Associations between the Combined Fat Mass Index and Fat-Free Mass Index with Carotid Intima-Media Thickness in a Japanese Population: The Tohoku Medical Megabank Community-Based Cohort Study

Abstract

Aim: Although many epidemiological studies have shown that obesity assessed by body mass index is associated with carotid intima-media thickness (cIMT), few studies have evaluated fat-free mass, which is a component of body composition. We investigated the associations between the combined fat mass index (FMI) and fat-free mass index (FFMI) with cIMT.

Methods: We conducted a cross-sectional study of 3,873 men and 9,112 women aged 20 years or older who lived in Miyagi prefecture, Japan. The FMI and FFMI were calculated as fat mass and fat-free mass divided by height squared, respectively. The indices were classified into sex-specific quartiles and were combined into 16 groups. The maximum common carotid artery was measured using high-resolution B-mode ultrasound. An analysis of covariance was used to assess associations between the combined FMI and FFMI with cIMT adjusted for age and smoking status. The linear trend test was conducted by stratifying the FMI and FFMI, scoring the categories from 1 (lowest) to 4 (highest), and entering the number as a continuous term in the regression model.

Results: In multivariable models, a higher FMI was not related to higher cIMT in men and women in most FFMI subgroups. Conversely, a higher FFMI was related to higher cIMT in all FMI subgroups (p＜0.001 for linear trend).

Conclusions: FMI was not associated with cIMT in most FFMI subgroups. Conversely, FFMI was positively associated with cIMT independently of FMI.

Introduction

Carotid intima-media thickness (cIMT) is considered a marker for atherosclerosis¹⁾. Many epidemiological studies have reported that an increase in cIMT is related to well-established cardiovascular disease (CVD) risk factors (i.e., age, smoking, systolic blood pressure (SBP), blood glucose, and total serum cholesterol)^1-4) and the incidence of myocardial infarction and stroke^5-9).

Obesity assessed by body mass index (BMI) is a well-known major risk factor for CVD^{10, 11)} and is associated with an increase in cIMT^12-14). However, a recent study showed that high blood pressure and serum glucose were related to the risk of incident CVD regardless of obesity assessed by BMI¹⁵⁾. Furthermore, previous epidemiological studies on the Japanese population have shown that over 60% of diabetic participants were not obese¹⁶⁾ and the prevalence of non-obese hypertension participants was higher than that of obese hypertension participants¹⁷⁾. Although BMI is often used as a proxy measure for adiposity, the major shortcoming of BMI measurement is that it cannot consider body composition (i.e., fat and fat-free mass)¹⁸⁾. Therefore, BMI may not be a good indicator for screening high-risk individuals with CVD. It is well-known that fat mass (FM) worsens plasma lipids, blood pressure, and glucose/insulin resistance¹¹⁾. Metabolically active fat-free mass (FFM) has been shown to be positively associated with stroke volume and cardiac output^{19, 20)}. These findings suggest that both FM and FFM might affect the increase in cIMT.

To clarify the association between body composition and cIMT, two different body compositions should be considered. Many previous studies have mutually adjusted FM and FFM to consider body composition^21-25). However, there is a correlation between FM and FFM because higher FFM is required to carry excess body fat. Therefore, it may be inappropriate to adopt FM and FFM simultaneously in the same statistical model.

Fat mass index (FMI) and fat-free mass index (FFMI) are indicators of body composition²⁶⁾. The FMI and FFMI are calculated as the FM and FFM in kilograms divided by the height in meters squared, respectively^{18, 26-28)}. Both indicators are useful for comparing individuals with different height measurements^{18, 26-28)}. In addition, combining FMI and FFMI not only avoids multicollinearity but also investigates the association between the FMI and cIMT in each FFMI subgroup and the association between FFMI and cIMT in each FMI subgroup. However, to the best of our knowledge, no study has examined the association between combined FMI and FFMI with cIMT.

Aim

We examined the association between combined FMI and FFMI with cIMT in a Japanese population.

Materials

Study Design and Population

We conducted a cross-sectional study using data from the Tohoku Medical Megabank Community-Based Cohort Study (TMM CommCohort Study). The design of the TMM CommCohort Study has been described in detail in a previous study²⁹⁾. Briefly, the source population for the survey comprised men and women aged ≥ 20 years living in Miyagi prefecture, northeastern Japan. All participants were recruited between May 2013 and March 2016 using the following three approaches. The type 1 survey (40,433 participants) was performed at a specific municipal health check-up site. The type 1 additional survey (664 participants) was conducted on different dates from specific municipal health check-ups. The type 2 survey (13,855 participants) was conducted in an assessment center. All surveys collected basic information from blood and urine, a questionnaire, and municipal health check-ups. In the type 2 survey, several physiological measurements (carotid echography, body composition, calcaneal ultrasound bone mineral density, etc.)²⁹⁾. Informed consent was obtained from a total of 54,952 participants. This study was approved by the Institutional Review Board of the Tohoku Medical Megabank Organization (approval number: 2021-4-028, approval date: May 31, 2021).

To be included in the analysis, participants were required to undergo several physiological measurements. Thus, 13,855 participants who underwent several physiological measurements in the type 2 survey were included. From the 13,855 participants, we excluded 870 participants for the following reasons: (1) those who withdrew from the study by July 13, 2021, failed to return the self-reported questionnaire or did not undergo physiological measurements (n=699); (2) data on body fat percentage (BF%), height or weight, and cIMT were missing (n=125); and (3) data on SBP, diastolic blood pressure (DBP), glucose, glycated hemoglobin A1c (HbA1c), total cholesterol (TC), triglyceride (TG), and high-density lipoprotein cholesterol (HDL-C) were missing (n=46). Therefore, 12,985 participants were analyzed for this study.

Anthropometry

Height was measured to the nearest 0.1 cm using a stadiometer (AD-6400; A&D Co, Ltd, Tokyo, Japan). Weight and BF% were measured using a body composition analyzer (InBody720; Biospace Co, Ltd, Seoul, Korea). Weight was measured in increments of 0.1 kg, and 1.0 kg was subtracted to account for the weight of the participant’s clothing. BMI was calculated as weight (kg) divided by height squared (m²). FM was calculated by multiplying the weight (kg) by BF%. FMI was calculated as FM (kg) divided by the height squared (m²). To calculate the FFMI, the FFM% was calculated by subtracting the BF% from 100%. The FFM was then calculated by multiplying the weight by the FFM%. Subsequently, FFMI was calculated as FFM (kg) divided by height squared (m²)^{18, 26-28, 30)}.

Carotid Intima-Media Thickness

Ultrasound imaging equipment (GM-72P00A; Panasonic Healthcare, Co, Ltd, Japan) was used to measure right and left cIMT at the common carotid artery. The UK Biobank study has shown an excellent reproducibility and validity of this automated device, and it has also been used in a previous study of the Japanese population^{31, 32)}. The left and right cIMT were measured at a plaque-free site 10 mm proximal to the carotid bifurcation. The Cardiovascular Health Study used maximum cIMT as a parameter since it is more closely associated with cardiovascular risk factors than the mean cIMT⁵⁾. Furthermore, the Shiga Epidemiological Study of Subclinical Atherosclerosis showed that the mean cIMT of maximum values had stronger associations with coronary artery calcification, which can predict CVD events than internal carotid artery and bifurcation³³⁾. Thus, we measured the left and right maximum common carotid arteries. The analysis used the average of the maximum IMT values of the left and right IMT.

Potential Confounders

We obtained information on the participants’ demographic characteristics, smoking status, and history of CVD using a self-reported questionnaire. Age was determined at the time of visiting the community support center. Smoking status was classified into the following three categories: never smoker, ex-smoker, and current smoker. Never smokers were defined as participants who had smoked ＜100 cigarettes during their lifetime. Ex-smokers were defined as participants who had smoked ≥ 100 cigarettes during their lifetime and indicated on the questionnaire that they no longer smoked. Current smokers were defined as participants who had smoked ≥ 100 cigarettes during their lifetime and indicated on the questionnaire that they currently smoke. FM has been associated with worsening blood pressure, glucose, and lipids¹²⁾. Obesity, the state of excessive fat accumulation, can cause hypertension, type 2 diabetes, and hyperlipidemia³⁴⁾. Thus, we considered blood pressure, blood glucose, and lipids to be intermediate factors in the association between combined FMI and FFMI with cIMT. Hence, we did not adjust for blood pressure, blood glucose, and lipid levels.

Measurement of other Variables

After resting in a sitting position for ≥ 2 min, blood pressure was measured twice in the upper right arm using a digital automatic blood pressure monitor (HEM-9000AI; Omron Healthcare Co, Ltd, Kyoto, Japan). The mean values of the two recorded measurements were used. Hypertension was defined as SBP ≥ 140 mmHg, DBP ≥ 90 mmHg, and/or self-reported treatment for hypertension. Before the survey, we did not restrict diet or drinking. Thus, we collected non-fasting blood samples. Non-fasting glucose levels were measured using the hexokinase method. HbA1c levels were measured using the latex agglutination turbidimetry method. Diabetes was defined as non-fasting glucose ≥ 200 mg/dL, HbA1c ≥ 6.5%, and/or self-reported treatment for diabetes. TC was measured using the Ultra-Violet-End method using cholesterol dehydrogenase. TG was measured using an enzymatic method. HDL-C levels were measured using the direct method. We did not obtain low-density lipoprotein cholesterol (LDL-C) levels. Although the Friedewald formula can be used to calculate LDL-C, this formula only holds for fasting blood samples³⁵⁾. Since we obtained a non-fasting blood sample, we did not calculate the LDL-C. Dyslipidemia was defined as TG ≥ 150 mg/dL or HDL-C ＜40 mg/dL and/or self-reported treatment for dyslipidemia. Information regarding hypertension, diabetes, and dyslipidemia was obtained using a self-reported questionnaire. Furthermore, participants answered whether they have a history of each disease (stroke and myocardial infarction).

Statistical Analyses

Data are presented as the mean (standard deviation [SD]) or median (interquartile range [IQR]) for continuous variables and as the number (%) for categorical variables. Since the distributions of the FMI and FFMI differed between men and women, all analyses were performed separately for men and women. FMI was categorized into the following quartile groups using the whole population: Q1 (lowest group), Q2, Q3, and Q4 (highest group). FFMI was also categorized into the following quartile groups using the whole population: Q1 (lowest group), Q2, Q3, and Q4 (highest group). We examined the association between FMI and FFMI using Pearson’s correlation coefficients.

Regarding the baseline characteristics of the FMI quartile, a trend test was performed to evaluate the linear relationships among FMI and age, height, BMI, BF%, FM, FFM, FFMI, cIMT, SBP, DBP, glucose, HbA1c, TC, TG, and HDL-C. During the trend test, a simple linear model was used to analyze age, height, BMI, BF%, FM, FFM, FFMI, cIMT, SBP, DBP, glucose, HbA1c, TC, TG, and HDL-C as continuous variables. The chi-square test was used to compare the treatment for hypertension, diabetes, dyslipidemia, the prevalence of hypertension, diabetes, dyslipidemia, smoking status, and history of CVD among the quartile groups for the FMI. A similar analysis was performed for the baseline characteristics of the FFMI quartile.

An analysis of covariance (ANCOVA) was used to test for associations between the FMI and least squares (LS) means of cIMT. The LS means and corresponding 95% confidence intervals (CIs) are presented. The multivariable-adjusted models included age (continuous) and smoking status (never smoker, ex-smoker, and current smoker). Similarly, we analyzed the association between FFMI and LS means of cIMT; p values for the analysis of linear trends were calculated by scoring the categories from 1 (the lowest category) to 4 (the highest category) and entering the number as a continuous term in the regression model.

The FMI and FFMI were combined and categorized into 16 groups. We also used ANCOVA to assess the association between the combined FMI and FFMI and LS means of cIMT. The p values for the analysis of linear trends were calculated by stratifying the FMI and FFMI, scoring the categories from 1 (the lowest category) to 4 (the highest category), and entering the number as a continuous term in the regression model.

