Trunk-to-Peripheral Fat Ratio Predicts Subsequent Blood Pressure Levels in Pubertal Children With Relatively Low Body Fat　– Three-Year Follow-up Study –

Abstract

Background: Only a few studies have examined the relationship between fat distribution measured by dual-energy X-ray absorptiometry (DXA) and blood pressure (BP), and no cohort study has targeted a pediatric population.

Methods and Results: The source population comprised all students registered as fifth graders in the 2 elementary schools in Hamamatsu, Japan. Of these, 258 children participated in both baseline (at age 11) and follow-up (at age 14) surveys. Body fat distribution was assessed using trunk-to-appendicular fat ratio (TAR) and trunk-to-leg fat ratio (TLR) measured by DXA. Relationships between BP levels and fat distribution (TAR or TLR) were examined after stratification by tertiles of whole-body fat. Systolic BP at follow-up was significantly (P<0.05) associated with both TAR (boys, β=0.33; girls β=0.36) and TLR (girls β=0.35) at baseline, after adjusting for confounding factors such as baseline BP in the lowest tertile of whole-body fat. Moreover, adjusted means of systolic and diastolic BPs in girls showed a significant increase from the lowest to highest tertile of TAR within the lowest tertile of whole-body fat.

Conclusions: Body fat distribution in childhood could predict subsequent BP levels in adolescence. Children with a relatively low body fat that is more centrally distributed tended to show relatively high BP later on. (Circ J 2016; 80: 1838–1845)

Obesity is generally defined as an abnormal accumulation of body fat and is associated with a high risk of developing cardiovascular disease.^1,2 It is a complex condition with a heterogeneous phenotype, and has puzzled clinicians in some cases.³ Obese individuals with the same amount of total body fat could present with markedly different risk factors.⁴ Such heterogeneity appears to be attributed to individual variation in regional body fat distribution.³ Based on this variation, obese individuals can be classified into 2 groups: (1) those with gluteal-femoral or peripheral fat accumulation, and (2) those with abdominal or central fat accumulation. The association between central body fat distribution and cardiovascular health risk has been previously suggested.⁵

Editorial p 1707

Individual variation in body fat distribution (body fat patterning) is generally expressed by the ratio of regional body fat accumulation. The ratio of waist-to-hip circumferences (WHR) has been used to determine the distribution pattern of body fat from the early 1980 s,^6,7 According to a recent systematic review and meta-analysis, WHR is a better discriminator than body mass index (BMI) of cardiovascular disease mortality risk.^8,9 However, waist and hip circumferences are both indirect indices of regional fat mass (FM), measuring both lean and adipose tissue. Therefore, WHR is also an indirect method of determining fat distribution.

Dual-energy X-ray absorptiometry (DXA) is a technique that enables direct determination of whole-body or regional (arm, leg, and trunk) FM.¹⁰ It is a non-invasive method requiring low radiation doses that is simple to perform, less dependent on operator skill and experience, and highly reproducible.¹¹ The radiation exposure from a whole-body scan is equal to or less than the accumulated daily natural background radiation dose to the body.^10,11 Therefore, DXA is particularly useful for evaluating body fat distribution because it provides an accurate measure of FM deposition. Various DXA-based fat distribution indices have been reported, including the android-to-gynoid FM ratio,^12,13 % regional FM in the trunk,¹⁴ trunk-to-leg FM ratio (TLR),^15–17 and trunk-to-peripheral FM ratio,^15,18,19 which is also referred to as the trunk-to-extremity FM ratio^20,21 or trunk-to-appendicular FM ratio (TAR).^22,23

Increasing evidence suggests that high blood pressure (BP) is a risk factor for hypertension in early adulthood.²⁴ A previous follow-up study found that childhood BP is a predictor of arterial stiffness in young adults.²⁵ To date, only a few epidemiological studies have examined the relationship between fat distribution measured by DXA and BP in adult populations.¹⁵ In pediatric populations, only 2 cross-sectional studies have been conducted,^22,26 and no cohort study has been reported. Accordingly, we conducted a 3-year follow-up study to examine whether childhood fat distribution measured by DXA (TAR or TLR) predicts BP levels in subsequent years.

Methods

Subjects

The source population consisted of all students registered as fifth graders in the 2 elementary schools in Hamamatsu, Japan (Aritama Elementary School and Sekishi Elementary School [521 children: 268 boys, 253 girls]), from November to December in 2010 and November to December in 2011.²² Of these, 413 children (205 boys, 208 girls) participated in the baseline survey, and 258 children (123 boys, 135 girls) participated in the follow-up survey in December 2013 or December 2014. The 258 children were analyzed in the present study.

All parents and guardians of the children received information regarding study procedures, including the dose of DXA radiation exposure, and provided written consent prior to the baseline and follow-up surveys. The children were also allowed to decline participation of their own accord. This study was approved by the Ethics Committee of the Kindai University Faculty of Medicine and performed in accordance with the ethical standards set out in the Declaration of Helsinki.

