Improvements in Glucose Metabolism and Insulin Sensitivity with a Low-Carbohydrate Diet in Obese Patients with Type 2 Diabetes

Read Also

: A low carbohydrate diet was well tolerated and achieved weight loss associated with a reduction in total energy intake over 24 weeks in subjects with type 2 diabetes.  Glycaemic control improved with a reduction in requirements for hypoglycaemic agents, and there were no clear overall adverse effects on cardiovascular risk factors or renal function.  This provides support for this dietary approach, at least in the short-term, for promoting weight loss in patients with type 2 diabetes. 

Although the “Atkins” dietary prescription does not include specific instruction to reduce total energy intake, this pilot study has confirmed the findings of previous studies that by removing carbohydrate from the diet, compensatory increases in fat and protein intake do not bring total energy intake back up to baseline.  This supports the conclusions of Boden et al that weight loss is not due to the generation of ketosis, but simply due to the reduction in energy intake [25].

There was a clear improvement in glucose metabolism, with reductions in hypoglycaemic medications, particularly insulin doses, and clinically and statistically significant reductions in HbA1c. This confirms the changes observed by Boden et al [25] over just a 7 day intervention and those of Nuttel and Gannon [22-24]. However, the reduction in HbA1c of 1.3% is particularly impressive in comparison, especially with the non-significant change seen by Davis et al [26]. There was a statistically significant improvement in fasting glucose, HOMA and AUC glucose during the IVGTT and non-significant improvements in fasting insulin.  These findings are similar to that of Nuttel et al [24] where a diet with carbohydrate reduced to 30% of energy intake, but specifically tailored to be weight neutral over 5 weeks, showed reduction in fasting glucose and integrated glucose over 24 hours with standard meals, but no changes in insulin.  The individuals in that study were poorly controlled with higher baseline fasting glucose and HbA1c and were not taking glucose lowering medication.  The explanation for the metabolic improvements was concluded to be related to the reduced absorbed carbohydrate, and possibly to reduced stored glycogen. No specific assessment of insulin sensitivity was made. It is important to note that whilst the weight loss was maintained at 24 weeks, the metabolic improvements in fasting glucose, AUC glucose and fasting insulin were more variable, and a larger study over a longer period of weight maintenance on this diet would be required to determine whether initial benefits can be sustained.

In the present study, not only were there changes in macronutrient composition of the diet, but subjects also lost weight.  Therefore, it is likely that improvements in glucose metabolism are a composite of the changes proposed by Nuttell et al but also changes in insulin sensitivity [24].  The lack of statistically significant improvements in fasting insulin and whole body insulin sensitivity (S_I) in this study may be due to several factors. In this very small sample the group data is skewed by 2 individuals. One (Number 8, Table 1) was considerably more insulin sensitive than the rest at baseline, had lower baseline insulin levels, was controlled with diet and was less overweight. This subject’s diabetes is likely to be driven more by reduced beta cell function than insulin resistance.  Another individual (Number 10, Table 1) was unable to successfully achieve or maintain a low carbohydrate diet.  Consequently, his total energy intake, fat and protein intakes all increased. This was associated with weight gain and worsening of glycaemic control, insulin sensitivity and lipid profile.  However, based on the changes seen in this small sample, a sample size of 35-50 subjects would have given 80-90% power to detect a significant improvement in insulin sensitivity.

The mean group data also hides somewhat greater individual variability in lipid profile changes. Although total cholesterol and LDL cholesterol increased, so did HDL cholesterol with no change in the ratio. Whilst there may be important changes in LDL particle size, not measured in this study, there is no evidence to suggest a major harmful effect on overall lipid profile. This supports the findings of Davis et al  who demonstrated a greater rise in HDL cholesterol with a low-carbohydrate diet compared with a low-fat diet, but no difference in other lipid parameters [26].  The reduction in triglycerides was non-significant unlike most reduced carbohydrate diets [22-24], but not seen by Davis [26], and probably reflects the small study cohort and the effect of non-adherence.   Some individuals had a deterioration in their overall lipid profile raising caution that low-carbohydrate high-fat diet may worsen lipid profiles. This effect can be particularly exacerbated if individuals are unsuccessful in losing weight, and as in an individual in this study, may be linked with further weight gain and risk of adverse health outcomes.

This study has confirmed that in people with type 2 diabetes an “Atkins style” low carbohydrate diet can achieve short-term weight loss through reduction in total energy intake, associated improvements in glucose metabolism without significant adverse effects. However, caution must remain over the long-term effects of a low carbohydrate diet, particularly if the source of protein and fat is animal based rather than vegetable based [21].  This concern is firstly whether individuals are able to maintain weight loss and health benefits by maintaining a low carbohydrate intake, and second whether adverse effects not seen in the short-term become apparent over a longer-term if individuals do maintain the diet.  The findings of this small study support the need for large randomised trials, specifically in those with diabetes, over a minimum of 2 years to examine the relative benefits and compliance of a low carbohydrate diet compared with the currently advised low fat diet.