We also conducted several sensitivity analyses to test the robustness of our findings. First, because age affects body composition and cIMT, we conducted a stratified analysis according to the following three age categories (20–39 years, 40–74 years, and 75 years or older). Second, several studies have shown that treatment for dyslipidemia, hypertension, and diabetes reduces cIMT^{1, 36-38)}. Furthermore, participants with cardiovascular disease have higher cIMT³⁹⁾. Therefore, to eliminate the effect of treatment for hypertension, diabetes, dyslipidemia, and history of CVD, we selected only individuals who did not undergo any treatment for hypertension, diabetes, dyslipidemia, and no history of CVD.

Statistical significance was set at p＜0.05. All analyses were performed using SAS version 9.4 for Windows (SAS Inc, Cary, NC, USA).

Results

Characteristics of the Study Population

A total of 3,873 men and 9,112 women fulfilled all the inclusion criteria, and their data were included in the analyses. The mean age (±SD) of the study participants was 59.9 years (±14.1 years) for men and 56.0 years (±13.5 years) for women. The median FMI (IQR) was higher for women (6.7 [5.1–8.6] kg/m²) than for men (5.5 [4.3–7.0] kg/m²). Conversely, the median FFMI was higher for men (18.0 [17.0–18.9] kg/m²) than for women (15.2 [14.5–16.0] kg/m²). The cIMT (SD) was higher in men (0.65 [0.14] mm) than in women (0.60 [0.12] mm). The percentages of treatment for hypertension, diabetes, dyslipidemia, and the percentage of current smokers were higher in men than in women. The correlations of the FMI and FFMI were r=0.39 for men and r=0.52 for women.

The FMI was categorized into the following sex-specific quartiles for men: Q1, ＜4.3 kg/m²; Q2, 4.3–5.5 kg/m²; Q3, 5.6–7.0 kg/m²; Q4, ≥ 7.0 kg/m². For women, they were as follows: Q1, ＜5.1 kg/m²; Q2, 5.1–6.7 kg/m²; Q3, 6.8–8.6 kg/m²; and Q4, ≥ 8.6 kg/m² (Supplemental Table 1). For both men and women, height and HDL-C were inversely associated with FMI (p＜0.001 for linear trend), and other variables were positively associated with FMI (p＜0.001 for linear trend). For both men and women, the prevalence of hypertension, diabetes, dyslipidemia, and smoking status were statistically different among the quartile groups (p＜0.05 for difference).

Supplemental Table 1. Characteristics of participants according to FMI

	Men	FMI				p valuea	Women	FMI				p valuea
		Q1 (＜4.3)	Q2 (4.3-5.5)	Q3 (5.6-7.0)	Q4 (≥ 7.0)			Q1 (＜5.1)	Q2 (5.1-6.7)	Q3 (6.8-8.6)	Q4 (≥ 8.6)
Number	3,873	969	967	968	969		9,112	2,277	2,279	2,293	2,278
Age,	59.9	56.4	60.4	61.6	61.3	＜0.001	56.0	51.5	55.2	58.7	58.7	＜0.001
years	(14.1)	(15.7)	(13.5)	(12.7)	(13.7)		(13.5)	(13.8)	(13.9)	(12.5)	(12.7)
Height,	167.5	168.7	167.6	167.2	166.7	＜0.001	155.8	157.5	156.2	155.1	154.3	＜0.001
cm	(6.3)	(6.6)	(6.2)	(6.0)	(6.3)		(5.8)	(5.5)	(5.7)	(5.7)	(5.8)
BMI,	23.8	20.6	22.7	24.4	27.4	＜0.001	22.3	18.8	20.9	22.8	26.9	＜0.001
kg/m2	(3.1)	(1.7)	(1.3)	(1.5)	(2.7)		(3.5)	(1.4)	(1.1)	(1.2)	(2.9)
BF%,	23.7	15.8	21.7	25.7	31.5	＜0.001	30.7	21.5	28.3	33.3	39.7	＜0.001
%	(6.3)	(3.1)	(1.7)	(1.9)	(3.6)		(7.3)	(3.5)	(2.0)	(2.0)	(3.6)
FM,	16.2	9.3	13.8	17.5	24.2	＜0.001	17.1	10.1	14.4	18.3	25.6	＜0.001
kg	(6.2)	(2.2)	(1.4)	(1.7)	(5.2)		(6.5)	(2.0)	(1.5)	(1.8)	(5.2)
FFM,	50.6	49.4	50.0	50.7	52.3	＜0.001	37.1	36.5	36.6	36.6	38.6	＜0.001
kg	(6.4)	(6.1)	(5.7)	(6.0)	(7.2)		(4.0)	(3.6)	(3.7)	(3.9)	(4.5)
FMI,	5.5	3.4	4.9	6.2	8.2	-	6.7	4.2	5.9	7.6	10.2	-
kg/m2	(4.3-7.0)	(2.8-3.8)	(4.6-5.2)	(5.9-6.6)	(7.5-9.2)		(5.1-8.6)	(3.5-4.7)	(5.5-6.3)	(7.1-8.0)	(9.3-11.5)
FFMI,	18.0	17.3	17.8	18.1	18.6	-	15.2	14.7	14.9	15.2	16.1	-
kg/m2	(17.0-18.9)	(16.3-18.3)	(17.0-18.6)	(17.1-19.0)	(17.6-19.8)		(14.5-16.0)	(14.0-15.3)	(14.3-15.6)	(14.5-15.9)	(15.3-16.9)
cIMT,	0.65	0.61	0.65	0.66	0.67	＜0.001	0.60	0.56	0.59	0.61	0.62	＜0.001
mm	(0.14)	(0.13)	(0.14)	(0.14)	(0.14)		(0.12)	(0.11)	(0.12)	(0.12)	(0.13)
SBP,	133.8	129.5	133.4	135.3	137.2	＜0.001	125.9	119.4	124.6	127.9	131.7	＜0.001
mmHg	(16.1)	(16.5)	(15.8)	(15.7)	(15.2)		(17.8)	(16.9)	(17.4)	(17.7)	(17.0)
DBP,	80.9	78.1	80.7	81.8	83.1	＜0.001	76.5	73.3	75.4	76.9	80.3	＜0.001
mmHg	(10.9)	(10.7)	(10.9)	(10.3)	(11.1)		(10.5)	(10.3)	(10.2)	(10.1)	(10.2)
Treatment for hypertension,	1,109	143	243	311	412	＜0.001	1,567	176	278	398	715	＜0.001
%	(28.6)	(14.8)	(25.1)	(32.1)	(42.5)		(17.2)	(7.7)	(12.2)	(17.5)	(31.4)
Hypertension, %	2,059	344	481	567	667	＜0.001	3,076	445	634	818	1,179	＜0.001
	(53.2)	(35.5)	(49.7)	(58.6)	(68.8)		(33.8)	(19.5)	(27.8)	(35.9)	(51.8)
Glucose,	92.6	88.1	92.0	93.6	96.8	＜0.001	86.6	83.2	85.3	87.6	90.5	＜0.001
mg/dL	(21.0)	(16.3)	(19.9)	(20.4)	(25.3)		(14.7)	(10.9)	(12.4)	(15.9)	(17.5)
HbA1c,	5.6	5.4	5.5	5.6	5.8	＜0.001	5.5	5.4	5.4	5.5	5.7	＜0.001
%	(0.6)	(0.4)	(0.6)	(0.6)	(0.7)		(0.5)	(0.3)	(0.4)	(0.5)	(0.6)
Treatment for diabetes,	307	59	59	86	103	＜0.001	320	45	43	70	162	＜0.001
%	(7.9)	(6.1)	(6.1)	(8.9)	(10.6)		(3.5)	(2.0)	(1.9)	(3.1)	(7.1)
Diabetes,	405	66	73	124	142	＜0.001	422	53	61	98	210	＜0.001
%	(10.5)	(6.8)	(7.6)	(12.8)	(14.7)		(4.6)	(2.3)	(2.7)	(4.3)	(9.2)
TC	201.2	195.7	201.3	202.9	204.9	＜0.001	212.1	205.4	212.8	215.8	214.5	＜0.001
(mg/dL)	(35.0)	(34.3)	(33.7)	(33.5)	(37.7)		(35.5)	(35.2)	(35.0)	(35.5)	(35.5)
TG,	101.0	75.0	97.0	110.5	129.0	＜0.001	79.0	61.0	73.0	87.0	104.0	＜0.001
mg/dL	(72.0-149.0)	(56.0-104.0)	(71.0-141.0)	(82.0-162.0)	(95.0-183.0)		(58.0-112.0)	(48.0-82.0)	(55.0-100.0)	(64.0-120.0)	(75.0-148.0)
HDL-C,	57.1	65.0	58.0	54.0	51.5	＜0.001	67.6	75.5	70.2	65.1	59.7	＜0.001
mg/dL	(15.1)	(15.9)	(15.3)	(12.9)	(12.5)		(16.3)	(16.2)	(15.9)	(14.6)	(14.0)
Treatment for dyslipidemia,	413	49	100	105	159	＜0.001	1,025	114	215	296	400	＜0.001
%	(10.7)	(5.1)	(10.3)	(10.9)	(16.4)		(11.3)	(5.0)	(9.4)	(13.0)	(17.6)
Dyslipidemia,	1,360	155	324	390	491	＜0.001	1,988	191	370	576	851	＜0.001
%	(35.1)	(16.0)	(33.5)	(40.3)	(50.7)		(21.8)	(8.4)	(16.2)	(25.3)	(37.4)
History of cardiovascular disease, %
Smoking status, %
Never smoker	1,107	291	298	257	261	0.018	7,170	1,740	1,820	1,826	1784	0.022
	(28.6)	(30.0)	(30.8)	(26.6)	(26.9)		(78.7)	(76.4)	(79.9)	(80.2)	(78.3)
Ex-smoker	1,944	451	473	501	519		1,238	322	300	296	320
	(59.2)	(46.5)	(48.9)	(51.8)	(53.6)		(13.6)	(14.1)	(13.2)	(13.0)	(14.1)
Current smoker	811	227	193	207	184		670	205	154	145	166
	(20.9)	(23.4)	(20.0)	(21.4)	(19.0)		(7.4)	(9.0)	(6.8)	(6.4)	(7.3)
Unknown	11	0	3	3	5		34	10	5	11	8
	(0.3)	(0.0)	(0.3)	(0.3)	(0.5)		(0.4)	(0.4)	(0.2)	(0.5)	(0.4)

Values are expressed as mean (standard deviation) or median (interquartile range) for continuous variables or as number (%) for categorical variables.

BF%, body fat percentage; BMI, body mass index; cIMT, carotid intima-media thickness; DBP, diastolic blood pressure; FFM, fat-free mass; FFMI, fat-free mass index; FM, fat mass; FMI, fat mass index; HbA1c, glycated hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; Q, quartile; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride

Hypertension was defined as SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg or receiving treatment for hypertension.

Diabetes was defined as non-fasting glucose ≥ 200 mg/dL and/or HbA1c ≥ 6.5% or receiving treatment for diabetes.

Dyslipidemia was defined as TG ≥ 150 mg/dL or HDL-C ＜40 mg/dL, and/or receiving treatment for dyslipidemia.

^a p value for trend test for continuous variables and chi-square test for categorical variables.

The FFMI was categorized into the following sex-specific quartiles for men: Q1, ＜17.0 kg/m²; Q2, 17.0–18.0 kg/m²; Q3, 18.1–18.9 kg/m²; and Q4, ≥ 18.9 kg/m². For women, they were as follows: Q1, ＜14.5 kg/m²; Q2, 14.5–15.1 kg/m²; Q3, 15.2–16.0 kg/m²; and Q4, ≥ 16.0 kg/m² (Supplemental Table 2). For both men and women, BMI, BF%, FM, FFM, DBP, and TG were positively associated with the FFMI (p＜0.001 for linear trend), and HDL-C was inversely associated with FFMI (p＜0.001 for linear trend). For men only, height was positively associated with FFMI, and age was inversely associated with FFMI (p＜0.001 for linear trend). For women only, IMT, SBP, glucose, and HbA1c were positively associated with FFMI (p＜0.001 for linear trend), and TC was inversely associated with FFMI (p＜0.001 for linear trend). For women only, the prevalence of hypertension and diabetes were statistically different among the quartile groups (p＜0.001 for difference). For both men and women, the prevalence of dyslipidemia and smoking status were statistically different among the quartile groups (p＜0.001 for difference).