Measurement of Whole-Body and Regional FM

A single DXA scanner (QDR-4500A, Hologic Inc, Bedford, MA, USA) in a mobile test room was used to measure whole-body and regional FM for both the baseline and follow-up survey. Quality control of the DXA scanner, which included a rigorous phantom scanning schedule, was performed throughout the surveys. Subjects were asked to wear light clothing without any metal objects. A single experienced medical radiology technician performed the scans and also performed all scan analyses. Arm, leg, and head FM volumes were isolated from trunk FM in the anterior scan image using the standard manufacturer-recommended analysis, as follows:^10,15,22,23 (a) the head was delineated at the horizontal shoulder line just below the chin, (b) each arm was delineated at the vertical shoulder line that bisected the head of the humerus at the glenoid fossa, and (c) the leg region was separated by angled lines that defined a pelvic triangle and bisected both femoral necks.

Appendicular FM was calculated as the sum of arm and leg FM volumes. TAR was calculated as trunk FM divided by appendicular FM,²² and TLR as trunk FM divided by leg FM.^22,23 The FM index (FMI) was calculated as total body FM (kg) divided by height squared (m²) and used as an appropriate index of whole-body adiposity independent of overall body size.^27,28

BP Measurements

BP measurements were performed during the same session as body FM measurement. Systolic BP (SBP) and diastolic BP (DBP) were measured by an experienced physician using an automated device (BP-103i2, OMRON COLIN, Tokyo, Japan) in both surveys. Appropriate cuff size was selected based on the subject’s arm circumference.²⁴ Measurements were performed with the subject in a seated position after a 5-mine rest, with the left arm supported at the level of the heart, legs uncrossed, and in a quiet and appropriate environment at room temperature.²² The mean value of 2 measurements was used for analysis.

Other Variables

Body weight, height, and waist circumference were measured in light clothing with no shoes. BMI (kg/m²) was calculated as weight (kg) divided by height squared (m²). International BMI cut-offs for child overweight and child underweight (based on adult BMI cut-offs of 25 kg/m² and 18.5 kg/m², respectively)^29,30 were used to assess baseline and follow-up subject characteristics. The first appearance of pubic hair and sedentary behavior such as watching television, video game playing, mobile phone use, and computer use were determined based on self-reported questionnaire responses.

Statistical Analysis

Baseline and follow-up changes were evaluated using a paired t-test. An unpaired t-test was used to examine differences in subject characteristics by sex. Other statistical analyses were performed separately by sex. Simple regression analysis was used to examine the trend from the lowest to highest tertile of FMI. The dependent variable was each measurement, and the independent variable was a tertile class of FMI. Simple regression analysis was also used to evaluate relationships between BP and body fat parameters. After stratification by tertiles of FMI, multiple regression analysis was performed to detect relationships between follow-up BP and baseline body fat parameters after adjusting for BP, height, weight, sedentary behavior, and pubic hair appearance at baseline. Mean values of BP at follow-up according to tertiles of TAR/TLR within each tertile of FMI at baseline were calculated using the general linear model after adjusting for BP, height, weight, sedentary behavior, and pubic hair appearance at baseline. For trend tests of BP and tertiles of TAR, multiple linear regression analysis was performed after adjusting for BP, height, weight, sedentary behavior, and pubic hair appearance at baseline. The dependent variable was either SBP or DBP in each individual, and independent variables were potential confounding factors and tertiles of TAR within each tertile of FMI. P<0.05 was considered statistically significant. Data were analyzed using the SPSS Statistics Desktop for Japan, Version 22 (IBM Japan, Ltd, Tokyo, Japan).

Results

Baseline and follow-up subject characteristics are summarized in Table 1. TAR and TLR significantly increased over the period of 3 years (from age 11 to 14 years) in both sexes. The increments in TAR and TLR were significantly larger for boys than girls. DBP significantly decreased from age 11 to 14 in girls. Baseline characteristics stratified by tertiles of FMI are shown in Table 2. TAR and TLR significantly increased from the lowest to highest tertile of FMI in both sexes.

