The value of neck adipose tissue as a predictor for metabolic risk in health and type 2 diabetes

Abstract

Upper-body adiposity is adversely associated with metabolic health whereas the opposite is observed for the lower-body. The neck is a unique upper-body fat depot in adult humans, housing thermogenic brown adipose tissue (BAT), which is increasingly recognised to influence whole-body metabolic health. Loss of BAT, concurrent with replacement by white adipose tissue (WAT), may contribute to metabolic disease, and specific accumulation of neck fat is seen in certain conditions accompanied by adverse metabolic consequences. Yet, few studies have investigated the relationships between neck fat mass (NFM) and cardiometabolic risk, and the influence of sex and metabolic status. Typically, neck circumference (NC) is used as a proxy for neck fat, without considering other determinants of NC, including variability in neck lean mass. In this study we develop and validate novel methods to quantify NFM using dual x-ray absorptiometry (DEXA) imaging, and subsequently investigate the associations of NFM with metabolic biomarkers across approximately 7000 subjects from the Oxford BioBank. NFM correlated with systemic insulin resistance (Homeostatic Model Assessment for Insulin Resistance; HOMA-IR), low-grade inflammation (plasma high-sensitivity C-Reactive Protein; hsCRP), and metabolic markers of adipose tissue function (plasma triglycerides and non-esterified fatty acids; NEFA). NFM was higher in men than women, higher in type 2 diabetes mellitus compared with non-diabetes, after adjustment for total body fat, and also associated with overall cardiovascular disease risk (calculated QRISK3 score). This study describes the development of methods for accurate determination of NFM at scale and suggests a specific relationship between NFM and adverse metabolic health.

Keywords: Adipose tissue; Cardiometabolic health; Fat distribution; Imaging; Neck; Type 2 diabetes mellitus.

Figure 1. NFM estimation pipeline and validation.

a) DEXA images were used as input to the NFM estimation model which produces probabilities of neck fat measurement estimates for each subject. An ensemble average of 10 models per measurement is used to make predictions more robust. Model activation maps are shown as an intermediate step, to enable visual assurance that the trained models are indeed focusing on the neck region. b) NFM average estimations and standard deviation across 10 models for NFM (n = 6955) c) Validation scatterplot. Ground truth compared to estimated values for NFM (n = 143). Mean absolute error (MAE) = 54.8g

Figure 2. Associations between neck fat parameters. NFM units are in grams, NFM_adjFM units are the proportion of NFM to total fat mass, and the remaining variables are measured in centimetres.

a) Correlations between predicted ND, NFM_adjFM, and NFM. Data is stratified and coloured by subgroup in scattergraph and density plots (n = 6934). NFM values are log-transformed. Pearson correlation co-efficients are coloured by strength of association and significance is indicated by asterisks (* p<0.05 ; ** p < 0.01 ; *** p < 0.001). b) Correlations between predicted neck fat parameters (NFM, ND), and actual measurements (NC, neck skinfold). Actual neck fat measurements were taken from a subset of the T2DM group (n= 41), as detailed in Section 2.3. Scatterplots and density plots for each variable are displayed on the lower-left and central line of the grid respectively, and Pearson correlation co-efficients are displayed in the upper-right panel, coloured by strength of association. Neck skinfold values are log-transformed. Asterisks indicate significance of association (* p<0.05; ** p < 0.01 ; *** p < 0.001).

Figure 3. Distribution of NFM_adjFM stratified by sex and T2DM.

Relative frequency histogram, illustrating differences in NFM_adjFM. Data are stratified and coloured by subgroup. Horizontal bars representing the interquartile range for each group are shown above the histogram, and the median values are represented by the point within the bar (n = 6394).

Figure 4. NFM associations with other adipose depot masses.

a) Correlations between NFM and other fat depots (n=6934). Scattergraphs and density plots are stratified and coloured by subgroup. Pearson correlation co-efficients are displayed in the upper-right panel, with asterisks indicating the significance of associations (* p<0.05 ; ** p < 0.01 ; *** p < 0.001). b) Partial correlation analysis between NFM and other adipose tissue masses, controlling for total fat mass. Analyses are stratified by group, and coloured by strength of association (n=6934). Asterisks indicate the significance of associations (* p<0.01 ; ** p < 0.001 ; *** p < 0.0001)

Figure 5. Correlation between NFM_adjFM and relevant cardiometabolic risk variables.

All values have been log-transformed. Scatterplots and density plots are stratified and coloured by sex, using the non-T2DM population. On the upper-right panel, Pearson correlation co-efficients are displayed, and boxes are coloured by strength of association. Asterisks indicate the significance of association (* p<0.05 ; ** p < 0.01 ; *** p < 0.001). HOMA-IR = Homeostatic Model Assessment for Insulin Resistance (glucose (mmol/L) x insulin (mmol/L) /22.5) ; TG = triglyceride ; Total/HDL cholesterol = total cholesterol / high-density lipoprotein cholesterol ; SBP = systolic blood pressure ; NEFA = non-esterified fatty acid ; hsCRP= high-sensitivity C-reactive protein.

Figure 6. QRISK3 scores, stratified by NFM_adjFM and VFM_adjFM.

Relative differences in QRISK3 score across the deciles is captioned in the coloured bubbles, and is relative to the score of the first decile. Data are stratified and coloured by sex, using the non-T2DM groups. a) QRISK3 score stratified by NFM_adjFM b) QRISK3 score stratified by VFM_adjFM