Supplemental Table 2. Characteristics of participants according to FFMI

	Men	FFMI				p valuea	Women	FFMI				p valuea
		Q1 (＜17.0)	Q2 (17.0- 18.0)	Q3 (18.1- 18.9)	Q4 (≥ 18.9)			Q1 (＜14.5)	Q2 (14.5- 15.1)	Q3 (15.2- 16.0)	Q4 (≥ 16.0)
Number	3,873	969	967	969	968		9,112	2,278	2,278	2,278	2,278
Age,	59.9	63.7	61.7	59.6	54.7	＜0.001	56.0	55.7	56.4	56.5	55.6	0.974
years	(14.1)	(14.3)	(13.4)	(13.4)	(13.6)		(13.5)	(14.0)	(13.5)	(13.4)	(13.2)
Height,	167.5	166.1	166.9	167.6	169.5	＜0.001	155.8	156.0	155.7	155.5	155.9	0.370
cm	(6.3)	(6.5)	(6.3)	(6.1)	(6.0)		(5.8)	(5.7)	(5.9)	(5.7)	(5.9)
BMI,	23.8	21.0	22.9	24.2	26.9	＜0.001	22.3	19.5	21.2	22.8	25.8	＜0.001
kg/m2	(3.1)	(2.1)	(1.9)	(1.9)	(3.1)		(3.5)	(2.0)	(2.1)	(2.4)	(3.6)
BF%,	23.7	22.6	23.3	23.7	25.1	＜0.001	30.7	28.5	29.5	31.0	33.8	＜0.001
%	(6.3)	(6.7)	(6.0)	(5.7)	(6.6)		(7.3)	(6.6)	(6.8)	(7.0)	(7.5)
FM,	16.2	13.4	15.1	16.4	19.9	＜0.001	17.1	13.7	15.4	17.4	21.8	＜0.001
kg	(6.2)	(5.0)	(4.9)	(5.1)	(7.6)		(6.5)	(4.3)	(4.9)	(5.5)	(7.7)
FFM,	50.6	44.6	48.8	51.8	57.4	＜0.001	37.1	33.7	36.0	37.7	40.9	＜0.001
kg	(6.4)	(4.0)	(3.7)	(3.9)	(5.3)		(4.0)	(2.7)	(2.8)	(2.8)	(3.8)
FMI,	5.5	4.7	5.3	5.7	6.7	-	6.7	5.6	6.2	7.0	8.8	-
kg/m2	(4.3-7.0)	(3.5-6.1)	(4.2-6.6)	(4.5-7.0)	(5.1-8.3)		(5.1-8.6)	(4.3-6.9)	(4.9-7.7)	(5.5-8.7)	(6.7-10.8)
FFMI,	18.0	16.3	17.5	18.4	19.7	-	15.2	13.9	14.8	15.5	16.6	-
kg/m2	(17.0-	(15.8-	(17.2-	(18.2-	(19.3-		(14.5-	(13.6-	(14.6-	(15.4-	(16.3-
	18.9)	16.7)	17.7)	18.6)	20.3)		16.0)	14.2)	15.0)	15.7)	17.2)
cIMT,	0.65	0.65	0.65	0.65	0.64	0.108	0.60	0.58	0.59	0.60	0.61	＜0.001
mm	(0.14)	(0.14)	(0.14)	(0.15)	(0.14)		(0.12)	(0.12)	(0.12)	(0.12)	(0.13)
SBP,	133.8	133.7	134.8	133.4	133.5	0.411	125.9	122.9	125.0	126.9	128.8	＜0.001
mmHg	(16.1)	(17.5)	(16.7)	(15.2)	(14.7)		(17.8)	(17.6)	(17.9)	(18.0)	(17.4)
DBP,	80.9	78.6	81.1	81.1	82.9	＜0.001	76.5	74.6	75.8	76.9	78.7	＜0.001
mmHg	(10.9)	(11.0)	(10.6)	(10.6)	(10.9)		(10.5)	(10.0)	(10.2)	(10.6)	(10.8)
Treatment for hypertension,	1,109	247	275	295	292	＜0.001	1,567	272	328	422	545	＜0.001
%	(28.6)	(25.5)	(28.4)	(30.4)	(30.2)		(17.2)	(11.9)	(14.0)	(18.5)	(23.9)
Hypertension,	2,059	490	516	518	535	0.224	3,076	598	691	812	975	＜0.001
%	(53.2)	(50.6)	(53.4)	(53.5)	(55.3)		(33.8)	(26.3)	(30.3)	(35.7)	(42.8)
Glucose,	92.6	92.3	92.0	93.3	92.9	0.289	86.6	85.0	85.7	87.0	88.9	＜0.001
mg/dL	(21.0)	(22.3)	(19.2)	(21.0)	(21.3)		(14.7)	(13.7)	(13.0)	(14.8)	(16.6)
HbA1c,	5.6	5.6	5.6	5.6	5.6	0.08	5.5	5.4	5.5	5.5	5.6	＜0.001
%	(0.6)	(0.7)	(0.5)	(0.6)	(0.6)		(0.5)	(0.4)	(0.4)	(0.4)	(0.6)
Treatment for diabetes,	307	94	62	83	68	＜0.001	320	62	56	77	125	＜0.001
%	(7.9)	(9.7)	(6.4)	(8.6)	(7.0)		(3.5)	(2.7)	(2.5)	(3.4)	(5.5)
Diabetes,	405	111	80	105	109	0.079	422	85	78	91	168	＜0.001
%	(10.5)	(11.5)	(8.3)	(10.8)	(11.3)		(4.6)	(3.7)	(3.4)	(4.0)	(7.4)
TC,	201.2	199.8	200.6	202.0	202.4	0.074	212.1	214.0	212.8	211.7	210.0	＜0.001
mg/dL	(35.0)	(34.5)	(34.5)	(34.4)	(36.5)		(35.5)	(36.2)	(35.2)	(35.3)	(35.3)
TG,	101.0	90.0	99.0	101.0	115.5	＜0.001	79.0	72.0	74.0	81.0	92.0	＜0.001
mg/dL	(72.0-	(65.0-	(71.0-	(74.0-	(80.0-		(58.0-	(54.0-	(56.0-	(58.0-	(64.0-
	149.0)	131.0)	140.0)	149.0)	176.0)		112.0)	98.0)	105.0)	114.0)	133.0)
HDL-C,	57.1	60.6	58.3	56.5	53.1	＜0.001	67.6	71.5	69.3	67.2	62.4	＜0.001
mg/dL	(15.1)	(15.7)	(14.7)	(14.6)	(14.4)		(16.3)	(16.3)	(15.9)	(15.9)	(15.7)
Treatment for dyslipidemia,	413	89	91	102	131	＜0.001	1,025	200	243	281	301	＜0.001
%	(10.7)	(9.2)	(9.4)	(10.5)	(13.5)		(11.3)	(8.8)	(10.7)	(12.3)	(13.2)
Dyslipidemia,	1,360	263	298	355	444	＜0.001	1,988	354	453	502	679	＜0.001
%	(35.1)	(27.1)	(30.8)	(36.6)	(45.9)		(21.8)	(15.5)	(19.9)	(22.0)	(29.8)
History of cardiovascular disease, %
Smoking status, %
Never smoker	1,107	299	293	271	244	＜0.001	7,217	1,813	1,833	1,817	1,707	＜0.001
	(28.6)	(30.9)	(30.3)	(28.0)	(25.2)		(78.7)	(79.6)	(80.5)	(79.8)	(74.9)
Ex-smoker	1,944	485	483	511	465		1,248	288	291	304	355
	(59.2)	(50.1)	(50.0)	(52.7)	(48.0)		(13.6)	(12.6)	(12.8)	(13.4)	(15.6)
Current smoker	811	181	189	186	255		672	172	143	152	203
	(20.9)	(18.7)	(19.5)	(19.2)	(26.3)		(7.3)	(7.6)	(6.3)	(6.7)	(8.9)
Unknown	11	4	2	1	4		34	5	11	5	13
	(0.3)	(0.4)	(0.2)	(0.1)	(0.4)		(0.4)	(0.2)	(0.5)	(0.2)	(0.6)

Values are expressed as mean (standard deviation) or median (interquartile range) for continuous variables or as number (%) for categorical variables.

BF%, body fat percentage; BMI, body mass index; cIMT, carotid intima-media thickness; DBP, diastolic blood pressure; FFM, fat-free mass; FFMI, fat-free mass index; FM, fat mass; FMI, fat mass index; HbA1c, glycated hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; Q, quartile; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride

Hypertension was defined as SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg or receiving treatment for hypertension.

Diabetes was defined as non-fasting glucose ≥ 200 mg/dL and/or HbA1c ≥ 6.5% or receiving treatment for diabetes.

Dyslipidemia was defined as TG ≥ 150 mg/dL or HDL-C ＜40 mg/dL and/or receiving treatment for dyslipidemia.

^a p value for trend test for continuous variables and chi-square test for categorical variables.

The characteristics of the participants according to the combined FMI and FFMI are shown in Table 1 for men and Table 2 for women. Participants with higher FMI and higher FFMI were more likely to have a higher BMI in both men and women. Men categorized as FMI Q1 and FFMI Q4 (the lowest FMI quartile and the highest FFMI quartile) had younger age, taller height, lower cIMT, lower SBP, and lower glucose and had a lower prevalence of hypertension, diabetes, and more current smokers. Men categorized as FMI Q4 and FFMI Q1 (the highest FMI quartile and the lowest FFMI quartile) had older age, shorter height, higher cIMT, higher SBP, higher glucose, higher HbA1c, and higher prevalence of hypertension and diabetes and were fewer current smokers. Women categorized as FMI Q1 and FFMI Q4 had a taller height, lower TC, and lower TG and were more current smokers. Women categorized as FMI Q4 and FFMI Q1 had older age, shorter height, higher SBP, higher glucose, and higher TC and were fewer current smokers.

Table 1. Characteristics of male participants according to the combined FMI and FFMI

