Lifestyle and medication interventions for the ... - Wiley Online Library

3 downloads 98165 Views 235KB Size Report
One reviewer (AY) used Adobe Photoshop CS312 to estimate the ..... T2DM drug classes. ... Additionally, none of the included studies gave resistance training.
Controversial issues

Article

Lifestyle and medication interventions for the prevention or delay of type 2 diabetes mellitus in prediabetes: a systematic review of randomised controlled trials Abstract

Agnes Yuen

Objective: To assess lifestyle and pharmacological interventions aiming to delay type 2 diabetes mellitus (T2DM) in prediabetes. Methods: We searched the Cochrane

Emergency Practice Innovation Centre, St Vincent’s Hospital Melbourne and Faculty of Medicine, Dentistry and Health Sciences, The University of Melbourne, Victoria

Yulia Sugeng Emergency Practice Innovation Centre, St Vincent’s Hospital Melbourne, Victoria

Register of Controlled Trials, MEDLINE,

Tracey J. Weiland and George A. Jelinek

EMBASE, CINAHL, PsycINFO, Web of Science, BIOSIS and LILACS

Emergency Practice Innovation Centre, St Vincent’s Hospital Melbourne and Faculty of Medicine, Dentistry and Health Sciences, The University of Melbourne, Victoria

databases, examined reference lists and contacted authors. We included randomised controlled trials (RCTs) on both lifestyle and medication interventions in prediabetes. These studies were at least 12 month duration and aimed to delay T2DM. Results: Four studies investigating lifestyle and medication with a total of 5,196 participants were identified. There was a high risk of bias in the studies and the interventions utilised varied considerably; thus, meta-analysis was not undertaken. The comparison between lifestyle and medication interventions was largely dependent on the intensity of the lifestyle program while we could not adequately assess their effects on cardiovascular morbidity. Adverse events with metformin and acarbose were common. Conclusion: There is substantial evidence that intensive lifestyle programs and medications delay T2DM in impaired glucose tolerance though it remains unclear which is more effective. Implications: Both interventions seem to be able to delay T2DM. However, both have issues with adherence and side effects and more RCTs are required. Key words: Diabetes mellitus, Type 2, prediabetic state, prevention and control, lifestyle intervention, pharmacological intervention. Aust NZ J Public Health. 2010; 34:172-8 doi: 10.1111/j.1753-6405.2010.00503.x

172

T

ype 2 diabetes mellitus (T2DM) is a worldwide health issue. Approximately 171 million people had T2DM in 2000 and a predicted 366 million will be affected by 2030. 1 Economically it accounts for 10% or more of healthcare budgets in developed countries,2 mostly due to the significantly higher risk of cardiovascular disease (CVD) as well as cerebrovascular and peripheral vascular disease. Additionally, T2DM is a major cause of blindness and renal failure and can necessitate amputation. Clearly, primary prevention of T2DM is needed and should target prediabetes defined by either impaired glucose tolerance (IGT) or impaired fasting glucose (IFG). Approximately 3 to 10% of people with prediabetes progress to T2DM annually3 but with lifestyle changes or medication, this can be delayed.4-9 A range of interventions aiming to delay or prevent diabetes has been examined, most notably medication or lifestyle changes or both.4 Several recent systematic reviews and meta-analyses on T2DM prevention provide indirect evidence for the comparison between Submitted: August 2008

medication and lifestyle interventions as all the reviews compared medication with placebo or lifestyle with standard care.4-11 This systematic review focuses on the effectiveness of these two interventions as well as the interventions combined.

Objectives To assess the effects of lifestyle compared to medication and the combination of lifestyle and medication (LSM) compared to medication or lifestyle alone where, in people with prediabetes, all interventions aimed to prevent, delay or revert T2DM to normoglycaemia.

Methods Inclusion criteria Study selection was restricted to RCTs that followed participants for at least one year, investigating IGT or IFG or both. As the criteria for IGT and IFG have changed over time, we accepted studies that specified their criteria for IGT or IFG in detail. We accepted studies that compared interventions involving medications such as oral hypoglycaemics and anti-obesity agents

Revision requested: May 2009

Accepted: September 2009

Correspondence to: Dr Tracey Weiland, Emergency Practice Innovation Centre, St. Vincent’s Hospital Melbourne, 41 Victoria Pde, Fitzroy VIC 3065. Fax: 03 9288 author please provide; e-mail: [email protected]

AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH © 2010 The Authors. Journal Compilation © 2010 Public Health Association of Australia

2010 vol. 34 no. 2

Controversial issues

Interventions for the prevention or delay of type 2 diabetes

with lifestyle interventions still on the market. Lifestyle interventions were defined as those offering more than what is considered ‘standard care’. Standard care usually involves the provision of general information, while we were interested in more intensive care such as individually tailoring information and goal setting. Our primary outcome measures were the incidence of T2DM and normal glucose tolerance (NGT). We included studies that identified T2DM or NGT using ranges of two hour plasma glucose (2hPG) and fasting plasma glucose (FPG) that were specified in detail. We extracted available data for secondary outcomes (Table 1).

Search methods for identification of studies We searched the Cochrane Register of Controlled Trials, MEDLINE, EMBASE, CINAHL, PsycINFO, Web of Science, BIOSIS and LILACS databases. We also identified ongoing trials by searching trial registries. The search strategies linked filters for RCTs, prediabetes, T2DM prevention and lifestyle interventions. We examined the reference lists of included studies and relevant reviews and meta-analyses. We also attempted to contact authors of potentially eligible trials where there was insufficient information provided for inclusion. Furthermore, we corresponded with authors of included trials and relevant systematic reviews to obtain information on unpublished and ongoing trials. No language restrictions were imposed. Where needed, one reviewer (AY) assessed the eligibility of the study and extracted data with the help of a translator. Table 1: Secondary outcomes assessed.

