Relationship between body type (somatotype) and bone density in perimenopause women

Along with increasing age in women, there will be a decrease of estrogen levels and changes in the size, structure, and composition of the body, so those will affect bone density. This study aimed to analyze the relationship between body type (somatotype) and bone density in perimenopause women. This study was a cross-sectional study held in Ngemplak, Simongan, West Semarang with forty-two women subjects aged 31-50 years. Their bone density was estimated using QUS on calcaneal. Somatotype was measured using skinfold caliper, medline and sliding caliper. Bivariate analysis used Pearson Correlation and Rank-Spearman correlation test. Multivariate analysis used multiple regression test. All body types were grouped as endomorph, mesomorph, and ectomorph. Most subjects have mesomorph type were 71.4% (n = 30). Subjects with osteopenia were 4.8% (n = 2). There was a positive correlation between endomorph values, mesomorph values and body fat percentage with bone density scores in bivariate analysis. There was correlation between mesomorph values, body fat percentage and body mass index with bone density scores in subjects (SC = 0.533; p = 0.037, SC = 0.590; p = 0.011, SC = 0.643; p = 0.037) in multivariate analysis. In conclusion, there were significant correlations between mesomorph values with bone density scores.


Introduction
The adult period is the period when a person reaches physical maturity, usually occurring from the age of 20 years. Women age 30 years and over, physiologically will experience a decrease in estrogen levels called perimenopause. Decreasing estrogen levels during perimenopause can increase the risk of bone fragility, resulting in a high potential for fractures (McGuire and Beerman, 2013). Osteoporosis is a disease characterized by a decrease in bone mass due to the inability of the body to regulate mineral content in the bone, resulting in decreased bone strength. Fractures caused by osteoporosis can lead to death (Dieny and Fitranti, 2017). The prevalence of osteoporosis and fractures caused by osteoporosis keep increasing from years to years and become one of the main health problems in Asian countries (Mithal and Kaur, 2012;Kementrian Kesehatan RI, 2015). Based on a study by Tirtarahardja et al. (2006), the prevalence of osteoporosis in Indonesian women (aged 50-80 years) was 22.5% and 32% on aged 60-80 years. In 2007, the Osteoporosis Association in Indonesia showed that there was an increase in the prevalence of osteoporosis, especially in women over 50 years by 32% (Kementrian Kesehatan RI, 2015). In addition, data from the Health Profile in Semarang showed that prevalence of osteoporosis increased from 2016 to 2017 was 1.03% (Badan Pusat Statistik Kota Semarang, 2018).
Risk factors that can cause osteoporosis are divided into two namely changeable and unchangeable risk factors. Risk factors that can be changed are nutrient intake, especially vitamin D and calcium intake, alcohol consumption, smoking, use of glucocorticoid drugs and physical activity. Meanwhile, unchangeable risk factors are age, sex and genetic factor related to bone density (Vijayakumar and Busselberg, 2016). Low body mass index (BMI) is often found in studies as a risk factor of osteoporosis (Sözen et al., 2017). However, high body mass index (BMI) in some studies is positively correlated with bone density. Obese women based on BMI are found to have protective factors against bone density. However, obesity as a protective factor for bone mass is still debated. Changes in the body both in terms of size, structure, proportion and body composition can affect body type (Mazocco and Chagas, 2016). Human body types have been described as endomorph, mesomorph, and ectomorph commonly called somatotype (Carter, 2002).
Somatotype is a quantitative description of a person's physique type introduced by Heath and Carter which consisted of three components: endomorph, mesomorph, and ectomorph (Galic et al., 2016). Endomorph has a fat body build; mesomorph has strong bones and ectomorph has a very thin build. Endomorph is characterized by a round body, big and round head, short bones, short neck, narrow shoulders, wide legs, and waist. Mesomorph type is characterized by a solid body, hard, muscular, large bones, relatively have a small waist, and broad shoulders, usually this type is found in athletes. The ectomorph type is characterized by a thin or slender body, small bones, uniform lumbar abdomen and arches, narrow shoulders and invisible muscles. In this type of muscle mass and fat is lower than the other two types (Saitoglu et al., 2007;Galić et al., 2016).
Research related to the relationship between somatotype and bone density is still limited. Previous studies have shown a correlation between somatotype and bone density in mid-adult male subjects that the endomorph and mesomorph types are positively and significantly correlated with high levels of bone density, from measurements of the proximal hip bone and femur density using DXA. Subjects with endomorph and mesomorph type are related to bone density (Saitoglu et al., 2007). However, it is not explained how somatotype relates directly to bone density. However, it is explained indirectly through the existence of a significant relationship between body mass index with the somatotype component. Body mass index with normal, overweight, and obese categories was found to be more dominant than the underweight category. Therefore, the dominant types of somatotype found in this study are mesomorph and endomorph. The results of this study stated that the higher BMI, the better bone density. However, the weakness of BMI cannot describe body composition (fat mass and muscle mass) (Galić et al., 2016). Research about association between somatotype and bone density has never been done in Indonesia. Based on this background, the researchers wanted to conduct research on the relationship between body shape (somatotype) and bone density in adult women aged 31-50 years.