	FMI	Q1				Q2				Q3				Q4
	FFMI	Q1	Q2	Q3	Q4	Q1	Q2	Q3	Q4	Q1	Q2	Q3	Q4	Q1	Q2	Q3	Q4
Number	3,873	395	253	193	128	246	299	255	167	208	224	281	255	120	191	240	418
Age,	59.9	59.3	57.2	54.2	49.5	64.9	61.2	59.0	54.7	66.6	63.6	61.0	56.4	70.7	66.3	63.1	55.2
years	(14.1)	(16.0)	(15.2)	(14.7)	(15.2)	(13.4)	(12.8)	(12.9)	(13.2)	(11.2)	(12.0)	(12.5)	(12.6)	(10.6)	(11.1)	(12.4)	(13.6)
Height,	167.5	167.8	168.1	169.7	171.1	165.8	168.0	167.6	169.6	165.4	166.1	167.5	169.2	162.9	164.6	166.1	169.2
cm	(6.3)	(6.8)	(6.1)	(6.6)	(6.6)	(6.2)	(6.3)	(5.9)	(5.8)	(5.9)	(5.9)	(5.9)	(5.8)	(5.9)	(6.0)	(5.7)	(5.9)
BMI,	23.8	19.1	20.8	21.8	23.0	21.1	22.4	23.3	24.6	22.5	23.7	24.7	26.1	24.6	25.7	26.7	29.4
kg/m2	(3.1)	(1.2)	(0.8)	(0.7)	(1.1)	(0.8)	(0.5)	(0.5)	(0.7)	(0.8)	(0.5)	(0.5)	(0.9)	(1.1)	(1.2)	(1.2)	(2.7)
BF%,	23.7	16.1	16.0	15.6	14.5	23.2	22.0	21.1	20.1	27.7	26.2	25.3	24.1	33.9	31.9	31.0	30.9
%	(6.3)	(3.5)	(2.9)	(2.5)	(3.0)	(1.5)	(1.4)	(1.3)	(1.3)	(1.6)	(1.4)	(1.3)	(1.4)	(2.8)	(2.8)	(2.9)	(4.1)
FM,	16.2	8.7	9.4	9.8	9.8	13.5	13.9	13.9	14.2	17.1	17.1	17.5	18.0	22.1	22.3	22.9	26.3
kg	(6.2)	(2.2)	(2.0)	(1.9)	(2.3)	(1.3)	(1.4)	(1.5)	(1.4)	(1.7)	(1.6)	(1.7)	(1.8)	(2.6)	(3.2)	(3.4)	(6.4)
FFM,	50.6	45.1	49.4	53.0	57.5	44.5	49.4	51.7	56.8	44.5	48.3	51.7	56.9	43.1	47.5	50.9	58.0
kg	(6.4)	(4.2)	(3.7)	(4.3)	(5.3)	(3.9)	(3.8)	(3.8)	(4.7)	(3.8)	(3.5)	(3.7)	(5.0)	(3.8)	(3.5)	(3.7)	(5.7)
FMI,	5.5	3.2	3.5	3.5	3.6	4.9	5.0	4.9	4.9	6.2	6.2	6.2	6.3	8.2	7.8	8.0	8.6
kg/m2	(4.3-	(2.6-	(2.9-	(3.0-	(2.9-	(4.6-	(4.6-	(4.6-	(4.6-	(5.9-	(5.8-	(5.9-	(5.9-	(7.7-	(7.4-	(7.5-	(7.7-
	7.0)	3.8)	3.9)	3.9)	3.9)	5.2)	5.3)	5.2)	5.2)	6.6)	6.6)	6.6)	6.7)	8.9)	8.8)	8.8)	10.0)
FFMI,	18.0	16.1	17.5	18.3	19.4	16.3	17.5	18.4	19.5	16.4	17.5	18.4	19.6	16.4	17.5	18.4	20.0
kg/m2	(17.0-	(15.6-	(17.2-	(18.1-	(19.1-	(15.9-	(17.2-	(18.1-	(19.2-	(16.0-	(17.2-	(18.2-	(19.3-	(15.8-	(17.3-	(18.2-	(19.4-
	18.9)	16.6)	17.7)	18.6)	19.9)	16.7)	17.7)	18.6)	20.0)	16.8)	17.7)	18.6)	20.2)	16.8)	17.7)	18.6)	20.8)
cIMT,	0.65	0.62	0.63	0.61	0.59	0.67	0.64	0.64	0.64	0.67	0.67	0.67	0.65	0.69	0.67	0.68	0.65
mm	(0.14)	(0.14)	(0.14)	(0.13)	(0.11)	(0.15)	(0.14)	(0.15)	(0.13)	(0.14)	(0.14)	(0.15)	(0.14)	(0.15)	(0.13)	(0.15)	(0.14)
SBP,	133.8	130.0	130.3	129.0	127.1	133.4	134.8	132.7	132.1	137.9	137.7	133.3	133.3	138.9	137.4	137.8	136.2
mmHg	(16.1)	(18.1)	(16.3)	(15.3)	(13.2)	(16.4)	(17.3)	(14.4)	(14.1)	(16.8)	(16.3)	(14.6)	(15.0)	(16.2)	(15.6)	(15.6)	(14.4)
DBP,	80.9	77.2	79.0	78.8	77.7	78.6	81.4	80.9	82.0	81.0	82.6	81.3	82.3	79.1	81.4	82.9	85.2
mmHg	(10.9)	(10.5)	(10.6)	(11.7)	(9.5)	(11.3)	(11.0)	(10.2)	(10.6)	(10.4)	(10.6)	(9.7)	(10.5)	(12.1)	(9.8)	(10.8)	(11.1)
Treatment for hypertension,	1,109	56	44	29	14	58	73	68	44	72	67	91	81	61	91	107	153
%	(28.6)	(14.2)	(17.4)	(15.0)	(10.9)	(23.6)	(24.4)	(26.7)	(26.4)	(34.6)	(29.9)	(32.4)	(31.8)	(50.8)	(47.6)	(44.6)	(36.6)
Hypertension,	2,059	143	98	71	32	125	152	123	81	134	134	154	145	88	132	170	277
%	(53.2)	(36.2)	(38.7)	(36.8)	(25.0)	(50.8)	(50.8)	(48.2)	(48.5)	(64.4)	(59.8)	(54.8)	(56.9)	(73.3)	(69.1)	(70.8)	(66.3)
Glucose,	92.6	88.7	88.6	87.4	86.1	92.5	91.2	93.4	90.8	94.0	93.8	94.1	92.4	100.7	95.8	96.9	96.1
mg/dL	(21.0)	(17.5)	(18.8)	(11.6)	(13.1)	(24.0)	(19.0)	(19.8)	(14.0)	(19.2)	(19.9)	(22.8)	(19.1)	(32.8)	(18.5)	(24.7)	(25.9)
HbA1c,	5.6	5.5	5.4	5.4	5.4	5.6	5.5	5.6	5.5	5.6	5.7	5.6	5.6	5.9	5.7	5.7	5.8
%	(0.6)	(0.5)	(0.4)	(0.4)	(0.3)	(0.7)	(0.4)	(0.6)	(0.6)	(0.6)	(0.8)	(0.5)	(0.6)	(1.2)	(0.5)	(0.6)	(0.7)
Treatment for diabetes,	307	31	16	10	2	23	9	20	7	20	21	26	19	20	16	27	40
%	(7.9)	(7.9)	(6.3)	(5.2)	(1.6)	(9.4)	(3.0)	(7.8)	(4.2)	(9.6)	(9.4)	(9.3)	(7.5)	(16.7)	(8.4)	(11.3)	(9.6)
Diabetes,	405	33	18	11	4	26	12	25	10	26	28	35	35	26	22	34	60
%	(10.5)	(8.4)	(7.1)	(5.7)	(3.1)	(10.6)	(4.0)	(9.8)	(6.0)	(12.5)	(12.5)	(12.5)	(13.7)	(21.7)	(11.5)	(14.2)	(14.4)
TC,	201.2	195.8	194.5	196.7	196.1	201.5	201.7	200.9	201.1	205.7	201.5	202.9	201.9	199.4	205.9	206.3	205.1
mg/dL	(35.0)	(35.5)	(32.5)	(34.1)	(34.4)	(33.3)	(33.5)	(34.5)	(33.6)	(32.5)	(34.4)	(32.4)	(34.8)	(35.7)	(37.7)	(36.5)	(39.0)
TG,	101.0	73.0	73.0	82.0	77.0	94.0	100.0	94.0	100.0	110.0	109.5	104.0	117.0	110.5	125.0	127.0	138.5
mg/dL	(72.0-	(56.0-	(55.0-	(57.0-	(56.5-	(68.0-	(74.0-	(70.0-	(69.9-	(78.5-	(84.0-	(81.0-	(83.0-	(89.0-	(93.0-	(94.0-	(98.0-
	149.0)	102.0)	97.0)	117.0)	103.0)	143.0)	137.0)	136.0)	152.0)	159.0)	163.0)	149.0)	176.0)	150.0)	160.0)	185.0)	198.0)
HDL-C,	57.1	66.1	64.3	63.8	64.6	58.9	58.9	57.5	56.0	55.2	54.7	54.6	51.7	55.7	53.3	51.7	49.3
mg/dL	(15.1)	(15.8)	(14.9)	(15.6)	(18.2)	(16.0)	(14.7)	(15.1)	(15.3)	(12.7)	(13.1)	(13.3)	(12.2)	(14.3)	(13.1)	(11.9)	(11.6)
Treatment for dyslipidemia,	413	23	17	4	5	27	26	28	19	19	26	32	28	20	22	38	79
%	(10.7)	(5.8)	(6.7)	(2.1)	(3.9)	(11.0)	(8.7)	(11.0)	(11.4)	(9.1)	(11.6)	(11.4)	(11.0)	(16.7)	(11.5)	(15.8)	(18.9)
Dyslipidemia,	1,360	56	43	37	19	83	83	90	68	79	91	107	113	45	81	121	244
%	(35.1)	(14.2)	(17.0)	(19.2)	(14.8)	(33.7)	(27.8)	(35.3)	(40.7)	(38.0)	(40.6)	(38.1)	(44.3)	(37.5)	(42.4)	(50.4)	(58.4)
History of cardiovascular	3.7	20	11	9	2	10	16	18	10	23	26	23	21	20	28	32	38
disease, %	(7.9)	(5.1)	(4.4)	(4.7)	(1.6)	(4.1)	(5.4)	(7.1)	(5.9)	(11.1)	(11.6)	(8.2)	(8.2)	(16.7)	(14.7)	(13.3)	(9.1)
Smoking status, %
Never smoker	1,107	118	79	58	36	81	96	78	43	59	66	68	64	41	52	67	101
	(28.6)	(29.9)	(31.2)	(30.1)	(28.1)	(32.9)	(32.1)	(30.6)	(25.8)	(28.4)	(29.5)	(24.2)	(25.1)	(34.2)	(27.2)	(27.9)	(24.2)
Ex-smoker	1,944	191	117	89	54	117	143	127	86	109	114	159	119	68	109	136	206
	(59.2)	(48.4)	(46.3)	(46.1)	(42.2)	(47.6)	(47.8)	(49.8)	(51.5)	(52.4)	(50.9)	(56.6)	(46.7)	(56.7)	(57.1)	(56.7)	(49.3)
Current smoker	811	86	57	46	38	47	59	50	37	38	44	53	72	10	29	37	108
	(20.9)	(21.8)	(22.5)	(23.8)	(29.7)	(19.1)	(19.7)	(19.6)	(22.2)	(18.3)	(19.6)	(18.9)	(28.2)	(8.3)	(15.2)	(15.4)	(25.8)
Unknown	11	0	0	0	0	1	1	0	1	2	0	1	0	1	1	0	3
	(0.3)	(0.0)	(0.0)	(0.0)	(0.0)	(0.4)	(0.3)	(0.0)	(0.6)	(1.0)	(0.0)	(0.4)	(0.0)	(0.8)	(0.5)	(0.0	(0.7)

Values are expressed as mean (standard deviation) or median (interquartile range) for continuous variables or as number (%) for categorical variables.

BF%, body fat percentage; BMI, body mass index; cIMT, carotid intima-media thickness; DBP, diastolic blood pressure; FFM, fat-free mass; FFMI, fat-free mass index; FM, fat mass; FMI, fat mass index; HbA1c, glycated hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; Q, quartile; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride

Hypertension was defined as SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg or receiving treatment for hypertension.

Diabetes was defined as non-fasting glucose ≥ 200 mg/dL and/or HbA1c ≥ 6.5% or receiving treatment for diabetes.

Dyslipidemia was defined as TG ≥ 150 mg/dL or HDL-C ＜40 mg/dL and/or receiving treatment for dyslipidemia.

