Efficacy of Primary Prevention Interventions When Fasting And

Clinical Care/Education/Nutrition/Psychosocial Research ORIGINAL ARTICLE

Efficacy of Primary Prevention Interventions When Fasting and Postglucose Dysglycemia Coexist Analysis of the Indian Diabetes Prevention Programmes (IDPP-1 and IDPP-2)

AMBADY RAMACHANDRAN, MD ANANTH SAMITH SHETTY, MBBS, MDRC done among subjects with IGT in differ- NANDITHA ARUN, MD CHAMUKUTTAN SNEHALATHA, DSC ent ethnic populations (9–14). Among these, only the Diabetes Reduction As- sessment with Ramipril and Rosiglitazone OBJECTIVE — Impaired glucose tolerance (IGT) and impaired fasting glucose (IFG) have Medication (DREAM) trial (12) recruited different pathophysiological abnormalities, and their combination may influence the effective- subjects with either isolated IFG (iIFG) or ness of the primary prevention tools. The hypothesis was tested in this analysis, which was done isolated IGT (iIGT) or both. Rosiglitazone in a pooled sample of two Indian Diabetes Prevention Programmes (IDPP-1 and IDPP-2). was found to be a potent agent in prevent- ing diabetes in this trial (12). The Diabe- RESEARCH DESIGN AND METHODS — Researchers analyzed and followed up on tes Prevention Programme (DPP) (9) the details of 845 of the 869 IGT subjects in the two studies for 3 years. Incidence of diabetes and recruited subjects with a fasting glucose reversal to normoglycemia (normal glucose tolerance [NGT]) were assessed in group 1 with baseline isolated IGT (iIGT) (n ϭ 667) and in group 2 with IGT ϩ IFG (n ϭ 178). The proportion in the range of 5.3–6.9 mmol/l (95–125 developing diabetes in the groups were analyzed in the control arm with standard advice mg/dl) and 2-h postglucose of 7.8–11 (IDPP-1) (n ϭ 125), lifestyle modification (LSM) (297 from both), metformin (n ϭ 125, IDPP-1), mmol/l (140–199 mg/dl) and nearly one- and LSM ϩ metformin (n ϭ 121, IDPP-1) and LSM ϩ pioglitazone (n ϭ 298, IDPP-2). Cox third of the participants had IFG by the regression analysis was used to assess the influence of IGT ϩ IFG versus iIGT on the effectiveness present criteria (15). of the interventions. Results of the Indian Diabetes Preven- tion Programme-1 (IDPP-1) showed that RESULTS — Group 2 had a higher proportion developing diabetes in 3 years (56.2 vs. 33.6% a moderate lifestyle modification (LSM) ϭ ϭ in group 1, P 0.000) and a lower rate of reversal to NGT (18 vs. 32.1%, P 0.000). Cox or a small dose of metformin (500 mg/ regression analysis showed that effectiveness of intervention was not different in the presence of day) reduced the risk of diabetes in a rel- fasting and postglucose glycemia after adjusting for confounding variables. atively nonobese but insulin resistant CONCLUSIONS — The effectiveness of primary prevention strategies appears to be similar Asian Indian population (13). In the in subjects with iIGT or with combined IGT ϩ IFG. However, the possibility remains that a larger IDPP-2 study, we noted that pioglitazone study might show that the effectiveness is lower in those with the combined abnormality. did not improve the efficacy of LSM in Asian Indians (14). In both studies, sub- Diabetes Care 33:2164–2168, 2010 jects with persistent IGT and fasting glucose levels below 6.9 mmol/l were mpaired glucose tolerance (IGT) and Analysis of six prospective studies recruited. Therefore, some participants impaired fasting glucose (IFG) have a among subjects with IGT showed that the also had IFG. In view of the higher degree I high potential to convert to type 2 dia- incidence of diabetes varied widely from of biochemical abnormalities occurring betes. While an elevated basal hepatic 23 to 62% within two to twenty-seven when fasting and postprandial dysglyce- glucose output and impaired early phase years of follow-up (3). The incidence was mia coexisted, it was considered impor- insulin secretion are the major abnormal- higher among populations with high tant to study whether the combined ities in IFG, IGT is characterized by more prevalence of diabetes than in white pop- abnormalities influenced the cumulative severe muscle insulin resistance (IR) and ulations. Incidence rates of diabetes in incidence of diabetes in comparison with defects in late insulin secretion (1). subjects with IFG or IGT or with a com- subjects with iIGT. To increase the sam- Among Asian Indians, higher degrees of bined abnormality were varied in differ- ple size, data from both IDPP studies were IR and ␤-cell dysfunction are seen in IFG ent populations (4–8). pooled. The participants’ baseline charac- than in IGT (2). Primary prevention studies have been teristics were identical in the two studies. ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● RESEARCH DESIGN AND From the India Diabetes Research Foundation and Dr. A. Ramachandran’s Diabetes Hospitals, Chennai, METHODS — IDPP-1 and IDPP-2 India. Corresponding author: Ambady Ramachandran, [email protected]. were 3-year prospective, randomized Received 25 June 2009 and accepted 26 May 2010. Published ahead of print at http://care.diabetesjournals. controlled studies among Asian Indian org on 2 June 2010. DOI: 10.2337/dc09-1150. subjects with persistent IGT (13,14). © 2010 by the American Diabetes Association. Readers may use this article as long as the work is properly IDPP-1 had four groups: 1) control with cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons. standard advice, 2) LSM, 3) treated with org/licenses/by-nc-nd/3.0/ for details. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby metformin (500 mg/day), and 4) a com- marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. bination of LSM and metformin (13).