Study selection Two reviewers (AY, YS) independently scanned titles and abstracts for potentially eligible trials. The full article was retrieved where the title and abstract alone were insufficient for exclusion or where there was disagreement. Articles describing trials of people at risk of T2DM were fully retrieved. Two reviewers independently screened full text articles for eligibility. They were not blinded to information such as journal names or authors. Disagreements were resolved through consensus.

Data extraction Two reviewers (AY, YS) independently extracted data from included studies using a pre-formed data collection tool except for secondary outcomes which were extracted by one reviewer (AY). Disagreements were discussed and, where necessary, authors were contacted. Data extraction was not blinded. Where data were presented in graphical form only, we determined point estimates from enlargements of the published graphs. One reviewer (AY) used Adobe Photoshop CS312 to estimate the coordinates of data points at 1600% magnification. The results for the primary outcomes were entered into Review Manager 5.0.113 and checked by two reviewers.

Assessment of risk of bias Two authors (AY, YS) independently assessed internal validity using The Cochrane Collaboration’s tool for assessing risk of bias.14 We only assessed blinding in the medication arms and chose not to consider blinding of outcome assessors. Disagreement was resolved through consensus, attempts to contact authors and consultation with a third reviewer (TW).

Secondary outcome

Measurements

Measures of treatment effect

• Mortality • Incidence of CVD morbidity • Glycaemic control • Plasma lipids • Blood pressure • Insulin • Body weight • Adverse effects • Compliance • Quality of life

• Total deaths or rate of death • Number or rate of events • Glycated haemoglobin levels (%) • Fasting blood glucose levels (mmol/L) • 2 hour plasma glucose levels (mmol/L) • Total cholesterol • High-density lipoprotein (HDL) (mmol/L) • Low-density lipoprotein (LDL) (mmol/L) • Triglycerides (mmol/L) • Diastolic blood pressure (mmHg) • Systolic blood pressure (mmHg) • Fasting insulin (pmol/L) • Postload insulin (pmol/L) • Body weight (kg) • BMI (kg/m2) • Waist circumference (cm) • Waist to hip ratio (WHR) • E.g. all adverse effects, gastrointestinal symptoms • E.g. percentage adherence to intervention, percentage taking required dosage of medication • Assessed with a validated instrument

We expressed T2DM incidence in risk differences and risk ratios with 95% confidence intervals (CI) and generated forest plots using Review Manager for primary outcomes.

2010 vol. 34 no. 2

Clarification of data We attempted to locate all publications arising from included studies and to contact authors for missing information or where clarification was required. We used open questions in order to avoid guiding authors’ responses.15

Assessment of heterogeneity Meta-analysis was not carried out due to the clinical diversity of trials.

Results Results of the search A total of 2,307 citations were retrieved (Figure 1). Of these, we assessed the full text of 158 articles (Figure 2). Four studies met our entry criteria, randomising a total of 5,196 participants.16-19 We found no unpublished studies. One study was translated from Chinese.16

Included studies For a summary of the included studies, Table 2 can be found at http://lifeinthefastlane.com/resources/research-tables/.

AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH © 2010 The Authors. Journal Compilation © 2010 Public Health Association of Australia

173

Yuen et al.

Article

The average length of these studies ranged from about three to five years with sample sizes ranging from 178 to 3,234 participants. All studies used IGT as an inclusion criterion while two studies also required a defined level of FPG and BMI.18-21 The intensity, complexity, content and delivery personnel involved in lifestyle interventions differed across studies. The lifestyle intervention of one trial16 consisted of standard T2DM and lifestyle education while the STOP-NIDDM19 participants were instructed on reducing weight via diet and regular exercise. In contrast, another study17 provided individualised advice on exercise and specific dietary changes. For the DPP, two arms met our inclusion criteria for lifestyle: the standard lifestyle advice (SL) given to the metformin and control groups and the intensive lifestyle (IL) intervention. For the IL intervention, participants were set individualised goals on weight reduction and exercise levels.18,20 In contrast, the SL intervention given to placebo and medication groups18 was similar to the STOP-NIDDM lifestyle intervention.19 The care given to the control groups of the IDPP-117 and the Chinese trial16 was considered ‘standard’. The medications also varied between studies: three investigated metformin16-18 while two acarbose.16,19 Three studies investigated the effects of LSM.17-19 In addition to incidence of T2DM, two studies also reported the number of participants reverting to normoglycaemia.16,19

Ongoing studies A total of 326 citations of ongoing trials were retrieved. Three further citations were received by contacting authors of a relevant review (Van de Laar, FA, pers. comm., 1 April 2008) and an included study17 (Ramachandran, A, pers. comm., 11 April 2008).

Eight ongoing trials seemed eligible. One allowed comparison between medication and lifestyle while six provided lifestyle advice to all participants. There is also the five to 11 year followup of the DPP18 however, the IL intervention was offered to all groups at the original study’s closure albeit in group settings rather than individually.22

Risk of bias in included studies The risk of bias for each of the included studies is presented in detail in Table 3. The overall risk of bias was high. Under the category of other sources of bias, two studies terminated early due to treatment effect.17,18 While this risks overestimation of treatment effects, we felt this had minimal impact on the results. In the DPP, the results were unlikely to be a chance finding as the T2DM incidence rates were similar at two time points. Furthermore, the results of both trials were not implausibly high compared to a similar trial23 and both trials had a considerable number of participants developing T2DM.

