CHAPTER 9. Anthropometry and Body Composition

CHAPTER 9 Anthropometry and Body Composition 9.1 INTRODUCTION Ageing is characterized by reduction in fat free mass (FFM), primarily via loss of muscle mass, loss of bone mineral in women, redistribution of body fat leading to increased deposition of body fat in 'internal fat' depots as opposed to subcutaneous depots (Durnin & Womersley, 1974; Allison et al, 1997; Allison et al, 1999; Andres et al, 1985). Though the general pattern of change in body composition in old age is clear, it is also clear that the rate, timing and extent of the changes varies between subjects and between sexes (Durenberg et al, 1998; Heymsfield et al 1999; Hirch et al, 1997). The purposes of this chapter are: a) To compare anthropometry and body composition of Indonesian elderly living in urban metropolitan Jakarta and in urban non-metropolitan Semarang. b) To investigate the prevalence of underweight and obesity in these two communities. 9.2 ANTHROPOMETRY Anthropometry is a simple, non-invasive, quick and reliable form of obtaining objective information on nutritional status. Longitudinal observations are particularly valuable in individual assessment and help to overcome the current lack of reference values in the cross sectional situation. 9.2.1 STATURE AND ARM SPAN Arm span is another substitute for height and happens to be the same as maximal height achieved. It is sometimes necessary to ask for maximum adult height to be recalled by the subject or by a carer. Gradual reduction in height may be an indicator of vertebral crush fractures due to osteoporosis, or may be due to loss of vertebral disc space. The high correlation between height and arm span of 0.83 for female and 0.81 for males, found in Indonesian elderly, confirmed that arm span approximates the same rank-ordering of a population as height (Rabe, 1996). 191

192 CHAPTER 9 Table 9.1 shows that Jakarta elderly were significantly taller (153cm) than their Semarang counterparts (150cm) (P<0.0001). The mean standing height (stature) of was 162cm and of 148cm (Table 9.2). In Semarang, average height for men was 158cm and for women 147cm (Table 9.3). The men were significantly taller than the women in both centres. The majority of the men (90%) in both communities had height 150cm or more (Figure 1). More women had height less than 150cm, particularly in Semarang. Arm span was assessed amongst Semarang elderly. The average arm span of Semarang elderly was 163cm for men and 153cm for women. The men had a greater arm span by 10cm compared to the women. A higher proportion of had arm span 150cm of more compared to (Figure 2). Figure 1. Percent distribution of height by gender and centre 63.4 36.6 11.1 88.9 50.7 49.3 2.6 97.4 <150 cm >=150 cm Figure 2. Percent distribution of armspan of Semarang elderly 28 72 6.3 93.7 < 150 cm >= 150 cm

Anthropometry and body composition 193 9.2.2 WEIGHT Interpretation of the weight of the elderly people should be done with circumspection. Increases in body weight may indicate overweight/ obesity or oedema. On the other hand, decreases in body weight can signify the correction of oedema, development of dehydration or emergence of nutritional disorder. Current weight may not be sufficient to identify those at health risk in old age. Weight change, however, is an indicator of health risk, particularly for thinner people who lost weight or heavier people who had gained weight. A major determinant of muscle mass in relatively sedentary population is body weight. Forbes (1987) suggests that lean mass is logarithmically related to body fat. Thus, heavier individuals of all ages are also generally stronger when asked to perform simple tests of muscle strength. Figure 3. Percent distribution of weight by gender and centre 57.1 40.6 2.3 61.9 36.5 1.6 33.1 64 2.9 10.3 75.6 14.3 < 50 kg 50-75 kg >=75 kg In Jakarta, mean body weight for men was 63 kg and for the women 54kg (Table 9.2). In Semarang, mean body weight for men was 51kg and for the women 47kg (Table 9.3). In both centres, men were significantly heavier than women. Centre differences in body weight were significant (P<0.0001). Figure 3 shows that more Jakarta elderly had body weight between 50kg to 75 kg compared to their Semarang counterparts.

194 CHAPTER 9 9.2.3 CIRCUMFERENCES (MID-UPPER ARM, ABDOMINAL, HIP AND CALF) 9.2.3.1 Mid-upper arm circumference (MUAC) Mid-upper arm circumference combined with triceps skinfold can be used to calculate midarm muscle area (MAMA) which is an index of total body protein mass. MAMA of less than 44cm 2 for men and less than 30cm2 for women may indicate protein malnutrition in Caucasian population. In both centres, men had an equal MUAC to women (Tables 9.2 and 9.3). The average circumference was 28cm for Jakarta subjects and 25cm for Semarang subjects (Table 9.1). Centre differences were significant (P<0.0001). Figure 4 shows that more Semarang elderly (40-64%) had MUACs less than 25cm than their Jakarta counterparts (13-18%). Figure 4. Percent distribution of MUAC by gender and centre 40 42.9 17.1 63.5 27 9.5 18.4 57.4 24.3 12.8 61.5 25.6 < 25 cm 25-30 cm >=30 cm

