International Journal of Technology Assessment in Health Care, 28:3 (2012), 228–234.

c Cambridge University Press 2012

doi:10.1017/S0266462312000256
Assessments
Cost-Effectiveness of Exercise
Programs in Type 2 Diabetes
Doug Coyle George A. Wells
email: dcoyle@uottawa.ca University of Ottawa; University of Ottawa Heart Institute
University of Ottawa; Ottawa Health Research Institute Michelle Fortier
Kathryn Coyle University of Ottawa
Applied Health Economic Research Unit Robert D. Reid
Glen P. Kenny University of Ottawa Heart Institute
University of Ottawa Penny Phillips
Normand G. Boulé Ottawa Health Research Institute
University of Ottawa; University of Alberta Ronald J. Sigal
University of Ottawa; University of Calgary

Background: A randomized controlled trial has shown that supervised, facility-based exercise training is effective in improving glycemic control in type 2 diabetes. However, these programs are
associated with additional costs. This analysis assessed the cost-effectiveness of such programs.
Methods: Analysis used data from the Diabetes Aerobic and Resistance Exercise (DARE) clinical trial which compared three different exercise programs (resistance, aerobic or a combination of both) of
6 months duration with a control group (no exercise program). Clinical outcomes at 6 months were entered for individual patients into the UKPDS economic model for type 2 diabetes adapted for the
Canadian context. From this, expected life-years, quality-adjusted life-years (QALYs) and costs were estimated for all patients within the trial.
Results: The combined exercise program was the most expensive ($40,050) followed by the aerobic program ($39,250), the resistance program ($38,300) and no program ($31,075). QALYs
were highest for combined (8.94), followed by aerobic (8.77), resistance (8.73) and no program (8.70). The incremental cost per QALY gained for the combined exercise program was $4,792
compared with aerobic alone, $8,570 compared with resistance alone, and $37,872 compared with no program. The combined exercise program remained cost-effective for all scenarios considered
within sensitivity analysis.
Conclusions: A program providing training in both resistance and aerobic exercise was the most cost-effective of the alternatives compared. Based on previous funding decisions, exercise training for
individuals with diabetes can be considered an efficient use of resources.

Keywords: Cost-effectiveness, type 2 diabetes mellitus, aerobic exercise, resistance exercise

Physical activity is important in the management of type 2 di-
abetes mellitus. Previous studies (2;12;21;23) have found that
structured aerobic exercise (walking, jogging, or cycling) and
resistance exercise (weightlifting) are effective in reducing ab-
solute hemoglobin A1c (HbA1c). Reductions in HbA1c are
associated with a decrease in major cardiovascular events (19)
and in microvascular complications (22).
The DARE (Diabetes Aerobic and Resistance Exercise)
clinical trial (20) was designed to determine the effects on
glycemic control and other risk factors for cardiovascular dis-
ease of aerobic and resistance training alone and in combination
compared with a sedentary control group. The study found that
Funding for this study was part of the funding of the DARE trial. The DARE trial was funded by the
Canadian Institutes of Health Research, the Canadian Diabetes Association and the University of
Ottawa Interfaculty Grants Program. Dr. Sigal is supported by a Health Senior Scholar award from
the Alberta Heritage Foundation for Medical Research. D.C. designed the analysis. K.C., G.A.W.,
and R.J.S. contributed to the design of the analysis. D.C., K.C., P.P., and R.J.S. researched data.
D.C. and K.C. conducted the analysis and wrote the manuscript. G.P.K., N.G.B., G.A.W., M.F.,
R.D.R., P.P., and R.J.S. contributed to the discussion and reviewed/edited the manuscript.
in comparison to the control group, HbA1c was lower in the aer-
obic training group (mean reduction = 0.51 percent: 95 percent
confidence interval [CI], –0.87 to –0.14) and in the resistance
group (mean reduction = 0.38 percent: 95 percent CI, –0.72
to –0.22). Furthermore, combined exercise training resulted in
an additional reduction in HbA1c of 0.46 percent (CI, –0.83
to –0.09) compared with aerobic training alone and 0.59 per-
cent (CI, –0.95 to –0.23) compared with resistance training
alone. Differences in blood pressure and lipid values between
groups were not statistically significantly. An hypothesis that
those randomized to combined aerobic and resistance exercise
training or resistance exercise alone would show greater change
in well-being and health status was only partially supported
(16).
Thus the DARE trial demonstrated that either aerobic or
resistance training alone improves glycemic control in type 2
diabetes, but the improvements are greatest with combined aer-
obic and resistance training. Stemming from this trial, the focus
of this study is to assess the cost-effectiveness of aerobic and
resistance exercise, both alone and in combination versus a

