1 INTRODUCTION

Both hypertension and obesity are important risk factors for cardiovascular disease.1, 2 The relationship between hypertension and obesity had been established3 and previous studies also demonstrated the positive association of body mass index (BMI) to both systolic and diastolic blood pressure (BP) levels.4, 5

Despite general obesity, central obesity which is defined by waist circumference (WC) also plays an important role in cardiometabolic diseases.6 Studies have shown a positive association of gain of WC with increased BP.7, 8 The relationship between different measurements of obesity and hypertension reported that even the WC predicts the risk of hypertension better than BMI.9

Sarcopenia was defined as age-related loss of muscle.10 Both an increase in adipose tissue and a decrease in muscle mass are key risk factors of cardiometabolic disease in the elderly.11 Recently, the concern for the relationship between sarcopenic obesity and BP has risen. Although earlier meta-analysis and systemic review demonstrated sarcopenia is associated with hypertension,12 the direct relationship between lean body mass and BP remains inconsistent since several studies had demonstrated a positive association of lean body mass to BP.13-15 On the other hand, few studies in Asian population use lean body mass to total body mass ratio as indicator and revealed an inverse relationship with BP.16, 17 In our study, we also included AMM to total body mass ratio and try to evaluate its relationship to BP.

The objective of this study was to evaluate the association of obesity, central obesity, AMM, and BP among the community-dwelling elderly population in Taiwan. Therefore, by understanding the relationship between body composition and BP, our strategy toward physical training health promotion among the elderly can be improved in the future.

2 MATERIALS AND METHODS 2.1 Study population

A Series of community-based surveys were conducted among elderly individuals in Chiayi County in Taiwan from 2017 to 2019. We invited those who were greater than and aged to 65 years and who have lived in Chiayi County for more than one year. The inclusion criteria were individuals aged 65–85 years and those free from infectious diseases or acute disorders in the past three weeks.

2.2 Questionnaire

General demographic data including gender, age, residency, education level, occupation, and the need for a caregiver were collected using a standard questionnaire. Lifestyle patterns including dietary habits, cigarette smoking, alcohol intake, and daily activity were also collected. History of chronic diseases including diabetes, hypertension, cardiovascular disease, chronic kidney disease, any type of cancer, and cerebrovascular disease as well as medication history was recorded by a research technician.

2.3 Anthropometric measurement

The anthropometric characteristics including body weight (BW), body height, and WC were determined using standard methods. The height was measured in meters using a digital stadiometer that recorded to the nearest 0.5 cm. BW was measured to an accuracy of 0.1 kg by a standard beam balance scale. During the above measurement, the persons were barefoot and they wore only light indoor clothing. The WC was measured using standard methods suggested by the World Health Organization. The WC was measured to the nearest 0.1 cm at the midpoint between the margin of the last rib and the iliac crest of the ilium. BMI was calculated by dividing BW (kg) by the square of the height (m2).

2.4 Appendicular muscle mass (AMM) measurement The equation to calculate AMM included handgrip strength (GS), BW, sex, and height: AMM=−9.833+0.397×weightkg+4.433×sex+0.121×heightcm+0.061×handGSkg.\begin } &=& - 9.833 + 0.397 \times }\left( }} \right) + 4.433 \times }\\ && +\, 0.121 \times }\left( }} \right) + 0.061 \times }\left( }} \right). \end

This equation best predicts AMM measured by dual-energy X-ray absorptiometry (adjusted R2 = 0.914, standard error of the estimate = 2.062, P < .001).18 As BMI, WHR, and AMM, may confound with each other. We found the AMM divided by the total body weight ratio as a surrogate marker of the percentage of body lean muscle (as AMMW) may demonstrate different body composition meaning. We also calculated the AMMW ratio (Consequently, the study population was divided into two subgroups based on the median of AMMW (0.647 for males, 0.570 for females). The group with a higher AMMW ratio was defined as AMMW (+) and the group with a lower AMMW ratio was defined as AMMW (−).

2.5 Blood pressure measurement

Blood pressures were measured after resting 5-10 minutes in a seated position. The right arm was asked to be positioned into cuffs of appropriate sizes at the same height as the heart. Two measurements were recorded; the mean value of the two recordings was used for data analysis. The pulse pressure (PP) was calculated by systolic blood pressure (SBP) minus diastolic blood pressure (DBP). Mean arterial pressure (MAP) was calculated by one third of SPP plus two thirds of DBP.