Anthropometry and body composition 195 9.2.3.2 Abdominal circumference had a higher abdominal circumference than their female counterparts (95 vs 83; P<0.0001) (Table 9.2). In Semarang, the average abdominal circumference for men was 73cm and for women 74cm (Table 9.3). Centre differences for abdominal circumference were observed (88 vs 74) (Table 9.1). Figure 5 shows that less Jakarta elderly (23% men, 46% women) had abdominal circumferences less than 80cm compared to Semarang elderly (78% men, 62% women). Figure 5. Percent distribution of abdominal circumferences by gender and centre 61.7 36 2.3 77.8 19 3.2 45.6 50.7 3.7 23.1 69.2 7.7 < 80 cm 80-100 cm >= 100 cm 9.2.3.3 Hip circumference Men had a smaller hip circumference than women, particularly in Semarang (Table 9.3). The average hip circumference of was 92cm, of 99cm (Table 9.2), of 84cm, of 88cm. Centre differences were significant (96 cm vs. 87cm, P<0.0001). The majority of the elderly in both places had hip circumferences between 80cm to 100 cm (Figure 6). Figure 6. Percent distribution of hip circumferences by gender and centre 21.1 64 14.9 33.3 63.5 3.2 7.4 71.3 21.3 3.8 85.9 10.3 <80 cm 80-100 cm >=100 cm

196 CHAPTER 9 9.2.3.4 Calf circumference Calf circumference (in the absence of lower limb oedema) can be used to calculate weight in bed bound patients. Calf circumference is expected to have increasing application for assessment of lean mass (Chumlea and Guo, 1999). It can also be used as a measure of physical fitness in the aged. In Jakarta, the average calf circumference for men was 35cm and for women 33cm (Table 9.2). In Semarang, men had a similar value to women (31cm) (Table 9.3). Gender differences were not significant in both places. Jakarta elderly, however, had a markedly higher calf circumference (Table 9.1). Figure 7 shows that a greater proportion of Jakarta elderly had calf circumferences 35 cm than Semarang elderly. Figure 7. Percent distribution of calf circumferences by gender and centre 42.3 45.7 12 41.3 42.9 15.9 19.1 56.6 24.3 5.1 42.3 52.6 < 30 cm 30-35 cm >=35 cm

Anthropometry and body composition 197 9.2.4 SKINFOLD THICKNESS (BICEPS, TRICEPS, SUBSCAPULAR, SUPRAILIAC, CALF) 9.2.4.1 Biceps skinfold The mean biceps skinfold for was 6.2mm and for the women 6.4mm (Table 9.2). In Semarang, the mean biceps for men was 4.4mm and for women 7.3mm (Table 9.3). Gender differences were only significant in Semarang. Centre differences in biceps skinfold were not observed (Table 9.1). Most of (75%) had biceps skinfolds less than 5mm (Figure 8). Figure 8. Percent distribution of bicep skinfolds by gender and centre 36.6 41.7 21.7 74.6 23.8 1.6 43.4 42.6 14 30.8 65.4 3.8 < 5 mm 5-10 mm >=10 mm 9.2.4.2 Triceps skinfold Men had a similar triceps skinfold to women in Jakarta (11.5mm vs. 11.9mm) (Table 9.2). In contrast, had a lower triceps skinfold than the women (10.6mm vs 15.1mm, (Table 9.2). Overall, Jakarta elderly had lower triceps skinfolds than their Semarang counterparts by 2 mm (11.8mm vs 13.8mm, P<0.01) (Table 9.1). Figure 9 shows that 57% of, 44% of, 35% of, and 25% of had triceps skinfolds 5mm or less.

198 CHAPTER 9 Figure 9. Percent distribution of tricep skinfolds by gender and centre 24.6 24 51.4 57.2 22.2 20.6 44.1 35.3 20.6 34.6 53.8 11.5 < 10 mm 10-15 mm >=15 mm 9.2.4.3 Subscapular skinfold The average subscapular skinfold of men was equal to that of women (14.6mm) in Jakarta (Table 9.2). On the other hand, had a markedly lower subscapular than (11.9mm vs 15.3mm, (Table 9.2). Centre differences for subscapular skinfolds were not significant (Table 9.1). The majority of the elderly (60-90%) had subscapular skinfolds 10 mm or more, except for. Figure 10. Percent distribution of subscapular skinfold by gender and centre 30.9 21.1 48 57.2 19 23.8 22.1 36 41.9 10.3 44.9 44.9 < 10 mm 10-15 mm >= 15 mm