2164 DIABETES CARE, VOLUME 33, NUMBER 10, OCTOBER 2010 care.diabetesjournals.org Ramachandran and Associates

Figure 1—The selection of original cohorts for the two studies, randomization, and the ﬁnal outcome in the available subjects at the 3rd-year follow-up are shown. The selection criteria for this analysis and the baseline distribution of subjects in group 1 (iIGT) and group 2 (IGT ϩ IFG) are also shown.

IDPP-2 was done in a different cohort of (28.5% with LSM, 26.4% with met- (2-h plasma glucose 7.8–11.0 mmol/l, IGT subjects using LSM and placebo as formin, and 28.2% with LSM and met- 140–199 mg/dl) and group 2 as those the control group and LSM and pioglita- formin) when compared with the control with combined IFG (fasting 6.1–6.9 zone (30 mg/day) as the intervention group (13). In IDPP-2, the cumulative in- mmol/l, 110–125 mg/dl) and IGT. Data group (14). cidences of diabetes at 36 months were were available for 667 subjects in group 1 Sample selection for this analysis is similar in the intervention (29.8%) and and 178 subjects in group 2 (Fig. 1). shown in the flow chart in Figure 1 (Fig. placebo groups (31.6%) (14). In this study, the group with standard 1). The two-stage selection procedure was BMI, waist circumference, and blood care from IDPP-1 was defined as the con- used for recruiting reproducible IGT pressure were measured at baseline and trol group. The LSM group included LSM only. No case of isolated IFG was selected, during each review. Plasma glucose was from IDPP-1 and the placebo group from and the presence of IFG was not an inclu- measured (glucose oxidase method) at IDPP-2. The effect of LSM ϩ drugs (met- sion criterion. fasting, 30 min, and 2 h during the formin and pioglitazone) was analyzed In both studies the primary outcome OGTT, and corresponding plasma insulin since the numbers with IGT ϩ IFG were was the development of diabetes detected was measured using the radioimmunoas- small in the individual drug group, and by a standard oral glucose tolerance test say kit of DiaSorin (Saluggia, Italy). In- the outcome measures were present in (OGTT) (fasting plasma glucose Ն7.0 dexes of insulin resistance (homeostasis numbers inadequate for statistical mmol/l and/or 2-h post glucose Ն11.1 model assessment of insulin resistance comparisons. mmol/l) (15). Reversal to normal glucose [HOMA-IR]) (16) and early insulin secre- tolerance (fasting plasma glucose Ͻ6.1 tion [(30-min fasting insulin [pmol/l]) Ϭ Statistical analysis mmol/l and 2-h plasma glucose tolerance 30-min glucose (mmol/l) (index of insu- Means and SD are shown for normally Ͻ7.8 mmol/l) was also considered as an lin secretion [␦ I/G])] were calculated distributed variables. Student t test was outcome. All subjects underwent annual (17). Subjects from both studies having used for intergroup comparison. Median OGTT. A semiannual postprandial capil- all the above measurements were in- values were used for skewed variables, lary glucose test was done. Diabetes de- cluded for this analysis. Among the total and Mann-Whitney U test was used for tected in any person was confirmed with of 869 followed up for 3 years in both the comparisons. Intergroup proportions an OGTT. studies, data on plasma insulin was not were compared using the ␹2 test. Kaplan- In IDPP-1, in a median follow-up of available for 24 subjects. Hence this anal- Meier survival analysis was used to cal- 30 months, the relative risk reductions ysis was done using the data of 845 sub- culate the probability of cumulative with the interventions were similar jects. Group 1 was defined as having iIGT incidence of diabetes in the groups. Log- care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 10, OCTOBER 2010 2165 Diabetes prevention in combined IGT and IFG