Effects of interventions For the DPP, the number of participants lost to follow-up for the lifestyle intervention was not reported and we received no reply from authors.18 Thus we decided to use the total number randomised initially.18,24 Additionally, the number of participants progressing to T2DM was after a follow-up period four months longer than the main report.24,25 We decided to combine the two arms investigating medication in the Chinese trial 16 as recommended in the Cochrane Handbook.26 Figure 2: Adapted Quorum Flowchart: Part 2. Articles describing a possible intervention trial on prediabetes in the title/abstract (n = 94)

Figure 1: Adapted Quorum flowchart: Part 1. Citations retrieved from MEDLINE (n = 240)

Citations retrieved from PsycInfo (n = 32)

Citations retrieved from The Cochrane Library (n = 297)

Citations retrieved from EMBASE (n = 696)

Citations retrieved from CINAHL (n = 105)

Citations retrieved from LILACS (n = 323)

Citations retrieved from WoS (n = 359)

Citations retrieved from BIOSIS (n = 255)

Articles excluded on basis of title and/or abstract (n = 1602)

Total citations retrieved (n = 2307)

Did not describe an intervention trial on humans (n = 1264)

Relevant reviews (n = 19)

Additional articles describing a possible intervention trial on prediabetes selected by second reviewer (n = 45)

Additional articles describing possible intervention trials on prediabetes (n = 19) Full text retrieved for further review (n = 158)

Duplicate citations (n = 598)

Articles describing a possible intervention trial on prediabetes in the title/abstract (n = 94)

Figure 1 – Adapted Quorum flowchart: Part 1

174

Relevant reviews (n = 14)

Reference lists checked (n = 33)

Total citations retrieved (n = 2307)

Remaining citations (n = 1709)

Citations from review search strategy in MEDLINE (n = 777)

Articles excluded on basis of title and/or abstract (n = 1615) Did not describe an intervention trial on humans (n = 1266) Described an intervention trial on participants with T2DM (n = 217) Described an intervention trial on participants with T1DM (n = 14) Described an intervention trial on participants with other/no condition(s) (n = 118)

Multiple reports on the same trial (n = 49) Of included studies (n = 23) Of excluded studies (n = 26)

Possibly eligible RCTs (n = 6) Inadequate reporting of results (2x2 factorial trial) and no response from authors (n = 2)

Articles excluded based on full text (n = 103) Did not describe a trial (reviews etc) (n = 24) Described an intervention trial on participants at risk of diabetes but not IGT/IFG specifically (n = 22) Described an intervention trial without a lifestyle intervention (n = 24) Described an intervention trial without a medication intervention (n = 27) Described an intervention trial without investigating development of T2DM or NGT (n = 1) Medication was withdrawn from market (n = 1) Problems with randomisation or allocation (n = 4)

RCTs included in the systematic review (n = 4)

Figure 2 – Adapted Quorum Flowchart: Part 2 AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH

© 2010 The Authors. Journal Compilation © 2010 Public Health Association of Australia

2010 vol. 34 no. 2

Controversial issues

Interventions for the prevention or delay of type 2 diabetes

Comparisons could not be made for LSM versus medication alone as there were insufficient trials. The comparisons are presented in Figures 3-7. The DPP18 did not investigate the effects of adding medication to the IL regime and thus, the comparison presented is between the IL intervention alone and the combination of medication and SL recommendations.

Similarly, results from the limited trials evaluating the effect of LSM versus lifestyle or medication alone seemed to be dependent on the intensity of the lifestyle regime implemented. In the DPP,18 the comparison between the combination of medication and the SL recommendations with the IL intervention suggested IL advice was more effective. In contrast, the same study,18 found benefit from the addition of medication to the SL intervention; a result comparable to that of the STOP-NIDDM trial.19

Secondary outcomes The results for the secondary outcomes are reported in Table 4, found at http://lifeinthefastlane.com/resources/research-tables/, in means and standard deviations unless otherwise specified.

Discussion Summary of main results From these four trials with an overall high risk of bias, it was not possible to draw any firm conclusions on which intervention was more effective in delaying T2DM. It was difficult to directly compare lifestyle and medication interventions in delaying T2DM from the limited studies.16,17 We can only speculate that a more intensive lifestyle intervention would be more effective by promoting patient adherence. It could be argued that lifestyle advice given in the IDPP-117 was reinforced more regularly and thus was more likely to change behaviour than that of the Chinese study.16 However, the difference between the two interventions was not statistically significant in the IDPP-1.17

Overall completeness and applicability of evidence Completeness of evidence

We were unable to fulfil all our objectives. There were insufficient studies investigating reversion to NGT; most relied heavily on the incidence of T2DM to demonstrate effect. In contrast to reporting changes in glucose concentration, solely reporting T2DM incidence could give healthcare professionals the impression of a greater effect27 and would not allow estimation of the effect on T2DM-related complications.28 Thus, we consider measuring reversion to NGT and glucose concentrations on a continuous scale could aid in result interpretation. Additionally, the inclusion criteria of all studies focused on IGT only or IGT with IFG. Therefore, none of the participants had isolated IFG. This would however, have limited impact on the relevance of this systematic review as isolated IGT is more prevalent than isolated IFG and the combination of these risk factors has the highest risk of progression.29 However, there are suggestions that people with IFG and IGT have different