Table 1. Baseline and Follow-up Characteristics of the Children

	Baseline		Dropped out		Follow-up		Difference between ages 11 and 14a		Change from age 11 to 14 years
	Boys, n=123	Girls, n=135	Boys, n=82	Girls, n=73	Boys, n=123	Girls, n=135	Boys	Girls	Boys	Girls	Difference between boys and girlsb
Age (years)	11.2±0.3	11.2±0.3	11.2±0.3	11.2±0.3	14.2±0.3	14.2±0.3
Height (cm)	141±6	143±7	142±6	143±7	162±6	155±5			21±3	11±4	<0.001
Weight (kg)	34.6±7.4	35.3±6.0	35.6±7.5	35.5±8.7	49.9±8.4	47.0±5.8			15.3±3.8	11.8±3.8	<0.001
BMI (kg/m2)	17.2±2.7	17.1±1.9	17.6±2.8	17.1±2.9	18.9±2.4	19.6±2.0			1.7±1.3	2.6±1.4	<0.001
WC (cm)	63.0±7.8	62.6±5.5	63.8±8.1	62.6±7.8	68.1±6.2	68.5±5.1			5.2±3.5	5.9±4.0	NS
Overweightc, n (%)	13 (11)	6 (4)	13 (15)	8 (11)	12 (10)	6 (4)
Underweightc, n (%)	19 (15)	16 (12)	9 (11)	13 (18)	14 (11)	11 (8)
Pubic hair appearance, n (%)	8 (7)	47 (35)
TAR	0.52±0.10	0.55±0.09	0.53±0.10	0.53±0.08	0.65±0.12	0.59±0.08	<0.001	<0.001	0.14±0.12	0.05±0.07	<0.001
TLR	0.70±0.14	0.72±0.14	0.72±0.15	0.71±0.12	0.88±0.19	0.77±0.13	<0.001	<0.001	0.18±0.19	0.05±0.11	<0.001
Trunk FM (kg)	2.2±1.6	2.5±1.2	2.4±1.7	2.5±1.6	2.3±1.1	3.7±1.2	NS	<0.001	0.1±0.9	1.2±1.0	<0.001
Trunk FMI (kg/m2)	1.1±0.7	1.2±0.5	1.2±0.8	1.2±0.7	0.9±0.4	1.5±0.5	<0.001	<0.001	−0.2±0.5	0.3±0.4	<0.001
Appendicular FM (kg)	4.0±2.2	4.5±1.6	4.3±2.2	4.6±2.3	3.6±1.8	6.2±1.8	<0.001	<0.001	−0.5±1.1	1.7±1.2	<0.001
Appendicular FMI (kg/m2)	2.0±1.0	2.2±1.0	2.1±1.0	2.2±0.9	1.4±0.7	2.6±0.7	<0.001	<0.001	−0.6±0.6	0.4±0.6	<0.001
Leg FM (kg)	3.0±1.5	3.4±1.2	3.2±1.6	3.4±1.8	2.7±1.4	4.8±1.3	<0.001	<0.001	−0.3±0.8	1.4±0.9	<0.001
Leg FMI (kg/m2)	1.5±0.7	1.7±0.5	1.6±0.7	1.6±0.7	1.0±0.5	2.0±0.6	<0.001	<0.001	−0.5±0.4	0.3±0.4	<0.001
Total FM (kg)	7.0±3.7	7.8±2.7	7.6±3.9	7.8±4.0	6.8±3.0	10.7±2.9	NS	<0.001	−0.3±1.9	3.0±2.2	<0.001
Total FMI (kg/m2)	3.5±1.7	3.7±1.2	3.7±1.8	3.7±1.6	2.6±1.1	4.5±1.2	<0.001	<0.001	−0.9±1.0	0.7±1.0	<0.001
Total lean mass (kg)	27.5±4.2	27.4±3.8	28.0±4.3	27.6±5.1	41.8±5.9	35.0±3.7	<0.001	<0.001	14.2±3.1	7.6±2.5	<0.001
Total lean mass index (kg/m2)	13.8±1.2	13.3±1.0	13.9±1.3	13.3±1.4	15.8±1.5	14.6±1.2	<0.001	<0.001	2.1±0.8	1.3±0.7	<0.001
SBP (mmHg)	105±11	107±11	106±10	107±9	107±11	105±10	NS	NS	2±13	−1±11	0.024
DBP (mmHg)	58±8	60±8	59±7	59±8	57±7	58±7	NS	0.027	0±9	−2±8	NS

TAR calculated as trunk FM divided by appendicular FM. TLR calculated as trunk FM divided by leg FM. Values represent mean±standared deviation, or n (%). P<0.05 was considered statistically significant. ^aP value calculated from the paired t-test. ^bP value calculated from the unpaired t-test. ^cInternational cut-offs for BMI were used to determine overweight and underweight children. BMI, body mass index; DBP, diastolic blood pressure; FM, fat mass; FMI, fat mass index; NS, not significant; SBP, systolic blood pressure; TAR, trunk-to-appendicular fat ratio; TLR, trunk-to-leg fat ratio; WC, waist circumference.