-Table 1—Comparison of characteristics of study subjects with iIGT (group 1) and IGT ؉ IFG reversal to normal glucose tolerance in re (group 2) lation to the control group (P ϭ 0.000). In group 2, none of the intervention meth- Variables Group 1 Group 2 ods produced a significant benefit. The crude effect of all interventions on the in- n (%) 667 (78.9) 178 (21.1) cidence of diabetes appeared to be stron- Men:Women 559:108 140:38 ger among the subjects in group 1 (hazard Age (years)* 45.5 Ϯ 6.0 46.0 Ϯ 5.8 ratio 0.547 [95% CI 0.400–0.747], P ϭ BMI (kg/m2)* 25.7 Ϯ 3.2 26.4 Ϯ 3.8 0.000) than in group 2 (0.792, [0.470– Waist circumference (cm)* 89.6 Ϯ 8.2 91.3 Ϯ 8.4 1.335], P ϭ 0.382) when compared with Blood pressure (mmHg)* the control group. Systolic 120.4 Ϯ 13.7 119.4 Ϯ 12.5 Table 3 shows the results of the Cox Diastolic 75.0 Ϯ 9.6 75.6 Ϯ 10.2 regression models. The first model Plasma glucose (mmol/l)* showed that the incidence of diabetes was Fasting 5.2 Ϯ 0.6 6.4 Ϯ 0.2** reduced by the interventions. The effec- 30 min 9.4 Ϯ 1.7 10.6 Ϯ 1.7** tiveness of intervention appeared to be 120 min 8.4 Ϯ 1.0 9.1 Ϯ 1.3** significantly more in the iIGT group than Plasma insulin (pmol/l)† in the group with IFG ϩ IGT. The second Fasting 108 114 model showed lack of interaction be- 30 min 480 420 tween the groups and intervention. The 120 min 618 612 results were similar even after adjusting HOMA-IR† 4.2 5.5** for age, sex, BMI, and IR. After adjusting ␦ I/G† 39.4 28.0** for the confounding variables, the effec- *Means Ϯ SD, †median values,**P ϭ 0.000 vs. group 1. tiveness of intervention did not differ between the two study groups. rank test was used for calculating P val- Comparative analysis of the outcomes CONCLUSIONS — In the Asian In- ues. Cox proportional hazard model in the groups in 3 years is shown in Table dian subjects, the prevention strategies (enter method) was used to study the ef- 2 in relation to the interventions. In significantly decreased the cumulative in- fect of independent variables on the inci- LSM ϩ drug group, the incidence of dia- cidence of diabetes in comparison with dence of diabetes. Initially, the crude betes was significantly lower in group 1 the control group both in the group with effect of all interventions versus the con- when compared with group 2 (P ϭ iIGT and in the group with IFG ϩ IGT as trol group was assessed separately by Cox 0.000). In group 1, LSM and LSM ϩ met- shown by the Cox regression analysis. regression analysis in group 1 and group formin significantly reduced the inci- The crude effect of all interventions on the 2. Three models were computed as fol- dence of diabetes and increased the incidence of diabetes appeared to be lows. In the first model, all interventions versus the control group (reference) and group 2 versus group 1 (reference) were Table 2—Glycemic outcome up to 3 years in relation to interventions entered as independent variables. In the second model, the interaction of the two Outcome iIGT (group 1) IGT ϩ IFG (group 2) variables was also included. In the third model, the variables in the second model Control were included after adjusting for age n 99 26 (years), sex, BMI (kg/m2), and IR. Fasting NGT 14 (14.1) 4 (15.4) values of glucose and insulin were not in- IGT 34 (34.3) 5 (19.2) cluded as HOMA-IR was calculated using Diabetes 51 (51.5) 17 (65.4) these parameters. Statistical analyses were LSM done using SPSS version 10.0 (SPSS, Chi- n 224 73 cago, IL). A P value of Յ0.001 was con- NGT 80 (35.7)* 15 (20.5) sidered significant. IGT 78 (34.8) 25 (34.2) Diabetes 66 (29.5)* 33 (45.2) RESULTS — Baseline characteristics of Drug (metformin) subjects in group 1 and group 2 are n 106 19 shown in Table 1. There was an excess of NGT 29 (27.4)* 1 (5.3) males in both the original studies; hence, IGT 36 (33.9) 4 (21.0) an overrepresentation of men is also seen Diabetes 41 (38.7) 14 (73.7) in this analysis. Subjects with combined LSM ϩ drug abnormalities (group 2) had higher n 238 60 plasma glucose concentrations (P ϭ NGT 91 (38.2)* 12 (20.0) 0.000), higher HOMA-IR, and lower ␦ I/G IGT 81 (34.0) 12 (20.0) values than the subjects with iIGT (group Diabetes 66 (27.7)* 36 (60.0)** 1) (P ϭ 0.000). Data are n (%). Intragroup comparison: *P ϭ 0.000 vs. control; intergroup comparison: **P ϭ 0.000.

