Anthropometric and Body Composition Measurements Related to Osteoporosis in Geriatric Population.

Objective: This study aimed to evaluate different anthropometric and body composition measurements, including weight, body mass index (BMI), body fat percentage (fat%), skeletal muscle index (SMI), a body shape index (ABSI), waist circumference (WC), and hip circumference (HC), in relation to bone mineral density.
Methods: This is a cross-sectional study of a total of 482 patients who consulted the geriatric outpatient clinic between 2018 and 2019. Patients were evaluated with dual-energy X-ray absorptiometry. Anthropometric measurements (HC, WC, weight, BMI, and ABSI), as well as body composition analysis (fat% and SMI) by bioimpedance analysis system, were performed. The patients were evaluated with the rapid Assessment of Physical Activity Index in terms of aerobic exercise habits.
Results: There was a significant correlation between lumbar spine (LS) T-score and SMI (r=0.36, p=0.000) and between LS T-score and weight (r=0.21, p=0.000), BMI (r=0.10, p=0.045) and WC (r=0.15, p=0.001). There was a statistically significant correlation between femur neck (FN) T-score and fat% (r=0.15, p=0.001), SMI (r=0.15, p=0.010), weight (r=0.22, p=0.000), BMI (r=0.20, p=0.000), WC (r=0.14, p=0.003), and HC (r=0.17, p=0.001). There was no statistically significant correlation between physical activity and LS T-score (n=353, r=0.08, p=0.16) and FN T-score (n=360, r=0.03, p=0.53). In multiple regression analysis, SMI contributes most in predicting FN and LS T-scores. Conclusions: Anthropometric measurements should be carefully selected in the geriatric population. Among the measurements, the strongest relationship was found between LS T-score and SMI. © Copyright Istanbul Medeniyet University Faculty of Medicine.

INTRODUCTION

Osteoporosis is a systemic skeletal disease characterized by low bone mineral density (BMD) and impaired bone microarchitecture that can lead to increased fracture risk[1]. In contrast with the general belief that obesity reduces the risk of fractures, recent studies have shown that a significant proportion of postmenopausal women with low-energy fractures are obese[2].

According to the results of Turkey Diabetes, Obesity and Hypertension Epidemiology Research I and II, which was obtained with intervals of twelve years, the prevalence of obesity increased by 40%. The rate of obesity is reported to be 44% in women and 27% in men[3]. Considering obesity as a protective factor in osteoporosis patients may delay the fight against an existing health problem.

The relationship between metabolic syndrome and osteoporosis has been the focus of many studies in recent years. Von Muhlen et al.[4] reported that metabolic syndrome is associated with low BMD values and an increased frequency of non-vertebral fractures. Also, it has been reported that high abdominal fat mass is associated with a decreased risk of fracture in elderly women[5].

Anthropometric measures are simple, inexpensive, and practical tools for population screening and early detection of several diseases[6]. Body mass index (BMI) is the most commonly used anthropometric measure that has been shown to be associated with several diseases, as well as mortality[7,8]. Recently, a new anthropometric measurement, known as a body shape index (ABSI), was developed to estimate the risks of diseases that cannot be predicted by BMI[9].

Many anthropometric measurements that are associated with osteoporosis have been proposed. There is no definite opinion on which anthropometric measurement has a higher association with osteoporosis[10,11,12].

In literature, both positive[13,14] and negative[15,16] correlations between anthropometric indices and BMD have been reported. Among these anthropometric indices, the waist circumference (WC) or waist-to-hip ratio has been reported to have a stronger association with BMD compared to BMI[16,17]. In addition, some studies have reported that fat mass can adversely affect skeletal health compared to lean mass[18,19]. Given the conflicting results in literature, there is a need to thoroughly investigate the association between anthropometric measurements and osteoporosis.

In this regard, we took different anthropometric and body composition measurements, such as body weight, BMI, body fat percentage (fat%), skeletal muscle index (SMI), ABSI, WC, and hip circumference (HC), in order to investigate the relationship with to lumbar spine (LS) and femur neck (FN) T-scores.