Table 2. Baseline Characteristics of the Children Stratified by Tertiles of FMI

	Boys, n=123				Girls, n=135
	FMI T1, n=41	FMI T2, n=41	FMI T3, n=41	P for trend	FMI T1, n=45	FMI T2, n=45	FMI T3, n=45	P for trend
Age (years)	11.2±0.3	11.2±0.3	11.1±0.3	NS	11.2±0.3	11.1±0.3	11.1±0.3	NS
Height (cm)	139±6	141±6	143±6	0.002	142±8	143±6	145±6	0.035
Weight (kg)	29.3±3.2	33.1±4.6	41.4±7.8	<0.001	31.5±4.5	34.0±4.4	40.4±5.2	<0.001
BMI (kg/m2)	15.1±0.8	16.6±1.1	20.0±2.8	<0.001	15.5±1.0	16.5±1.1	19.1±1.5	<0.001
WC (cm)	57.7±2.5	60.6±3.4	70.6±8.5	<0.001	58.8±3.0	60.9±3.6	68.0±4.9	<0.001
TAR	0.50±0.09	0.48±0.07	0.57±0.10	<0.001	0.51±0.07	0.54±0.08	0.59±0.11	<0.001
TLR	0.68±0.13	0.64±0.09	0.78±0.16	<0.001	0.67±0.10	0.71±0.12	0.78±0.16	<0.001
Trunk FM (kg)	1.1±0.2	1.7±0.4	3.8±1.8	<0.001	1.6±0.3	2.2±0.4	3.8±1.1	<0.001
Trunk FMI (kg/m2)	0.6±0.1	0.8±0.1	1.8±0.8	<0.001	0.8±0.1	1.1±0.1	1.8±0.5	<0.001
Appendicular FM (kg)	2.2±0.4	3.5±0.7	6.4±2.0	<0.001	3.1±0.6	4.1±0.5	6.3±1.2	<0.001
Appendicular FMI (kg/m2)	1.1±0.2	1.7±0.3	3.1±0.9	<0.001	1.5±0.2	2.0±0.2	3.0±0.5	<0.001
Leg FM (kg)	1.7±0.3	2.6±0.5	4.7±1.4	<0.001	2.4±0.5	3.1±0.5	4.8±0.9	<0.001
Leg FMI (kg/m2)	0.9±0.2	1.3±0.2	2.3±0.6	<0.001	1.2±0.2	1.5±0.2	2.3±0.4	<0.001
Total FM (kg)	4.1±0.5	5.9±1.0	11.1±3.8	<0.001	5.4±0.9	7.1±0.8	10.9±2.2	<0.001
Total FMI (kg/m2)	2.1±0.2	3.0±0.4	5.4±1.7	<0.001	2.6±0.3	3.4±0.3	5.1±0.9	<0.001
Total lean mass (kg)	25.3±2.8	27.2±3.8	30.1±4.6	<0.001	26.2±3.7	26.8±3.6	29.4±3.5	<0.001
Total lean mass index (kg/m2)	13.0±0.8	13.7±0.9	14.6±1.4	<0.001	12.9±0.8	13.0±0.9	13.9±1.0	<0.001
SBP (mmHg)	103±9	104±11	109±11	0.009	102±10	107±10	110±11	<0.001
DBP (mmHg)	56±8	58±7	59±8	NS	60±7	59±9	61±8	NS

TAR calculated as trunk FM divided by appendicular FM. TLR calculated as trunk FM divided by leg FM. Values represent mean±standard deviation. Simple regression analysis was used to examine the trend from the lowest (T1) to highest (T3) tertile of FMI. The dependent variable was each measurement, and independent variable was a tertile class of FMI. P<0.05 was considered statistically significant. T, tertile. Other abbreviations as in Table 1.

Table 3 shows the relationships between body fat parameters at age 11 and BP at ages 11 and 14. TAR and TLR at age 11 were significantly and positively related to SBP and DBP at age 14 in boys. On the other hand, no significant relationships were observed between regional fat volumes at age 11 and BP levels at age 14.

Table 3. Relationships Between Body Fat Parameters at Baseline and BP at Baseline or Follow-up

	SBP at age 11		DBP at age 11		SBP at age 14		DBP at age 14
	β	P value	β	P value	β	P value	β	P value
Boys, n=123
TAR	0.362	<0.001	0.359	<0.001	0.336	<0.001	0.294	<0.001
TLR	0.353	<0.001	0.353	<0.001	0.290	0.001	0.261	0.004
Trunk FM (kg)	0.335	<0.001	0.211	0.019	0.137	NS	0.090	NS
Trunk FMI (kg/m2)	0.304	<0.001	0.203	0.024	0.138	NS	0.091	NS
Appendicular FM (kg)	0.317	<0.001	0.172	NS	0.078	NS	0.032	NS
Appendicular FMI (kg/m2)	0.272	0.002	0.153	NS	0.075	NS	0.029	NS
Leg FM (kg)	0.312	<0.001	0.162	NS	0.084	NS	0.035	NS
Leg FMI (kg/m2)	0.266	0.003	0.142	NS	0.081	NS	0.031	NS
Total FM (kg)	0.331	<0.001	0.193	0.032	0.105	NS	0.058	NS
Total FMI (kg/m2)	0.288	0.001	0.177	NS	0.102	NS	0.056	NS
Girls, n=135
TAR	0.103	NS	0.099	NS	0.132	NS	0.125	NS
TLR	0.098	NS	0.090	NS	0.094	NS	0.080	NS
Trunk FM (kg)	0.278	0.001	0.109	NS	0.043	NS	−0.032	NS
Trunk FMI (kg/m2)	0.259	0.002	0.090	NS	0.040	NS	−0.011	NS
Appendicular FM (kg)	0.332	<0.001	0.103	NS	0.039	NS	−0.060	NS
Appendicular FMI (kg/m2)	0.305	<0.001	0.069	NS	0.029	NS	−0.047	NS
Leg FM (kg)	0.331	<0.001	0.104	NS	0.056	NS	−0.041	NS
Leg FMI (kg/m2)	0.306	<0.001	0.069	NS	0.048	NS	−0.027	NS
Total FM (kg)	0.317	<0.001	0.112	NS	0.046	NS	−0.047	NS
Total FMI (kg/m2)	0.291	<0.001	0.081	NS	0.035	NS	−0.029	NS