Table 3: Risk of bias of included studies. Study

Sequence generation

Allocation Blinding concealment

Incomplete outcome data

Selective outcome reporting

Other sources of bias

• [18] • Low risk of bias • Low risk of bias • Low risk of bias • Unclear risk of bias • High risk of bias • High risk of bias • Losses to follow-up • Some outcomes were • Ended before • DPP not reported not reported in any of the planned papers or were reported termination selectively for some groups • [19] • Low risk of bias • Low risk of bias • Low risk of bias • High risk of bias • High risk of bias • Unclear risk of bias • STOP- • Excluded on the basis • Changes to the endpoints • Information NIDDM of a second OGTT published in study protocol available in the Participants excluded • Unclear risk of bias past was not were different to those from their reporting accessible to followed up of outcomes judge the validity of criticisms of the study [24, 25] • [17] • High risk of bias • High risk of bias • High risk of bias • High risk of bias • High risk of bias • High risk of bias • IDPP-1 • Inadequate • Not blinded • More people not • Missing outcomes • Ended before sequence followed up in lifestyle • Some outcomes were not planned generation group in total and due sufficiently reported to termination to being “not willing” allow meta-analysis • [16] • Low risk of bias • Unclear risk of • High risk of bias • High risk of bias • Low risk of bias • High risk of bias bias • Not blinded • Not intention-to-treat • Baseline analysis, exclusion imbalance though based on medication probably not adherence in acarbose leading to and metformin group significant bias as it favoured the groups which showed a smaller treatment effect

2010 vol. 34 no. 2

AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH © 2010 The Authors. Journal Compilation © 2010 Public Health Association of Australia

175

Yuen et al.

Article

alterations in glucose homeostasis30-32 and reflect different stages of abnormal glucose metabolism.33 Thus, we are uncertain whether our conclusions can be applied to patients with isolated IFG. Furthermore, the studies located investigated metformin and acarbose only. Thus, we could not examine the effects of other T2DM drug classes. Fortunately, there are numerous ongoing trials assessing antihypertensives, statins, meglitinides, sulfonylureas and thiazolidinediones with lifestyle interventions in prediabetes. Additionally, none of the included studies gave resistance training. This is unfortunate as both aerobic and resistance exercises are

recommended to reduce the risk of CVD having different benefits on health and fitness such as in the risk of musculoskeletal injuries.34 We were also unable to adequately investigate the effect of interventions on CVD morbidity and mortality. The positive results from the STOP-NIDDM trial19 were based on a limited number of events and high discontinuation rates of acarbose. This study also had a high risk of bias conferred from the 61 participants lost to follow-up who had a greater risk of CVD than those followed up: they were older, had greater FPG levels and more were already on CVD medication. Other studies also had a limited number of

Figure 3: Lifestyle versus medication: T2DM conversion.

Lifestyle Medication Risk Ratio Lifestyle Medication Risk Ratio Lifestyle Medication Risk Ratio95% CI Fixed, Total M-H, Total Events Study or Subgroup Events Lifestyle Medication Risk Ratio Fixed, 95% CI CI Total Total Study or or Subgroup M-H, Fixed, 95% Total M-H, Events TotalEvents Events Study Subgroup Events 0.88 [0.64,95% 1.21]CI 128 M-H, 51 Total 120 Events 42 Total IDPP-1 Fixed, Events Study or Subgroup [0.64, 1.21] 5151 128128 0.88 4242 120120 IDPP-1 0.88 IDPP-1 1.98 [0.64, [1.03, 1.21] 3.80] 89 15 36 12 Fang 0.88 1.21] 128 51 120 42 IDPP-1 1.98 [1.03, 3.80] 89 15 36 12 Fang 1.98 [0.64, [1.03, 3.80] 89 15 36 12 Fang 1.98 [1.03, 3.80] 89 15 36 12 Fang

Figure 4: Lifestyle versusversus medication: Reversion to NGT. Figure 3 – Lifestyle medication: T2DM conversion

Figure 3 –3 Lifestyle versus medication: T2DM conversion Figure – Lifestyle versus medication: T2DM conversion Figure 3 – Lifestyle versus medication: T2DM Medication Lifestyle conversion Risk Ratio

Risk Ratio Risk Ratio Ratio M-H,Risk Fixed, 95% CI Risk Ratio M-H, Fixed, 95% CI CI M-H, Fixed, 95% M-H, Fixed, 95% CI 0.5 0.7 1 1.5 2 0.50.50.7 1.51.52 2medication 0.7 1 1 Favours Favours lifestyle Favours lifestyle Favours 0.5 0.7 1 1.5 medication 2medication Favours lifestyle Favours Favours lifestyle Favours medication

Medication Lifestyle Risk Ratio Medication Lifestyle Risk Ratio95% CI Study or Subgroup Events Total Events Total M-H, Fixed, Medication Lifestyle Risk Ratio Study or Subgroup Events Total Events Total M-H, Fixed, 95% CI CI Study or Subgroup Events Total Events Total M-H, Fixed, 95% Fang 60 Total 89 Events 12 Total 36 M-H, 2.02 [1.25,95% 3.28]CI Study or Subgroup Events Fixed, Fang 6060 8989 1212 3636 2.02 [1.25, 3.28] Fang 2.02 [1.25, 3.28] Fang 60 89 12 36 2.02 [1.25, 3.28]

Figure 4 – Lifestyle versus medication: Reversion to NGT

Risk Ratio Risk Ratio Ratio M-H,Risk Fixed, 95% CI Risk Ratio M-H, Fixed, 95% CI CI M-H, Fixed, 95% M-H, Fixed, 95% CI 0.5 0.7 1 1.5 2 0.5 1.51.52 2medication 0.50.7 0.7 1 1 Favours Favours lifestyle 0.5 0.7 Favours 1 Favours 1.5medication 2medication Favours lifestyle Favours lifestyle Favours lifestyle Favours medication