Table 2. Characteristics of female participants according to the combined FMI and FFMI

	FMI	Q1				Q2				Q3				Q4
	FFMI	Q1	Q2	Q3	Q4	Q1	Q2	Q3	Q4	Q1	Q2	Q3	Q4	Q1	Q2	Q3	Q4
Number	9,112	931	666	449	231	691	668	563	357	516	605	656	501	140	339	610	1,189
Age,	56.0	52.4	51.1	50.7	51.0	55.8	55.5	54.6	54.6	59.1	59.8	58.9	56.6	64.2	62.2	60.1	56.3
years	(13.5)	(14.1)	(13.8)	(13.8)	(12.3)	(14.3)	(13.7)	(13.5)	(14.0)	(12.2)	(11.8)	(12.6)	(13.2)	(12.1)	(11.2)	(12.0)	(13.0)
Height,	155.8	157.4	157.5	157.3	158.4	156.0	155.9	156.5	156.5	154.4	154.7	155.2	156.0	152.4	153.3	153.7	155.1
cm	(5.8)	(5.4)	(5.6)	(5.6)	(5.5)	(5.5)	(5.7)	(5.7)	(5.9)	(5.5)	(5.7)	(5.5)	(5.9)	(5.4)	(5.9)	(5.4)	(5.9)
BMI,	22.3	17.7	18.9	19.7	20.7	19.8	20.7	21.5	22.5	21.4	22.4	23.2	24.3	23.5	24.8	25.9	28.4
kg/m2	(3.5)	(1.0)	(0.8)	(0.8)	(0.9)	(0.7)	(0.5)	(0.5)	(0.7)	(0.7)	(0.6)	(0.6)	(0.8)	(1.0)	(1.2)	(1.5)	(2.9)
BF%,	30.7	22.0	21.5	21.0	20.4	29.8	28.3	27.6	26.4	35.0	33.8	32.9	31.6	40.5	39.9	39.5	39.6
%	(7.3)	(3.8)	(3.3)	(3.0)	(2.9)	(1.8)	(1.6)	(1.5)	(1.6)	(1.8)	(1.6)	(1.6)	(1.5)	(2.3)	(2.6)	(3.2)	(4.1)
FM,	17.1	9.7	10.1	10.3	10.7	14.3	14.3	14.6	14.6	17.8	18.1	18.4	18.7	22.1	23.3	24.2	27.4
kg	(6.5)	(2.1)	(2.0)	(1.9)	(1.9)	(1.4)	(1.4)	(1.6)	(1.6)	(1.7)	(1.8)	(1.8)	(1.9)	(2.4)	(3.2)	(3.7)	(5.9)
FFM,	37.1	34.2	36.8	38.5	41.4	33.8	36.1	38.1	40.6	33.1	35.5	37.5	40.5	32.4	35.0	36.9	41.2
kg	(4.0)	(2.7)	(2.7)	(2.8)	(3.0)	(2.6)	(2.7)	(2.8)	(3.4)	(2.8)	(2.7)	(2.7)	(3.6)	(2.5)	(2.8)	(2.6)	(4.1)
FMI,	6.7	4.1	4.2	4.3	4.4	5.9	5.8	6.0	5.9	7.4	7.5	7.6	7.7	9.2	9.6	9.9	10.7
kg/m2	(5.1	(3.4-	(3.6-	(3.7-	(3.9-	(5.5-	(5.5-	(5.6-	(5.5-	(7.0-	(7.1-	(7.1-	(7.2-	(8.8-	(9.1-	(9.1-	(9.6-
	-8.6)	4.6)	4.7)	4.8)	4.8)	6.3)	6.3)	6.3)	6.4)	7.9)	8.0)	8.1)	8.1)	9.9)	10.5)	11.0)	12.4)
FFMI,	15.2	13.9	14.8	15.5	16.3	14.0	14.8	15.5	16.4	14.0	14.8	15.5	16.5	14.0	14.9	15.6	16.8
kg/m2	(14.5-	(13.5-	(14.6-	(15.3-	(16.1-	(13.6-	(14.6-	(15.3-	(16.2-	(13.6-	(14.6-	(15.4-	(16.2-	(13.8-	(14.7-	(15.4-	(16.4-
	16.0)	14.2)	15.0)	15.7)	16.7)	14.2)	15.0)	15.7)	16.8)	14.2)	15.0)	15.7)	16.8)	14.3)	15.1)	15.8)	17.5)
cIMT,	0.60	0.56	0.56	0.56	0.57	0.58	0.59	0.59	0.60	0.60	0.61	0.62	0.61	0.62	0.63	0.62	0.62
mm	(0.12)	(0.11)	(0.10)	(0.11)	(0.12)	(0.13)	(0.11)	(0.12)	(0.11)	(0.12)	(0.12)	(0.13)	(0.13)	(0.13)	(0.12)	(0.12)	(0.13)
SBP,	125.9	119.1	118.8	119.6	122.2	124.2	124.6	125.5	123.8	125.7	128.5	128.5	128.6	132.0	132.0	132.0	131.5
mmHg	(17.8)	(16.9)	(16.7)	(16.6)	(17.5)	(17.3)	(17.3)	(18.1)	(16.9)	(17.7)	(17.7)	(17.3)	(18.0)	(17.6)	(17.6)	(17.6)	(16.4)
DBP,	76.5	73.0	73.0	73.5	74.8	75.4	75.5	75.7	74.9	75.5	77.2	77.4	77.6	77.4	79.5	79.8	81.1
mmHg	(10.5)	(9.9)	(10.0)	(10.5)	(11.8)	(9.9)	(9.7)	(11.0)	(10.4)	(9.8)	(9.7)	(10.1)	(10.9)	(9.9)	(10.5)	(10.0)	(10.1)
Treatment for hypertension,	1,567	71	54	31	20	80	73	71	54	75	106	128	89	46	95	192	382
%	(17.2)	(7.6)	(8.1)	(6.9)	(8.7)	(11.6)	(10.9)	(12.6)	(15.1)	(14.5)	(17.5)	(19.5)	(17.8)	(32.9)	(28.0)	(31.5)	(32.1)
Hypertension,	3,076	179	128	82	56	185	172	172	105	159	223	251	185	75	168	307	629
%	(33.8)	(19.2)	(19.2)	(18.3)	(24.2)	(26.8)	(25.8)	(30.6)	(29.4)	(30.8)	(36.9)	(38.3)	(36.9)	(53.6)	(49.6)	(50.3)	(52.9)
Glucose,	86.6	82.9	82.9	83.7	84.2	84.8	85.4	85.6	85.5	87.2	87.2	87.5	88.6	91.3	89.1	90.0	91.0
mg/dL	(14.7)	(9.9)	(10.2)	(13.4)	(10.7)	(12.4)	(12.6)	(12.7)	(11.3)	(19.1)	(14.3)	(14.6)	(15.7)	(15.0)	(15.0)	(16.8)	(18.8)
HbA1c,	5.5	5.4	5.3	5.4	5.4	5.4	5.4	5.4	5.4	5.5	5.5	5.5	5.5	5.6	5.6	5.6	5.7
%	(0.5)	(0.4)	(0.3)	(0.4)	(0.3)	(0.5)	(0.4)	(0.4)	(0.3)	(0.5)	(0.5)	(0.4)	(0.5)	(0.5)	(0.5)	(0.5)	(0.7)
Treatment for diabetes,	320	21	10	10	4	12	11	13	7	18	15	20	17	11	20	34	97
%	(3.5)	(2.3)	(1.5)	(2.2)	(1.7)	(1.7)	(1.7)	(2.3)	(2.0)	(3.5)	(2.5)	(3.1)	(3.4)	(7.9)	(5.9)	(5.6)	(8.2)
Diabetes,	422	24	12	11	6	18	21	15	7	29	21	24	24	14	24	41	131
%	(4.6)	(2.6)	(1.8)	(2.5)	(2.6)	(2.6)	(3.1)	(2.7)	(2.0)	(5.6)	(3.5)	(3.7)	(4.8)	(10.0)	(7.1)	(6.7)	(11.0)
TC,	212.1	208.1	205.2	202.9	200.3	215.2	212.7	213.0	208.0	221.2	219.3	213.6	208.9	221.1	216.4	215.1	212.9
mg/dL	(35.5)	(35.8)	(34.0)	(35.4)	(34.6)	(36.0)	(34.0)	(36.6)	(31.7)	(35.9)	(36.6)	(34.1)	(34.3)	(34.4)	(34.3)	(34.3)	(36.5)
TG,	79.0	62.0	60.5	60.0	59.0	74.0	73.0	71.0	73.0	65.0	85.0	89.0	91.0	95.5	98.0	100.0	108.0
mg/dL	(58.0-	(49.0-	(47.0-	(47.0-	(44.0-	(56.0-	(56.0-	(53.0-	(54.0-	(56.0-	(64.0-	(66.0-	(65.0-	(75.0-	(72.0-	(74.0-	(78.0-
	112.0)	83.0)	80.0)	84.0)	76.0)	100.0)	99.0)	95.0)	103.0)	77.0)	115.0)	123.5)	128.0)	132.5)	138.0)	142.0)	152.0)
HDL-C,	67.6	75.7	75.6	75.0	75.3	70.9	69.5	70.8	69.5	67.0	66.2	64.9	62.1	63.8	62.1	60.8	58.0
mg/dL	(16.3)	(16.4)	(16.0)	(16.3)	(15.8)	(16.5)	(15.1)	(15.7)	(16.3)	(14.5)	(15.0)	(14.4)	(14.2)	(13.3)	(14.4)	(13.9)	(13.8)
Treatment for dyslipidemia,	1,025	58	31	16	9	60	70	54	31	49	87	103	57	33	55	108	204
%	(11.3)	(6.2)	(4.7)	(3.6)	(3.9)	(8.7)	(10.5)	(9.6)	(8.7)	(9.5)	(14.4)	(15.7)	(11.4)	(23.6)	(16.2)	(17.7)	(17.2)
Dyslipidemia,	1,988	87	59	31	14	104	122	83	61	109	158	170	139	54	114	218	465
%	(21.8)	(9.3)	(8.9)	(6.9)	(6.1)	(15.1)	(18.3)	(14.7)	(17.1)	(21.1)	(26.1)	(25.9)	(27.7)	(38.6)	(33.6)	(35.7)	(39.1)
History of cardiovascular		25	10	7	2	24	17	13	6	20	19	19	12	10	15	13	42
disease, %		(2.7)	(1.5)	(1.6)	(0.9)	(3.5)	(2.5)	(2.3)	(1.7)	(3.9)	(3.1)	(2.9)	(2.4)	(7.1)	(4.4)	(2.1)	(3.5)
Smoking status, %
Never smoker	7,217	733	514	328	165	550	546	454	270	412	492	540	382	118	281	495	890
	(78.7)	(78.7)	(77.2)	(73.1)	(71.4)	(79.6)	(81.7)	(80.6)	(75.6)	(79.8)	(81.3)	(82.3)	(76.3)	(84.3)	(82.9)	(81.2)	(74.9)
Ex-smoker	1,248	115	91	75	41	89	81	76	54	68	76	79	73	16	43	74	187
	(13.6)	(12.4)	(13.7)	(16.7)	(17.8)	(12.9)	(12.1)	(13.5)	(15.1)	(13.2)	(12.6)	(12.0)	(14.6)	(11.4)	(12.7)	(12.1)	(15.7)
Current smoker	672	80	57	45	23	51	38	32	33	35	33	37	40	6	15	38	107
	(7.3)	(8.6)	(8.6)	(10.0)	(10.0)	(7.4)	(5.7)	(5.7)	(9.2)	(6.8)	(5.5)	(5.6)	(8.0)	(4.3)	(4.4)	(6.2)	(9.0)
Unknown	34	3	4	1	2	1	3	1	0	1	4	0	6	0	0	3	5
	(0.4)	(0.3)	(0.6)	(0.2)	(0.9)	(0.1)	(0.5)	(0.2)	(0.0)	(0.2)	(0.7)	(0.0)	(1.2)	(0.0)	(0.0)	(0.5)	(0.4)

Values are expressed as mean (standard deviation) or median (interquartile range) for continuous variables or as number (%) for categorical variables.

BF%, body fat percentage; BMI, body mass index; cIMT, carotid intima-media thickness; DBP, diastolic blood pressure; FFM, fat-free mass; FFMI, fat-free mass index; FM, fat mass; FMI, fat mass index; HbA1c, glycated hemoglobin A1c; HDL-C, high-density lipoprotein cholesterol; Q, quartile; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride

Hypertension was defined as SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg or receiving treatment for hypertension.

Diabetes was defined as non-fasting glucose ≥ 200 mg/dL and/or HbA1c ≥ 6.5% or receiving treatment for diabetes.

Dyslipidemia was defined as TG ≥ 150 mg/dL or HDL-C ＜40 mg/dL and/or receiving treatment for dyslipidemia.

Associations between FMI and cIMT

FMI was associated with cIMT, even after adjusting for potential confounders for both men and women (p＜0.001 for linear trend). For men, the cIMT LS means (95% CI) were 0.628 (0.609–0.647) for Q1, 0.641 (0.622–0.659) for Q2, 0.649 (0.631–0.668) for Q3, and 0.655 (0.637–0.674) for Q4. For women, the cIMT LS means (95% CI) were 0.584 (0.575–0.592) for Q1, 0.590 (0.581–0.599) for Q2, 0.593 (0.585–0.602) for Q3, and 0.600 (0.591–0.609) for Q4 (Supplemental Table 3).

