Aging Clinical and Experimental Research

Relationship between the abdominal wall fat index and blood pressure in elderly women: a comparison with the body mass index Kazuhiko Kotani1,2,3, Seiji Adachi3,4, Kokoro Tsuzaki1 and Naoki Sakane1 1Division

of Preventive Medicine, Clinical Research Institute for Endocrine and Metabolic Disease, National Hospital Organization Kyoto Medical Center, Kyoto, 2Department of Clinical Laboratory Medicine, Jichi Medical University, Tochigi, 3Tottori University Faculty of Medicine, Yonago, 4Division of General Practice and Community Medicine, Fujii Masao Memorial Hospital, Kurayoshi, Japan

ABSTRACT. Background and aims: Several studies have examined the associations between measurements used to assess obesity, such as the body mass index (BMI), waist circumference (WC) and intra-abdominal fat area, and cardiometabolic abnormalities. However, the application of these measures in clinical practice requires more detailed examination in older individuals. The abdominal wall fat index (AFI) is ultrasonographically determined via a vertical scan along the upper abdominal median, to measure the maximum thickness of pre-peritoneal fat at the liver surface and the minimum thickness of subcutaneous fat. Few w studstudth thee B MI as ies, however, have compared the AFI with BMI a measure of obesity. Older women en were wer eree examined exami xamine ned d to determine the associations carns among BM BMI, MI, AFI aand nd car diometabolic variables. 86 asymptomatbles. Methods: ble Meth Me thods: s In 8 6 asymp as SD D wiith BMIs MIs of 18.518.5-29 29 kg k g/m2 ((mean age±SD ic women with kg/m ) w wee me mea meaa7 6 years; yea s; mean mea ean n BMI±SD BMI± MI SD 22. 2 7 7± 77±6 22.7±2.5 kg/m2), sured the ffollowing ollowi owing cardiometabolic cardi variab vari ables: les: systo systol lic sured variables: systolic blood pr pressur ess e (SB ((SBP), diastolicc bl blo ood pre press ssure ure (D blood pressure blood pressure (DBP), press sma gl lucose, serum seru total cholespulse pressure (PP), plasma glucose, terol, triglycerides, dees, aand nd hig highhigh-density h densi en t lipoprotein choless: In a mu multi lti terol. Results: multiple regression analysis adth above cardiometabolic variables, justed for all the BMI showed a significant negative correlation with age alone, whereas AFI showed a significant positive correlation with DBP and PP. Conclusions: Our results suggest that, compared with BMI, AFI may be useful in identifying blood pressure-related abnormalities, which represent an atherosclerotic risk in older Japanese women. (Aging Clin Exp Res 2009; 21: 349-352) ©2009,

Editrice Kurtis

INTRODUCTION Obesity is a major risk factor for cardiometabolic disease (1, 2). More specifically, the type of obesity is rmali thought to be an important determinant of abnormalities see m e ab et abolis olis in blood pressure (BP) and lipid/glucose metabolism. era rall o besity, be sity, rrather ath the Moreover, recently, abdominal or visce visceral obesity, e repo een eporrted to o be closely closel aasthan overall obesity, has bbeen reported se ca cardi rdiome mettabolicc ddisorders isorder (1, 2). sociated with these cardiometabolic istribution ch chan angges with age ((e.g., greater deBody fat distribution changes sit i ion in th he abdo abd min nal cavity) position the abdominal cavity), and research into such aage-related ge-rrellat geated ed changes cha hang nges e is re tee understand oberequired to better ssity-related ity-rel ity-r elated ated health h lth risks in the elderly hea derlly (3). der (3) 3).. Currently, C it is debbated atedd whether wh dy m ass in ind dex (BMI) (BM is an appro(B the body mass index pria measuree of obesity obe besit sityy in older old people (3-5), and priate ns hav havee bbeen een rais ais d regarding the relationship questions raised bbetween etween en measurements meaasure me applied to assess obesity and car arddiomet etab ab cardiometabolic abnormalities. M Measurements used to assess obesity include the BMI (the standard measure of overall obesity), waist circumference (WC; a more sensitive measure of abdominal obesity) and intra-abdominal fat area as determined by computed tomography (CT). BMI and WC are advantageous in that they are convenient even in primary health care settings, and BMI is also useful as an epidemiologyfriendly measure (6). However, WC is not always accurate (7, 8) and CT has several drawbacks, including exposure to radiation, lack of ease and simplicity, and cost (7). However, measurements confirmed by ultrasonography, termed the abdominal wall fat index (AFI) (9-11), may overcome some of these disadvantages. To date, AFI has proven useful in the evaluation of some metabolic abnormalities and atherosclerotic risk factors (9-11). BMI and AFI commonly include a subcu-

