glycemic index

as i've mentioned in several prior emails, there are pretty impressive data on the positive effects of a low glycemic index (GI) diet (ie one which documents the actual effect of different foods on the actual blood glucose levels). in general, high glycemic foods are associated with higher blood sugar levels and higher insulin levels in the early postprandial period, subsequently followed by low blood sugar levels (below the basal preprandial level), and associated increased hunger, food intake, and possibly weight.  also, high glycemic meals are associated with high triglycerides and lower HDL levels.  for an old but useful article, see glycemic index ludwig jama 2002 in dropbox (or JAMA 2002; 287:2414-2423) for a more detailed physiologic description.   subsequent data have consistently found that low glycemic diets improve cholesterol/HDL ratios and are at least as good as low fat diets in decreasing weight (interestingly, several studies have shown that an ad lib low glycemic index diet leads to the same calorie ingestion as a restricted-calorie low fat diet -- ie, as per the physiology noted above, the low fat diet, which typically is high in carbs, leads to increased appetite, so there needs to be calorie restriction to achieve the same calorie intake as with an unrestricted low glycemic index diet).  david ludwig (at boston childrens hosp and leader of their OWL weight loss clinic) is a major investigator and proponent of the glycemic index. another recent article by his group found that low fat diets tend to lower the basal metabolic state (ie, high carb foods make your body burn less fuel in a resting state), further tipping the scale in favor of a low glycemic diet.  and, now there is a recent article in am journal of clinical nutrition (see glycemic index food craving am j clin nutr 2013 in dropbox, or doi: 10.3945/ajcn.113.064113.) which looked at only a few subjects (12 overweight or obese men aged 18-35yo) in a crossover design and gave them high vs low GI diets, controlling for calories, macronutrients and palatability,  and looked at cerebral blood flow in MRI imaging 4 hours after the test meal, finding:

--as anticipated, the low GI meal was associated with lower blood sugar and lower serum insulin levels in the first 2-3 hours, then the high GI meal subjects had lower levels (ie, higher peaks early and lower troughs later with high GI meal), and the hunger rating was consistently higher with the high GI meal.

--MRI after the high GI meal showed higher activity in the right nucleus accumbens that then spread to other areas of the right striatum and to the olfactory area.  this area is an important reward and craving region (this area is also involved in substance abuse and dependence). 

of note, prior studies have also shown more activity in this area of the brain when obese vs lean individuals viewed or consumed palatable foods.  the authors also noted lower striatal dopamine D2 receptor activity in obese vs lean individuals, suggesting that overeating may compensate for low dopaminergic activity.

another angle, which might be important in the long run, is that insulin is a powerful trophic hormone, which has several different physiologic effects, including stimulating HMG CoA reductase activity (and thereby increasing LDL cholesterol levels locally), stimulating smooth muscle proliferation, increasing plasminogen activator inhibitor levels (and therefore being prothrombotic), increasing inflammation, enhanced sympathetic activity, enhanced renal tubular sodium retention -- all bad stuff for the heart (for review, see insulin resistance and heart disease arterioscl thromb vasc 2012 in dropbox, or doi: 10.1161/ATVBAHA.111.241885).  there have been several epidemiologic studies (eg the Quebec Study about 10 years ago) which found that blood insulin levels are at least as predictive of heart disease as blood sugar levels, and independent of them. i think this is the reason that metformin (which improves insulin action and decreases circulating insulin levels) is really the drug of choice for diabetics (sulfonylureas and insulin do not have the same cardioprotective effect).