Supplemental Table 3. Association between FMI and cIMT

Men	Multivariable model
Quartiles of FMI	LS meansc IMT (mm)	95% CI	p for trend
Q1 (lowest)	0.628	(0.609-0.647)	＜0.001
Q2	0.641	(0.622-0.659)
Q3	0.649	(0.631-0.668)
Q4 (highest)	0.655	(0.637-0.674)
p for ANCOVA	＜0.001
Women
Q1 (lowest)	0.584	(0.575-0.592)	＜0.001
Q2	0.590	(0.581-0.599)
Q3	0.593	(0.585-0.602)
Q4 (highest)	0.600	(0.591-0.609)
p for ANCOVA	＜0.001

Adjusted for age (continuous), smoking status (never-smoker, ex-smoker, current smoker, and unknown).

p-values for the analysis of linear trends were calculated by scoring the FMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

ANCOVA, analysis of covariance; CI, confidence interval; cIMT carotid intima-media thickness; FMI fat mass index; LS, least squares; Q, quartile

Associations between FFMI and cIMT

A higher FFMI was associated with a higher cIMT in both men and women adjusting for potential confounder (p＜0.001 for linear trend). For men, the cIMT LS means (95% CIs) were 0.626 (0.607–0.644) for Q1, 0.637 (0.618–0.655) for Q2, 0.651 (0.632–0.669) for Q3, and 0.667 (0.648–0.685) for Q4. For women, the cIMT LS means (95% CI) were 0.579 (0.570–0.588) for Q1, 0.586 (0.577–0.595) for Q2, 0.593 (0.584–0.602) for Q3, and 0.606 (0.597–0.615) for Q4 (Supplemental Table 4).

Supplemental Table 4. Association between FFMI and cIMT

Men	Multivariable model
Quartiles of FFMI	LS means cIMT (mm)	95% CI	p for trend
Q1(lowest)	0.626	(0.607-0.644)	＜0.001
Q2	0.637	(0.618-0.655)
Q3	0.651	(0.632-0.669)
Q4(highest)	0.667	(0.648-0.685)
p for ANCOVA	＜0.001
Women
Q1(lowest)	0.579	(0.570-0.588)	＜0.001
Q2	0.586	(0.577-0.595)
Q3	0.593	(0.584-0.602)
Q4(highest)	0.606	(0.597-0.615)
p for ANCOVA	＜0.001

Adjusted for age (continuous), smoking status (never-smoker, ex-smoker, current smoker, and unknown).

p-values for the analysis of linear trends were calculated by scoring the FFMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

ANCOVA, analysis of covariance; CI, confidence interval; cIMT, carotid intima-media thickness; FFMI, fat-free mass index; LS, least squares; Q, quartile

Associations between Combined FMI and FFMI with cIMT

For both men and women, the combined FMI and FFMI were associated with cIMT, even after adjusting for potential confounders (Table 3). Higher FMI did not tend to be associated with higher cIMT in most FFMI subgroups, except for men categorized as FMI Q4 (p=0.008 for linear trend) and women categorized as FMI Q2 (p＜0.030 for linear trend). Conversely, higher FFMI tended to be associated with higher cIMT in all FMI subgroups for both men and women (p＜0.05 for linear trend).

Table 3. Adjusted least square means of cIMT associated with FMI and FFMI

Men	FMIQ1	FMIQ2	FMIQ3	FMIQ4	p for linear trend among FFMI subgroupsa
LS means cIMT (mm), 95%CI
FFMIQ1	0.614 (0.593-0.634)	0.639 (0.617-0.661)	0.628 (0.605-0.651)	0.627 (0.600-0.653)	0.004
FFMIQ2	0.639 (0.616-0.661)	0.629 (0.607-0.650)	0.646 (0.623-0.669)	0.631 (0.608-0.655)	0.019
FFMIQ3	0.641 (0.617-0.664)	0.642 (0.620-0.664)	0.657 (0.636-0.679)	0.656 (0.633-0.679)	0.001
FFMIQ4	0.644 (0.618-0.671)	0.664 (0.640-0.688)	0.661 (0.639-0.683)	0.675 (0.655-0.695)	＜0.001
p for linear treand among FMI subgroupsb	0.130	0.966	0.139	0.008
Women	FMIQ1	FMIQ2	FMIQ3	FMIQ4	p for linear trend among FFMI subgroupsa
LS means cIMT (mm), 95%CI
FFMIQ1	0.579 (0.569-0.589)	0.580 (0.569-0.590)	0.582 (0.570-0.593)	0.567 (0.550-0.585)	0.004
FFMIQ2	0.583 (0.572-0.593)	0.585 (0.575-0.596)	0.588 (0.577-0.599)	0.591 (0.578-0.604)	＜0.001
FFMIQ3	0.585 (0.573-0.597)	0.598 (0.587-0.609)	0.597 (0.586-0.607)	0.591 (0.580-0.602)	＜0.001
FFMIQ4	0.597 (0.583-0.612)	0.604 (0.591-0.616)	0.603 (0.592-0.614)	0.610 (0.600-0.619)	＜0.001
p for linear trend among FMI subgroupsb	0.568	0.030	0.837	0.067

Adjusted for age (continuous), smoking status (never-smoker, ex-smoker, current smoker, and unknown).

^a p values for the analysis of linear trends were calculated by stratifying FMI, scoring the FFMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

^b p values for the analysis of linear trends were calculated by stratifying FFMI, scoring the FMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

ANCOVA, analysis of covariance; CI, confidence interval; cIMT, carotid intima-media thickness; FFMI, fat-free mass index; FMI, fat mass index; LS, least squares; Q, quartile

We conducted a stratified analysis according to the three age categories (20–39 years, 40–74 years, and 75 years or older). Although FFMI was not statistically significantly associated with cIMT among 75 years or older, the results were substantially unchanged compared with those using all participants (Supplemental Tables 5, 6, 7). Furthermore, the results of excluding participants’ treatment for hypertension, diabetes, and dyslipidemia and participants with a history of CVD were also substantially unchanged compared with those using all participants (Supplemental Table 8).

Supplemental Table 5. Adjusted least square means of cIMT associated with FMI and FFMI among 20-39 years old

Men	FMIQ1	FMIQ2	FMIQ3	FMIQ4	p for linear trend among FFMI subgroupsa
LS means cIMT (mm), 95%CI
FFMIQ1	0.473 (0.446-0.500)	0.465 (0.433-0.497)	0.478 (0.448-0.508)	0.490 (0.436-0.544)	0.467
FFMIQ2	0.510 (0.477-0.543)	0.471 (0.442-0.501)	0.485 (0.455-0.515)	0.472 (0.438-0.506)	0.084
FFMIQ3	0.494 (0.461-0.527)	0.494 (0.466-0.523)	0.483 (0.452-0.514)	0.510 (0.479-0.542)	0.367
FFMIQ4	0.524 (0.478-0.570)	0.507 (0.474-0.541)	0.497 (0.464-0.530)	0.515 (0.488-0.542)	0.924
p for linear trend among FMI subgroupsb	0.001	0.013	0.417	0.027
Women	FMIQ1	FMIQ2	FMIQ3	FMIQ4	p for linear trend among FFMI subgroupsa
LS means cIMT (mm), 95%CI
FFMIQ1	0.453 (0.435-0.470)	0.452 (0.434-0.471)	0.463 (0.432-0.468)	0.440 (0.414-0.467)	0.252
FFMIQ2	0.461 (0.443-0.479)	0.457 (0.439-0.475)	0.469 (0.445-0.481)	0.470 (0.447-0.492)	0.376
FFMIQ3	0.457 (0.438-0.476)	0.469 (0.440-0.477)	0.463 (0.451-0.488)	0.459 (0.441-0.478)	0.492
FFMIQ4	0.467 (0.442-0.492)	0.450 (0.448-0.491)	0.469 (0.444-0.482)	0.483 (0.467-0.499)	0.006
p for linear trend among FMI subgroupsb	0.191	0.041	0.072	0.002

Adjusted for age (continuous), smoking status (never-smoker, ex-smoker, current smoker, and unknown).

^b p-values for the analysis of linear trends were calculated by stratifying FFMI, scoring the FMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

^b p values for the analysis of linear trends were calculated by stratifying FFMI, scoring the FMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

ANCOVA, analysis of covariance; CI, confidence interval; cIMT, carotid intima-media thickness; FFMI, fat-free mass index; FMI, fat mass index; LS, least squares; Q, quartile

Supplemental Table 6. Adjusted least square means of cIMT associated with FMI and FFMI among 40-74 years old

Men	FMIQ1	FMIQ2	FMIQ3	FMIQ4	p for linear trend among FFMI subgroupsa
LS means cIMT (mm), 95%CI
FFMIQ1	0.619 (0.592-0.647)	0.642 (0.613-0.671)	0.634 (0.604-0.664)	0.648 (0.612-0.683)	0.066
FFMIQ2	0.638 (0.609-0.667)	0.641 (0.613-0.670)	0.660 (0.630-0.690)	0.632 (0.601-0.663)	0.777
FFMIQ3	0.645 (0.614-0.675)	0.652 (0.623-0.682)	0.66 (0.632-0.688)	0.662 (0.632-0.692)	0.158
FFMIQ4	0.633 (0.598-0.668)	0.673 (0.642-0.705)	0.674 (0.645-0.704)	0.685 (0.659-0.712)	0.001
p for linear trend among FMI subgroupsb	0.038	0.025	0.005	＜0.001
Women	FMIQ1	FMIQ2	FMIQ3	FMIQ4	p for linear trend among FFMI subgroupsa
LS means cIMT (mm), 95%CI
FFMIQ1	0.593 (0.582-0.605)	0.590 (0.578-0.602)	0.596 (0.583-0.609)	0.575 (0.555-0.595)	0.359
FFMIQ2	0.595 (0.583-0.607)	0.599 (0.587-0.611)	0.606 (0.593-0.618)	0.602 (0.587-0.616)	0.061
FFMIQ3	0.600 (0.587-0.614)	0.613 (0.600-0.626)	0.609 (0.597-0.621)	0.606 (0.593-0.618)	0.923
FFMIQ4	0.613 (0.597-0.628)	0.617 (0.600-0.631)	0.620 (0.607-0.633)	0.624 (0.613-0.635)	0.104
p for linear trend among FMI subgroupsb	0.008	＜0.001	0.001	＜0.001

Adjusted for age (continuous), smoking status (never-smoker, ex-smoker, current smoker, and unknown).

^b p-values for the analysis of linear trends were calculated by stratifying FFMI, scoring the FMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

^b p values for the analysis of linear trends were calculated by stratifying FFMI, scoring the FMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

ANCOVA, analysis of covariance; CI, confidence interval; cIMT, carotid intima-media thickness; FFMI, fat-free mass index; FMI, fat mass index; LS, least squares; Q, quartile

Supplemental Table 7. Adjusted least square means of cIMT associated with FMI and FFMI among 75 years or older

Men	FMIQ1	FMIQ2	FMIQ3	FMIQ4	p for linear trend among FFMI subgroupsa
LS means cIMT (mm), 95%CI
FFMIQ1	0.756 (0.700-0.811)	0.771 (0.699-0.844)	0.758 (0.692-0.824)	0.733 (0.660-0.806)	0.680
FFMIQ2	0.767 (0.701-0.834)	0.782 (0.724-0.840)	0.690 (0.631-0.749)	0.745 (0.675-0.815)	0.112
FFMIQ3	0.794 (0.728-0.860)	0.774 (0.711-0.837)	0.789 (0.729-0.850)	0.754 (0.688-0.819)	0.406
FFMIQ4	0.770 (0.689-0.851)	0.806 (0.738-0.875)	0.766 (0.702-0.830)	0.821 (0.764-0.879)	0.224
p for linear trend among FMI subgroupsb	0.400	0.535	0.199	0.019
Women	FMIQ1	FMIQ2	FMIQ3	FMIQ4	p for linear trend among FFMI subgroupsa
LS means cIMT (mm), 95%CI
FFMIQ1	0.744 (0.683-0.806)	0.759 (0.690-0.824)	0.725 (0.651-0.799)	0.707 (0.629-0.786)	0.258
FFMIQ2	0.768 (0.698-0.838)	0.737 (0.668-0.806)	0.754 (0.688-0.820)	0.760 (0.685-0.836)	0.938
FFMIQ3	0.779 (0.710-0.848)	0.763 (0.692-0.834)	0.750 (0.683-0.817)	0.734 (0.666-0.802)	0.253
FFMIQ4	0.791 (0.710-0.873)	0.777 (0.705-0.849)	0.748 (0.676-0.819)	0.757 (0.693-0.820)	0.275
p for linear trend among FMI subgroupsb	0.126	0.461	0.745	0.247

Adjusted for age (continuous), smoking status (never-smoker, ex-smoker, current smoker, and unknown).

