Association of Body Shape Index with Cerebral Small Vessel Disease Open Access

Cerebral small vessel disease (cSVD) is a group of subclinical pathologies and is increasing rapidly with the aging of the world population [1]. So far, cSVD has received clinical attention because it increases the risk of stroke and dementia [2, 3]. Recently, it has also been reported that cSVD itself can cause symptoms such as cognitive impairment, dysphagia, and gait disturbance [4]. cSVD has subtypes with different pathologies and are often found co-existing on brain magnetic resonance imaging (MRI) [5]. Therefore, studies have been conducted to find the common pathological mechanisms or risk factors penetrating these subtypes [3, 6]

Obesity is a major health problem worldwide, a well-known risk factor for cerebrovascular disease [7]. Recently, researchers have stated that the distribution of fat centered on visceral fat (i.e., central obesity) has a greater effect on obesity-related complications, rather than an increase in total fat represented by a high body mass index (BMI) [8, 9]. Since central obesity is a modifiable risk factor, it is important for physicians to accurately measure visceral fat and evaluate the actual risk. The most accurate way to measure visceral fat is to directly measure the fat tissue area using MRI, but this is expensive and not a method that can be easily used in most clinical situations [10]. There is a simple method using the waist circumference (WC), but this does not consider the differences in height or overall body shape, and there may be limitations in accuracy [11]. Therefore, an index that reflects visceral fat more simply and accurately is needed.

A body shape index (ABSI) was proposed as a novel anthropometric index in 2012 [12]. ABSI was designed in the form of a WC standardized by height and BMI, so it was able to accurately reflect visceral fat without being affected by BMI [12]. Based on these characteristics, ABSI was able to independently measure the risk of visceral fat for obesity complications while minimizing the influence of height and weight, and several studies have shown a close relationship between ABSI and arteriosclerosis, metabolic diseases, and cardiovascular diseases [11, 13‒15]. However, relatively few studies have been conducted in the area of cerebrovascular disease [16, 17].

In this study, we evaluated the association between ABSI and cSVD in health check-up participants. In addition, by analyzing the relationship between ABSI and each subtype of cSVD, it was confirmed whether ABSI was a risk factor that penetrated all cSVD pathologies or a risk factor that was only related to a specific cSVD subtype.

We constructed a consecutive health check-up registry at the Seoul National University Hospital Health Promotion Center based on medical records between January 2006 and December 2013. Our health promotion center conducts broad evaluations as part of the routine health check-up, including demographic, clinical, laboratory, and radiological evaluations. We retrospectively assessed the participants who underwent brain MRI (n = 3,257). Participants who met the following criteria were excluded: (1) history of stroke or severe neurological diseases (n = 11); (2) age under 30 years (n = 20); or (3) missing covariates data (n = 7). Finally, 3,219 health check-up participants were included in the analysis.

We assessed the demographic, clinical, and anthropometric factors, including age, sex, hypertension, diabetes, hyperlipidemia, ischemic heart disease, current smoking, systolic and diastolic blood pressure (BP), height, weight, WC, BMI, and ABSI [18]. The ABSI was calculated using the following formula: ABSI (m^11/6/kg^−2/3) = WC (m)/(BMI [kg/m²]^2/3 × height [m]^1/2) [12]. Participants also performed laboratory evaluations after 12 h of overnight fasting. Laboratory evaluations included glucose profiles, lipid profiles, and inflammatory markers (e.g., white blood cell [WBC] counts) [18].

In this study, all participants underwent brain MRI and magnetic resonance angiography using 1.5-T MR scanners (SIGNA, GE Healthcare, Milwaukee, WI, USA, or MAGNETOM Sonata, Siemens, Munich, Germany). The detailed MRI parameters were as follows: basic slice thickness = 5 mm, T1-weighted images (repetition time [TR]/echo time [TE] = 500/11 ms), T2-weighted images (TR/TE = 5,000/127 ms), T2-gradient echo images (TR/TE = 57/20 ms), T2 fluid-attenuated inversion recovery images (TR/TE = 8,800/127 ms), and three-dimensional time-of-flight magnetic resonance angiography images (TR/TE = 24/3.5 ms, slice thickness = 1.2 mm) [18].