Materials and methods
This study was an observational study with a crosssectional study design in which each variable was measured once at a time. This study was conducted in Ngemplak Region, Simongan, West Semarang. The variables were components of body type (endomorph, mesomorph, and ectomorph), bone density, age, physical activity, nutrient intake (carbohydrate, fat, protein, caffeine, vitamin D, calcium intake), BMI and body fat percentage. Criteria for inclusion of the sample included women aged 31-50 years, not menopause yet, not pregnant or breastfeeding, not suffering from chronic diseases, not taking drugs such as glucocorticoids, steroids, and hormone drugs, not smoking, not consuming alcohol and willing to be the subject of research by filling out the Informed Consent. Exclusion criteria in this study included not being present during data collection, resigning before the study was completed, and passing away. The selection of research subjects was done by purposive sampling method.
Screening was done through the interview to the subjects, adjusting the inclusion criteria. Ninety-three subjects from screening process, obtained 50 person who met the inclusion criteria. Fifty subjects become 42 people due there were subjects who were absent during data collection and were excluded during data processing. The significance of the study (Zα) was 95%, while the research power (Zβ) was 80%, and the correlation coefficient (r) was 0.48 (Saitoglu et al., 2007). The result of calculating the sample size by considering the drop out was 37 subjects. Data of anthropometric measurements were body weight, height, skinfolds (triceps, calf, subscapular and supraspinal), bone width measurement (humerus and femur), girth (bicep girth and calf girth). In addition, bone density data, food intake and physical activity. This study was approved by The Ethical Committee of Medical Research of Faculty of Medicine, Universitas Diponegoro (No.143/EC/KEPK/FK/UNDIP/V/2019).
Anthropometry data of body type was done by direct measurements, such as body weight, height, skinfolds (triceps, calf, subscapular and supraspinal), width (humerus and femur), and girth (bicep girth and calf girth). Body weight was measured using digital weight scale (OMRON) with an accuracy of 0.1 kg. Height was measured using microtoise with an accuracy of 0.1 cm. Jamar type skinfold caliper, to measure the body skinfolds with an accuracy of 1 mm. Girth measurement using medline and body width using sliding caliper with an accuracy of 0.02 mm. Measurement of body fat percentage using the Bioimpedance Analysis (BIA) (OMRON). Anthropometry measurements were done twice for each variable.
Measurement of bone density using bone densitometry OsteoSys SONOST 2000 and categorized by T-score. Bone density measurements were carried out FULL PAPER eISSN: 2550-2166 © 2020 The Authors. Published by Rynnye Lyan Resources by experts from the KALBE Company. We used QUS because it was inexpensive, easy to carry, easily available, and easily accessible as an alternative to measuring Dual-X absorptiometry energy (DXA) measurements of bone mineral density (BMD) to assess fracture risk in many studies (Hans and Baim, 2017). Among the quantitative ultrasound measurements of the heel is considered one of the best alternative methods at present because it is available for assessment of fracture risk. The advantage of this technique is the possibility of detecting structural bone characteristics (elastic properties, trabecular orientation) thus enabling better quantification of bone quality (Quiros-Roldan et al., 2017). Normal bone density if it has T-score ≥-1.0 SD and has low bone density if T-score <-1.0 SD. Food intake data was done by interview using the Semi-Quantitative FFQ method and food photographic instruments. Food intake data were analyzed using the Nutrisurvey 2004 program. Physical activity data was measured using the short-form International Physical Activity Query (IPAQ) form.
Univariate analysis was performed to describe the characteristics of the subject, the variables studied included independent (somatotype measurement), dependent variable (measurement of bone density) and confounding variables (age, carbohydrate intake, fat intake, caffeine intake, calcium intake, vitamin D intake, BMI, percentage body fat, and physical activity) by entering data in the frequency distribution table. Normality test used Saphiro-Wilk because the number of samples is less than 50. Bivariate analysis was done to analyze the relationship between somatotype and bone density in adult women aged 31-50 years. Moreover, bivariate analysis was also conducted to determine the relationship of each confounding variable with bone density scores. Bivariate analysis was performed using the Spearman Rank test on which data were abnormally distributed and the Pearson Correlation test on normally distributed data. Multivariate analysis using the Multiple Linear Regression test.