2166 DIABETES CARE, VOLUME 33, NUMBER 10, OCTOBER 2010 care.diabetesjournals.org Ramachandran and Associates

Table 3—Results of Cox regression analyses (dependent variable: diabetes) IDPP studies were originally designed to analyze the impact of prevention strat- Hazard egies in subjects with IGT. Subjects with Independent variable ␤ ratio 95% CI P isolated IFG were not selected, and we did not aim to confirm the presence of IFG in Model 1 both steps of screening. Intervention vs. control 1 Ϫ0.5 0.61 0.47–0.80 0.000 The original IDPP cohorts had a male Group 2 vs. group 1 0.69 1.99 1.57–2.52 0.000 excess. Therefore, possible differences in Model 2 the compliance to LSM among men and Intervention vs. control Ϫ0.60 0.55 0.40–0.75 0.000 women could not be assessed. Group 2 vs. group 1 0.4 1.49 0.86–2.58 0.16 In summary, interventions for the pri- Intervention vs. control ϫ mary prevention of diabetes work effec- group 2 vs. group 1 0.36 1.44 0.78–2.64 0.24 tively in subjects having either iIGT or a Model 3 combination of IGT ϩ IFG. Although this Age (years) 0.001 1.00 0.98–1.02 0.97 study cannot confirm that the primary Sex Ϫ0.45 0.64 0.46–0.89 0.01 prevention of diabetes is more effective in BMI (kg/m2) 0.02 1.02 0.99–1.06 0.173 people with iIGT than in those with HOMA-IR 0.01 1.01 0.97–1.05 0.65 IFG ϩ IGT, the possibility remains that Intervention vs. control Ϫ0.64 0.53 0.39–0.72 0.000 such an effect may become apparent in a Group 2 vs. group 1 0.32 1.38 0.78–2.43 0.27 larger study or a meta-analysis. Intervention vs. control ϫ group 2 vs. group 1 0.45 1.58 0.85–2.92 0.15 Acknowledgments— No potential conflicts of interest relevant to this article were stronger among those with iIGT than higher baseline IR and/or low ␤-cell func- reported. ϩ A.R. researched, analyzed, discussed, and those with IFG IGT. However, a test of tion had poor outcomes (18). It had been edited the manuscript; N.A. contributed to the the interaction of the intervention effect suggested that the differences in insulin discussion and reviewed the manuscript; by glycemic status was not statistically sensitivity and insulin secretion between A.S.S. contributed to the discussion and re- significant. IGT and IFG and the greater severity of viewed the manuscript; and C.S. researched, Among the primary prevention stud- the abnormalities when both coexist analyzed, discussed, and prepared the ies, the DPP (9) had nearly one-third of its might predict different rates of progres- manuscript. participants having IFG by the present di- sion to diabetes, and different pharmaco- We acknowledge the valuable help given by agnostic criteria (15). In the placebo logical agents might be needed to treat the the following Indian Diabetes Prevention Pro- group, cumulative incidence of diabetes pathophysiology (19). gramme team members in conducting the field was more than threefold higher when the Prospective studies in different popu- survey: Mary Simon, C.K. Sathish Kumar, S. Selvam, A. Catherin Seeli, M. Muruganandam, participants had basal fasting plasma glu- lations have demonstrated higher rates of V. Lalitha