2.6 Grip strength measurement

The GS was measured using digital dynamometers (TKK5101). All persons were in a seated position with fully extended elbows. After 2-3 minutes of rest, we measured the GS for right or left hands two times. Two values for the GS were recorded, and the mean value of the two recordings was used for analysis.19

2.7 Approval of the IRB

All participants provided written informed consent and agreed to have their general demographic data, questionnaire, and anthropometric data taken for this study. The institutional review board of the Tri-service General Hospital (Number: TSGHIRB-1-108-05-073) approved this study.

2.8 Statistical methods

We used SPSS v. 22 to conduct all statistical analyses. Continuous variables, such as anthropometric measures, blood pressure, and grip strength were presented as sample mean and SD. The Mann-Whitney U test was used to compare the differences between the groups. The Kruskal-Wallis H test was used to compare > 3 groups. The Pearson's correlation coefficient was applied for correlation of each variable. The categorical variables were described as numbers and percentages. The chi-square test was performed to compare the differences among ≥ 2 groups. Multivariant regression analyses and logistic regression analyses were used for further statistical inference and a two-tailed P-value < .05 was considered statistically significant.

3 RESULTS

Table 1 demonstrates the general characteristics of anthropometric measures, BP, chronic disease status, and adverse lifestyle behaviors according to AMMW subgroups among the study male and female population, respectively. In this study, there was about 87.3% (603/691) male and 87.5 % (832/951) female with hypertension that took anti-hypertension medications regularly. The study population was divided into two subgroups based on the median of AMMW with gender specification. The persons with higher AMMW showed lower SBP (136.8 ± 19.1 to 140.6 ± 17.0, P < .001 for male; 137.8 ± 18.3 to 142.7 ± 17.5, P < .001 for female); lower DBP (80.6 ± 10.9 to 82.8 ± 11.1, P < .001 for male; 77.4 ± 10.3 to 80.0 ± 10.0, P < .001 for female); lower MAP (99.3 ± 12.3 to 102.1 ± 11.5, P < .001 for male; 97.5 ± 11.4 to 100.9 ± 11.0, P < .001 for female), and lower prevalence of hypertension (35.6% to 50.7% in male and 36.6% to 52.3% in female, both P < .001).

TABLE 1. Characteristics of anthropometric measures, grip strength, blood pressure, chronic disease, and adverse behaviors among community-based population (no. = 3739) AMMW(-) AMMW(+) Variables Mean±SD Mean±SD P Male (no. = 1600) Age (years)a 72.9±5.9 73.6±6.2 .038 Body height (cm) 163.0 6.1 161.7   6.0 <.001 Body weight (kg) 73.0  7.9 58.5   6.1 <.001 BMI (kg/m2)a 27.5±2.6 22.4±19 <.001 Body waist (cm)a 94.2±7.2 82.5±6.7 <.001 Grip strength (kg) 33.1±7.4 32.5±6.9 .123 AMM (kg) 45.3±3.8 39.4±3.2 <.001 AMMW 0.622±0.018 0.675±0.024 <.001 SBP (mmHg) 140.6±17.0 136.8±19.1 <.001 DBP (mmHg) 82.8±11.1 80.6±10.9 <.001 PP (mmHg) 57.9±14.3 56.2±14.9 .023 MAP (mmHg) 102.1±11.5 99.3±12.3 <.001 CVD (no.,%)b 140  17.5 112   14.0 .055 CVA (no.,%)b 38   4.8 24    3.0 .070 Hypertension (no.,%)b 406  50.7 285   35.6 <.001 DM (no.,%)b 201  25.1 133   16.6 <.001 CKD (no.,%)b 35   4.4 30   3.8 .527 Smoking (no.,%)b 94  11.8 119   14.9 .066 Alcohol drinking (no.,%)b 146   18.3 131   16.4 .322 Female (no. = 2139) Age (years)a 72.5±5.9 72.9±6.1 .118 Body height (cm) 150.5  5.6 151.3   5.8 .001 Body weight (kg) 63.5   7.8 50.8    6.0 <.001 BMI (kg/m2)a 28.0±2.9 22.1±1.9 <.001 Body waista 88.8±7.9 77.5±6.9 <.001 Grip strength (kg) 21.4±4.8 21.8±4.8 .098 AMM (kg) 34.9±3.6 30.0±3.1 <.001 AMMW 0.551±0.014 0.591±0.017 <.001 SBP (mmHg) 142.7±17.5 137.8±18.3 <.001 DBP (mmHg) 80.0±10.0 77.4±10.3 <.001 PP (mmHg) 62.7±14.7 60.4±15.2 .001 MAP (mmHg) 100.9±11.0 97.5±11.4 <.001 CVD (no.,%)b 172   16.1 150    14.0 .180 CVA (no.,%)b 20   1.9 16    1.5 .500 Hypertension(no.,%)b 559   52.3 392    36.6 <.001 DM (no.,%)b 239   22.4 184   17.2 .003 CKD (no.,%)b 30   2.8 22     2.1 .260 Smoking (no.,%)b 8   0.7 11     1.0 .491 Alcohol drinking (no.,%)b 27   2.5 20     1.9 .300 Abbreviations: AMM, appendicular muscle mass; AMMW, appendicular muscle mass divided by body weight; BMI, body mass index; CKD, chronic kidney disease.; CVA, cerebrovascular disease; CVD, cardiovascular disease; DBP, diastolic blood pressure; DM, diabetes mellitus; MAP, mean arterial pressure; PP, pulse pressure; SBP, systolic blood pressure.