Figure 5:4Lifestyle andversus medication versus lifestyle alone:to T2DM conversion. Figure –4 Lifestyle medication: Reversion NGT Figure – Lifestyle versus medication: Reversion to NGT

Figure 4 – Lifestyle versus Reversion NGT Lifestyle andmedication: medication Lifestyle alone to R isk Ratio

Risk Ratio andand medication Risk isk Ratio Ratio Lifestyle medication Lifestyle alone Risk RiFsk Ratio95% CI Ratio M-H, M-H, Total R Events Eventsalone ixed, 95% CI Study or Subgroup Lifestyle TotalLifestyle Fixed, M-H, Fixed, Total Events Events 95% CI CI Study or Subgroup Total Fixed, 95% CI Lifestyle and medication RiFsk isk Ratio Ratio M-H, M-H, To talM-H, Events Events ixed, 95% Study Total Fixed, 95% DPP or Subgroup 1073 Lifestyle [1.23, 1.78]CI 1R.48 1079 225 153 alone M-H, F M-H, To tal Events Events ixed, 95% CI Study or Subgroup Total Fixed, 95% CI 225225 DPP 1073 [1.23, 1.78] 10.48 10 DPP 1073 [1.23, 1.78] 1079 82 1.48 225 304 1079 DPP - control lifestyle 1073 153153 .75 [0.64, 0.87] 0 10 82 225 304 DPP control lifestyle 1073 .75 [0.64, 0.87] 100.48 10668 79 225 153 DPP 1073 82 304 - control lifestyle .75 [0.64, 0.87] 221 285 STOP-NIDDM 656 .79 [1.23, [0.69, 1.78] 0.91] 221225 STOP-NIDDM 656 [0.69, 0.91] 00.75 10668 82 0.79 DPP - control lifestyle 1073 [0.64, 0.87] 221 285 STOP-NIDDM 656 .79 [0.69, 0.91] 42 120 42 668 IDPP-1 121 285304 .99 [0.70, 1.40] 42221 42285 IDPP-1 121656 [0.70, 1.40] 0.79 STOP-NIDDM [0.69, 0.91] 120 0.99 42 42 120668 IDPP-1 121 .99 [0.70, 1.40] 0.7 1 1.5 2 0.99 [0.70, 1.40] 0.5 120 42 42 IDPP-1 121 0.5 0.70.7 and med 1 1 Favours lifestyle 1.51.5 alone 2 2 0.5 Favours lifestyle lifestyle alone Favours andand med lifestyle alone 2 Favours lifestyle medFavours 0.5 lifestyle 0.7 1 Favours 1.5 Figure 5 – Lifestyle and medication versus lifestyle alone: T2DMFavours conversion lifestyle and med Favours lifestyle alone

Figure –5 Lifestyle medication versus lifestyle alone: T2DM conversion Figure 6:5Lifestyle andand medication versus lifestyle alone: Reversion to NGT. Figure – Lifestyle and medication versus lifestyle alone: T2DM conversion Figure 5 – Lifestyle and alone medication lifestyle alone: T2DM conversion Lifestyle Lifestyleversus and medication Risk Ratio

Risk Ratio andand medication Ratio Risk Ratio Lifestyle alone Lifestyle medication Risk Ratio95% CI Risk Ratio Study or Subgroup Lifestyle Eventsalone TotalLifestyle Events Total Risk M-H, Fixed, M-H, Fixed, 95% CI Study or Subgroup Events Total Events Total M-H, Fixed, 95% CI CI M-H, Fixed, 95% CI CI Lifestyle alone Lifestyle and medication Risk Ratio Risk Ratio Study or Subgroup Events Total Events Total M-H, Fixed, 95% M-H, Fixed, 95% STOP-NIDDM 212 668 241 656 0.86 [0.74, 1.00] Study or Subgroup Events Total Events Total M-H, Fixed, 95% M-H, Fixed, 95% CI STOP-NIDDM 212212 668 241241 656 [0.74, 1.00] STOP-NIDDM 668 656 0.86 0.86 [0.74, 1.00]CI 0.7 1 1.5 2 STOP-NIDDM 212 668 241 656 0.86 [0.74, 1.00] 0.5 0.5 0.70.7 and med 1 1 Favours lifestyle 1.51.5 alone 2 2 0.5 lifestyle Favours Favours andand med lifestyle alone Favours lifestyle medFavours lifestyle alone 2 0.5 lifestyle 0.7 1 Favours 1.5 Figure 6 – Lifestyle and medication versus lifestyle alone: Reversion NGT Favours to lifestyle and med Favours lifestyle alone

Figure 6 –6 Lifestyle and medication versus lifestyle alone: Reversion toto NGT Figure – Lifestyle and medication versus lifestyle alone: Reversion NGT

Figure 6 – Lifestyle andand medication versus Reversion to NGT Figure 7: Lifestyle and medication versus medication alone:alone: T2DM conversion. Lifestyle medication Medicationlifestyle alone Risk Ratio

Risk Ratio andand medication Ratio Risk Ratio Lifestyle medication Medication alone Risk Ratio95% CI Risk Ratio Study or Subgroup Lifestyle Events TotalMedication Eventsalone Total Risk M-H, Fixed, M-H, Fixed, 95% CI Study or Subgroup Events Total Events Total M-H, Fixed, 95% CI M-H, Fixed, 95% CI CI Study or Subgroup Lifestyle Events Total Events Total Fixed, 95% M-H, Fixed, 95% and Risk Ratio Risk Ratio IDPP-1 42 medication 121 Medication 51 alone 128 M-H, 0.87 [0.63, 1.20]CI Study or Subgroup Events Total Events Total M-H, Fixed, 95% CI M-H, Fixed, 95% CI IDPP-1 42 121 51 128 0.87 [0.63, 1.20] IDPP-1 42 121 51 128 0.87 [0.63, 1.20] 0.7 1 1.5 2 IDPP-1 42 121 51 128 0.87 [0.63, 1.20] 0.5 0.50.5 1 1 Favours medication 1.51.5 alone 2 2 0.7 and med Favours 0.7 lifestyle Favours lifestyle med medication Favours lifestyle medFavours medication alone2 0.5 0.7andand 1 Favours 1.5 alone Figure 7 – Lifestyle and medication versus medication alone: T2DM conversion Favours lifestyle and med Favours medication alone