Manjula, and L. Subhashini. We also cose in the range of 6.1–6.9 mmol/l ver- development of diabetes in subjects hav- thank L. Vijaya for helping with the statistical sus those who had lower values. In the ing combined IFG and IGT than in sub- analysis and providing secretarial assistance. former group, the relative risk reductions jects having either of the glycemic with interventions were also lower, more abnormalities (4–6). The highest propor- so with metformin (9). tion of diabetes development among sub- References In the DREAM trial, the annual con- jects with IFG and IGT (72.7%) was 1. DeFronzo RA. Pharmacologic therapy for version to diabetes in the placebo group reported in a Brazilian Japanese popula- type 2 diabetes mellitus. Ann Intern Med was almost double in the participants tion (7). Differences in follow-up periods 1999;131:281–303 with combined glycemic abnormalities and racial/ethnic differences among the 2. Snehalatha C, Ramachandran A, Sivasan- than those with isolated abnormalities study subjects account for the varied kari S, Satyavani K, Vijay V. Insulin se- (12). However, the primary outcome with results. cretion and action show differences in impaired fasting glucose and in impaired rosiglitazone was much the same, irre- The strength of the analysis lies in the glucose tolerance in Asian Indians. Dia- spective of the glycemic abnormality prospective nature of our studies. By bete Metab Res Rev 2003;19:329–332 present at randomization. combining the results of two studies using 3. Edelstein SL, Knowler WC, Bain RP, An- We have not studied the effectiveness different interventions, we could assess dres R, Barrett-Connor EL, Dowse GK, of pioglitazone in isolation. In our study, the impact of LSM and insulin sensitizers Haffner SM, Pettitt DJ, Sorkin JD, Muller pioglitazone did not improve the effec- on the incidence of diabetes. However, DC, Collins VR, Hamman RF. Predictors tiveness of LSM (14). It might be that LSM the small sample sizes in the control and of progression from impaired glucose tol- had produced the maximum possible metformin groups posed some limita- erance to NIDDM: an analysis of six pro- benefit on the glycemic status, and hence tions. Moreover, due to the small num- spective studies. Diabetes 1997;46:701– no further improvement was seen by add- bers of subjects with combined IGT and 710 4. de Vegt F, Dekker JM, Jager A, Hienkens E, ing an insulin sensitizer (13,14). IFG, the analysis could not be done sepa- ϩ ϩ Kostense PJ, Stehouwer CD, Nijpels G, The beneficial outcomes caused by rately in LSM metformin and LSM Bouter LM, Heine RJ. Relation of impaired the intervention strategies in the IDPP-1 pioglitazone. The small number in group fasting and postload glucose with incident occurred due to improved insulin action 2 might have influenced the results of uni- type 2 diabetes in a Dutch population: the and insulin sensitivity. Subjects with variate analyses. Hoorn Study. JAMA 2001;285:2109–2113 care.diabetesjournals.org DIABETES CARE, VOLUME 33, NUMBER 10, OCTOBER 2010 2167 Diabetes prevention in combined IGT and IFG