We assessed three cSVD subtypes: white matter hyperintensity (WMH), silent brain infarct (SBI), and cerebral microbleed (CMB). WMH quantitatively evaluated the volume: thus, the Medical Imaging Processing, Analysis, and Visualization software (version, 11.0.0, National Institutes of Health, Bethesda, MD, USA) program was used [18]. After obtaining data from the converted DICOM file and outlining the lesion area by slice, the computer-assisted semiautomated technique calculated the volume. SBI was defined as an asymptomatic, well-defined lesion ≥3 mm in size with the same signal characteristics, such as cerebrospinal fluid on T1- or T2-weighted images [6]. CMBs were defined as focal round lesions <10 mm in size and had low signal characteristics on T2-gradient echo images [6]. To estimate the burden of cSVD, we measured the number of corresponding lesions along with the prevalence of SBI and CMB [18]. Intracranial atherosclerosis (ICAS) and extracranial atherosclerosis (ECAS) were defined as occlusion or ≥50% stenosis of the intracranial and extracranial vessels on time-of-flight magnetic resonance angiography images [19, 20]. Radiological parameters were rated by two well-trained neurologists (K.-W.N. and H.-Y.J.), and disagreements were resolved by discussion with a third rater (H.-M.K.).

To show the characteristics of ABSI, which may be somewhat unfamiliar to neurologists, we divided the study population into four groups (quartiles) according to ABSI values. We then compared the baseline characteristics, burden of metabolic syndrome, and burden of cSVD lesions between quartiles. The χ² test and Jonckheere-Terpstra test were used for this analysis.

To perform univariate analysis, we used simple linear regression analysis of WMH volume. Continuous variables with skewed data were transformed into log scales, except for WMH volume. Because the WMH volume had many zero values, it was transformed to a square-root scale. For univariate analysis of SBI and CMB, Student’s t test, Mann-Whitney U-test, and χ² test were used accordingly. Variables with a univariate analysis result of p < 0.05, along with age and sex, were included as confounders in multivariable analysis. If multicollinearity was highly suspected, variables considered to be more clinically relevant were included in the multivariable analysis model (e.g., BP parameters and hypertension). For multivariable analysis, we used linear regression analysis for WMH volume and logistic regression analysis for SBI and CMBs. For clinical use, we used the cut-off point of 0.083 m^11/6/kg^−2/3, which is the reference point used in the previous study and the boundary point of quartile 4 in our data [21]. Based on this cut-off point, “ABSI >0.083 m^11/6/kg^−2/3” was defined as a new categorical variable, and the correlation with cSVD was additionally analyzed.

To compare the effect of WC, a well-known indicator of central obesity, and ABSI on cSVD incidence, we compared the β or odds ratio values of multivariable analysis and 95% confidence intervals (CIs). Because WC and ABSI have quite different units and result values, the result values for one standard deviation change were compared. All statistical analyses were conducted using SPSS version 21.0 (IBM SPSS, Chicago, IL, USA), and statistical significance was set at p < 0.05.

A total of 3,219 health check-up participants were assessed (median [interquartile] age, 56 [50–63] years; male sex, 54.0%). The median ABSI value was 0.080 (0.078–0.083) m^11/6/kg^−2/3. The median WMH volume was 1.08 (0.20–2.68) mL, and the prevalence of SBI and CBMs was 267 (8.3%) and 131 (4.1%), respectively. Other detailed baseline characteristics are shown in online supplementary Table 1 (for all online suppl. material, see www.karger.com/doi/10.1159/000528701).

When comparing the baseline characteristics according to ABSI quartile groups, ABSI showed a positive quantitative association with age, BMI, WC, hypertension, diabetes, hyperlipidemia, WBC counts, ICAS, ECAS, WMH volume, SBI, and CMBs (Table 1). In addition, participants in the ABSI quartile 4 group had a significantly higher number of metabolic syndrome items than the other participants and showed a positive quantitative relationship as the quartile increased (Fig. 1). ABSI also showed a dose-response relationship with the disease burden of each pathology of cSVD (Fig. 2).