Anthropometry and body composition 199 9.2.4.4 Suprailiac skinfold The average suprailiac skinfold of was 7.7mm and of women 7.2mm (Table 9.2). In Semarang, the value was 8.2mm for men and 13.2 for women (Table 9.3). Overall, suprailiac skinfold for Jakartans was 7.4mm and for Semarangs 11.8mm. Gender differences for suprailiac skinfolds were seen in Semarang, and centre differences were also significant (P<0.0001) (Table 9.1). Most of the elderly (75-85%) had suprailiac skinfolds 5 mm or more, except for (Figure 11). Figure 11. Percent distribution of suprailiac skinfolds by gender and centre 14.9 24.6 22.9 37.7 38.1 36.5 14.3 11.1 27.2 58.1 12.5 2.2 14.1 64.1 21.8 <5 mm 5-10 mm 10-15 mm >=15 mm 9.2.4.5 Total skinfold Total skinfolds (biceps + triceps + subscapular + subscapular) averaged 40mm both for men and women in Jakarta (Table 9.2). In Semarang, total skinfolds averaged 35mm for men and 51mm for women (Table 9.3). Overall, the average total skinfolds of Jakarta elderly were 40mm and of Semarang elderly of 46mm. Gender differences within centre were significant in Semarang. Centre differences in total skinfolds were significant (P<0.06) (Table 9.1). Most of the elderly (80-88%) had total skinfolds less than 50mm, except for (Figure 12).

200 CHAPTER 9 Figure 12. Percent distribution of total skinfolds by gender and centre 25.7 9.7 16.6 48 58.7 9.5 11.1 20.6 26.5 28.7 24.3 20.6 14.1 35.9 37.2 12.8 <30 mm 30-40 mm 40-50 mm >=50 mm 9.2.4.6 Calf skinfold The average calf skinfold of was 7.7mm and of women 7.2mm (Table 9.2). In Semarang, the value was 6.7mm for men and 8.7 for women (Table 9.3). Overall, calf skinfold for Jakartans was 7.4mm and for Semarangs 8.1mm. Gender and centre differences were not seen. Most of the elderly in both communities (70-90%) had calf skinfold less than 10mm (Figure 13). Figure 13. Percent distribution of calf skinfold by gender and centre 19.4 51.4 29.2 33.3 55.6 11.1 27.2 58.1 14.7 14.1 64.1 21.8 < 5 mm 5-10 mm >=10 mm

Anthropometry and body composition 201 Table 9.1 Anthropometric characteristics of Indonesian elderly by centre Jakarta (N=212) Semarang (N=238) Mean ± std Median Range Mean ± std Median Range Weight, kg 57±11**** 57 49-65 48±11 47 40-55 Height, cm 153±10 **** 152 147-160 150±7 150 145-155 BMI 24±6 **** 24 22-27 21±5 21 18-24 MUAC, cm 28±4 **** 28 26-30 25±4 25 23-28 MUAMC, cm 28±3**** 28 25-29 21±3 21 19-22 MAMA, cm 2 48±11 **** 47 40-54 37±9 35 29-41 Abdominal circumference, cm Hip circumference, cm 88±56 83 76-90 74±13 73 63-83 96±59 **** 92 87-98 87±10 86 80-93 A/H ratio 1.0±1.4 **** 0.9 0.8-0.9 0.8±0.1 0.8 0.8-0.9 Calf circumference, cm 33±5 **** 33 31-36 31±3 31 29-33 Biceps skinfold, mm 6.4±2.9 5.7 4.2-8.1 6.5±4.1 5.2 3.4-8.0 Triceps skinfold, mm 11.8±4.1 11.4 9.0-14.1 13.8±6.5 ** 13.4 8.2-19.2 Subscapular skinfold, mm 14.6±4.9 14.2 11.1-17.2 14.3±7.4 13.4 8.0-19.5 Suprailiac skinfold, mm 7.4±2.8 7.2 5.3-8.9 11.8±7.7 **** 10.0 5.4-16.0 Total skinfold, mm 40.1±12.7 39.4 31.4-47.2 46.4±23.4 43.4 25.4-63.4 Calf skinfold, mm 7.4±2.8 7.2 5.3-8.9 8.1±4.5 6.8 5.1-9.8 Significant differences from zero: * P<0.005; ** P<0.01; *** P<0.001; **** P<0.0001; P<0.06; P<0.07. BMI: Body mass index. MUAC: Mid-upper arm circumference. MUAMC: Mid-upper arm muscle circumference. MAMA: Mid-upper arm muscle area. A/H ratio: Abdominal/ Hip ratio. Range: 25 th 75 th percentile