Table 2 shows the correlation between anthropometric measures, AMM, AMMW, and different types of BP measurements among study population with gender specifications. For example, the BMI was positively correlated to both SBP (r = 0.166 in male, r = 0.176 in female, both P < .001) and DBP (r = 0.149 in male, r = 0.172 in female, both P < .001). The WC showed positive correlation to SBP (r = 0.158 in male, r = 0.153 in female, both P < .001) and DBP (r = 0.116 in male, r = 0.123 in female, both P < .001). The AMM was positively correlated to both SBP (r = 0.126 in male, r = 0.124 in female, both P < .001) and DBP (r = 0.213 in male, r = 0.181 in female, both P < .001). The AMMW ratio was negatively correlated to SBP (r = −0.168 in male, r = −0.167 in female, both P < .001) and DBP (r = −0.140 in male, r = −0.150 in female, both P < .001).

TABLE 2. Correlation between blood pressure and anthropometric measures, AMM, and AMM/W among community-based population with gender specifications SBP (mmHg) DBP (mmHg) PP (mmHg) MAP (mmHg) coefficient coefficient coefficient coefficient Male (no. = 1600) Age (years) 0.142*** −0.143*** 0.284*** −0.017 BMI (kg/m2) 0.166*** 0.149*** 0.093*** 0.175*** Body waist (cm) 0.158*** 0.116*** 0.109*** 0.151*** Grip strength (kg) −0.003 0.195*** −0.151*** 0.118*** AMM (kg) 0.126*** 0.213*** −0.005 0.194*** AMMW −0.168*** −0.140*** −0.103*** −0.170*** Female (no. = 2139) Age (years) 0.188*** −0.089*** 0.286*** 0.046* BMI (kg/m2) 0.176*** 0.172*** 0.094*** 0.197*** Body waist (cm) 0.153*** 0.123*** 0.100*** 0.155*** Grip strength (kg) 0.022 0.094*** −0.038 0.069*** AMM (kg) 0.124*** 0.181*** 0.026 0.175*** AMMW −0.167*** −0.150*** −0.098*** −0.179*** Abbreviations: AMM, appendicular muscle mass; AMMW, appendicular muscle mass divided by body weight;BMI, body mass index; DBP, diastolic blood pressure; MAP, mean arterial pressure.; PP, pulse pressure; SBP, systolic blood pressure. ***P < .001, **P < .01, *P < .05.

Table 3 summarizes the distributions of BP among different WC and AMMW subgroups with gender specifications. Persons were further divided based on their WC (as central obesity, OB+: WC ≥ 90 cm for male, WC ≥ 80 cm for female; non-central obesity, OB-: WC < 90 cm for male, < 80 cm for female). AMMW was also divided into two subgroups according to the median of AMMW (0.647 for male and 0.570 for female). Then, the study persons were divided into four subgroups based on different AMMW and WC statuses. The subgroup with both central obesity and lower AMMW (OB(+), AMMW (−)) had the highest SBP (141.1 ± 17.1 for male,143.8 ± 18.2 for female), DBP (83.0 ± 11.4 for male, 80.4 ± 10.1 for female), and MAP (102.3 ± 11.8 for male, 101.5 ± 11.3 for female).