MATERIALS and METHODS

Patients who consulted to the geriatric outpatient clinic between 2018 and 2019 were included in the study. The inclusion criteria were: patients aged ≥65 years and diagnosed with osteoporosis according to the definition of World Health Organization[20]. The exclusion criteria were: vertebral fracture due to known accidental traumas, history of drug therapy in the past year (bisphosphonate, estrogen replacement therapy, and glucocorticoids), history of comorbidities, such as malignancy, radiotherapy or chemotherapy, renal failure, hyperthyroidism, primer hyperparathyroidism, and rheumatic disease or adrenal diseases.

Initially, a demographic form, including age, sex, and comorbidities, was filled by the participants. The BMI, which is a ratio of weight and height squared (kg/m2), was calculated. The patient’s WC and HC were also measured in centimeter and their fat% and skeletal muscle mass (SMM) were calculated by bioimpedance analysis (BIA) (Tanita TBF 300; Tanita Corp., Tokyo, Japan). SMI was calculated from the BIA-based SMM with the formula SMM/height (m2)[21]. ABSI was measured with the formula WC (m)/[Weight (kg)2/3×Height (m)5/6][9].

Osteoporosis was evaluated by dual-energy X-ray absorptiometry (DXA). According to the T-scores calculated from FN and LS, the participants were divided into three groups according to FN score as follows: normal: T-score >-1, osteopenia: T-score is between -1 and -2.5, and osteoporosis: T-score <-2.5[20]. Also, vitamin D, calcium, phosphorus, parathyroid hormone (PTH), and alkaline phosphatase (ALP) levels were recorded.

Physical activity levels of the patients were evaluated with the rapid assessment of physical activity (RAPA) aerobic assessment, which is a valid and reliable tool used in the elderly[22]. Accordingly, patients were divided into five groups as follows: 1= sedentary, 2= underactive, 3= regular underactive (light activities), and 4= regular underactive, and 5= regular active).

The study protocol was approved by the Clinical Research Ethics Committee of the Istanbul Medeniyet University Goztepe Training and Research Hospital (decision no: 2018/0478, date: 12.12.2018) and all participants gave written informed consent. This study is registered to clinicaltrials.gov p(NCT04255173).

Statistical Analysis

All statistical analyses were performed with SPSS Statistics for Windows, version 25.0 (IBM corps., Chicago, IL., USA). Pearson correlation (r) and Spearman correlation (rho) were used to determine the relationship between variables. Multiple regression was used to determine the best linear combination of variables. ANOVA was used to examine the differences in variables between more than two groups. P values less than 0.05 were considered statistically significant.

A post hoc power analysis was done using G*Power 3 (Faul et al.[1]) to 1) determine the relationship between LS T-score and FN T-score with other variables (fat%, SMI, BMI, WC, HC, and weight) using a one-tailed test, 2) conduct a multiple regression analysis of six predictor variables, 3) determine the relationship between RAPA aerobic physical activity and four variables (fat%, SMI, BMI, and HC) using a one-tailed test, and 4) determine the relationship between RAPA aerobic physical activity and a laboratory test P using a one-tailed test.

RESULTS

During data collection, a total of 571 participants were included in the study. Individuals who did not meet the inclusion criteria and had missing DXA values were excluded. After the exclusion of non-eligible individuals, a total of 482 patients (mean age =74.27±7.40 years), which comprise 331 females (68.7%) and 151 males (31.3%), participated in the study. Among, 482 patients, 134 patients were diagnosed of osteoporosis (30.39%) and 254 patients were diagnosed of osteopenia (57.60%). According to RAPA aerobic index, the 482 patients were classified as follows: sedentary patients: 16 (3.3%), underactive patients: 12 (2.5%), regular underactive (light activities): 136 (28.2%), regular underactive: 187 (38.8%), and regular active: 131 (27.2%). The other characteristics of the patients are presented in Table 1.

Table 1

Descriptive statistics of patients’ characteristics.