202 CHAPTER 9 Table 9.2 Anthropometric data of Jakarta subjects Men (N=78) Women (N=135) Mean ± std Median Range Mean ± std Median Range Weight, kg 63±10 **** 61 56-70 54±11 54 46-61 Height, cm 162±6 **** 163 158-166 148±9 150 145-152 BMI 24±3 24 22-26 25±7 24 21-27 MUAC, cm 28±3 28 26-30 28±4 28 26-30 MUAMC, cm 28±3 28 26-30 28±4 27 25-29 MAMA, cm 2 49±10 48 43-54 47±12 46 40-53 Abdominal circumference, cm Hip circumference, cm 95±82 **** 87 80-93 83±34 81 74-87 92±7 92 88-96 99±7 93 87-98 A/H ratio 1.0±0.9 0.9 0.9-1.0 1.0±1.6 0.9 0.8-0.9 **** Calf circumference, cm 35±3 35 33-37 33±5 33 30-35 Biceps skinfold, mm 6.2±2.1 5.9 4.7-7.3 6.4±3.2 5.4 4.1-8.5 Triceps skinfold, mm 11.5±3.1 11.6 9.7-13.2 11.9±4.5 11.1 8.6-14.4 Subscapular skinfold, mm Suprailiac skinfold, mm 14.6±4.1 14.3 11.5-16.9 14.5±5.3 13.8 10.4-17.2 7.7±2.5 7.5 5.9-9.3 7.2±3.0 6.4 4.9-8.7 Total skinfold, mm 40.2±9.6 39.6 33.7-46.4 40.1±14.2 39.1 29.7-48.0 Calf skinfold, mm 7.7±2.5 **** 7.5 5.9-9.3 7.2±3.0 6.4 4.9-8.7 Significant differences from zero: * P<0.005; ** P<0.01; *** P<0.001; **** P<0.0001; P<0.06. BMI: Body mass index. MUAC: Mid-upper arm circumference. MUAMC: Mid-upper arm muscle circumference. MAMA: Mid-upper arm muscle area. A/H ratio: Abdominal/ Hip ratio. Range: 25 th 75 th percentile

Anthropometry and body composition 203 Table 9.3 Anthropometric data of Semarang subjects Men (N=62) Women (N=176) Mean ± std Median Range Mean ± std Median Range Weight, kg 51±11 * 48 44-61 47±12 47 39-54 Height, cm 158±6 **** 159 154-163 147±5 147 144-151 BMI 20±4 ** 19 18-23 22±5 22 18-25 Arm span, cm 163±13 **** 165 158-170 153±6 153 149-157 MUAC, cm 25±3 25 23-28 25±4 26 23-28 MUAMC, cm MAMA, cm 2 38±7 38 30-42 35±10 34 28-40 Abdominal circumference, cm Hip circumference, cm 73±12 72 65-80 74±14 74 63-83 84±8 83 79-89 88±10 ** 88 80-94 A/H ratio 0.9±0.1** 0.9 0.8-0.9 0.8±0.1 0.8 0.7-0.9 Calf circumference, cm 31±3 31 29-33 31±3 30 28-33 Biceps skinfold, mm 4.4±2.2 3.6 3.0-5.1 7.3±4.4 6.5 4.0-9.1 Triceps skinfold, mm 10.6±5.8 8.8 5.8-14.6 15.1±6.4 **** 15.2 10.0-20.1 **** Subscapular skinfold, mm Suprailiac skinfold, mm 11.9±6.6 9.4 7.2-16.0 15.3±7.5 **** 14.3 8.4-20.4 8.2±6.2 6.3 4.6-10.4 13.2±7.7 **** 12.0 6.4-19.0 Total skinfold, mm 35.1±18.0 28.8 20.8-46.9 51.0±23.8 **** 49.1 31.2-67.5 Calf skinfold, mm 6.7±4.0 5.6 4.4-8.0 8.7±4.5 7.8 5.4-10.8 Significant differences from zero: * P<0.005; ** P<0.01; *** P<0.001; **** P<0.0001; P<0.06; P<0.07. BMI: Body mass index. MUAC: Mid-upper arm circumference. MUAMC: Mid-upper arm muscle circumference. MAMA: Mid-upper arm muscle area. A/H ratio: Abdominal/ Hip ratio. Range: 25 th 75 th percentile

