Effects of Isocaloric Energy‐Restricted Diets with Different Levels of Carbohydrate on Body Composition and Metabolism in Rats

Previous studies on low carbohydrate diets, which are a popular approach to weight loss, have employed carbohydrate (C) levels that range from 5% to ~35% of energy intake. However, the effects of carbohydrate intake on body composition and metabolism remain unclear because differences in the subjects' total energy intake can confound interpretation of the effects of diet composition alone. We, therefore, compared the effects of energy‐restricted diets with different levels of carbohydrate but the same total energy intake, on male Sprague‐Dawley rats who were first made obese by 8 weeks of ad lib consumption of a high fat diet (HF, 60% of kcal as fat). When compared with control rats fed a low‐fat diet (LF, 15% fat), HF rats at wk 8 had greater % body fat (19± 2 vs. 11±1, p<0.001) and higher plasma levels of leptin (p < 0.001). Three modestly energy‐restricted (~ 70% of ad lib LF intake) diets were then fed for an additional 8 wk to 3 weight‐matched groups of the obese rats (10/group): 1) a Very Low Carbohydrate diet (VLC, 5 % starch, 60% fat), 2) a Low Carbohydrate diet (LC, 19% starch, 55% fat), or 3) a High Carbohydrate diet (HC, 55% starch, 15% fat); LC and VLC diets were similarly high in fat. An additional group of 10 obese rats continued to consume the HF diet ad lib (HF16). Body Composition Body weights of all energy‐restricted groups remained essentially constant during wk 8 – 16, whereas HF rats continued to gain weight (VLC~LC~HC~HF8 < HF16, p < 0.02); plasma leptin levels returned to that of the LF control in all restricted groups. By wk 16, total body fat and percent body fat decreased in the LC and HC but not VLC rats (LC~HC < VLC~HF8, p < 0.01). In contrast, liver lipid concentration decreased in VLC and HC but not LC rats (VLC~HC < LC~HF8, p < 0.05). Metabolic Indices Fasting (16 h) levels of beta‐hydroxybutyrate were similar in all groups, indicating that all diets were equally ketogenic (VLC~LC~HC~HF16). Although fasting plasma glucose levels did not differ among the four groups, fasting plasma insulin levels were lower in LC and VLC, but not HC, than in HF rats (HF8~HF16~HC > VLC~LC, p< 0.05). Fasting plasma triglyceride (TG) levels were higher in HC than VLC or LC rats (HC > VLC~LC, p < 0.05). Three hours after an oral glucose load, 1) plasma glucose levels remained elevated only in the HF group (VLC~LC~HC < HF16), 2) plasma insulin levels remained higher than in the fasted state only in the VLC group (p < 0.05), and 3) all groups showed similarly elevated plasma TG levels. The above findings show that the HC diet produced better body composition results than the VLC or LC diets, but that the VLC and LC diets produced somewhat better metabolic results than the HC diet, although all groups displayed some evidence of insulin resistance. The results also suggest that efforts to nearly eliminate carbohydrate intake offer no advantage over more moderate carbohydrate restriction when energy and fat intakes are similar. Support or Funding Information NIGMS SC3GM086298 PSC‐CUNY 4346100‐41