The relationship between of LS and FN T-scores with the anthropometric measurements was investigated using Pearson correlation coefficient (r). Table 2 shows the results of Pearson correlation (r) analysis. There was a weak positive relationship between LS T-score and SMI (r=0.36, p=0.000) and between LS T-score and weight (r=0.21, p=0.000), BMI (r=0.10, p=0.045), and WC (r=0.15, p=0.001). There was a weak relationship between FN T-score and fat% (r=0.15, p=0.001), SMI (r=0.15, p=0.010), weight (r=0.22, p=0.000), BMI (r=0.20, p=0.000), WC (r=0.14, p=0.003), and HC (r=0.17, p=0.001).

Table 2

Pearson correlation (r) analysis of variables with LS and FN T-scores.

The above correlation results indicate that the variables fat%, SMI, BMI, WC, HC, and weight are associated with LS and FN T-score. To ascertain the highest possible multiple correlations of these variables with LS and FN T-score and to elaborate these relationships, multiple regression analysis was performed.

A post hoc power analysis was conducted using G*Power 3[1] to determine the relationship between LS T-score and FN T-score with other variables (fat%, SMI, BMI, WC, HC, and weight) using a one-tailed test, with a total sample of 482 participants and an alpha of 0.05. Results show that the achieved power was more than 0.99 for the LS T-score and SMI and that it was 0.70 for the other significant relationships.

Multiple regression was conducted to determine the best linear combination of fat%, BMI, WC, HC, weight, and SMI for predicting LS T-scores. This combination of variables significantly predicted LS T-score (F [6,352]=12.20, p<0.001), with four variables (fat%, WC, weight, and SMI) significantly contributing to the prediction. The adjusted R squared value was 0.16, which indicates that 16% of the variance in the LS T-score was explained by the model. According to Cohen[2] this is a small effect. The beta weights, presented in Table 3, suggest that SMI contributes most to the prediction of LS T-score and that WC, weight, and fat% also contribute to the prediction.

Table 3

Multiple regression analysis for LS T-score.

A post hoc power analysis was conducted using G*Power 3[1] to determine the multiple regression analysis of six predictor variables with 0.16 adjusted R square, a total sample size of 482, and an alpha of 0.05. The results show that the achieved power was 0.99.

Multiple regression was conducted to determine the best linear combination of fat%, BMI, WC, and HC, weight, and SMI for predicting FN T-scores (Table 3). This combination of variables significantly predicted FN T-score (F [6,357]=5.46, p<0.001), with two variables (fat% and SMI) significantly contributing to the prediction. The adjusted R squared value was 0.07, which indicates that 7% of the variance in FN T-score was explained by the model. According to Cohen[2] this is also a small effect. The beta weights, presented in Table 3, suggest that SMI contributes most to the prediction of FN T-score and that fat% also contributes to this prediction. A post hoc power analysis was conducted using G*Power 3 (1) to determine the multiple regression analysis of six predictor variables with 0.07 adjusted R square, a total sample size of 482, and an alpha of 0.05. The results show that the achieved power was 0.99.

The relationship between aerobic physical activity (as measured by RAPA index) and the variables (fat%, SMI, ABSI, weight, BMI, WC, and HC) was investigated using Spearman’s correlation coefficient (rho). Table 4 shows the results of Spearman’s correlation (rho) analysis. There were weak negative correlations between physical activity and fat% (r=-0.23, p=0.000), BMI (r=-0.12, p=0.031), and HC (r=-0.10, p=0.050) and a weak positive relationship between physical activity and appendicular muscle mass (r=0.11, p=0.030).

Table 4

Multiple regression analysis for FN T-score.

A post hoc power analysis was conducted using G*Power 3[1] to determine the relationship between RAPA aerobic physical activity with four variables (fat%, SMI, BMI, and HC) using a one-tailed test, with a total sample of 482 participants and an alpha of 0.05. Results show that the achieved power was more than 0.99 for fat%, 0.78 for SMI, 0.84 for BMI, and 0.71 for HC.