Key words: Blood pressure, geriatric practice, hypertension, obesity, pulse pressure, visceral fat. Correspondence: Kazuhiko Kotani, MD, PhD, Division of Preventive Medicine, Clinical Research Institute for Endocrine and Metabolic Disease, National Hospital Organization Kyoto Medical Center, 1-1 Mukaihata-cho, Fushimi-ku, Kyoto 612-8555, Japan. E-mail: [email protected] Received July 3, 2008; accepted in revised form October 21, 2008.

Aging Clin Exp Res, Vol. 21, No. 4/5 349

K. Kotani, S. Adachi, K. Tsuzaki et al.

taneous fat element, so that the comparison is of interest. Although research has compared BMI with WC in assessing cardiometabolic variables (12-14), few studies have compared the relevance of AFI or BMI with an indicator of cardiometabolic state, among older people in specific. AFI may be more indicative of age-related obesity and its pathophysiology than BMI. Our aim was to investigate the relationship between cardiometabolic variables and AFI in comparison with BMI in older Japanese women. SUBJECTS AND METHODS Altogether, 86 Japanese women (mean age±standard deviation [SD] 77±6 years; range 65-87 years) were consecutively recruited from among outpatients and community-living volunteers during routine health check-ups. This study was approved by the institutional ethics committees and all subjects provided their informed consent. Eligible subjects were asymptomatic, without any features of ischemic heart, cerebrovascular, kidney, liver, cognitive or malignant disease. Interviews were conducted to confirm that all subjects were non-smokers and nonsedentary (not home-bound every day). In terms of nutritional disorders, subjects with BMI of 18.5-29 kg/m2 and without a significant body weight change (±5%) in the king ng previous year were enrolled. Subjects currently taking etabolism abolism drugs known to affect BP and lipid/glucose met metabolism ssume ss um the th were excluded. Those who were not able to aassume supine position (e.g., due to severe osteoporosis) evere osteo eop poro rosis) sis) were wer allow win ingg cardiometabo etabolic lic variab so excluded. The following cardiometabolic variables sub ubj bjec ect in the morn were measuredd in each subject morning after an g t fast: gh fas ast: t: seated seated syst systo olic blood bloo ood pressure (SBP), SBP), SBP ), didiovernight systolic astolic pressure (measured asto as to olic ic bblood lood p res ure (DBP) ress (D ( using a standard stan sta ndard sphy sph ygm momanomet anom er) plasma glucose, ose,, serum ose serum total cholescho chol sphygmomanometer), tterol (TC) TC) and triglyceridess (TG) G) (m (mea easured sured by enzymatic (measured d and serum rum um high high-dens ensity ity lipoprotein li methods), high-density cholesterol (HDL-C; ddetermined etermined using usin a homogeneous method). (PP)) w Pulse pressure (PP was also determined by subtracting DBP from SBP. Ultrasonographic evaluation with a 7.5-MHz linear type B-mode probe was performed by a single specialist to determine AFI, as described previously (10). The coefficient of variation for intraobserver reproducibility in our preliminary study was 10.1% (data not shown). As the subject rested in the supine position, the probe was held perpendicular to the skin at the epigastrium, and a vertical scan was performed along the abdominal median from the xiphoid process to the umbilicus, to measure the maximum thickness of pre-peritoneal fat at the anterior surface of the liver and the minimum thickness of the subcutaneous fat of the abdomen. The ratio of these two values was calculated to determine AFI. Data are expressed as mean ± SD. To determine the relationships among BMI, AFI and other cardiometabolic variables, Pearson’s rank coefficient test and multiple re-