204 CHAPTER 9 9.3 BODY COMPOSITION 9.3.1 BODY MASS INDEX (BMI) Being overweight is an important contributor to risk of morbidity in younger people, particularly coronary heart disease. In people around age 65, heavier weight is associated with a modest increase in the risk of coronary heart disease. Harris et al (1997a) reported that BMI 27 in late middle age was associated with increased risk of coronary heart disease in later life (relative risk=1.7, 95% confidence interval 1.3 to 2.1). Thinner older people who lost weight and heavier people who had gained weight showed increased risk of CHD compared with thinner people with stable weight. The association between BMI and health risk remains controversial. Matilla and co-workers (1986) reported that BMI between 27 to 30 in the elderly was associated with less mortality. Numerous methods are also available to assess nutritional status by way of body composition, applicable for the elderly (Deurenberg et al, 1998; Lukaski, 1997). Body mass index has usually been the index of choice in studies, because its components - height and weight- are rapid and simple to measure. The robustness of the BMI as an indicator of chronic energy deficiency has been vigorously championed and to a lesser extent, defended as an index of overweight as well. Furthermore both extremes of BMI have been proven to be strong predictors of increased mortality and morbidity (Harris, 1997b). Although not significant, men had a lower BMI than women in Jakarta (24 vs 25). In Semarang, men had a markedly lower BMI than women (20 vs 22, P<0.01). When men and women were pooled together, Jakarta elderly had a significantly higher BMI than Semarang elderly (24 vs 21) (Table 9.4). Figure 14 shows that more Semarang elderly (30%) had BMI less than 18.5 compared to Jakarta elderly (<7%). A small proportion of the elderly in both communities had BMI 30 or greater. The figure indicates the problems of undernutrition are more prevalent compared to that of overnutrition.

Anthropometry and body composition 205 Table 9.4 Distribution of BMI by gender and centre Gender Jakarta Semarang Mean ± std g/d Median g/d Range Mean ± std g/d Median g/d Range Men 24±3 24 22-26 20±4 19 18-23 Women 25±7 24 21-27 22±5 ** 22 18-25 Men + Women 24±6 **** 24 22-27 21±5 21 18-24 Range: 25 th 75 th percentile Figure 14. Percent distribution of BMIs by gender and centre 29.1 8 36.6 18.9 7.4 36.5 19 31.7 11.11.6 6.6 7.4 45.6 30.9 9.6 5.1 7.7 47.4 35.9 3.8 <18.5 18.5-19.9 20.0-24.9 25.0-29.9 >=30.0 9.3.2 BODY MASS ARM SPAN Using height in calculating BMI could be inappropriate because of height loss with aging due to the compression of vertebrae, kyposis and osteoporosis. Solomon and his colleagues (1993) have raised the question of an appropriate weight to cover a frame in which the vertical dimension (height) is shrinking. Correction of the measured height by an estimate of the stature that existed in adulthood was suggested. Arm span is relatively independent of aging and is highly related to the height of an individual (Rabe et al, 1996). To predict the prevalence of CHD in elderly it may be more reliable to use arm span in the determination

206 CHAPTER 9 of weight related to body stature. Body mass arm span (BMA) can be calculated by dividing body weight (kg) to arm span (m 2 ). Men had a similar body mass arm span to women (20) in Semarang (Table 9.5). Figure 15 shows that a greater proportion of men (61%) had BMA <18.5 compared to women (41%), but a reversed figure for BMA between 18.5 to 25. Table 9.5 Body mass arm span of Semarang subjects by gender Mean ± std g/d Median g/d Range (25 th -75 th percentile) Men (n=62) 20±13 18 16-23 Women (n=176) 20±5 20 16-23 Men + Women (n=238) 20±8 19 16-23 P<0.07. Figure 15. Percent distribution of BMAs of Semarang elderly <18.5 18.5-19.9 20.0-24.9 25.0-29.9 >=30.0 9.3.3 ABDOMINAL TO HIP RATIO The average ratio of abdominal to hip in both and women was 1.0. The ratio for was 0.9 and for 0.8. The ratio amongst Jakarta elderly was 1.0 and amongst Semarang elderly 0.8. Gender and centre differences were significant (Table 9.6). Figure 16 indicates that more women were abdominally obese compared to their male counterparts.

Anthropometry and body composition 207 Studies have found that abdominal circumference is a stronger predictor of health than the abdominal to hip ratio. Abdominal circumference greater than 100cm for men and 90cm for women is associated with greater health risks. This study found that a small proportion of the elderly in both communities (<8%) had abdominal circumference 100cm or greater (Figure 5). Table 9.6 Abdominal to hip ratio by gender and centre Gender Mean ± std g/d Jakarta (n=) Median g/d Range Mean ± std g/d Semarang (n=) Median g/d Range Men (n=) 1.0±0.9 **** 0.9 0.9-1.0 0.9±0.1 ** 0.9 0.8-0.9 Women (n=) 1.0±1.6 0.9 0.8-0.9 0.8±0.1 0.8 0.7-0.9 Total 1.0±1.4 **** 0.9 0.8-0.9 0.8±0.1 0.8 0.8-0.9 Range: 25 th 75 th percentile Figure 16. Percent distribution of A/H ratio by gender and centre 58.3 41.7 92.1 7.9 45.6 54.4 62.8 37.2 =<0.85 women, =< 0.95 men >0.85 women,>0.95 men