TAR calculated as trunk FM divided by appendicular FM. TLR calculated as trunk FM divided by leg FM. Simple regression analysis was used to examine the relationships between BP at age 11 or 14 (dependent variable) and body fat variables at age 11 (independent variables). P<0.05 was considered statistically significant. Abbreviations as in Table 1.

Table 4 shows the relationships between BP levels at age 14 and body fat variables at age 11 stratified by whole-body fat. In the lowest tertile of FMI (FMI T1), TAR were significantly and positively associated with SBP in both sexes after adjusting for confounding factors including BP at age 11. In addition, TAR was significantly and positively related to DBP in girls. Table 5 shows the relationships between changes in BP from age 11 to age 14 and body fat variables at age 11 stratified by whole-body fat. In the lowest tertile of FMI (FMI T1), TAR was significantly and positively associated with changes in SBP and DBP in girls after adjusting for confounding factors.

Table 4. Relationships Between BP at Follow-up and Body Fat Variables at Baseline Stratified by Whole-Body Fat

Whole-body fat	Body fat variables at age 11 (baseline)	Boys, n=123		Girls, n=135
		β	P value	β	P value
SBP at age 14 (follow-up)
FMI T1	Trunk fat	0.324	NS	−0.067	NS
	Appendicular fat	−0.182	NS	−0.533	0.002
	TAR	0.333	0.031	0.362	0.006
	TLR	0.292	NS	0.347	0.007
FMI T2	Trunk fat	0.404	NS	−0.112	NS
	Appendicular fat	−0.106	NS	−0.322	NS
	TAR	0.349	NS	0.136	NS
	TLR	0.350	NS	0.033	NS
FMI T3	Trunk fat	0.962	0.013	−0.179	NS
	Appendicular fat	0.678	NS	−0.195	NS
	TAR	0.361	NS	−0.018	NS
	TLR	0.282	NS	−0.052	NS
DBP at age 14 (follow-up)
FMI T1	Trunk fat	0.281	NS	0.000	NS
	Appendicular fat	−0.019	NS	−0.541	0.005
	TAR	0.222	NS	0.409	0.005
	TLR	0.181	NS	0.362	0.012
FMI T2	Trunk fat	0.262	NS	−0.014	NS
	Appendicular fat	−0.154	NS	−0.123	NS
	TAR	0.303	NS	0.086	NS
	TLR	0.318	NS	−0.013	NS
FMI T3	Trunk fat	0.813	0.035	−0.071	NS
	Appendicular fat	0.540	NS	−0.051	NS
	TAR	0.388	NS	0.027	NS
	TLR	0.345	NS	−0.003	NS

TAR calculated as trunk FM divided by appendicular FM. TLR calculated as trunk FM divided by leg FM. FMI calculated as total FM divided by height squared. Multiple regression analysis was used to examine the relationships between BP at age 14 (dependent variable) and BP, height, weight, sedentary behavior, pubic hair appearance, and each of the body fat variables (trunk fat, appendicular fat, TAR or TLR) at baseline (independent variables). P<0.05 was considered statistically significant. Abbreviations as in Tables 1,2.

Table 5. Relationships Between Change in BP From Baseline to Follow-up and Body Fat Variables at Baseline Stratified by Whole-Body Fat

Whole-body fat	Body fat variables at age 11 (baseline)	Boys, n=123		Girls, n=135
		β	P value	β	P value
Change in SBP from 11 to 14 years
FMI T1	Trunk fat	0.346	NS	−0.071	NS
	Appendicular fat	−0.175	NS	−0.510	0.012
	TAR	0.284	NS	0.337	0.028
	TLR	0.249	NS	0.317	0.037
FMI T2	Trunk fat	0.150	NS	0.000	NS
	Appendicular fat	0.041	NS	−0.184	NS
	TAR	0.011	NS	0.124	NS
	TLR	−0.001	NS	0.066	NS
FMI T3	Trunk fat	0.975	0.007	−0.154	NS
	Appendicular fat	0.666	NS	−0.412	NS
	TAR	0.365	NS	0.108	NS
	TLR	0.295	NS	0.076	NS
Change in DBP from 11 to 14 years
FMI T1	Trunk fat	0.238	NS	−0.068	NS
	Appendicular fat	0.226	NS	−0.477	0.018
	TAR	−0.098	NS	0.302	0.047
	TLR	−0.122	NS	0.272	NS
FMI T2	Trunk fat	0.127	NS	−0.006	NS
	Appendicular fat	0.219	NS	0.092	NS
	TAR	−0.120	NS	−0.046	NS
	TLR	−0.128	NS	−0.079	NS
FMI T3	Trunk fat	0.597	NS	−0.091	NS
	Appendicular fat	0.222	NS	−0.085	NS
	TAR	0.363	NS	0.012	NS
	TLR	0.300	NS	−0.015	NS