350 Aging Clin Exp Res, Vol. 21, No. 4/5

gression analysis, adjusted for all measured cardiometabolic variables (SBP, DBP, PP, TC, TG, HDL-C, glucose), were used. Because of the skewed distribution, TG values were log-transformed. A p<0.05 was considered significant. RESULTS The clinical characteristics of our study population are listed in Table 1. In the simple correlation test, BMI had a significant positive correlation with AFI (r=0.30, p=0.005), and a significant negative correlation with age (r=−0.31, p=0.004). In contrast, AFI was not correlated with any variable. In the multiple regression analysis (Table 2), BMI showed a significant correlation with age alone (in Model 1, SBP and DBP were considered; in Model 2, PP was substituted for SBP and DBP). In contrast, AFI had a significant positive correlation with DBP in Model 1 andd w with PP in Model 2. DISCUSSION Our analysis in older der Jap Japan Japanese nese women wom rrevealed ev evealed that ficantly corr orreela late tedd wit with h DBP DBP B aand PP among AFI was significantly correlated ardiomet ar metab abo olic va i our measuredd ccardiometabolic variables, whereas BMI sh howe ow d n o ssignificant ignific ignif icant a relationship. relation showed no Substances that induce ccardiometabolic ardi ar diome omettab abollic i abnormalities, abn uch as iinsulin resistance such and iincreases an n rease in BP, are pr nc produ oduce cedd ffrom rom visceral fat (15produced 1 17 mor re aappropriate ppro ropr pria iatt measure of fat dis17),) so AFI may bee a mo more han BMI BMI,, wh which ich may partly explain our retribution th than sults. ts. A Also, lso ls o, o our ur re res results sult ultss support the claim that BMI is not alw ays a rreliable ay eliiab el abll indicator of obesity and its related patholoways gie gi es aamong older people, most probably due to body fat gies redistribution with age (4, 5). In addition, because PP is an indicator of the widespread arterial stiffening often observed in older people (18-20), the positive relationship between AFI and PP suggests that AFI is a better indica-

Table 1 - Clinical characteristics of study population. Variables Age (years) BMI (kg/m2) AFI Total cholesterol (mmol/L) Triglycerides (mmol/L) HDL cholesterol (mmol/L) Glucose (mmol/L) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Pulse pressure (mmHg)

Mean ± standard deviation (range) 77±6 (65-87) 22.7±2.5 (18.5-29) 0.7±0.3 (0.2-1.5) 5.3±0.8 (3.1-7.7) 1.3±0.6 (0.5-3.0) 1.4±0.4 (0.8-2.6) 5.6±0.9 (3.9-8.2) 142.8±13.4 (102-162) 79.1±7.6 (60-98) 63.7±13.2 (24-98)

BMI: body mass index; AFI: abdominal wall fat index; HDL: high-density lipoprotein.

AFI and blood pressure in older women

Table 2 - Multiple regression analysis for BMI or AFI against cardiometabolic variables. Variables

BMI

Model 1 (considering systolic and diastolic blood pressure) Age (years) −0.298 (0.006**) Total cholesterol (mmol/L) 0.083 (0.480) Log-triglycerides (mmol/L) −0.263 (0.134) HDL cholesterol (mmol/L) 0.306 (0.077) Glucose (mmol/L) 0.140 (0.266) Systolic blood pressure (mmHg) 0.132 (0.235) Diastolic blood pressure (mmHg) 0.137 (0.216) Model 2 (pulse pressure substituted for systolic and diastolic blood pressure) Age (years) −0.318 (0.004**) Total cholesterol (mmol/L) 0.103 (0.390) Log-triglycerides (mmol/L) −0.255 (0.153) HDL cholesterol (mmol/L) 0.292 (0.097) Glucose (mmol/L) 0.132 (0.303) Pulse pressure (mmHg) 0.078 (0.472)