228

sedentary waiting list control in the treatment of nonexercising
persons with type 2 diabetes.
METHODS
Patient Population
The patient population for the model was the same as the DARE
clinical trial (20). In the trial, 251 subjects aged 39 to 70 with
a diagnosis of type 2 diabetes were randomly assigned to one
of three exercise programs (aerobic exercise, resistance exer-
cise, or combined) or a sedentary control. Exercise training was
performed three times weekly at community based facilities.
The clinical trial included type 2 diabetes, as defined by
the American Diabetes Association (17). Exclusion criteria re-
lated to current insulin therapy, previous exercise and resistance
training, changes during the previous 2 months in oral medi-
cations, and significant changes in body weight, serum creati-
nine, proteinuria, or blood pressure. Patients in the trial had a
mean age of 54.2 with an average duration of diabetes of 5.4
years. There were more males (65.1 percent) than females. Mean
HbA1c was 7.7 percent, and mean body mass index (BMI) was
33.5 kg/m2 . The mean SF-36 physical and mental component
summary scores were 49.2 and 51.0, respectively. Full inclusion
and exclusion criteria as well as more detailed information on
the patient population are available from the original trial paper
(20).
Treatment Comparators
Analysis compared the three exercise programs from DARE
with the control group with no supervised exercise program:
aerobic exercise, resistance exercise, and a combination of aer-
obic and resistance exercise. Subjects exercised three times per
week and training progressed gradually in length and intensity.
Individual supervision was provided by an exercise specialist.
After the prerandomization initial run-in period, eleven personal
training sessions were provided by the exercise specialist over
the course of the first 6 months. At the end of the 6-month period,
subjects were given a maintenance program and re-examined 6
months later. During this period, all patients had the option to
follow whichever exercise program they wanted and there were
no further restrictions on changes in medication. Therefore, the
effect of the exercise programs is reflected to a greater extent in
the 6-month results rather than the 12-month results. Adherence
with exercise training within the clinical trial was high–median
exercise training attendance was 86 percent in the combined ex-
ercise training group, 80 percent in the aerobic training group,
and 85 percent in the resistance training group (20).
Analytical Framework
Analysis was conducted to estimate both the mean lifetime costs
of each of the exercise programs and the mean effectiveness,
which was expressed as both life-years and quality-adjusted life-
years (QALYs) (7). Lifetime costs, life expectancy, and QALYs
229
Cost-effectiveness of exercise in diabetes
were estimated for each patient within the DARE clinical trial
using a computer simulation model described below. Cost-
effectiveness was assessed by incremental cost-effectiveness
ratios (ICER). The primary perspective of the study was soci-
etal in that it incorporates both the costs of healthcare services
and the costs of the exercise program. Future costs and benefits
were discounted at 5 percent per annum (7).
Decision Model
Analysis was conducted by adapting the United Kingdom
Prospective Diabetes Study (UKPDS) Health Outcomes Model
(9) to the Canadian context. The UKPDS model predicts the
transition of patients into different health states which are com-
bination states incorporating seven diabetes-related complica-
tions; myocardial infarction (MI), other ischemic heart disease
(IHD), stroke, heart failure, amputation, renal failure, and eye
disease as well as death. The model was developed using data
from 3642 patients with newly diagnosed type 2 diabetes who
participated in the UKPDS and were followed up for between 6
and 20 years. It incorporates information regarding risk factors
which include age, sex, ethnicity, duration of diabetes, height,
weight, smoking status, total cholesterol, HDL cholesterol, sys-
tolic blood pressure, and HbA1c to predict the likely outcomes.
The cycle time for the model is 12 months with a 40-year simu-
lation duration which is assumed to approximate the maximum
life expectancy of the patient population.
The UKPDS model estimates the discounted time an indi-
vidual will spend in each health state (9). Thus, the model can be
used to estimate discounted life expectancy. By weighting life
expectancy by appropriate costs and utility values, the model can
estimate both discounted lifetime costs and discounted quality-
adjusted life-years.
For each individual patient, both the individual baseline
data for the patients from the trial and the 6-month follow-
up data are entered into the model. The model is then used
to predict lifetime costs, life expectancy, and QALYs for the
cohort of patients in the clinical trial. The model was adapted
to the Canadian context by estimating costs associated with
the specific diabetes complications modeled within the UKPDS
model, as described below under “Costs.”
The predicted outcomes for each cohort are obtained
through Monte Carlo simulation where repeated estimates of
the outcomes based on the individual patient characteristics are
obtained (11). In this analysis, 1,000 estimates of outcomes per
patient were obtained to derive expected lifetime costs, life ex-
pectancy, and QALYs. From this, mean estimates for all four
treatment comparators were obtained.
Information on some risk factors, specifically ethnicity and
presence of atrial fibrillation and peripheral vascular disease
(PVD) was not collected systematically as part of the clinical
trial. For the base case scenario, all patients were assumed to be
Caucasian nonsmokers with neither atrial fibrillation nor PVD.
These assumptions were tested within the sensitivity analysis.
INTL. J. OF TECHNOLOGY ASSESSMENT IN HEALTH CARE 28:3, 2012
Coyle et al.