TAR calculated as trunk FM divided by appendicular FM. TLR calculated as trunk FM divided by leg FM. FMI calculated as total FM divided by height squared. Multiple regression analysis was used to examine the relationships between BP (dependent variable) and height, weight, sedentary behavior, pubic hair appearance, and each of the body fat variables (trunk fat, appendicular fat, TAR or TLR) at baseline (independent variables). P<0.05 was considered statistically significant. Abbreviations as in Tables 1,2.

The Figure shows the adjusted means of BP at age 14 according to tertiles of TAR within each tertile of FMI at age 11. In the lowest tertile of FMI, SBP showed a significant increase from the lowest to highest tertile of TAR in girls after adjusting for confounding factors, including BP at age 11. In the lowest tertile of FMI, DBP also showed a significant increase from the lowest to highest tertile of TAR in girls.

Figure.

Adjusted means and standard errors of blood pressure (BP) at age 14 according to tertiles (T1, T2, T3) of TAR within each tertile class of FMI at age 11 years. Mean values were calculated from the general linear model after adjusting for potential confounding factors, including BP, height, weight, sedentary behavior, and pubic hair appearance at age 11. The arrows show significant (P<0.05) trends. For trend tests of BP, multiple linear regression analysis was performed after adjusting for potential confounding factors. The dependent variable was either systolic BP (SBP) or diastolic BP (DBP), and independent variables were tertiles of TAR within each tertile of FMI and potential confounding factors. TAR, trunk-to-appendicular fat ratio; FMI, fat mass index.

Discussion

Our results demonstrate that body fat distribution measured by DXA (TAR or TLR) at age 11 could predict BP levels at age 14 after adjusting for confounding factors, including baseline BP. TAR and TLR were significantly and positively related to subsequent BP change, particularly in the lowest tertile of FMI (ie, relatively low-weight children). Children with relatively low body fat with a more centralized distribution tended to have relatively high levels of BP later on. A large volume of body fat (or trunk fat) may not be a prerequisite for the development of high BP levels in childhood. Body fat distribution provides additional insight into the evaluation of body fat, and is a meaningful measure, particularly for children who are relatively underweight.

In a cross-sectional study of black and white children aged 9–17 years, greater central fat deposition (an android fat pattern) was found to be an important predictor of less favorable plasma lipid and lipoprotein concentrations as well as BP.²⁶ Another cross-sectional study of Japanese children aged 11 years, which served as the baseline analysis of the present follow-up study, also reported that an excessive proportion of trunk fat is related to high BP levels, and that the relationship between fat distribution and BP was independent of the relationship between fat volume and BP.²² Consistent with those findings, the present 3-year follow-up study demonstrated that the proportion of trunk fat at age 11 could predict BP levels at age 14, after adjusting for baseline BP.

Although the biochemical background underlying the relationship between fat distribution and BP levels was not explored in the present study, a recent epidemiological study examined the relationship between increased TAR (ie, more centralized fat distribution) and decreased serum adiponectin levels.²³ Subcutaneous and visceral adipose tissues are hormonally active and secrete hormones and cytokines, such as adiponectin.³¹ Adiponectin is currently considered a beneficial hormone that has antiatherogenic properties at the cellular level.³² Cellular and molecular mechanisms connecting hypertension and hypoadiponectinemia have also been suggested.³³ A recent study reported that the release of omental adipocyte adiponectin (found in visceral adipocytes) is reduced to a greater extent compared with subcutaneous adipocyte adiponectin in women with visceral obesity.³⁴ Lower adiponectin levels derived from increasing TAR²³ might be an interpretation of the relationship between TAR and BP.

Study Limitations

First, follow-up data were obtained from 49.5% of the source population and 62.5% of the baseline population. The reason for the high drop-out rate is unclear. Loss of subjects to follow-up may lead to selection bias, and may have affected the results of the present study. However, there were no marked differences in characteristics between those who participated and those who dropped out (Table 1). Second, there might be selection bias in the present study, as subjects were enrolled from 2 elementary schools in 2 city in Japan, rather than from the entire country. However, the anthropometric values of the subjects in the follow-up survey were similar to those reported in a Japanese national survey (boys and girls at a mean age of 14.2 years; mean height, 162.6 cm and 155.8 cm; mean weight, 52.5 kg and 49.3 kg, respectively).³⁵ Third, the association between fat distribution and BP might be confounded by factors other than age, sex, height, weight, puberty, and sedentary behavior, such as dietary intake and family history, which were not assessed in this study.