Figure 7 –7 Lifestyle and medication versus medication alone: T2DM conversion Figure – Lifestyle and medication versus medication alone: T2DM conversion Figure 7 – Lifestyle and medication versus medication alone: T2DM conversion 176

AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH © 2010 The Authors. Journal Compilation © 2010 Public Health Association of Australia

2010 vol. 34 no. 2

Controversial issues

Interventions for the prevention or delay of type 2 diabetes

CVD events, highlighting the need for longer follow-up periods. Hopefully this will be addressed by the DPPOS.22

Applicability of findings The external validity of the findings of this review is uncertain. It is likely that study participants were more motivated than other patients being self-selected in the two largest studies18,19 and DPP participants were even given honoraria.18 DPP groups also reported being more physically active than the nationally representative sample with IGT in NHANES III.35 Similarly, none of the IL participants were in the precontemplation stage of behavioural change in physical activity levels,36 a result markedly different to other populations.37,38 Due to these differences, delivery personnel in the real world may require more time to successfully initiate changes in lifestyle. Furthermore, interventions causing substantial side effects would be impractical. Acarbose in particular is of concern as 98% of participants reported adverse events and nearly 15% ceased the medication in the STOP-NIDDM trial,19 a cessation rate similar to the Chinese trial.16 As people with prediabetes are often asymptomatic, there is a low threshold for adverse effects. Given the side effects of acarbose were significant enough to affect compliance in a volunteer population, it is not surprising that the American Diabetes Association (ADA) currently considers metformin the only suitable medication for prediabetes.39 However, metformin also demonstrated substantial side effects. The rate of gastrointestinal symptoms in the DPP18 was significantly higher with metformin than with placebo while the side effects in the Chinese study16 made three participants stop taking the medication. Twenty-two participants on metformin in the IDPP-117 also reported symptoms of hypoglycaemia. Thus, the future role of medications in prediabetes is uncertain. We are also unsure whether the effects from these interventions last following cessation of active treatment. Both trials with washout periods19,24 demonstrated a greater incidence of T2DM during the washout period with medication than with placebo thus suggesting medication conceals T2DM via its glucose lowering properties. Therefore to ‘prevent’ T2DM, lifelong treatment would be required. However, both washout periods were relatively short and therefore it is unclear whether cumulative incidence rates would have eventually converged. In addition, long-term follow-up of lifestyle groups in included studies is incomplete and we await the results of the DPPOS22 with interest. Results from a similar trial were promising: the benefits were sustained following a median of three years since active treatment.40 Furthermore, the capability of healthcare systems around the world to provide adequate lifestyle interventions is unclear. It seems that lifestyle advice needs to be reinforced regularly and yet in the IDPP-117 that alone was not be enough. Further components such as enhancing social support, behaviour modification and having face to face contact may be required. The DPP18 found that initial attainment of exercise and weight loss goals predicted future success.41 Thus, it may be beneficial for interventions to employ an initial period of regular contact. 2010 vol. 34 no. 2

Additionally, the costs associated with implementing such intensive lifestyle programs are barriers to its widespread use. The IL regime in the DPP18 is unlikely to be replicated outside research settings due to financial and time constraints, a view shared by the majority of nurses and general practitioners.42 However, the expenses of lifestyle interventions can be minimised by group counselling.43 We should also consider the costs to patients participating in such regimes such as that associated with purchasing related materials and travelling to and from counselling sessions.44 A survey found that individuals at risk of T2DM were reluctant to pay for a program similar to the DPP IL intervention with the $65 out-of-pocket costs per month.45 Thus, in the community, it is likely that associated expenses would need to be subsidised.

Potential biases in the review process Possible limitations

Internal validity could be a limitation of this review. It can be argued that publication bias is a possibility as we only identified published studies. However, we attempted to avoid this by searching trial registries, conducting an extensive search and contacting international experts. Furthermore, our requirement for trials to have both lifestyle and medication arms may have limited studies to those with large sample sizes. Additionally, it was inevitable that the classification of lifestyle interventions was subjective. Of the trials that fulfilled all other inclusion criteria, only the IDPP-1 control group may have also been included. However, we ultimately considered their recommended 30 minutes of daily physical activity was not substantial enough.17 Strengths

This review’s strengths lie in its rigorous methodology. We also aimed to keep the review transparent, extensively reporting our methods and results. Additionally, articles were not limited to those reported in English. Citation bias was also unlikely in this review as three studies17-19 were retrieved by searching databases and the other16 was retrieved from the reference list of another systematic review, which employed slightly different search filters.4

Conclusions The decision whether to implement a lifestyle intervention or begin medication in patients with IGT to delay T2DM should balance the advantages and side effects of each method as well as integrate patients’ values. Motivated patients may opt for an intensive lifestyle program with frequent follow-up sessions thereafter due to health benefits beyond lowering their risk of T2DM. The option of metformin could also be discussed though its common gastrointestinal side effects must be considered. The use of acarbose should be precluded by its frequent adverse events. There is currently insufficient evidence to make recommendations in choosing between lifestyle and medication as well as LSM against a lifestyle or medication intervention alone. Furthermore,

AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH © 2010 The Authors. Journal Compilation © 2010 Public Health Association of Australia

177

Yuen et al.