^b p-values for the analysis of linear trends were calculated by stratifying FFMI, scoring the FMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

^b p values for the analysis of linear trends were calculated by stratifying FFMI, scoring the FMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

ANCOVA, analysis of covariance; CI, confidence interval; cIMT, carotid intima-media thickness; FFMI, fat-free mass index; FMI, fat mass index; LS, least squares; Q, quartile

Supplemental Table 8. Adjusted least square means of cIMT associated with FMI and FFMIT excluding treatment for hypertension, diabetes, dyslipidemia, and participants with history of cardiovascular disease

Men	FMIQ1	FMIQ2	FMIQ3	FMIQ4	p for linear trend among FFMI subgroupsa
LS means cIMT (mm), 95%CI
FFMIQ1	0.594 (0.574-0.615)	0.602 (0.579-0.626)	0.598 (0.574-0.622)	0.602 (0.573-0.631)	0.610
FFMIQ2	0.613 (0.590-0.637)	0.610 (0.588-0.632)	0.624 (0.602-0.647)	0.609 (0.584-0.635)	0.768
FFMIQ3	0.622 (0.597-0.647)	0.621 (0.598-0.644)	0.628 (0.606-0.650)	0.645 (0.622-0.668)	0.102
FFMIQ4	0.618 (0.591-0.646)	0.638 (0.612-0.663)	0.630 (0.606-0.654)	0.655 (0.635-0.675)	0.002
p for linear trend among FMI subgroupsb	0.011	0.007	0.023	＜0.001
Women	FMIQ1	FMIQ2	FMIQ3	FMIQ4	p for linear trend among FFMI subgroupsa
LS means cIMT (mm), 95%CI
FFMIQ1	0.560 (0.549-0.571)	0.557 (0.545-0.569)	0.556 (0.544-0.568)	0.557 (0.539-0.574)	0.436
FFMIQ2	0.562 (0.551-0.574)	0.567 (0.556-0.579)	0.572 (0.560-0.584)	0.570 (0.556-0.583)	0.052
FFMIQ3	0.565 (0.553-0.578)	0.574 (0.562-0.586)	0.575 (0.563-0.587)	0.575 (0.563-0.586)	0.291
FFMIQ4	0.570 (0.554-0.585)	0.576 (0.563-0.589)	0.583 (0.571-0.595)	0.588 (0.577-0.598)	0.004
p for linear trend among FMI subgroupsb	0.104	＜0.001	＜0.001	＜0.001

Adjusted for age (continuous), smoking status (never-smoker, ex-smoker, current smoker, and unknown).

^b p-values for the analysis of linear trends were calculated by stratifying FFMI, scoring the FMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

^b p values for the analysis of linear trends were calculated by stratifying FFMI, scoring the FMI categories, from 1 for the lowest category to 4 for the highest category, entering the number as a continuous term in the regression model.

ANCOVA, analysis of covariance; CI, confidence interval; cIMT, carotid intima-media thickness; FFMI, fat-free mass index; FMI, fat mass index; LS, least squares; Q, quartile

Discussion

In this cross-sectional study, we analyzed 12,985 Japanese participants aged ≥ 20 years to investigate the association between combined FMI and FFMI with cIMT. When the FMI and FFMI were combined, a higher FMI was not related to higher cIMT in men and women in most FFMI subgroups. Conversely, a higher FFMI was related to a higher cIMT in all FMI subgroups.

Several previous studies have shown an association between body composition (i.e., fat and fat-free mass) and cIMT. In a study of 421 obese middle-aged European men and women, FFM contributed to the increased cIMT independent of FM and other atherosclerotic risk factors²²⁾. The China Kadoorie Biobank study showed that FFM was more strongly associated with cIMT than FM²⁴⁾. The findings of the Southampton Women’s Survey showed that when FMI and FFMI were mutually adjusted, FFMI was associated with cIMT, and FMI was not significantly associated with cIMT⁴⁰⁾. Furthermore, the Avon Longitudinal Study of Parents and Children study showed that elevated levels of FFM were associated with increases in cIMT independent of FM²⁵⁾. Several previous studies used absolute FM and BF% as an index of body fat^21-25). However, the absolute FM is correlated with height. Because the weight is equal to the sum of FM and FFM, the BF% is affected by FFM¹⁸⁾. Thus, we used the FMI and FFMI, which are not affected by FM or FFM and allowed for comparisons of individuals with different height measurements. Additionally, to clarify the association between body composition and cIMT, two different body compositions should be considered. Since a higher BMI resulted in not only a higher FMI but also a higher FFMI, the correlation between FMI and FFMI was observed (r=0.39 for men; r=0.52 for women). Therefore, it may be inappropriate to use FMI and FFMI simultaneously in the same statistical model. To consider both FMI and FFMI, we examined the association between combined FMI and FFMI with cIMT. Our findings support previous studies^{22, 25, 40)}.

In this study, a higher FFMI was related to a higher cIMT in all FMI subgroups. A potential mechanism may be that the high metabolic demands of FFM require an increase in blood flow²⁵⁾. A previous study has shown that stroke volume and cardiac output were more strongly associated with FFM than adipose mass, diabetes, and age¹⁹⁾. Further study has reported that the increment of the FFM by exercise training was associated with an increment of the left ventricle mass and the wall thickness⁴¹⁾. Similarly, the findings of our previous study investigating the association between combined FMI and FFMI with hypertension have shown that FFMI was associated with hypertension even when FMI was considered and the Assessment Prognostic Risk Observational Survey showed that left ventricular hypertrophy caused by hypertension was significantly associated with an increase in cIMT^{30, 42)}. Therefore, greater FFMI might lead to an increased cIMT via hypertension and left ventricle hypertrophy.

We observed that a higher FMI was not related to a higher cIMT in most FFMI subgroups. Several studies have shown that cIMT is associated with well-known cardiovascular risk factors such as blood pressure, glucose, and lipids^{1, 4)}. In addition, adipose tissue is known to affect elevated blood pressure, glucose, and lipids¹⁰⁾. Therefore, greater FM may be associated with increased cIMT via diseases resulting in atherosclerosis, such as hypertension, diabetes, and dyslipidemia. However, our findings showed that harmful effects of FM were not observed in most FFMI subgroups. Conversely, we showed that a higher FMI was positively related to a higher cIMT in FFMI Q4. Participants classified under FFMI Q4 had a greater increase in FFMI with increasing FMI than in other FFMI subgroups. Therefore, these results may reflect the effect of FFMI on cIMT classified in FFMI Q4. Further studies are required to elucidate the underlying mechanism.

The strength of this study, to our knowledge, is that this study is the first to show the association between combined FMI and FFMI with cIMT. Since this study enrolled a large population of approximately 13,000 participants, we were able to classify 16 groups of combined FMI and FFMI sex-specific quartiles. Therefore, we were able to show a relationship with cIMT due to differences in body composition. However, our study has several limitations. First, because we used the bioelectric impedance analysis (BIA) method to measure BF%, a measurement error may have occurred. However, a high correlation between whole-body FM measured using the BIA method and whole-body FM measured using the dual-energy X-ray absorptiometry method has been verified (men: r=0.95, women: r=0.92)⁴³⁾. Second, although plaque is also considered a marker of atherosclerosis, it is known to be related to the risk of CVD^{1, 7-9)}. Our study did not measure plaque. Thus, we could not show the association between body composition and plaque. Third, our study enrolled only the Japanese population. It is well-known that body composition varies due to race. A study showed that the FFMI differed among the four ethnic groups (Caucasian, African American, Hispanic, and Asian), with African Americans having the highest FFMI and Asians having the lowest FFMI⁴⁴⁾. Furthermore, South Asians have more body fat and less skeletal muscle mass than Caucasians⁴⁵⁾. Therefore, similar investigations are required for other populations. Fourth, residual or unmeasured confounding factors may exist. Although we used combined FMI and FFMI to avoid collinearity, even in the same FMI subgroup, higher FMI tended to have higher FFMI. Therefore, we could not completely rule out the effects of FMI and FFMI. Fifth, participants who voluntarily underwent several physiological measurements may have higher health consciousness than those who did not, which could have caused a volunteer bias in our study. Finally, this study had a cross-sectional design and could not definitively establish a causal relationship between combined FMI and FFMI with cIMT. To clarify the causal relationship, prospective cohort studies are required.

Conclusion

When FFMI was not considered, a higher FMI was associated with higher cIMT. When FMI was not considered, a higher FFMI was associated with higher cIMT. When the FMI and FFMI were combined, the FMI did not tend to be associated with higher cIMT in most FFMI subgroups. However, FFMI tended to be associated with higher cIMT in all FMI subgroups. These findings suggest that because higher FFM could induce adaptive changes in systemic hemodynamics, cIMT might represent not only atherosclerosis but also physiological adaptations in hemodynamics such as thickening of the vascular wall with elevated blood pressure. Further studies are warranted to elucidate the association between body composition and the incidence of cardiovascular diseases, such as stroke and myocardial infarction, and the potential mechanisms of the relationship between body composition and cIMT.

Acknowledgements

The authors thank the members of the Tohoku Medical Megabank Organization, including the Genome Medical Research Coordinators and the office and administrative personnel for their assistance. A complete list of members is available at

https://www.megabank.tohoku.ac.jp/english/a200601/.

Conflicts of Interest

None.

Funding

This work was supported by grants from the Japanese Society for the Promotion of Science [JSPS; Grant-in-Aid for Science Research (C), number 19K10637], Tohoku Medical Megabank Project from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT), the Japan Agency for Medical Research and Development [AMED; JP20km0105001] and JST SPRING, [Grant number JPMJSP2114].