AFI

−0.174 (0.122) −0.075 (0.543) 0.203 (0.271) −0.039 (0.830) −0.147 (0.267) 0.200 (0.090) −0.243 (0.039*) −0.161 (0.150) −0.088 (0.480) 0.198 (0.284) −0.030 (0.869) −0.142 (0.285) 0.229 (0.043*) (0.04

s i t r

BMI: body mass index; AFI: abdominal wall fat index; HDL: high-density lipoprotein. Data represent multiple regression coefficient (β, β, adjusted forr all all above abo cardiometabolic variables), followed by p-value in brackets. In Model 1 and Model 2, r2 = 0.2 for BMI and 0.1 for AFI. *p<0.05, 05, **p **p<0.01. p<0 <0.0 .01 1.

tor of vascular aging and increased atherosclerotic risk than BMI in older women. In primary health care settings and geriatric practice,, the he treatment of hypertension is considered an important port r ant target targ tar rget to prevent atherosclerotic disease, especially pec eciial ally ly in i Japan Japan (21(2121 23). In this case, convenient and appropriate app propr ropria iate te measuremeas me asurements for obesity assessment essm smen entt are required. requuir ired ed. Our Our results r demonstrate that which a AFI at FI aassessment ssessm sse ssment ent by uultrasonography, ltrrasonog is a rapid, api p d, non-in non-invas non-invasive, vasive, ve, and wide widely ly available procedure, roc oced edure ure,, would provide appropriate wou wo ud p uld rovidee an appr rovid ppropr o iate measure of obesity. obeesi sitty. Our however, First, O urr sstudy, tuddy, h owev ow w ver had some limitations. mitatio mit tations ns.. F irst, on only ly older women men e were wer analyzed, and nd further fuurtther investig investigation atio is required equal qually uuseful seful in sefu ind d to determine if AFI iss an eq equally indicator in men. The cross-sectional desi precluded causal inferences. Aldesign gn als also op re l rec though comparison with rison ris o w it waist-related measures, such as ith WC, may provide further useful information, these measures were not available in the present study. More detailed data on weight changes may also be helpful, because subjects with chronic malignant disease, which can affect AFI values, may have been included in the present short-term observations. Prospective studies with larger samples among various populations and with additional markers, including waist- and weight-related measures, are required. CONCLUSIONS Our results suggest that, compared with BMI, AFI is more useful in identifying cardiometabolic abnormalities such as increases in DBP and PP, which are atherosclerotic risk factors, in older Japanese women. Obesity assessment based on AFI is also easily applicable in geriatric practice. Further investigation is needed to confirm our findings and clarify the biological mechanism.

ACKNOWLEDGEMENTS CKNOWLE WLEDG D EME MEN NTS T This T h study his study was partly par ly supported part suppor by a grant-in-aid from the Founddation ation on for fo or the the Developm Development of the Community, mmuni munity, ty Japan. We have no co con conflicts fllict ictss off inter interest to declare.

REFERENCES NC CES 1. Reaven eav aven v n GM. Role Role of insulin insul resistance in human disease. Diabetes 1988 88;; 37: 37: 1595-607. 1595 15 95 1988; 2. Ca C labro b P, Yeh ET. Intra-abdominal adiposity, inflammation, Calabro and cardiovascular risk: new insight into global cardiometabolic risk. Curr Hypertens Rep 2008; 10: 32-8. 3. Zamboni M, Mazzali G, Zoico E et al. Health consequences of obesity in the elderly: a review of four unsolved questions. Int J Obes 2005; 29: 1011-29. 4. Baumgartner RN, Heymsfield SB, Roche AF. Human body composition and the epidemiology of chronic disease. Obes Res 1995; 3: 73-95. 5. Seidell JC, Visscher TL. Body weight and weight change and their health implications for the elderly. Eur J Clin Nutr 2000; 54: S33-9. 6. Mahabir S, Baer DJ, Johnson LL et al. Usefulness of body mass index as a sufficient adiposity measurement for sex hormone concentration associations in postmenopausal women. Cancer Epidemiol Biomarkers Prev 2006; 15: 2502-7. 7. Depres JP, Prud’homme D, Pouliot MC, Tremblay A, Bouchard C. Estimation of deep abdominal adipose-tissue accumulation from simple anthropometric measurements in men. Am J Clin Nutr 1991; 54: 471-7. 8. Leite CC, Wajchenberg BL, Radominski R, Matsuda D, Cerri GG, Halpern A. Intra-abdominal thickness by ultrasonography to predict risk factors for cardiovascular disease and its correlation with anthropometric measurements. Metabolism 2002; 51: 1034-40. 9. Suzuki R, Watanabe S, Hirai Y et al. Abdominal wall fat index, estimated by ultrasonography, for assessment of the ratio of visceral fat to subcutaneous fat in the abdomen. Am J Med 1993; 95: 309-14.