208 CHAPTER 9 9.3.4 ARM MUSCLE CIRCUMFERENCE AND ARM MUSCLE AREA 9.3.4.1 Mid-upper arm muscle circumference (MUAMC) Jakarta elderly had a significantly higher mean of MUAMCs (28cm) than Semarang elderly (21cm) (Table 9.1). In Jakarta, men had an equal MUAMCs to their female counterparts (Table 9.2). Nevertheless, had a significantly higher MUAMCs than (Table 9.3). A higher proportion of Jakarta elderly (80%) than Semarang elderly (65%) had MUAMCs of 20-30cm (Figure 17). Figure 17. Percent distribution of MUAMCs by gender and centre 40.6 58.9 0.6 30.2 69.8 1.5 76.5 22.1 1.3 80.8 17.9 < 20 cm 20-30 cm >=30 cm

Anthropometry and body composition 209 9.3.4.2 Mid-upper arm muscle areas (MAMA) Jakarta elderly had significantly higher MAMAs (48cm 2 ) than Semarang elderly (37cm 2 ) (Table 9.1). In Jakarta, men had similar MAMAs to women (Table 9.2); in had a significantly higher MAMAs than women ((Table 9.3). The majority of Jakarta elderly (75%) had MAMA 40cm 2 or more; but the majority of Semarang elderly (70%) had MAMA less than 40cm 2 (Figure 18). Figure 18. Percent distribution of MAMA by gender and centre 72 22.3 5.7 68.3 30.2 1.6 25.7 39.7 34.6 16.7 39.7 43.6 < 40 cm2 40-50 cm2 >=50 cm2 9.4 BODY FAT AND FAT MASS Great interest has focused recently on patterns of body fat distribution rather than amounts of fat per se as the real risk factor for cardiovascular disease. A variety of determinants influence regional body fat distribution, such as obesity, smoking and age (Bouchard et al, 1990). Because it has been shown that body weight varies with age, first increasing up to middle age and then declining in both sexes, it is important to distinguish between the effects of age and the effect of changes in body weight on body fat distribution. Cross sectional and longitudinal studies have shown age related changes in the pattern of body fat distribution evaluated by measuring waist to hip ratio or skinfold thickness (Noppa et al, 1980; Forbes, 1990).

210 CHAPTER 9 9.4.1 TOTAL BODY FAT Using bioimpedance assessment, total body fat was 17kg (27%) for, 8kg (14%) for, 12kg (23%) for, and 6kg (13%) for. Using skinfold measurements with Durnin equation, total body fat was 14kg (23%) for, 16kg (30%) for, 10kg (19%) for, and 16kg (32%) for (Table 9.7). Men in both centres had a higher total body fat calculated by bioimpedance assessment compared to skinfold measurements. However, it was found to be the reverse in women. Figure 19. Percent distribution of total body fat using skinfold measurements by gender and centre 6.3 28.6 65.1 60.3 27 12.7 4.3 48.9 46.8 23.1 74.4 2.6 < 20 % 20-30 % >= 30 %

Anthropometry and body composition 211 Table 9.7 Distribution of total body fat by gender and centre Total body fat Jakarta Semarang Men Women Men Women Total body fat (BIA), kg 17.5±4.8 8.0±2.5 12.1±5.1 6.5±2.6 Total body fat (BIA), % 27 14 23 13 Total body fat (skinfold), kg 14.5±4.3 16.4±5.4 10.2±6.1 16.2±6.7 Total body fat (skinfold), % 23 30 19 32 BIA: bio-impedance assessment Figure 19 shows that the majority (87%) of men in both communities had total body fat less than 30% using skinfold measurements. On the other hand, 47-65% of women had total body fat 30% or more using skinfold measurements. Figure 20 shows that all the women and the majority of men (80%) had total body fat 20% or less. Figure 20. Percent distribution of total body fat using BIA by gender and centre 100 15.9 79.4 4.8 100 2.6 75.6 21.8 < 20 % 20-30 % >= 30 %

212 CHAPTER 9 9.4.2 FAT FREE MASS Using bioimpedance assessment, fat free mass was 45kg (73%) for, 46kg (86%) for, 38kg (77%) for, and 41kg (87%) for. Using skinfold measurements with Durnin equation, fat free mass was 48kg (77%) for, 38kg (71%) for, 40kg (81%) for, and 32kg (68%) for (Table 9.8). Table 9.8 Distribution of fat free mass by gender and centre Jakarta Semarang Men Women Men Women Fat free mass (BIA), kg 45.3±5.1 46.4±8.1 38.2±6.5 41.2±9.4 Fat free mass (BIA), % 73 86 77 87 Fat free mass (skinfold), kg 48.4±5.9 37.9±5.7 40.1±6.5 32.0±5.6 Fat free mass (skinfold), % 77 71 81 68 Figure 21. Percent distribution of fat free mass using skinfold measurements by gender and centre 67.4 28.6 4 14.3 27 58.7 48.9 48.9 2.2 2.6 74.4 23.1 < 70 % 70-80 % >= 80 %