5. Eschwe`ge E, Charles MA, Simon D, Thib- kumaki M, Mannelin M, Martikkala V, tiveness of lifestyle modification in pre- ult N, Balkau B, Paris Prospective Study. Sundvall J, Uusitupa M. The Finnish Dia- venting conversion of impaired glucose Reproducibility of the diagnosis of diabe- betes Prevention Study Group. Preven- tolerance to diabetes in Asian Indians: re- tes over a 30-month follow-up: the Paris tion of type 2 diabetes mellitus by changes sults of the Indian Diabetes Prevention Prospective Study. Diabetes Care 2001; in lifestyle among subjects with impaired Programme-2 (IDPP-2). Diabetologia 24:1941–1944 glucose tolerance N Engl J Med 2001;344: 2009;52:1019–1026 6. Vaccaro O, Ruffa G, Imperatore G, Iovino 1343–1350 15. World Health Organization. Definition, V, Rivellese AA, Riccardi G. Risk of diabe- 11. Pan XR, Li GW, Hu YH, Wang JX, Yang Diagnosis and Classification of Diabetes tes in the new diagnostic category of WY, An ZX, Hu ZX, Lin J, Xiao JZ, Cao Mellitus and Its Complications. Part 1: Diag- impaired fasting glucose: a prospective HB, Liu PA, Jiang XG, Jiang YY, Wang JP, nosis and Classification of Diabetes Mellitus. analysis. Diabetes Care 1999;22:1490– Zheng H, Zhang H, Bennett PH, Howard WHO/NCD/NCS/99.2, Geneva: World 1493 BV. Effects of diet and exercise in prevent- Health Organization, 1999 7. Gimeno SG, Ferreira SR, Franco IJ, Iunes ing NIDDM in people with impaired glu- 16. Matthews DR, Hosker JP, Rudenski AS, M. Comparison of glucose tolerance cate- cose tolerance: the Da Qing IGT and Naylor BA, Treacher DF, Turner RC. Ho- gories according to World Health Organi- Diabetes Study. Diabetes Care 1997;20: meostasis model assessment: insulin re- zation and American Diabetes Association 537–544 sistance and ␤-cell function from fasting diagnostic criteria in a population based 12. DREAM (Diabetes REduction Assessment plasma glucose and insulin concentra- study in Brazil. The Japanese-Brazilian Di- with ramipril and rosiglitazone Medica- tions in man. Diabetologia 1985;28:412– abetes Study Group Diabetes Care 1998; tion) Trial Investigators, Gerstein HC, 419 21:1889–1892 Yusuf S, Bosch J, Pogue J, Sheridan P, 17. Wareham NJ, Philips DIW, Byrne CD, 8. Gabir MM, Hanson RL, Dabelea D, Im- Dinccag N, Hanefeld M, Hoogwerf B, Hales CN. The 30 minute insulin incre- peratore G, Roumain J, Bennett PH, Laakso M, Mohan V, Shaw J, Zinman B, mental response in an OGTT as a measure Knowler WC. The 1997 American Diabe- Holman RR. Effect of rosiglitazone on the of insulin secretion. Diabet Med 1995;12: tes Association and 1999 World Health frequency of diabetes in patients with im- 931 Organization criteria for hyperglycemia in paired glucose tolerance or impaired fast- 18. Snehalatha C, Mary S, Selvam S, Sathish the diagnosis and prediction of diabetes. ing glucose: a randomised controlled trial. Kumar CK, Shetty SB, Nanditha A, Ram- Diabetes Care 2000;23:1108–1112 Lancet 2006;368:1096–1105 achandran A. Changes in insulin secre- 9. Knowler WC, Barrett-Connor E, Fowler 13. Ramachandran A, Snehalatha C, Mary S, tion and insulin sensitivity in relation to SE, Hamman RF, Lachin JM, Walker EA, Mukesh B, Bhaskar AD, Vijay V, Indian the glycemic outcomes in subjects with Nathan DM. Reduction in the incidence of Diabetes Prevention Programme (IDPP). impaired glucose tolerance in the Indian type 2 diabetes with lifestyle intervention The Indian Diabetes Prevention Pro- Diabetes Prevention Programme-1 (IDPP- or metformin. Diabetes Prevention Pro- gramme shows that lifestyle modification 1). Diabetes Care 2009;32:1796–1801 gram Research Group N Engl J Med 2002; and metformin prevent type 2 diabetes in 19. Abdul-Ghani MA, Jenkinson CP, Rich- 346:393–403 Asian Indian subjects with impaired glu- ardson DK, Tripathy D, DeFronzo RA. In- 10. Tumoilehto J, Lindstrom J, Eriksson JG, cose tolerance (IDPP-1). Diabetologia sulin secretion and action in subjects with Valle TT, Hamalainen H, Ilanne-Parikka 2006;49:289–297 impaired fasting glucose and impaired P, Keinanen-Kiukanniemi S, Laakso M, 14. Ramachandran A, Snehalatha C, Mary S, glucose tolerance: results from the Veter- Lauheranta A, Rastas M, Salminen V, Selvam S, Kumar CK, Seeli AC, Shetty AS. ans Administration Genetic Epidemiol- Aunola S, Cepaitis Z, Moltchanov V, Ha- Pioglitazone does not enhance the effec- ogy Study. Diabetes 2006;55:1430–1435

2168 DIABETES CARE, VOLUME 33, NUMBER 10, OCTOBER 2010 care.diabetesjournals.org