TABLE 3. Distributions (Mean±SD) of blood pressure among study population by specific anthropometrics, AMMW, and gender groups Obesity and AMMW Status OB(−), AMMW(+) Group 1 OB(−), AMMW(−) Group 2 OB(+), AMMW(+) Group 3 OB(+), AMMW(−) Group 4 P-valuea Post hoc Male(no. = 1600) (no. = 702) (no. = 201) (no. = 98) (no. = 599) SBP(mmHg) 136.5±19.1 139.3±16.6 138.9±18.7 141.1±17.1 <.001 1 < 4a DBP(mmHg) 80.6±10.9 82.2±10.3 80.9±11.3 83.0±11.4 .001 1 < 4a PP(mmHg) 55.9±15.0 57.1±14.2 58.0±14.3 58.1±14.4 .353 MAP(mmHg) 99.2±12.3 101.2±10.8 100.2±12.5 102.3±11.8 <.001 1 < 4a Female(no. = 2139) (no. = 1022) (no. = 617) (no. = 48) (no. = 452) SBP(mmHg) 137.8±18.1 141.9±16.9 138.1±20.8 143.8±18.2 <.001 1 < 2,4a DBP(mmHg) 77.4±10.2 79.8±10.0 76.6±12.7 80.4±10.1 <.001 1 < 2,4a PP(mmHg) 60.4±15.2 62.2±14.6 61.4±15.9 63.4±15.0 .058 MAP(mmHg) 97.5±11.3 100.5±10.7 97.1±14.0 101.5±11.3 <.001 1 < 2,4a Abbreviations: DBP, Diastolic blood pressure; MAP, Mean arterial pressure.; PP, Pulse pressure;SBP, Systolic blood pressure. OB(−), body waist < 90 cm for male or body waist < 80 cm for female; OB(+), body waist > = 90 cm for male or body waist > = 80 cm for female; AMMW(+), AMMW > = median for male and female; AMMW(−), AMMW < median for male and female, the median value of AMMW is 0.647 for male and 0.570 for female.

Table 4 shows the association between AMMW on different BP measurements after adjusting for potential confounds with gender specification. In this study, we could not obtain the information on the type of anti-hypertension medications. Anti-hypertensive medicine may be potential cofounder and we only controlled this as with and without anti-hypertensive medicine in the model. In the final model, after adjusting for age, body waist, disease status, and anti-hypertensive medications, the AMMW was negatively associated with blood pressure. For example, every increase of 10% of AMMW is associated with decrease of 6.4 mmHg SBP and decrease of 3.1 mmHg DBP in male. The regression coefficients all are with statistical significance, such as, for SBP (β = −0.641 in male, P < .01; β = −0.780 in female, P < .001), DBP (β = −0.312 in male, P < .05; β = −0.438 in female, P = .001), and MAP (β = −0.422, P < .01 in male; β = −0.552, P < .001 in female).

TABLE 4. Multivariate regression analyses of AMMW (for every 1% change) on different blood pressure dependent variables with gender specifications Dependent Model Ia Model IIb variables β se β P-value Β se β P-value Male(no. = 1600) SBP(mmHg) −0.903 0.133 <.001 −0.641 0.236 .007 DBP(mmHg) −0.457 0.081 <.001 −0.312 0.146 .033 PP(mmHg) −0.445 0.108 <.001 −0.329 0.187 .078 MAP(mmHg) −0.606 0.088 <.001 −0.422 0.158 .008 Female(no. = 2139) SBP(mmHg) −1.169 0.149 <.001 −0.780 0.219 .001 DBP(mmHg) −0.598 0.085 <.001 −0.438 0.128 .001 PP(mmHg) −0.571 0.125 <.001 −0.342 0.180 .058 MAP(mmHg) −0.788 0.093 <.001 −0.552 0.140 <.001 Abbreviations: AMMW, appendicular muscle mass divided by body weight.; DBP, diastolic blood pressure; MAP, mean arterial pressure; PP, pulse pressure; SBP, systolic blood pressure; se, standard error;β, regression coefficient. 4 DISCUSSION

The present study demonstrated that obese persons with higher AMMW are associated with lower BP, and the percentage of lean muscle mass (AMMW) is a good marker associated with BP (negative association) among the community-dwelling elderly population in Taiwan.

To the best of our knowledge, this is the first study toward establishing the relationship between body composition, especially AMM to weight ratio, and BP in the community-dwelling Taiwanese population. Ho