Anthropometry and body composition 213 Figure 21 shows that most men (85%) and 30-50% women in both ethnic groups had fat free mass 70 % or more using skinfold measurements. By using BIA, most men and women (80%) in both places had fat free mass 70% or more (Figure 22) Figure 22. Percent distribution of fat free mass using BIA by gender and centre 96.6 3.4 4.8 79.4 15.9 2.2 97.8 21.8 75.6 2.6 < 70% 70-80 % 80-90 % >=90 % 9.4.3 PREVALENCE OF OBESITY 9.4.3.1 General obesity Obesity is a risk factor for many chronic diseases. Although studies give varying values for ideal weights in older persons, extremes of body mass index {very high (BMI 30) and very low (BMI<18.5) are adversely associated with health in most populations (Khaw, 1997). The BMI was used in this study to classify subjects as being underweight (BMI<18.5), desirable weight (BMI 18.5-25), overweight (BMI 25-30), and obese (BMI 30). Figure 23 shows that 4%, 10% of, 2% of and 7% of were obese using the WHO criteria of BMI 30 or more. A working party with representation from WHO, the International Society for the Study Obesity, and the International Obesity Task Force have recently recommended new criteria for defining obesity in the Asian population. In Asians, a BMI range of 23-24.9kg/m2 has an equivalent risk of type-2 diabetes, hypertension, and dyslipidemia as a BMI range of 25-

214 CHAPTER 9 29.0kg/m2 in Europeans (Zimmet et al, 1998 ). Figure 23 shows that 40% of and women, 13% of, and 26% of had BMI 25 or more. Figure 23. Percent distribution of obesity by gender and centre 45 40 35 30 25 20 15 10 5 0 3.8 39.7 40.5 9.6 1.6 12.7 7.4 26.3 Semarang women BMI >=30 BMI >=25 9.4.3.2 Central obesity Central obesity was defined using abdominal to hip ratio with cut-off point of 0.95 for men and 0.85 for women. The prevalence of central obesity in Jakarta was 37% for men and 54% for women. The prevalence in was 8% for men and 42% for women (Figure 24). Figure 24. Percent distribution of central obesity by gender and centre 60 54.4 50 40 37.2 41.7 30 20 10 7.9 0 Semarang women

Anthropometry and body composition 215 9.5 DISCUSSION Obtaining the correct height of an elderly individual may be complicated due to the physical and anatomical changes that occur with aging. The most obvious physical alteration is a change in body stature, which is primarily caused by a progressive decline in absolute height. This decrease in height that occurs with age has been reported in several studies. Shortening of the spinal column is primarily responsible for the loss of the stature that occurs with aging since it has shown that the long bones do not undergo a significant reduction in size with age. Another factor affecting height is the kyphosis, which gives an individual the hunchback appearance that is present in many elderly people. It is believed that this condition is due to generalized osteoporosis. Severe osteoporosis may cause the bones in the legs to bow. This bowing, in combination with a curvature of the spine, makes it difficult to obtain the correct height. Body weight is the most commonly used indicator of the body size, which reflects adequate nutriture. It, however, has limited use because it does not distinguish between body size and composition. Master et al reported data on height and weight and found that the percentage of overweight males declines steadily over the 30-years span to constitute only 10%. Simultaneously, the number of underweight men increased from 20 to 50%. For the same agegroup, the percentage of overweight females decreased from 40 to 10%, with underweight females increasing from 20 to 55%. Most males continue to gain weight until about age 42, while women do not attain their maximum weight until age 50. Thereafter, weight usually remains fairly constant up to about age 65 or 70, whereupon it will progressively decline. Most studies documented a cumulative decrease in lean body mass with age accompanied by an increase in adipose tissue. Evidence suggests that the increased amounts of fat be deposited primarily around internal organs, especially females, while the amount of subcutaneous fat increases only slightly. The end result is a reduced ratio of subcutaneous fat to total body fat. After age 65 to70, total body weight decreases without an increase in fat tissue. This study found that Jakarta elderly were taller by 3 cm and heavier by 4 kg compared to Semarang elderly. Men tended to be taller and heavier than women in both communities; height differences were 14cm in Jakarta and 11 cm in Semarang, while weight differences were 9 kg and 4 kg respectively. Despite genetic factors, food and nutrient intakes may cause differences in weight and height of the elderly people in these communities. It was reported before (chapter 8) that Jakarta elderly had significantly higher energy and nutrient intakes. Similarly, BMIs of Jakarta elderly were higher than that of Semarang elderly. In Jakarta, the men had a similar BMI compared to the women, despite differences in their stature and