The relationship between aerobic physical activity (as measured by RAPA index) and laboratory assays (vitamin D, PTH, calcium, phosphorus, and ALP) was investigated by Spearman’s correlation analysis. There was a weak negative relationship between physical activity and P (r=-0.12). A post hoc power analysis was conducted using G*Power 3 (Faul et al.[1]) to determine the relationship between RAPA aerobic physical activity with phosphorus using a one-tailed test, with a total sample of 482 participants and an alpha of 0.05. The achieved power was 0.84.

According to Spearman’s correlation analysis results, there was no relationship between physical activity and LS T-score (n=353, r=0.08, p=0.16) and FN T-score (n=360, r=0.03, p=0.53) (Table 5). In order to investigate the impact of physical activity on LS and FN T-score variables, ANOVA was performed conducted. The results show that there was no significant difference in LS T-score (F [2, 350]=0.77, p=0.47) and FN T-score (F [2, 357]=0.52, p=0.59) between the physical activity groups.

Table 5

Spearman’s correlation (rho) analysis of variables with RAPA aerobic physical activity.

DISCUSSION

Anthropometric measurements are often preferred because they are cost-effective and can be applied quickly and easily in clinics. Herein, we investigated the relationship between osteoporosis and various anthropometric and body composition measurements (BMI, ABSI, fat%, SMI, WC, and HC). Among these variables, all measurements, except ABSI, were correlated with LS and FN T-scores.

Previously, obesity was thought to protect against osteoporosis[23]. However, it was recently shown that high fat% adversely affects the bone health[24]. Although it was not statistically significant, the increase in fat% was found to be correlated with a decrease in LS T-score and a high fat% significantly contributes to the DXA scores in the regression analysis. While the cortical bone mainly assumes a mechanical and protective function, the trabecular bone is responsible for metabolic function. Osteoporosis is more common in areas rich in trabecular bone, such as vertebrae[25]. Therefore, changes in body composition may mostly affect the vertebrae. Although the risk of vertebral osteoporosis increases as the fat% increases, degenerative changes may also increase with age, thereby making accurate assessment difficult.

Although the index most commonly used in obesity screening is BMI, a new obesity index, ABSI, has been developed, considering that BMI does not distinguish between muscle and fat mass. ABSI has been shown by several studies to be a predictor of several diseases and mortality[9,26,27]. To the best of our knowledge, there is no study investigating the relationship between ABSI and osteoporosis in literature. The ABSI score is an obesity scale based on WC. However, in our study, although BMI and WC alone were correlated with BMD, ABSI was not correlated BMD. Also, in a recently published review, ABSI was shown to be outperformed by BMI and WC in predicting all-cause mortality but underperformed in predicting chronic diseases, such as osteoporosis[28].

Fat% and SMI were calculated by BIA. The calculation based on BIA has been shown to correlate with the calculation with DXA[29]. In this study, SMI has the strongest relationship with LS and FN T-scores. Moreover, SMI and osteoporosis have been associated in a recent study[30]. SMI is also used in the diagnosis of sarcopenia. BIA measurements in the geriatric population can be an easy and practical way of screening patients at risk of osteoporosis and sarcopenia.

The positive effects of aerobic physical activity on bone health have been shown in many studies[31,32,33]. In our study, aerobic physical activity was not directly related to the T-scores, which may be due to the low number of sedentary and underactive patients. However, aerobic physical activity was associated with fat%, SMI, BMI, and HC. In addition, fat%, SMI, BMI, HC, and WC were associated with both LS and FN T-scores. Therefore, we can conclude that aerobic physical activity is indirectly related to the severity of osteoporosis.

The limitations of the study were the determination of the severity of osteoporosis only according to the T-score, absence of a radiographic evaluation, and possible wrong lumbar DXA results due to degenerative changes.

CONCLUSIONS

The relationship between ABSI and osteoporosis severity has not been demonstrated. Among the anthropometric and body composition measurements, BMI, WC, HC, fat%, and SMI were found to be associated with BMD, with the strongest relationship found between SMI and BMD.