In the univariate linear regression analysis, WMH volume was related to age, hypertension, diabetes, ischemic heart disease, current smoking, systolic and diastolic BP, fasting glucose, hemoglobin A1c, total cholesterol, WBC counts, ICAS, ECAS, and ABSI (online suppl. Table S2). In the multivariable linear regression analysis, ABSI was significantly associated with WMH volume after adjusting for confounders (β = 0.107, 95% CI = 0.013–0.200, per 100 m^11/6/kg^−2/3). Even when analyzed using a cut-off point of 0.083 m^11/6/kg^−2/3, ABSI still showed a significant relationship with WMH volume (β = 0.144, 95% CI = 0.057–0.230) (Table 2).

In the univariate analyses, SBI was related to age, height, WC, hypertension, diabetes, systolic and diastolic BP, fasting glucose, hemoglobin A1c, total cholesterol, WBC counts, ICAS, and ABSI (online suppl. Table S3). CMBs were associated with age, WC, hypertension, diabetes, current smoking, systolic BP, and ABSI (online suppl. Table S4). In the multivariable logistic regression analysis, ABSI was significantly associated with both SBI (adjusted odds ratio = 1.62, 95% CI = 1.14–2.31, per 100 m^11/6/kg^−2/3) and CMB (adjusted odds ratio = 1.64, 95% CI = 1.16–2.33, per 100 m^11/6/kg^−2/3) after adjusting for confounders. These close associations were maintained even when using a categorical variable based on the cut-off point (Table 3).

To determine whether WC or ABSI showed a stronger association with cSVD, the same comparison was made based on 1 standard deviation. ABSI showed a stronger association than WC with all cSVD subtypes, including WMH volume, SBI, and CMBs (online suppl. Table S5).

In the current study, ABSI was found to be associated with cSVD in health check-up participants. These close associations were found across all cSVD subtypes, and quantitative relationships were also observed according to cSVD burden. This suggests that the pathology indicated by ABSI may be related to a common pathological mechanism that penetrates these three cSVD lesions.

Although an association between ABSI and cSVD was clearly shown in our data, the exact pathological mechanisms explaining the relationship between the two are unclear. We offer only a few plausible hypotheses. First, we can consider vasculo-metabolic diseases associated with visceral fat. Abdominal obesity is a key component of metabolic syndrome and usually coexists with other components [22]. In our data, ABSI showed a clear dose-response relationship with the burden of metabolic syndrome and showed significant associations with individual factors of hypertension, diabetes, and hyperlipidemia that constituted it. This high cardiometabolic burden can promote atherosclerosis in cerebral large and small vessels [23]. Indeed, ABSI has shown a close relationship with atherosclerosis in various parts of the body in previous studies, and our data also showed a statistically significant relationship with both ICAS and ECAS [11, 13, 24]. The diffuse hypoperfusion caused by ICAS and ECAS formed in this way is sufficient to develop cSVD and can also cause cSVD when micro-atheroma occurs in the cerebral perforating arteries [1, 25]. Second, the endothelial dysfunction mechanism can be considered. Visceral fat is not just a storage tissue but also an organ that secretes hormones [26]. Various pro-inflammatory cytokines secreted here induce systemic subclinical inflammation, which can damage the cerebral endothelium [27]. Indeed, some studies have reported a close association between ABSI and endothelial dysfunction [28]. When endothelial dysfunction occurs, toxic metabolites are introduced due to blood-brain-barrier breakdown, or the clearing of wastes through the glymphatic pathway is impaired, leading to cSVD [18, 25]. Finally, ABSI can indirectly reflect the skeletal muscle mass. In previous studies, it was reported that people with high ABSI had relatively less fat-free muscle mass, which was called “sarcopenic obesity” [21, 29, 30]. Because skeletal muscle has a vascular protective effect by acting as a metabolic sink that clears several circulating metabolites, this may explain the close association between high ABSI and cSVD.