Subject characteristics
The means age of the study subjects was 40 years old as seen in Table 1. The minimum value of the subject's bone density score was -1.3, which indicated that there were subjects who had osteopenia. Body type categories can be divided into two, namely based on 13 categories (category 1) and 4 categories (category 2).  Table 2 shows that the adequate levels of carbohydrate, protein, fat, calcium, vitamin D and caffeine intake. Most of the subjects were categorized as low intake. Endomorph body type values (r = 0.38; p = 0.013) were related to bone density as in Table 3. The higher the endomorph value, the greater the bone density score. The mesomorph type variable (r = 0.31; p = 0.045) was positively related to bone density. The variable body fat percent (r = 0.26; p = 0.004) was positively related to bone density. The higher the percentage of body fat, the greater the bone density score.
The body mass index, percent body fat, and mesomorph values had a relationship with bone density after being analyzed using linear regression tests as shown in Table 4. The adjusted R square value was 0.241 = 24.1%. The result showed that bone density was simultaneously influenced by BMI, value mesomorph and body fat percent were 24.1% while the rest were influenced by variables not examined.

Discussion
Based on the subject characteristics, there were two subjects with osteopenia, while most subjects had normal bone density. The mean age of the subjects was 40 years. This is consistent with the previous study by Dieny and Fitranti (2017) (Berger et al., 2008). This study showed that most subjects had mesomorph body types. This is in accordance with a study by Kalichman and Kobyliansky (2006) in which mesomorph body types were more common in subjects with the age group 31-40 years and 41-50 years compared with the age group 18-30 years.18 Based on the results of bivariate analysis test found a correlation between endomorph values, mesomorph and body fat percent with bone density. Variables that have a p value <0.25 followed by the analysis of the relationship test using multiple linear regression test. The results of multiple linear regression tests showed that the variables that correlated with bone density were mesomorph values, body fat percent, and Body Mass Index.

Relationship between body type (somatotype) with bone density
This study showed a correlation between mesomorph body type score and bone density. The relationship between the two had a positive correlation, indicating that the greater the value of the mesomorph body type components, the better bone density. Mesomorph type is characterized by a solid body, hard, muscular, large bones, relatively have a small waist, and broad shoulders, usually this type is found in athletes. Mesomorph type is usually dominated by muscle mass. The positive correlation between mesomorph and bone density in this study, in accordance by Patel et al. (2018) that lean mass can improve bone size and strength.

FULL PAPER
Muscles and bone are related to each other from the beginning of embryonic development to adulthood, thus affecting bone function and structure. Muscles and bones will get bigger and stronger due to the growth process, physical activity and adequate intake of nutrient that support bone health, especially vitamin D and calcium (Patel et al., 2018). Lean mass can improve bone density through the body's mechanical load. Bone strength can be affected by mechanical muscle movements and hormonal factors. Bone strength is affected by the amount of lean mass in body composition. Lean mass in human body can be obtained through physical activity. Most of the subjects had moderate physical activity, thus it is possible to have a large mechanical burden as well. This has resulted in the bone density score in most subjects belonging to the normal category. Muscle strength and physical activity have a more significant impact on bone health compared to hormonal factors, especially in premenopausal women (Ho-Pham et al., 2014). Moreover, lean mass has more significant effect on increasing bone density than body fat mass.

Relationship between body fat percentage with bone density
The result of this study was similar with study by Moayyeri et al. (2012) that there is correlation between body fat percentage and bone density in subjects (bone density was measured by Quantitative Ultrasound on calcaneal). Women with body fat percentage <20%, have 3.5 times the risk of bone fracture on the waist and 2.2 times increase bone fracture risk on the other areas compared to women with body fat percentage of about 40%.
Fat mass and lean mass in the body have a role to maintain bone density through the mechanism of mechanical load on the bone. Abdominal fat consists of subcutaneous fat and intraabdominal fat. Intraabdominal fat consists of visceral fat which composes fat mass in small intestine and kidney (Wajchenberg, 2000). There are two types of fat associated with bone, namely visceral fat and subcutaneous fat. Both have different effects on bones. Some studies explained that visceral fat has negative impact on bones through increased proinflammatory cytokines, such as TNF-α and IL-6. Increasing the amount of visceral fat in the body can cause the risk of arteriosclerosis, insulin resistance, and type 2 diabetes mellitus so that it can reduce bone density (Gilsanz et al., 2009;Campos et al., 2012). Subcutaneous fat has protective effect on bones.
However, the mechanism of the protective effect of subcutaneous fat on bone still remain unclear.
Anthropometry measurement that can describe visceral fat is waist-hip ratio, whereas to describe subcutaneous fat can use somatotype. Initially, fat is stored in subcutaneous fat tissue, but if the capacity of subcutaneous fat tissue has been obtained, it will be stored in visceral fat tissue and non-fat tissue such as skeleton muscle. This study showed that mesomorph was the dominant body type. This is consistent with the study by Galić et al. (2016) that the mesomorph value is higher than the endomorph and ectomorph values caused by fat deposits in skeletal muscle so that the width of the femur and humerus is greater. In addition to the mesomoporh value reflecting muscle mass, the mesomorph value also consists of several measurements of peripheral or subcutaneous fat stored under the skin. Women tend to have total fat deposits in the gluteal or femoral area (thigh area) and subcutaneous area of the skin. Body weight and abdominal fat distribution differ in women of childbearing age and women who have gone through menopause. Decreased estrogen levels in menopausal women are followed by a decrease in subcutaneous fat and an increase in abdominal or visceral fat (Lizcano and Guzmán, 2014). Women that included in this research were a perimenopause woman, it was assumed that subcutaneous fat in the body was still in large amounts and had a positive impact on bone density.

