Introduction
Obesity is closely associated with various metabolic disorders [
1]. In particular, visceral adipose tissue (VAT) accumulation, which often accompanies obesity, is associated with hypertension, dyslipidemia, type 2 diabetes, and cardiovascular disease (CVD)[
2,
3]. Although subcutaneous adipose tissue (SAT) also contributes to obesity-related insulin resistance and metabolic disease [
4,
5], most previous studies reported that VAT has a stronger association with metabolic risk factors as compared to SAT [
6,
7].
Recently, high-accuracy imaging techniques, such as computed tomography (CT) and magnetic resonance imaging (MRI), have attracted much attention in the estimation of abdominal adipose tissue. These techniques allow VAT and subcutaneous SAT to be easily distinguished [
8]. Consecutive images, in regular succession along the spine, are required for precise measurement of abdominal adipose tissue accumulation [
8,
9]. Unfortunately, taking large numbers of images per patient is not practical in a clinical setting because of the increased cost and time required as the number of images and analysis required increases [
10,
11]. Accordingly, most researchers use a single-slice image at L4–L5 to estimate the VAT [
12‐
16]. Recent studies [
17,
18] demonstrated that the most suitable site representing the total VAT volume is 5–10 cm above L4–L5. This site was better correlated with the total VAT volume than a single-slice image taken at the level of L4–L5. Additionally, several studies [
17,
19,
20] have shown that the VAT area measured in the upper abdomen (i.e., L1–L2 or L2–L3) is a better predictor of metabolic syndrome than the most common measurement site of L4–L5.
Weight loss is associated with improvements in risk factors for chronic disease [
21,
22] and all-cause mortality [
23]. Thus, the identification of the measurement site best representing VAT volume is particularly important for estimating the change (Δ) in VAT due to weight loss and for evaluating the effects of weight loss programs. An early study [
24] with a small sample size showed different relative changes in several VAT depots (i.e., L2–L3, L3–L4, L4–L5) in response to a 4.4 kg weight reduction induced by a liquid formula diet. Another prior study [
25] suggested that the single-slice VAT area with the highest correlation with Δtotal VAT volume was located at the L2–L3 level (5–6 cm above L4–L5) after two interventions (hypocaloric diet with or without resistance training).
These previous studies [
24,
25] were performed with Spanish and Scandinavian populations. Thus, the previous findings were unable to adequately address the possibility of racial differences. Asian populations have lower average body mass indices but higher body fat percentages than people of European ancestry [
26], and Japanese have a significantly greater amount of VAT than Caucasians [
27,
28] even at the same levels of waist circumference. Despite emerging evidence regarding racial differences in body composition, many races and ethnic groups are still poorly represented in a clinical setting. A person’s race or ethnicity may influence the measurement site that is best for determining VAT, and to date, there are no data showing the best measurement site for predicting changes in VAT in the Japanese. Thus, the objective of this study was to investigate the relationship between single-slice MRI images and abdominal fat tissue volume. Our focus in this longitudinal study was to identify the single-slice measurement site having the strongest association with VAT volume.
Discussion
The primary findings of this study were that single-slice image areas taken at different measurement sites have different associations with VAT volume and changes in VAT volume. Furthermore, images taken at 5–6 cm above L4–L5 had a better association with VAT volume and its changes in obese, Japanese men than images taken at L4–L5. The disc level that corresponds to 5–6 cm above L4–L5 was close to L2–L3 in 10.8% of participants and precisely at L2–L3 in 89.2% of participants in our study. Furthermore, we found that the relations between single-slice image changes at various measurement sites and the volume change were much more variable for VAT than for SAT.
Before the program, the image located at 5 cm above the L4–L5 image explained 94% of the variance in VAT volume (Figure 2), a correlation that was significantly greater than that for L4–L5 (67%). Our results are in agreement with those of recent studies [
25,
36] suggesting that positions higher on the abdomen (L2–L3 site and/or images located ~5–10 cm above the L4–L5) are better choices for assessing the VAT (L2–L3: 0.95–0.96, L4–L5: 0.81–0.89).
Dermerath et al.[
38] reported that, in 820 subjects, single-slice images taken 6 cm above L4–L5 (near the L3 vertebra) accurately predicted the total VAT volume in both black and white adults (
r = 0.98). Similarly, Kuk et al. [
19] indicated that the VAT areas at L1–L2 and L2–L3 (e.g., 8–9 cm and 5–6 cm above L4–L5) were more strongly correlated with VAT volume than the VAT areas obtained at all other anatomical landmarks (
r = 0.98 for L1–L2 and L2–L3). In the present study, we demonstrated for the first time the relationships between single-slice image areas relative to L4–L5 and VAT volume in obese Japanese men. These findings suggest that, although Asians have different body build and frame size than Caucasians [
39] and Europeans [
40], there may not be remarkable differences among races with respect to abdominal adipose tissue distribution patterns.