Interestingly, in our data, the association between ABSI and cSVD was stronger than that between WC and cSVD. WC is also an indicator of central obesity. However, in addition to visceral fat, various components, such as muscle and subcutaneous fat, also affect the measurement of WC; and some of them have a protective effect on cerebrovascular diseases. In addition, WC generally increases with increasing body weight [10, 11]. Although controversy remains, there is a concept called the “obesity paradox” that, in the case of the elderly, a high BMI has a protective effect on cerebrovascular disease [31]. Therefore, the opposing influence of deleterious and protective factors may have made the association between WC and cSVD appear relatively small. Meanwhile, ABSI was designed to reflect pure visceral fat that minimized the influence of height and weight by standardizing it with height and BMI [12], and this was confirmed through studies using imaging such as dual-energy X-ray absorptiometry scanning [32]. Due to this design, ABSI could be included in the analysis of the effect on the outcome of cardiometabolic disease as an independent contributor without interaction with BMI in several studies [13]. In fact, even in our data, ABSI did not significantly change the main results depending on whether or not BMI was corrected. These points may explain why cSVD showed a higher correlation with ABSI than with WC, as well as why ABSI showed significant results in previous cerebrovascular disease studies.

There was no difference in ABSI values between sexes in our data. This is a different phenomenon compared to previous studies that showed markedly high ABSI values in men [16, 33]. The reason why there was no difference in ABSI values between men and women in our data can be considered as follows: first, by default, ABSI is calibrated with BMI and height values. Second, our study population consists only of Asians with low WC, and they are relatively young and have few comorbidities. However, currently, there are not many studies related to ABSI, and there are other studies that do not show differences between sexes, so these issues related to gender require more research [13].

There are several limitations in interpreting our findings. First, this was a retrospective cross-sectional study. Our results only demonstrate the association between ABSI and cSVD lesions but do not guarantee a causal relationship. Second, we conducted the analysis using a single ABSI value measured at the time of the health check-up. cSVD is a chronic subclinical pathology that has progressed slowly over several decades. Therefore, if the occurrence and progression of cSVD are analyzed using the average or variation of the ABSI values measured over several periods, the causal relationship between the two will be clearer. Third, as revealed by our data, ABSI values increased with age. As age is the most potent risk factor for cSVD, the influence of ABSI may have been overestimated. Finally, our study population was relatively young, and the prevalence of cardiovascular risk factors was low. Therefore, the influence of age and other risk factors may have been underestimated.

We demonstrated that ABSI is significantly associated with cSVD in health check-up participants. ABSI can be obtained simply by measuring height, weight, and WC; therefore, it is an index that is not burdensome even for periodic measurements. We expect that classifying high-risk groups using ABSI and conducting early neuroimaging evaluation will be helpful in formulating a primary prevention plan for cerebrovascular disease. As studies related to ABSI in patients with cerebrovascular disease are still lacking, our findings should be verified in future studies.

The Institutional Review Board (IRB) of the Seoul National University Hospital approved this study (IRB number: 1502-026-647). The requirement for informed consent was waived by the IRB because of the retrospective study design and the use of de-identified information. All experiments were performed in accordance with the Declaration of Helsinki and the relevant guidelines.

The authors declare no competing financial interests.

This study received no funding.

Ki-Woong Nam and Jin-Ho Park designed the study. Ki-Woong Nam, Han-Yeong Jeong, and Hyuktae Kwon contributed to data acquisition. Ki-Woong Nam performed statistical analysis and drafted the manuscript. Ki-Woong Nam, Hyung-Min Kwon, Han-Yeong Jeong, Jin-Ho Park, and Hyuktae Kwon contributed to the discussion. Hyung-Min Kwon and Jin-Ho Park edited the manuscript. All authors read and approved the final manuscript.

All data and materials related to the article are included in the main text and supplementary materials. Further inquiries can be directed to the corresponding author.

Hyung-Min Kwon and Jin-Ho Park contributed equally to this work.