Article

our findings might not be applicable to patients with isolated IFG. We suggest further studies on more intensive lifestyle modification, incorporating measures to maximise compliance. Future trials should have post-intervention follow-up periods and report glucose measurements on a continuous scale as well as progression to T2DM and NGT.

References 1. World Health Organization [homepage on the Internet]. Geneva (CHE): WHO; 2008 [cited 2008 May 20]. Diabetes Programme Facts and Figures: Country and Regional Data. Available from: http://www.who.int/diabetes/facts/world_figures/en/ 2. ADA. Economic consequences of diabetes mellitus in the U.S. in 1997. Diabetes Care. 1998;21(2):296-309. 3. Twigg SM, Kamp MC, Davis TM, Neylon EK, Flack JR. Prediabetes: a position statement from the Australian Diabetes Society and Australian Diabetes Educators Association. Med J Aust. 2007;186(9):461-5. 4. Gillies CL, Abrams KR, Lambert PC, Cooper NJ, Sutton AJ, Hsu RT, et al. Pharmacological and lifestyle interventions to prevent or delay type 2 diabetes in people with impaired glucose tolerance: systematic review and meta-analysis. BMJ. 2007;334(299). 5. Norris SL, Zhang X, Avenell A, Gregg E, Schmid CH, Lau J. Long-term non-pharmacological weight loss interventions for adults with prediabetes (Cochrane Review): In: The Cochrane Database Systematic Review, 2, 2005. p. 1-80. Oxford (UK): Update Software; 2005. 6. Padwal R, Majumdar SR, Johnson JA, Varney J, McAlister FA. A systematic review of drug therapy to delay or prevent type 2 diabetes. Diabetes Care. 2005;28(3):736-44. 7. Yamaoka K, Tango T. Efficacy of Lifestyle Education to Prevent Type 2 Diabetes: a meta-analysis of randomized controlled trials. Diabetes Care. 2005;28(11):2780-6. 8. Yates T, Khunti K, Bull F, Gorely T, Davies MJ. The role of physical activity in the management of impaired glucose tolerance: a systematic review. Diabetologia. 2007;50:1116-26. 9. Van de Laar FA, Lucassen PLBJ, Akkermans RP, Van de Lisdonk EH, De Grauw WJC. Alpha-glucosidase inhibitors for people with impaired glucose tolerance or impaired fasting blood glucose (Cochrane Review): In: The Cochrane Database Systematic Review, 4, 2006. Oxford (UK): Update Software; 2006. 10. Padwal R, Laupacis A. Antihypertensive Therapy and Incidence of Type 2 Diabetes: A systematic review. Diabetes Care. 2004;27(1):247-55. 11. Norris SL, Zhang X, Avenell A, Gregg E, Bowman B, Schmid CH, et al. LongTerm Effectiveness of Weight-Loss Interventions in Adults with Pre-Diabetes. Am J Prev Med. 2005;28(1):126-39. 12. Knoll T, Narayan S, Coven A, Kmoch J, Byer S, Williams R, et al. Adobe Photoshop CS3. 10 th ed. San Jose (CA): Adobe Systems; 2007. 13. RevMan: Review Manager [computer program]. 5.0 ed. Copenhagen (DNK): The Nordic Cochrane Centre, The Cochrane Collaboration; 2008. 14. Higgins J, Altman D. Chapter 8: Assessing risk of bias in included studies. In: Higgins J, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions. Version 5.0.0 [updated February 2008]. Cambridge (UK): The Cochrane Collaboration; 2008. 15. Haahr MT, Hrobjartsson A. Who is blinded in randomized clinical trials? A study of 200 trials and a survey of authors. Clin Trials. 2006;3(4):360-5. 16. Fang Y, Li T, Chen S. Effect of medicine and non-medicine intervention on the outcomes of patients with impaired glucose tolerance - 5-year follow-up. Zhongguo Linchuang Kangfu. 2004;8:6562-3. 17. Ramachandran A, Snehalatha C, Mary S, Mukesh B, Bhaskar AD, Vijay V. The Indian Diabetes Prevention Programme shows that lifestyle modification and metformin prevent type 2 diabetes in Asian Indian subjects with impaired glucose tolerance (IDPP-1). Diabetologia. 2006;49(2):289-97. 18. Knowler WC, Barrett-Connor E, Fowler S, Hamman RF, Lachin JM, Walker E, et al. Reduction in the Incidence of Type 2 Diabetes with Lifestyle Intervention or Metformin. N Engl J Med. 2002;346(6):393-403. 19. Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M, et al. Acarbose for prevention of type 2 diabetes mellitus: the STOP-NIDDM randomised trial. Lancet. 2002;359(9323):2072-7. 20. Diabetes Prevention Program [homepage on the Internet]. Rockville (MD): The Biostatistics Center, The George Washington University; 2001 May 18 [cited 2008 Apr 14]. Protocol for the Diabetes Prevention Program. Version 4.4. Available from: http://www.bsc.gwu.edu/dpp/3armprot.htmlvdoc 21. Chiasson JL, Gomis R, Hanefeld M, Josse RG, Karasik A, Laakso M. The STOP-NIDDM Trial: an international study on the efficacy of an alphaglucosidase inhibitor to prevent type 2 diabetes in a population with impaired glucose tolerance: rationale, design, and preliminary screening data. Study to Prevent Non-Insulin-Dependent Diabetes Mellitus. Diabetes Care. 1998;21(10):1720-5.