216 CHAPTER 9 body weight. In Semarang mean BMI of the men was significantly lower than that of the women (20 vs 22). Moreover, less Jakarta elderly had BMI <18.5 compared to Semarang elderly (6% vs 31%). On the other hand, overweight was more prevalent in Jakarta. Differences in nutritional status agreed with their differences in energy intake. It has been reported that less Jakarta elderly consumed energy <1000 kcal/d and more Jakarta elderly consumed energy 2000 kcal/d. Energy intake may in part explain their nutritional status. Several studies have documented that extreme BMIs were predictive of CVD risk. Both low and high BMI increase risk of CVD and mortality. Semarang elderly may have a greater risk of CVD due to their low BMI. Body circumference includes MUAC, abdominal, hip and calf circumferences were higher in Jakarta than in Semarang. Their value differed by 3, 14, 9, and 2cm respectively. In Jakarta, the men had a higher abdominal circumference than the women; other circumferences were similar. In Semarang, the men had a smaller hip circumference. Abdominal to hip ratio of Jakarta subjects was higher than that of Semarang subjects (1.0 vs 0.8). In both communities, the men tended to have a higher ratio than their female counterparts. Overall, the prevalence of central obesity was higher in Jakarta. In both places, the women tended to be more abdominally obese compared to the men. This suggests that Jakarta elderly, especially the women may be at a greater risk of CVD. Jakarta elderly had lower total skinfolds than Semarang elderly (40mm vs 46mm). Biceps and subscapular skinfolds were equal in these two populations, however, triceps and suprailiac skinfolds were significantly lower in Jakarta. It is interesting that Semarang elderly had more body fat deposited on their triceps and on their suprailiac. 9.6 SUMMARY Jakarta elderly were significantly taller (153cm) than their Semarang counterparts (150cm) (P<0.0001). The men in both centres tended to be taller than the women. Mean standing height (stature) of was 162cm and the women 148cm. In Semarang, average height for men was 158cm and for women 147cm. Arm span was assessed amongst Semarang elderly. The average arm span of Semarang elderly was 163cm for men and 153cm for women. Jakarta elderly had a higher body weight than their Semarang counterparts by 9kg. The men were heavier than the women in both centres. The average body weight was 63kg for, 54kg for, 51kg for, and 47kg for.

Anthropometry and body composition 217 Although it was not significant, men had a lower BMI than women in Jakarta (24 vs 25). In Semarang, men had a markedly lower BMI than women (20 vs 22, P<0.01). When men and women were pooled together, Jakarta elderly had a significantly higher BMI than Semarang elderly (24 vs 21). It was also found that 6% Jakarta elderly and 31% Semarang elderly were underweight (BMI <18.5). Furthermore, 33% of Jakarta elderly and 17% Semarang elderly were overweight (25 BMI<30). Finally, the prevalence of obesity (BMI 30) in both centres was similar (6 to 7%). Men had a similar body mass arm span (20) to women in Semarang. A greater proportion of men (61%) had BMA <18.5 compared to women (41%), but a reversed figure for BMA between 18.5 to 25.A greater proportion of Semarang elderly was underweight using BMA compared to BMI. Of the body circumferences, Jakarta elderly had markedly higher circumferences. These include MUAC, abdominal, hip, and calf circumference. Jakarta elderly had higher circumferences by 3, 14, 9, and 2cm respectively. In both centres, the men had an equal MUAC and calf to the women. had a higher abdominal circumference than (95cm vs 83cm). had a smaller hip circumference than (84cm vs 88cm). The average ratio of abdominal to hip was higher in Jakarta (1.0) than in Semarang (0.8). Men in both centres had a higher ratio of abdominal to hip. The average ratio was 1.0 for and women, 0.9 for, and 0.8 for. Central obesity was defined using abdominal to hip ratio with cut-off point of 0.95 for men and 0.85 for women. The prevalence of central obesity in Jakarta was 37% for men and 54% for women. The prevalence in was 11% and 41%. The women in both communities had been found to be more abdominally obese compared to the men. Of the skinfold, Jakarta elderly had a significantly lower total skinfold than Semarang elderly (40.1mm vs 46.4mm). Biceps and subscapular skinfold were similar (6.4mm and 14.6mm), but, triceps and suprailiac were lower in Jakarta. Averaged triceps was 11.8mm in Jakarta and 13.8mm in Semarang, while suprailiac was 7.4mm in Jakarta and 11.8mm in Semarang. In Jakarta, men and women had a similar skinfold, but in Semarang, the men had a lower skinfold than the women.

218 CHAPTER 9