In conclusion, childhood body fat distribution measured by DXA (TAR or TLR) could predict subsequent BP in adolescence, and children who have relatively low body fat distributed more centrally tended to show relatively high BP later on. Our findings suggest that TAR is a meaningful parameter in the evaluation of body fat, particularly in children who are relatively underweight.

Declaration of Interests

The authors have no conflicts of interest to declare.

Acknowledgments

The authors thank the teaching staff of Aritama Elementary School, Sekishi Elementary School, and Sekishi Junior High School, and Dr Toshiko Okamoto for their support.

Grants

Grants-in-Aid for Scientific Research (KAKENHI Grant Numbers 21657068, 22370092, 24370101, 26291100) from the Japan Society for the Promotion of Science.

References

• 1.    Yatsuya  H,  Li  Y,  Hilawe  EH,  Ota  A,  Wang  C,  Chiang  C, et al. Global trend in overweight and obesity and its association with cardiovascular disease incidence. Circ J 2014; 78: 2807–2818.
• 2.    Huxley  RR,  Hirakawa  Y,  Hussain  MA,  Aekplakorn  W,  Wang  X,  Peters  SA, et al. Age- and sex-specific burden of cardiovascular disease attributable to 5 major and modifiable risk factors in 10 Asian countries of the Western Pacific Region. Circ J 2015; 79: 1662–1674.
• 3.    Despres  JP. Body fat distribution and risk of cardiovascular disease: An update. Circulation 2012; 126: 1301–1313.
• 4.    Despres  JP,  Moorjani  S,  Lupien  PJ,  Tremblay  A,  Nadeau  A,  Bouchard  C. Regional distribution of body fat, plasma lipoproteins, and cardiovascular disease. Arteriosclerosis 1990; 10: 497–511.
• 5.    Pi-Sunyer  FX. The epidemiology of central fat distribution in relation to disease. Nutr Rev 2004; 62: S120–S126.
• 6.    Krotkiewski  M,  Bjorntorp  P,  Sjostrom  L,  Smith  U. Impact of obesity on metabolism in men and women: Importance of regional adipose tissue distribution. J Clin Invest 1983; 72: 1150–1162.
• 7.    Larsson  B,  Svardsudd  K,  Welin  L,  Wilhelmsen  L,  Bjorntorp  P,  Tibblin  G. Abdominal adipose tissue distribution, obesity, and risk of cardiovascular disease and death: 13 year follow up of participants in the study of men born in 1913. Br Med J (Clin Res Ed) 1984; 288: 1401–1404.
• 8.    Czernichow  S,  Kengne  AP,  Stamatakis  E,  Hamer  M,  Batty  GD. Body mass index, waist circumference and waist-hip ratio: Which is the better discriminator of cardiovascular disease mortality risk? Evidence from an individual-participant meta-analysis of 82 864 participants from nine cohort studies. Obes Rev 2011; 12: 680–687.
• 9.    Coutinho  T,  Goel  K,  Correa de Sa  D,  Kragelund  C,  Kanaya  AM,  Zeller  M, et al. Central obesity and survival in subjects with coronary artery disease: A systematic review of the literature and collaborative analysis with individual subject data. J Am Coll Cardiol 2011; 57: 1877–1886.
• 10.    Laskey  MA. Dual-energy X-ray absorptiometry and body composition. Nutrition 1996; 12: 45–51.
• 11.    Albanese  CV,  Diessel  E,  Genant  HK. Clinical applications of body composition measurements using DXA. J Clin Densitom 2003; 6: 75–85.
• 12.    Walton  C,  Lees  B,  Crook  D,  Worthington  M,  Godsland  IF,  Stevenson  JC. Body fat distribution, rather than overall adiposity, influences serum lipids and lipoproteins in healthy men independently of age. Am J Med 1995; 99: 459–464.
• 13.    Niederauer  CM,  Binkley  TL,  Specker  BL. Effect of truncal adiposity on plasma lipid and lipoprotein concentrations. J Nutr Health Aging 2006; 10: 154–160.
• 14.    Demura  S,  Sato  S. Nonlinear relationships between visceral fat area and percent regional fat mass in the trunk and the lower limbs in Japanese adults. Eur J Clin Nutr 2008; 62: 1395–1404.
• 15.    Peppa  M,  Koliaki  C,  Hadjidakis  DI,  Garoflos  E,  Papaefstathiou  A,  Katsilambros  N, et al. Regional fat distribution and cardiometabolic risk in healthy postmenopausal women. Eur J Intern Med 2013; 24: 824–831.
