The authors searched the following databases: PubMed (through July 2010), Cochrane Library, Natural Standard, and Natural Medicines. They excluded studies with healthy subjects, type 1 diabetics, or patients with average glycosylated hemoglobin A1c (A1C) levels below 7%. They recovered 144 studies, and seven met the inclusion criteria, including three randomized crossover studies and four randomized parallel group studies, three of which were blinded. Five studies enrolled patients with T2DM, and two enrolled patients with T2DM and dyslipidemia.

A placebo-controlled, randomized crossover study (n=18) examined the effect of psyllium fiber on PPG. The subjects had T2DM controlled by diet (n=6) or oral hypoglycemic medication (n=12). They took two 6.8 g psyllium fiber packets or placebo packets dissolved in water before standardized breakfasts and dinners containing 14 g of dietary fiber. Compared to baseline fasting blood glucose (FBG) values, PPG increased by 107.7 ± 11.6 mg/dL for the psyllium treatment and by 125.4 ± 11.1 mg/dL for the placebo after breakfast (P=0.08). Similarly, PPG increased compared to FBG levels by 53.2 ± 7.5 mg/dL for psyllium and 67.1 ± 11.1 mg/dL for the placebo after dinner (P=0.06). Serum insulin levels after breakfast and the area under the curve (AUC) were significantly lower for psyllium compared to the placebo (P<0.05 for both). A "second meal effect" was observed at lunchtime when the patients continued to show lower AUC and PPG values after psyllium at breakfast compared to the placebo (P<0.05 for both). 

Another randomized crossover study (n=12) compared the effects of psyllium and acarbose on PPG and the glycemic index of bread. Seven subjects were taking glibenclamide, and five were taking tolbutamide. The researchers gave the patients 90 g of white bread alone (control), as well as white bread preceded by a 15 g dose of psyllium and a 200 mg dose of acarbose (twice the recommended doses for both). The glycemic index of the bread was 26.1 ± 13.4 for acarbose and 58.9 ± 10.1 for psyllium (P<0.05). Both psyllium and acarbose reduced PPG levels compared to the control (P<0.05 for both). The third crossover trial (n=13, d=6 weeks) compared the effects of a diet with the American Diabetes Association (ADA)-recommended level of dietary fiber (8 g/day soluble fiber, 16 g/day insoluble fiber) and a high fiber diet (25 g/day soluble fiber, 25 g/day insoluble fiber) on average plasma glucose (APG) levels. The researchers enrolled three patients treated by diet alone and 10 who took glyburide. The high fiber diet resulted in lower APG values than the ADA-recommended diet (P<0.05), as well as "slightly lower" A1C levels that were not statistically significant.  

A randomized, controlled, parallel group study (n=40, d=6 months) compared the effects of the ADA-recommended diet with a low glycemic index diet (total daily glycemic index score: 55) on A1C levels in T2DM patients on oral hypoglycemic medications or injectable insulin. There was no significant intergroup difference for the change in the glycemic index. Both groups showed a significant decrease in A1C levels from baseline (P<0.001 for both) with no significant difference between the diets. The reviewers commented that the sample size was too small to detect a significant difference in A1C levels and that there was a significant age difference between the groups at baseline. A randomized blinded study (d=24 weeks) compared the effect of a low glycemic index diet (glycemic index<62, 18.7 g fiber/1,000 kcal) with a high fiber cereal diet (mean glycemic index: 86, 15.7 g fiber/1,000 kcal) on A1C values in T2DM patients (n=210) treated with oral hypoglycemic drugs or α-glucosidase inhibitors. After adjusting for intergroup differences in fiber intake, average A1C levels were significantly lower for the low glycemic index diet compared to the high fiber cereal diet (P<0.05). Average fasting plasma glucose (FPG) levels decreased from baseline for both diets (P<0.05 for both). The reviewers noted a confounding factor in this study: the subjects in the low glycemic index diet group were allowed to eat All Bran Buds cereal, which contains psyllium fiber.

A randomized blinded study (d=8 weeks) compared the effects of 10.2 g/day psyllium with a cellulose placebo taken 20-30 minutes before breakfast and dinner on PPG in men (n=34) with T2DM and dyslipidemia controlled by diet or oral sulfonylurea drugs. Psyllium decreased all-day PPG values compared to baseline levels while the placebo increased them (P<0.05 intergroup difference). The PPG after lunch also decreased compared to baseline levels in the psyllium group and increased in the placebo group (P<0.05 intergroup difference). A1C decreases observed in both groups were not statistically significant. The reviewers commented that the short duration may have caused the lack of statistical difference in A1C levels. Another randomized blinded study (d=8 weeks) compared the effects of 10.2 g/day of psyllium and a cellulose placebo taken 20-30 minutes before breakfast and dinner on A1C levels in men (n=49) with T2DM and dyslipidemia controlled by diet and/or medication. The study showed significant decreases in A1C levels for the psyllium group and increases for the placebo group (P<0.05 intergroup difference). Similarly, FPG levels decreased in the psyllium group and increased in the placebo group (P<0.05 intergroup difference). Five patients in the placebo group dropped out due to adverse effects. No serious adverse effects were reported in the other studies. 

The authors conclude, "Results from seven randomized, controlled clinical trials suggest that psyllium fiber or a low glycemic index diet may improve glycemic risk factors in persons with T2DM." The studies on psyllium's effect on A1C levels showed "inconsistent" results, with two studies showing A1C decreases and one showing no significant effect, although the last study may have been too short in duration to reveal significant changes in A1C levels. The studies on low glycemic index diets and A1C levels were "inconclusive" because one was underpowered and one included a cereal containing psyllium in the diet. Additional research is needed on the effects of psyllium and low glycemic diets on A1C levels. Additional research is also needed to determine if there is a significant difference between the effects of acarbose and psyllium on PPG levels. Research is warranted on the effect of psyllium supplementation on PPG levels in patients with pre-diabetes.

The reviewers recommend that consumers gradually add psyllium to their diets to decrease adverse gastrointestinal effects and that patients allergic to psyllium avoid taking it. They also recommend waiting at least four hours after taking psyllium before taking medications because psyllium can decrease their absorption. Adverse effects are rarely associated with low glycemic index diets, but hypoglycemia has been reported.