References

• 1)  Nezu T, Hosomi N, Aoki S, Matsumoto M. Carotid intima-media thickness for atherosclerosis. J Atheroscler Thromb, 2016; 23: 18-31
• 2)  Al-Shali K, House AA, Hanley AJ, Khan HM, Harris SB Mamakeesick M, Zinman B, Fenster A, Spence JD, Hegele RA. Differences between carotid wall morphological phenotypes measured by ultrasound in one, two and three dimensions. Atherosclerosis, 2005; 178: 319-325
• 3)  Lee YH, Shin MH, Kweon SS, Nam HS, Park KS Choi JS, Choi SW, Kim HY, Oh GJ, Ahn HR, Oh HS, Jeong SK. Normative and mean carotid intima-media thickness values according to metabolic syndrome in Koreans: The Namwon study. Atherosclerosis, 2014; 234: 230-236
• 4)  Mannami T, Konishi M, Baba S, Nishi N, Terao A. Prevalence of asymptomatic carotid atherosclerotic lesions detected by high-resolution ultrasonography and its relation to cardiovascular risk factors in the general population of a Japanese city: The Suita study. Stroke, 1997; 28: 518-525
• 5)  O’Leary DH, Polak JF, Kronmal RA, Manolio TA, Burke GL, Wolfson-Jr SK. Carotid-artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. Cardiovascular Health Study Collaborative Research Group. N Engl J Med, 1999; 340: 14-22
• 6)  Lorenz MW, Markus HS, Bots ML, Rosvall M, Sitzer M. Prediction of clinical cardiovascular events with carotid intima-media thickness: A systematic review and meta-analysis. Circulation, 2007; 115: 459-467
• 7)  Polak JF, Pencina MJ, Pencina KM, O’Donnell CJ, Wolf PA, D’Agostino-Sr RB. Carotid-wall intima-media thickness and cardiovascular events. N Engl J Med, 2011; 365: 213-221
• 8)  Kokubo Y, Watanabe M, Higashiyama A, Nakao YM, Nakamura F, Miyamoto Y. Impact of intima-media thickness progression in the common carotid arteries on the risk of incident cardiovascular disease in the Suita study. J Am Heart Assoc, 2018; 7: e007720
• 9)  Shimoda S, Kitamura A, Imano H, Cui R, Muraki I, Yamagishi K, Umesawa M, Sankai T, Hayama-Terada M, Kubota Y, Shimizu Y, Okada T, Kiyama M, Iso H. Associations of carotid intima-media thickness and plaque heterogeneity with the risks of stroke subtypes and coronary artery disease in the Japanese general population: The circulatory risk in communities study. J Am Heart Assoc, 2020; 9: e017020
• 10)  Ortega FB, Lavie CJ, Blair SN. Obesity and cardiovascular disease. Circ Res, 2016; 118: 1752-1770
• 11)  Powell-Wiley TM, Poirier P, Burke LE, Despres JP, Larsen PG, Lavie CJ, Lear SA, Ndumele CE, Neeland IJ, Sanders P, St-Onge MP, American Heart Association Council on Lifestyle and Cardiometabolic Heath; Council on Cardiovascular and Stroke Nursing; Council on Clinical Cardiology; Council on Epidemiology and Prevention; and Stroke Council. Obesity and cardiovascular disease: A scientific statement from the American Heart Association. Circulation, 2021; 143: e984-e1010
• 12)  Lo J, Dolan SE, Kanter JR, Hemphill LC, Connelly JM, Lees RS, Grinspoon SK. Effects of obesity, body composition, and adiponectin on carotid intima-media thickness in healthy women. J Clin Endocrinol Metab, 2006; 91: 1677-1682
• 13)  Oren A, Vos LE, Uiterwaal CS, Grobbee DE, Bots ML. Cardiovascular risk factors and increased carotid intima-media thickness in healthy young adults: The Atherosclerosis Risk in Young Adults (ARYA) Study. Arch Intern Med, 2003; 163: 1787-1792
• 14)  Jin Y, Kim D, Cho J, Lee I, Choi K, Kang H. Association between obesity and carotid intima-media thickness in Korean office workers: The mediating effect of physical activity. Biomed Res Int, 2018; 2018: 4285038
• 15)  Matsumura T, Sankai T, Yamagishi K, Kubota Y, Hayama-Terada M, Muraki I, Umesawa M, Cui R, Imano H, Ohira T, Kitamura A, Okada T, Kiyama M, Iso H. Impact of Major Cardiovascular Risk Factors on the Incidence of Cardiovascular Disease among Overweight and Non-Overweight Individuals: The Circulatory Risk in Communities Study (CIRCS). J Atheroscler Thromb, 2022; 29: 422-437
• 16)  Kashima S, Inoue K, Matsumoto M, Akimoto K. Prevalence and characteristics of non-obese diabetes in Japanese men and women: the Yuport Medical Checkup Center Study. J Diabetes, 2015; 7(4): 523-530
• 17)  Hori M, Kitamura A, Kiyama M, Imano H, Yamagishi K, Cui R, Umesawa M, Muraki I, Okada T, Sankai T, Ohira T, Saito I, Tanigawa T, Iso H, CIRCS Invcestigators. Fifty-year Time Trends in Blood Pressure, Body Mass Index and their Relations in a Japanese Community: The Circulatory Risk in Communities Study (CIRCS). J Atheroscler Thromb, 2017; 24: 518-529
• 18)  Schutz Y, Kyle UU, Pichard C. Fat-free mass index and fat mass index percentiles in Caucasians aged 18-98 y. Int J Obes Relat Metab Disord, 2002; 26: 953-960
• 19)  Collis T, Devereux RB, Roman MJ, de Simone G, Yeh J, Howard BV, Fabsitz RR, Welty TK. Relations of stroke volume and cardiac output to body composition: The strong heart study. Circulation, 2001; 103: 820-825
• 20)  Kozakova M, Palombo C, Paterni M, Anderwald CH, Konrad T, Colgan MP, Flyvbjerg A, Dekker J, Relationship between Insulin Sensitivity Cardiovascular risk Investigators. Relationship between Insulin Sensitivity Cardiovascular risk Investigators. Body composition and common carotid artery remodeling in a healthy population. J Clin Endocrinol Metab, 2008; 93: 3325-3332
• 21)  Czernichow S, Bertrais S, Oppert JM, Galan P, Blacher J, Ducimetiere P, Hercberg S, Zureik M. Body composition and fat repartition in relation to structure and function of large arteries in middle-aged adults (the SU.VI.MAX study). Int J Obes (Lond), 2005; 826-832
• 22)  Moreno M, Puig J, Moreno-Navarrete JM, Xifra G, Ortega F, Ricart W, Fernandez-Real JM. Lean mass, and not fat mass is an independent determinant of carotid intima media thickness in obese subjects. Atherosclerosis, 2015; 243: 493-498
• 23)  Al-Shorman A, Al-Domi H, Al-Atoum M. The associations of body composition and anthropometric measures with carotid intima-media thickness in obese and non-obese schoolchildren: A possible predictor for cardiovascular diseases. Vascular, 2018; 26: 285-290
• 24)  Arnold M, Linden A, Clarke R, Guo Y, Du H, Bian Z, Wan E, Yang M, Wang L, Chen Y, Chen J, Long H, Gu Q, Collins R, Li L, Chen Z, Parish S, China Kadoorie Biobank Collaborative Group. Carotid intima-media thickness but not carotid artery plaque in healthy individuals is linked to lean body mass. J Am Heart Assoc, 2019; 8: e011919
• 25)  Chiesa ST, Charakida M, Georgiopoulos G, Dangardt F, Wade KH, Rapala A, Bhowruth DJ, Nguyen HC, Muthurangu V, Shroff R, Smith GD, Lawlor DA, Sttar N, Timpson NJ, Hughes A, Deanfield JE. Determinants of intima-media thickness in the young: The ALSPAC study. JACC Cardiovasc Imaging, 2021; 14: 468-478
• 26)  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
• 27)  Kyle UG, Schutz Y, Dupertuis YM, Pichard C. Body composition interpretation. Contributions of the fat-free mass index and the body fat mass index. Nutrition, 2003; 19: 597-604
• 28)  Dulloo AG, Jacquet J, Solinas G, Montani JP, Schutz Y. Body composition phenotypes in pathways to obesity and the metabolic syndrome. Int J Obes (Lond), 2010; 34 Suppl 2: S4-17
• 29)  Hozawa A, Tanno K, Nakaya N, Nakamura T, Tsuchiya N, Hirata T, Narita A, Kogure M, Nochioka K, Sasaki R, Takanashi N, Otsuka K, Sakata K, Kuriyama S, Kikuya M, Tanabe O, Sugawara J, Suzuki K, Suzuki Y, Kodama EN, Fuse N, Kiyomoto H, Tomita H, Uruno A, Hamanaka Y, Metoki H, Ishikuro M, Obara T, Kobayashi T, Kitatani K, Takai-Igarashi T, Ogishima S, Satoh M, Ohmomo H, Tsuboi A, Egawa S, Ishii T, Ito K, Ito S, Taki Y, Minegishi N, Ishii N, Nagasaki M, Igarashi K, Koshiba S, Shimizu R, Tamiya G, Nakayama K, Motohashi H, Yasuda J, Shimizu A, Hachiya T, Shiwa Y, Tominaga T, Tanaka H, Oyama K, Tanaka R, Kawame H, Fukushima A, Ishigaki Y, Arai T, Kawaguchi Y, Higuchi S, Sakaida M, Endo R, Nishizuka S, Tsuji I, Hitomi J, Nakamura M, Ogasawara K, Yaegashi N, Kinoshita K, Kure S, Sakai A, Kobayashi S, Sobue K, Sasaki M, Yamamoto M. Study profile of the Tohoku Medical Megabank Community-Based Cohort Study. J Epidemiol, 2021; 31: 65-76
• 30)  Takase M, Nakamura T, Tsuchiya N, Kogure M, Itabashi F Narita A, Hirata T, Nakaya N, Hamanaka Y, Sugawara J, Suzuki K, Fuse N, Uruno A, Kodama EN, Kuriyama S, Tsuji I, Kure S, Hozawa A. Association between the combined fat mass and fat-free mass index and hypertension: The Tohoku Medical Megabank Community-based Cohort Study. Clin Exp Hypertens, 2021; 43: 610-621
• 31)  Coffey S, Lewandowski AJ, Garratt S, Meijer R, Lynum S, Bedi R, Paterson J, Yaqub M, Noble JA, Neubauer S, Petersen SE, Allen N, Sudlow C, Collins R, Matthews PM, Leeson P. Protocol and quality assurance for carotid imaging in 100,000 participants of UK Biobank: Development and assessment. Eur J Prev Cardiol, 2017; 24: 1799-1806
• 32)  Miyake Y, Tanaka K, Senba H, Hasebe Y, Miyata T, Higaki T, Kimura E, Matsuura B, Kawamoto R. Education and household income and carotid intima-media thickness in Japan: Baseline data from the Aidai Cohort Study in Yawatahama, Uchiko, Seiyo and Ainan. Environ Health Prev Med, 2021; 26: 88
• 33)  Zaid M, Fujiyoshi A, Hisamatsu T, Kadota A, Kadowaki S, Satoh A, Sekikawa A, Barinas-Mitchell E, Horie M, Miura K, Ueshima H et al. A Comparison of Segment-Specific and Composite Measures of Carotid Intima-Media Thickness and their Relationships with Coronary Calcium. J Atheroscler Thromb, 2022; 29: 282-295
• 34)  Jiang SZ, Lu W, Zong XF, Ruan HY, Liu Y. Obesity and hypertension. Exp Ther Med, 2016; 12: 2395-2399
• 35)  Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem, 1972; 18: 499-502
• 36)  Huang Y, Li W, Dong L, Li R, Wu Y. Effect of statin therapy on the progression of common carotid artery intima-media thickness: An updated systematic review and meta-analysis of randomized controlled trials. J Atheroscler Thromb, 2013; 20: 108-121
• 37)  Wang JG, Staessen JA, Li Y, Van-Bortel LM, Nawrot T, Fagard R, Messerli FH, Safar M. Carotid intima-media thickness and antihypertensive treatment: A meta-analysis of randomized controlled trials. Stroke, 2006; 37: 1933-1940
• 38)  Geng DF, Jin DM, Wu W, Fang C, Wang JF. Effect of alpha-glucosidase inhibitors on the progression of carotid intima-media thickness: A meta-analysis of randomized controlled trials. Atherosclerosis, 2011; 218: 214-219
• 39)  Burke GL, Evans GW, Riley WA, Sharrett AR, Howard G, Barnes RW, Rosamond W, Crow RS, Rautaharju PM, Heiss G. Arterial wall thickness is associated with prevalent cardiovascular disease in middle-aged adults. The Atherosclerosis Risk in Communities (ARIS) Study. Stroke, 1995; 26(3): 386-391
• 40)  Sletner L, Mahon P, Crozier SR, Inskip HM, Godfrey KM, Chiesa S, Bhowruth DJ, Charakida M, Deanfield J, Cooper C, Hanson M. Childhood fat and lean mass: Differing relations to vascular structure and function at age 8 to 9 years. Arterioscler Thromb Vasc Biol, 2018; 38: 2528-2537
• 41)  Maron BJ, Pelliccia A. The heart of trained athletes: cardiac remodeling and the risks of sports, including sudden death. Circulation, 2006; 114: 1633-1644
• 42)  Cuspidi C, Mancia G, Ambrosioni E, Pessina A, Trimarco B, Zanchetti A, APROS Investigators. Left ventricular and carotid structure in untreated, uncomplicated essential hypertension: Results from the Assessment Prognostic Risk Observational Survey (APROS). J Hum Hypertens, 2004; 18: 891-896
• 43)  Kim M, Shinkai S, Murayama H, Mori S. Comparison of segmental multifrequency bioelectrical impedance analysis with dual-energy X-ray absorptiometry for the assessment of body composition in a community-dwelling older population. Geriatr Gerontol Int, 2015; 15: 1013-1022
• 44)  Hull HR, Thornton J, Wang J, Pierson RN, Kaleem Z, Pi-Sunyer X, Heymsfield S, Albu J, Fernandez JR, Vanitallie TB, Gallagher D. Fat-free mass index: Changes and race/ethnic differences in adulthood. Int J Obes (Lond), 2011; 35: 121-127
• 45)  Misra A, Khurana L. Obesity-related non-communicable diseases: South Asians vs White Caucasians. Int J Obes (Lond), 2011; 35: 167-187