As for the ΔVAT volume in response to weight loss, the influence of the measurement site on ΔVAT estimation became greater after weight loss (Figure 2(a)). The image located at 6 cm above L4–L5 explained 80% (r = 0.90) of the variance in the ΔVAT volume, but the image located at L4–L5 explained only 35% (r = 0.59) of the variance. This result suggests that estimation of the ΔVAT volume using a single-slice images required the use if an image taken at 5–6 cm above the L4–L5 (the highest correlation coefficients at baseline and changes). In agreement with our findings, Kamel et al. [
41] reported that, in 44 subjects (23 women and 21 men), a measurement site 5–10 cm above L4–L5 showed the best correlation (women;
r = 0.92, men;
r = 0.94) with the changes in VAT volume after a program period (weight loss treatment resulted in a comparable mean weight loss of 9.5 kg in men). On the other hand, in a longitudinal study on obese, postmenopausal women (6-month exercise intervention) Kuk et al. [
42] examined the influence that measurement site may have on the association between VAT and SAT volumes and the metabolic syndrome. They reported that VAT at L1–L2 and L2–L3 were significantly stronger correlates of VAT volume as compared to L4–L5 in both baseline levels and changes that occurred with the intervention. As for correlation between VAT areas and VAT volume in the present study, we found a similar tendency. These findings support the statement that a single-slice image at L4–L5 can lead to a misestimation of the VAT accumulation and changes in VAT accumulation that occur with an intervention, which would likely lead to a misunderstanding of the clinical effects of the intervention. Kuk et al. also concluded that, although measurement sites have an impact on the prediction of VAT volume, this does not translate into an improved prediction for the metabolic syndrome. However, the Kuk study’s conclusions on this matter need careful interpretation because there was little weight loss (−0.44 kg over 6-months) during the intervention and no data on the energy balance (intake and expenditure).
In our study, single-slice images taken 5–6 cm above L4–L5 had the highest correlation with VAT volume and changes in VAT volume. The reasons why such images have a strong relationship with VAT volume remains unclear, and this should be investigated in the future. However, the single-slice image having the highest association with VAT volume was not the slice having the largest absolute VAT area and largest change in VAT area secondary to the weight loss intervention. This observation suggests the need to think about a metabolic difference not a quantity difference. Shen et al. [
36] reported a possible explanation pertaining to VAT’s 2 compartments: intraperitoneal and extraperitoneal adipose tissue (IPAT and EPAT, respectively [
33]). Whereas EPAT components serve primarily as mechanical cushions for organs, IPAT depots have high metabolic activity, and thus, they may account for a large proportion of the intersubject variation in observed VAT volume [
36]. Consequently, a slice that contains mostly IPAT may show the highest correlation with VAT volume even though the slice does not have the largest VAT area. Our study found that the single-slice areas 5–6 cm above L4–L5 had larger decreases in VAT volume after an intervention than areas at L4–L5, and these results support previous studies [
43,
44]. This observation supports the hypothesis that slices above L4–L5 contain more metabolically active VAT and it is also consistent with the observation that IPAT is located primarily in the upper abdomen [
44]. Ross et al. [
45] reported that VAT reduction in the upper abdomen (+5 cm, +10 cm, +15 cm above L4–L5) was greater than that observed in the lower abdomen (−5 cm below L4–L5). This study examined VAT reduction by the different weight loss method (diet and aerobic exercise vs. diet and resistance exercise). This suggests possibility of higher metabolic activity of IPAT in upper abdomen in both methods.
According to these findings, the use of conventional L4–L5 single-slice images can lead to misestimating the distribution of abdominal adipose tissue and the changes in abdominal adipose tissue with weight loss, which may lead to an over- or underestimation of weight-loss effects. Therefore, it is important to determine the best measurement site to evaluate the effects of weight loss. The advantage of the present study was that this was the first study to show the distribution patterns of abdominal adipose tissue changes and their relations to VAT or SAT volumes exclusively in obese Japanese men. Racial differences in VAT have been reported between Asian and European populations, indicating that the former have relatively greater amounts of VAT than the latter even at the same level of waist circumference [
28]. Moreover, a recent study suggests that, in men, racial differences in VAT varied significantly depending on measurement location [
46]. However, until now, possible racial differences in the pattern of VAT deposition across different anatomical locations have not been clearly understood [
47]. Although we did not make direct racial comparisons in our study, by using the same analytical methods as previous studies [
32,
38] and having similar results, we found there may not be many substantial differences between races on these participants.
There are several noteworthy limitations of this study. First, the sample size was relatively small. An insufficient sample size may be concomitant with type II error. Second, the present study included only obese Japanese men; thus, our results should be applied cautiously to women and non-obese individuals. Future studies are needed to determine whether changes in adipose tissue distribution can be modified by demographic variables such as sex and age. Moreover, the variability of the association between the measurement site and metabolic risk factors (i.e., glucose, triglyceride, and HDL and LDL cholesterol levels and blood pressure) should be investigated in Japanese. Finally, although we used that VAT volume in intra-abdominal, we did not include pelvic bottom. So, our result should be interpreted with caution in terms of VAT volume definition.
In conclusion, our hypothesis was supported by demonstrating that the individual measurement sites have different abilities to estimate VAT volume and changes in VAT volume in obese Japanese men. The findings of the present study showed that located at 5–6 cm above L4–L5 were optimal zone have highest association with VAT volume both at baseline and for the changes occurring as the result of the program. As for SAT, more caudal slices provided better association with subcutaneous fat, not only at baseline but for the changes in the SAT. Taking our data as a whole, our findings showed that the measurement site should be taken into consideration to accurately estimate VAT volume before and changes resulting from weight loss. Optimal measurement site of abdominal adipose tissue will serve as an important tool in the analysis of effects resulting from lifestyle interventions such as diet or exercise. Estimating not only the reduction but also the redistribution of VAT by MRI might be helpful for the search for individual predictive factors for potential metabolic diseases in clinical study and field. However, for accurate determination of abdominal adipose tissue and intervention effect comparison, we need more study for multiple-slices in abdominal region.
Competing interests
The authors declare no conflict of interest.
Authors’ contributions
RS was designed study and interpretation of data in addition to drafting the manuscript; TM and HS was instrumental in the study`s inception, design and approval while providing critical analysis of data interpretation and manuscript review. ME and TT made significant contributions to the acquisition of the data. The final manuscript have been read and approved by all authors.