Fasting and Postprandial Activity of Brown Adipose Tissue in Healthy Men

DISCUSSION

Physiologically active BAT was found in 60% of the subjects at a room temperature of 16°C–18°C. Although more subjects would probably have shown BAT uptake if the temperature had been lower, a lower temperature would probably have caused shivering, thus making our assessment of nonshivering thermogenesis less reliable. Most studies performed to date have used similar temperatures, and the prevalence of 60% is in line with the literature (2).

In contrast to results found in rats, the present study showed that glucose uptake by BAT in humans is more pronounced in the fasting state. The different biodistribution pattern of ¹⁸F-FDG in the postprandial state may be explained by increased ¹⁸F-FDG retention by muscle because of the meal-induced insulin secretion, thus reducing the availability of ¹⁸F-FDG for uptake by BAT

In contrast, Orava et al. found an increase in BAT uptake of glucose after insulin stimulation. This finding might be attributed to the fact that they used arterial ¹⁸F-FDG sampling and dynamic scanning to estimate ¹⁸F-FDG uptake—a method that yields uptake rates per minute—whereas our method yields the cumulative uptake in the first hour (no absolute qualification was performed). However, comparison of the cumulative uptake in the first hour should be a reliable indicator of total activity since uptake is still ongoing at that time. Moreover, Orava et al. studied subjects in a warm environment and found an approximately 5-fold increase after insulin in cold, and in fasting subjects they found an approximately 12-fold increase in ¹⁸F-FDG uptake, which supports the notion that insulin is not the principal determinant of BAT activity (8). Finally, in contrast to the controlled intervention of a hyperinsulinemic euglycemic clamp, postprandial BAT metabolism and energy expenditure are under the control of a complicated physiologic response to the meal. Multiple mechanisms may affect energy expenditure, such as diet-induced thermogenesis and the thermic effect of food itself (e.g., metabolic, endocrine, cardiovascular, sympathetic, and gut responses) (10).

Another explanation for the lower glucose uptake in the postprandial state might be a relatively higher contribution of lipid oxidation to total BAT energy expenditure. Although ¹⁸F-FDG uptake has been considered and shown to be a good reflection of BAT activity in most studies performed to date, fatty acids rather than glucose are believed to be the main metabolic substrate for BAT (9). The effect of fatty meals on BAT metabolic activity is interesting, as fatty acids are presumed to be the physiologic activators of uncoupling protein 1 in rodents (9). Williams et al. compared the ¹⁸F-FDG PET/CT scans of 741 patients who had taken a high-fat, very-low-carbohydrate, protein-permitted meal with those of 1,229 patients who had fasted (11). The patients who had taken the high-fat meal had a significantly lower frequency of ¹⁸F-FDG uptake in BAT and lower blood glucose levels than the patients who had fasted. Although this study was retrospective, it did support the findings of the present study.

It is tempting to speculate that BAT-mediated glucose clearance was responsible for the observed decreased postprandial glucose levels, compared with baseline levels. However, this speculation is not supported by our ¹⁸F-FDG PET data. Moreover, lean and healthy subjects often have lower plasma glucose levels after ingestion of glucose than do obese volunteers (12). Whether in humans BAT functions as an organ that significantly contributes to substrate clearance, as has been shown for animals (13), remains to be established. The BAT-negative volunteers seemed to have lower postprandial insulin levels. These apparent differences in postprandial insulin levels were driven primarily by 1 patient with curiously low insulin levels after the meal.