kg) for the past 3 months, limbs completely unable to resist gravity, severe cognitive impairments (clinical dementia rating ≧ 3), and severe disease status such as severe cardiopulmonary liver and kidney diseases, severe malignancy, and undernutrition (Mini Nutritional Assessment scale of 0–7 points). The study received support from the senior administrator of the LTC facilities involved. During preparation, the senior health leader of the LTC facilities helped screen or introduce elderly residents for this study. All subjects also gave oral and written informed consent prior to inclusion in the study.

The settings included six LTC facilities in Taichung City, Taiwan, a city (2215 km²) with an elderly population of nearly 310,000. We selected three facilities in urban areas and three in suburban areas. Participants were recruited from three Ren-ai Senior Citizens’ Homes and three nursing homes. The resistance training was implemented between October 2015 and March 2016. Blinding of the two groups for participants was not established, but the subjects and professional health trainer had been blinded to either SO or non-SO residents. Before assessment, the individuals making the relevant measurements had been instructed with the study’s training guide, ensuring better interrater and intrarater reliability. Each training session had 5–15 elderly participants and was held in a lounge or an activity room depending on the setting situation of the institution.

The identification of SO was from a combination of sarcopenia and obesity. For sarcopenia, the method was based on the equation: skeletal muscle mass ÷ body mass × 100 [23]. For men and women, the cutoff values for sarcopenia (two standard deviations (SDs) below the mean for the young reference group) were 37.15 and 32.26%, respectively, as previously reported in a Shanghai population study [24]. For obesity, the body mass index corresponding to the 90th percentile of body fat was 25.4–26.1 kg/m²; the obesity cutoff points by percent of body fat in men and women were 29 and 40%, respectively, following a Hong Kong population study [25].

For the intervention group, the resistance training intervention program emphasized chair muscle strength training using 2–5 lbs. sandbags on wrist or ankle joints and a grip ball. The comparison group comprised participants who signed the consent letter for join the study but as unwilling to participate in resistance training. Design ideas were drawn from resistance exercises for physical fitness enhancement of sarcopenia and program reference supported from Chang [26]. To motivate the participants, a professional health trainer added nostalgic music at a tempo matched with resistance training and played games with them. The training program had consulted with a multidisciplinary healthcare team that included nurses, a physiotherapist, an occupational therapist, and sports experts. During the first day of training, the respondents did the exercises without load to adjust themselves to the movements and posture.

All intervention groups trained with two sessions each week for a total of 3 months. An interval of at least 48 h between exercise sessions was arranged. The entire, the approximately 60-min exercise consists of the following three parts: warm-up stage, muscle resistance training while sitting, and relaxation stage. The intervention program comprised both upper body and lower extremities training. Upper body exercises included training that targeted the biceps, deltoids, grip, and pinch. Lower extremities consisted of leg extension, leg flexion, calf raises, stepping forward and sideward, and others. All exercises were performed for three sets of 4–10 repetitions with about 30 s rest between each set (Table 1).

Before giving them weights, we assessed the residents’ condition through their vital signs, investigated their health history (such as skin wounds, disease, and any operations), and checked their grip strength. During the training process, we encouraged participants to provide feedback and other information. After the session, we recorded the load level of the rating scale of perceived exertion (RPE) and respected residents’ decisions regarding the suitability of each sandbag weight. RPE is an objective method for quantifying the intensity of resistance exercise. RPE is a psychophysical scaling with a score between 6 and 20 points, with 9 points standing for very light, 11 for fairly light, 13 for somewhat hard, 15 for hard, and 17 for very hard [27].

The Cumulative Illness Rating Scale (CIRS) is an objective and simple method of assessing physical impairment. This index measures chronic medical illness burden while taking into consideration the severity of chronic diseases. The scale format provides for 13 relatively independent areas grouped under body systems. Ratings are made on a 5-point degree of severity scale, ranging from “none” to “extremely severe” [28]. On this scale, 0 = no problem affecting that system, 2 = moderate disability or morbidity or requires first line therapy, 4 = extremely severe problem or immediate treatment required or organ failure or severe functional impairment [29].

As a portable, noninvasive method appropriate for persons with disabilities, bioimpedance analysis (BIA) estimates body composition of fat and lean tissue [30]. The measurements were made according to InBody’s instruction manual for BIA (InBody S10, Biospace, Seoul, Republic of Korea). BIA was performed with eight surface electrodes placed on a resident’s thumbs, middle fingers, and both sides of each ankle with six different frequencies of 1, 5, 50, 250, 500, and 1000 kHz [31]. The InBody uses segmental analysis can determine differences caused by gender, aging, disease, and ethnicity without any empirical estimation. Exclusion criteria were the contraindications of the device: presence of an electronic implant, such as a pacemaker, artificial heart/lung, or portable electrocardiograph [31]. One resident with a heart pacemaker was excluded from the study.

Isometric hand grip strength was measured using a Jamar Lafayette hydraulic hand dynamometer (J00105) with residents in a sitting position holding the dynamometer at approximately 90° flexion to their elbow. The hand dynamometer used is reliable and widely cited in the literature [32]. The finger pinch gauge can be used to measure pinch strength [33] by applying pinch force at the pinch groove while holding the pinch gauge between one’s thumb and finger(s) using a Lafayette Hydraulic pinch gauge (5030P1). For grip and pinch strength, maximal readings of three measurements from both the left and right hands were recorded [34, 35]. Total handgrip strength was summed up from readings of both hands. Grip and pinch strength were presented in kilograms of force units.

The FIM scale was designed to measure one’s ability to function with independence [36]. FIM included 18 items, covering six situations in the burden of personal care, specifically self-care, mobility, transfers, sphincter management, communication, and social cognition. A high score indicates better function. If the subjects’ functional status varies in different settings or at different times of the day, the lower rating was recorded in accordance to the principles outlined in the FIM system clinical guide [36]. A high test–retest reliability for residents repeating the FIM assessment for the motor intraclass correlation (ICC = 0.9) and cognitive subscales (ICC = 0.8) was demonstrated along with construct validity of FIM [37].

The sample sizes for the study was calculated based on the muscle strength variable using G-Power 3.1 software with 80% power and an alpha of 0.05. Effect size was found with the standardized mean difference of 0.68 (95% CI 0.52 to 0.84) from systematic literature [38]. Each group included 35 participants with a total of 70 subjects at baseline. After 3 months, 91.4% subjects (n = 64) had completed the study. Our research team recalculated the effect size to be 0.76 and a power of 0.85 based on the grip strength variable.

All the collected data were subjected to descriptive and frequency analyses, using the baseline of the two groups’ differences through chi-squared test and t-test. Additionally, generalized estimating equation (GEE) was used to analyze the repeated measurements of independent effectiveness on the independent variables [39]. All data were processed in SPSS version 17.0 (SPSS Inc., Chicago, IL, U.S.A.). A p < 0.05 was considered statistically significant. For missing data, we used simple imputation to deal with this issue, including last or baseline observation carried forward as close to real-time as possible during the study [40].