Aging Clin Exp Res, Vol. 21, No. 4/5 351

K. Kotani, S. Adachi, K. Tsuzaki et al.

10. Kondo I, Mizushige K, Hirao K et al. Ultrasonographic assessment of coronary flow reserve and abdominal fat in obesity. Ultrasound Med Biol 2001; 27: 1199-205. 11. Kawamoto R, Kajiwara T, Oka Y, Takagi Y. Association between abdominal wall fat index and carotid atherosclerosis in women. J Atheroscler Thromb 2002; 9: 213-8. 12. Onat A, Avci GS, Barlan MM, Uyarel H, Uzunlar B, Sansoy V. Measures of abdominal obesity assessed for visceral adiposity and relation to coronary risk. Int J Obes Relat Metab Disord 2004; 28: 1018-25. 13. Bosy-Westphal A, Geisler C, Onur S et al. Value of body fat mass vs anthropometric obesity indices in the assessment of metabolic risk factors. Int J Obes 2006; 30: 475-83. 14. Rezende FA, Rosado LE, Ribeiro Rde C et al. Body mass index and waist circumference: association with cardiovascular risk factors. Arq Bras Cardiol 2006; 87: 728-34. 15. Matsuzawa Y, Nakamura T, Shimomura I, Kotani K. Visceral fat accumulation and cardiovascular disease. Obes Res 1995; 3: S645-7. 16. Sironi AM, Gastaldelli A, Mari A et al. Visceral fat in hypertension: influence on insulin resistance and beta-cell function. Hypertension 2004; 44: 127-33. 17. Kanai H, Tokunaga K, Fujioka S, Yamashita S, Kameda-

352 Aging Clin Exp Res, Vol. 21, No. 4/5

Takemura K, Matsuzawa Y. Decrease in intra-abdominal visceral fat may reduce blood pressure in obese hypertensive women. Hypertension 1996; 27: 125-9. 18. Sleight P. Blood pressure, hearts, and U-shaped curves. Lancet 1988; 1: 235. 19. van Bortel LM, Spek JJ. Influence of aging on arterial compliance. J Hum Hypertens 1998; 12: 583-6. 20. Dart AM, Kingwell BA. Pulse pressure - a review of mechanisms and clinical relevance. J Am Coll Cardiol 2001; 37: 975-84. 21. Ishizaka N, Ishizaka Y, Toda E, Hashimoto H, Nagai R, Yamakado M. Hypertension is the most common component of metabolic syndrome and the greatest contributor to carotid arteriosclerosis in apparently healthy Japanese individuals. Hypertens Res 2005; 28: 27-34. 22. NIPPON DATA Research Group. Impact of elevated blood pressure on mortality from all causes, cardiovascular diseases, heart disease and stroke among Japanese: 14 year follow-up of randomly selected population from Japanese - Nippon Data 80. J Hum Hypertens 2003; 17: 851-7. 23. Asia Pacific Cohort Studies Collaboration. Blood pressure and caryperte r ns 20 diovascular disease in the Asia Pacific region. J Hype Hypertens 2003; 21: 707-16.