Relationship between BMI, physical activity, age with bone density
In this study, BMI was negatively correlated with bone density. This means that the greater BMI will reduce the bone density. In addition, this is consistent with research according to Silva et al. (2015) that overweight and obese women tend to have low bone density (osteopenia and osteoporosis). The majority of research states that a greater body mass index will increase bone density through the body's mechanical load mechanism. However, Compston et al. (2011) said that obesity does not have a protective factor in postmenopausal women who experience fractures, although they have high BMI value. Some studies said that obesity can increase bone density through hormonal mechanisms and mechanical loads. However, abdominal obesity can lead to some diseases such as insulin resistance, diabetes mellitus, metabolic syndrome so that it can disrupt bone metabolism through inflammation (Gürlek et al., 2018). Our research showed that two subjects who had overweight BMI were included in osteopenia category. This means that BMI is negatively correlated with bone density.
Physical activity has a role as a mechanical load to FULL PAPER eISSN: 2550-2166 © 2020 The Authors. Published by Rynnye Lyan Resources stimulate bone formation so that it affects bone size, shape and strength. The majority of research shows that physical activity is positively correlated with bone density (Alghadir et al., 2015). However, there is no relationship between physical activity and bone density in this study. Some studies state that physical activity and bone density were not related because of the intensity and frequency of low physical activity (Tan et al., 2014). Based on interviews on subjects regarding their physical activity, the results obtained that most subjects have a moderate physical activity. However, they did not do a specific physical activity that can improve bone density significantly such as washing, sweeping, drying and so forth. Most of the subjects are textile factory workers that most of the time they work is sitting. The type of physical activity mentioned in several studies that has a significant impact on bone density is regular exercise. The American College of Sports Medicine (ACSM) states that to improve bone density, a person must have an active lifestyle such as aerobic exercise and moderate to high-intensity sports such as jumping, squats, weightlifting and so on (Arazi and Eghbali, 2017; Benedetti et al., 2018).
One factor that affects bone density is age. Women will experience a decrease in bone mass greater than men when entering into menopause phase due to a decrease in the hormone estrogen. Estrogen plays a role in vitamin D regulation where vitamin D will help reabsorption of calcium in the kidneys and affect the process of bone formation (Aminah et al., 2009). Study explained that low bone density occurs with increasing age (Ho-Pham et al., 2014;Mishra et al., 2016). Along with increasing age, the bone will become thin and more porous about 4-10% since the age of 40 years. After the age of 40 years, bone-thinning will decrease to 3% per decade and increase up to 9% per decade in menopause women (Aminah et al., 2009). There was no association between age and bone density in our study both in 31-40 years and 41-50 years groups. This was because the mean age in this study was 40 years and in perimenopause period where it is possible that estrogen levels have not to decrease yet, thus the bone density in subjects is still in good condition.
The limitation of our study is that we used QUS to measure bone mass density, where QUS is not a gold standard tool. The limitation of QUS is a less accurate tool to predict bone mass density because it only measures calcaneal compared to DXA which predicts bone density through several bones located in the human body (lumbar spine, femoral neck).

Conclusion
Increased mesomorph component and body fat percentage values can increase bone density. However, high BMI in this study can decrease bone density in perimenopause women. Further research should examine the body type and bones mass density in adult women with inflammatory biomarkers corrections such as cytokines interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α) and inflammatory marker of C-reactive protein (CRP).
Physical activity and lifestyle effects on bone mineral density among young adults: sociodemographic and biochemical analysis. Journal of Physical Therapy Science, 27 (7)