• 16.    Aasen  G,  Fagertun  H,  Tonstad  S,  Halse  J. Leg fat mass as measured by dual X-ray absorptiometry (DXA) impacts insulin resistance differently in obese women versus men. Scand J Clin Lab Invest 2009; 69: 181–189.
• 17.    Zillikens  MC,  Uitterlinden  AG,  van Leeuwen  JP,  Berends  AL,  Henneman  P,  van Dijk  KW, et al. The role of body mass index, insulin, and adiponectin in the relation between fat distribution and bone mineral density. Calcif Tissue Int 2010; 86: 116–125.
• 18.    Novotny  R,  Daida  YG,  Grove  JS,  Le Marchand  L,  Vijayadeva  V. Asian adolescents have a higher trunk: Peripheral fat ratio than Whites. J Nutr 2006; 136: 642–647.
• 19.    Morimoto  Y,  Maskarinec  G,  Conroy  SM,  Lim  U,  Shepherd  J,  Novotny  R. Asian ethnicity is associated with a higher trunk/periphery fat ratio in women and adolescent girls. J Epidemiol 2012; 22: 130–135.
• 20.    Savgan-Gurol  E,  Bredella  M,  Russell  M,  Mendes  N,  Klibanski  A,  Misra  M. Waist to hip ratio and trunk to extremity fat (DXA) are better surrogates for IMCL and for visceral fat respectively than for subcutaneous fat in adolescent girls. Nutr Metab (Lond) 2010; 7: 86.
• 21.    Ackerman  KE,  Davis  B,  Jacoby  L,  Misra  M. DXA surrogates for visceral fat are inversely associated with bone density measures in adolescent athletes with menstrual dysfunction. J Pediatr Endocrinol Metab 2011; 24: 497–504.
• 22.    Kouda  K,  Nakamura  H,  Fujita  Y,  Ohara  K,  Iki  M. Increased ratio of trunk to appendicular fat and increased blood pressure: Study of a general population of Hamamatsu children. Circ J 2012; 76: 2848–2854.
• 23.    Kouda  K,  Nakamura  H,  Ohara  K,  Fujita  Y,  Iki  M. Increased ratio of trunk-to-appendicular fat and decreased adiponectin: A population-based study of school children in Hamamatsu, Japan. J Clin Densitom 2015 December 2, doi:10.1016/j.jocd.2015.10.004.
• 24.   National High Blood Pressure Education Program Working Group on High Blood Pressure in Children, Adolescents. The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents. Pediatrics 2004; 114: 555–576.
• 25.    Li  S,  Chen  W,  Srinivasan  SR,  Berenson  GS. Childhood blood pressure as a predictor of arterial stiffness in young adults: The bogalusa heart study. Hypertension 2004; 43: 541–546.
• 26.    Daniels  SR,  Morrison  JA,  Sprecher  DL,  Khoury  P,  Kimball  TR. Association of body fat distribution and cardiovascular risk factors in children and adolescents. Circulation 1999; 99: 541–545.
• 27.    VanItallie  TB,  Yang  MU,  Heymsfield  SB,  Funk  RC,  Boileau  RA. Height-normalized indices of the body’s fat-free mass and fat mass: Potentially useful indicators of nutritional status. Am J Clin Nutr 1990; 52: 953–959.
• 28.    Weber  DR,  Moore  RH,  Leonard  MB,  Zemel  BS. Reply to RF Burton. Am J Clin Nutr 2013; 98: 1368–1369.
• 29.    Cole  TJ,  Bellizzi  MC,  Flegal  KM,  Dietz  WH. Establishing a standard definition for child overweight and obesity worldwide: International survey. BMJ 2000; 320: 1240–1243.
• 30.    Cole  TJ,  Flegal  KM,  Nicholls  D,  Jackson  AA. Body mass index cut offs to define thinness in children and adolescents: International survey. BMJ 2007; 335: 194.
• 31.    Scherer  PE,  Williams  S,  Fogliano  M,  Baldini  G,  Lodish  HF. A novel serum protein similar to C1q, produced exclusively in adipocytes. J Biol Chem 1995; 270: 26746–26749.
• 32.    Balsan  GA,  Vieira  JL,  Oliveira  AM,  Portal  VL. Relationship between adiponectin, obesity and insulin resistance. Rev Assoc Med Bras 2015; 61: 72–80.
• 33.    Rojas  E,  Rodriguez-Molina  D,  Bolli  P,  Israili  ZH,  Faria  J,  Fidilio  E, et al. The role of adiponectin in endothelial dysfunction and hypertension. Curr Hypertens Rep 2014; 16: 463.
• 34.    Drolet  R,  Belanger  C,  Fortier  M,  Huot  C,  Mailloux  J,  Legare  D, et al. Fat depot-specific impact of visceral obesity on adipocyte adiponectin release in women. Obesity (Silver Spring) 2009; 17: 424–430.
• 35.    Suwa  S,  Tachibana  K. Standard growth charts for height and weight of Japanese children from birth to 17 years based on cross-sectional survey of national data. Clin Pediatr Endocrinol 1993; 2: 87–97.