178

22. Diabetes Prevention Program [homepage on the Internet]. Rockville (MD): The Biostatistics Center, The George Washington University; 2006 July 5 [cited 2008 Apr 18]. Protocol for the Diabetes Prevention Program Outcomes Study (DPPOS). Available from: http://www.bsc.gwu.edu/dpp/p2_1.pdf 23. Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen H, Ilanne-Parikka P, et al. Prevention of Type 2 Diabetes Mellitus by Changes in Lifestyle among subjects with Impaired Glucose Tolerance. N Engl J Med. 2001;344(18):1343-50. 24. DPPRG. Effects of withdrawal from metformin on the development of diabetes in the diabetes prevention program. Diabetes Care. 2003;26(4):977-80. 25. Hamman RF, Wing RR, Edelstein SL, Lachin JM, Bray GA, Delahanty L, et al. Effect of weight loss with lifestyle intervention on risk of diabetes. Diabetes Care. 2006;29(9):2102-7. 26. Lefebvre C, Manheimer E, Glanville J. Chapter 6: Searching for studies. In: Higgins J, Green S, editors. Cochrane Handbook for Systematic Reviews of Intervention. Version 5.0.0. [updated February 2008]. Cambridge (UK): The Cochrane Collaboration; 2008. 27. Mühlhauser I, Kasper J, Meyer G, Federation of European Nurses in D. Understanding of diabetes prevention studies: questionnaire survey of professionals in diabetes care. Diabetologia. 2006;49(8):1742-6. 28. Tuomilehto J, Wareham N. Glucose lowering and diabetes prevention: are they the same?[see comment]. Lancet. 2006;368(9543):1218-9. 29. Unwin N, Shaw J, Zimmet P, Alberti KGMM. Impaired glucose tolerance and impaired fasting glycaemia: the current status on definition and intervention. Diabet Med. 2002;19(9):708-23. 30. Weyer C, Bogardus C, Pratley RE. Metabolic characteristics of individuals with impaired fasting glucose and/or impaired glucose tolerance. Diabetes. 1999;48(11):2197-203. 31. Tripathy D, Carlsson M, Almgren P, Isomaa B, Taskinen MR, Tuomi T, et al. Insulin secretion and insulin sensitivity in relation to glucose tolerance: lessons from the Botnia Study. Diabetes. 2000;49(6):975-80. 32. Novoa FJ, Boronat M, Saavedra P, Diaz-Cremades JM, Varillas VF, La Roche F, et al. Differences in Cardiovascular Risk Factors, Insulin Resistance, and Insulin Secretion in Individuals With Normal Glucose Tolerance and in Subjects With Impaired Glucose Regulation: The Telde Study. Diabetes Care. 2005;28(10):2388-93. 33. Guerrero-Romero F, Rodríguez-Morán M. Impaired glucose tolerance is a more advanced stage of alteration in the glucose metabolism than impaired fasting glucose. J Diabetes Complicat. 2001;15(1):34-7. 34. Braith RW, Stewart KJ. Resistance Exercise Training: Its Role in the Prevention of Cardiovascular Disease. Circulation. 2006;113(22):2642-50. 35. Kriska AM, Edelstein SL, Hamman RF, Otto A, Bray GA, Mayer-Davis EJ, et al. Physical Activity in Individuals at Risk for Diabetes: Diabetes Prevention Program. Med Sci Sports Exerc. 2006;38(5):826-32. 36. Delahanty LM, Conroy MB, Nathan DM. Psychological Predictors of Physical Activity in the Diabetes Prevention Program. J Am Diet Assoc. 2006;106(5):698705. 37. Marcus BH, Selby VC, Niaura RS, Rossi JS. Self-efficacy and the stages of exercise behavior change. Res Q Exerc Sport. 1992;63(1):60-6. 38. Nishida Y, Suzuki H, Wang DH, Kira S. Psychological determinants of physical activity in Japanese female employees. J Occup Health. 2003;45(1):15-22. 39. ADA. Standards of Medical Care in Diabetes--2008. Diabetes Care. 2008;31 Suppl 1:12-54. 40. Lindstrom J, Ilanne-Parikka P, Peltonen M, Aunola S, Eriksson JG, Hemio K, et al. Sustained reduction in the incidence of type 2 diabetes by lifestyle intervention: follow-up of the Finnish Diabetes Prevention Study. Lancet. 2006;368:1673-9. 41. DPPRG. Achieving Weight and Activity Goals Among Diabetes Prevention Program Lifestyle Participants. Obesity. 2004;12(9):1426-34. 42. Williams R, Rapport F, Elwyn G, Lloyd B, Rance J, Belcher S. The prevention of type 2 diabetes: general practitioner and practice nurse opinions. Br J Gen Pract. 2004;54(504):531-5. 43. DPPRG. Within-Trial Cost-Effectiveness of Lifestyle Intervention or Metformin for the Primary Prevention of Type 2 Diabetes. Diabetes Care. 2003;26(9):251823. 44. DPPRG. Costs Associated With the Primary Prevention of Type 2 Diabetes Mellitus in the Diabetes Prevention Program. Diabetes Care. 2003;26(1):36-47. 45. Johnson FR, Manjunath R, Mansfield CA, Clayton LJ, Hoerger TJ, Zhang P. High-Risk Individuals’ Willingness to Pay for Diabetes Risk-Reduction Programs. Diabetes Care. 2006;29(6):1351-6.

AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH © 2010 The Authors. Journal Compilation © 2010 Public Health Association of Australia

2010 vol. 34 no. 2