Table 1. Annual Costs and Utility Values for Disease State For the time when patients experienced diabetes-related
complications, the baseline utility was reduced by an estimated
Costs10, 13, 15 utility deficit for the relevant complications. The utility deficits
associated with the seven diabetes-related complications had
First year been previously estimated from the UKPDS study (8) (Table 1).

Sensitivity Analysis
Complication Fatal Non-fatal Subsequent years Utility values8, 20
Univariate sensitivity analysis was conducted with respect to
the following factors: (4)
Diabetes management $1, 264 $1, 264 0.773
Ischemic heart disease $3, 744 $1, 403 0.683 • Patients’ smoking status, ethnicity, and the presence of atrial fibrillation or
Myocardial infarction $5, 059 $22, 057 $8, 159 0.718 PVD.
Heart failure $6, 109 $6, 109 $2, 142 0.665 • Based on outcomes at 12 months (6 months after the end of the randomized
Stroke $20, 535 $46, 505 $4, 567 0.609 portion of the trial) rather than 6 months.
Amputation $28, 401 $1, 200 0.493 • Different durations of gym membership and number of regular sessions with
Blindness $1, 519 $1, 519 0.699 an exercise specialist.
Renal failure $60, 608 $60, 608 $60, 608 0.510 • Increasing the costs of gym membership and exercise specialist sessions.
• Increasing and decreasing the cost of complications by 50 percent.
Note. The table provides the annual costs and utility values associated with each • Excluding the costs of the exercise programs.
complication of diabetes reported in 2008 Canadian dollars. Different costs are provided • Alternative discount rates for future costs and benefits.
for an incident event, subsequent years post event and a fatal event. Costs for each
complication are additional to the costs associated with diabetes management. Utilities
In addition to univariate sensitivity analysis, probabilistic
are obtained by subtracting the decrement associated with each complication from
sensitivity analysis was conducted through Monte Carlo simu-
the mean utility value from the DARE study which is used as the utility value for no
lation using the 1,000 estimates of lifetime costs and outcomes
complications.
for each patient (11). This allowed presentation of the second
order uncertainty around the results through cost-effectiveness
Costs acceptability curves (5;24). A cost-effectiveness acceptability
Both the healthcare costs associated with diabetes complica- curve shows the probability that each intervention is cost-
tions and the cost of the exercise program were incorporated effective based on alternative values for a QALY, given the
within the model. Costs that were the same for all four programs available data and their uncertainty.
(e.g., management during the run in period) were excluded from
the analysis. All costs are reported in 2008 Canadian dollars. RESULTS
A detailed literature review was conducted to determine the
costs to the healthcare system for each of the disease states Base Results
within the model: that is, the costs of managing diabetes with- In terms of total lifetime costs, including the costs of the exercise
out complications and the incremental costs associated with programs and downstream healthcare resource usage, the com-
each diabetes-related complication. The most recent relevant bined exercise program was the most expensive ($40,050), fol-
Canadian estimates were used in the analysis (10;13;15) lowed by the aerobic program ($39,250), the resistance program
(Table 1). ($38,300), and no program ($31,075) (Table 2). Both life ex-
The costs of the exercise program were based on the clini- pectancy and quality-adjusted life expectancy were highest for
cal trial program. Analysis assumed life time membership to a the combined exercise program (life-years = 11.79, QALYs =
health club, the YMCA, at a cost of $449 per annum and a cost 8.94) compared with aerobic training (life-years = 11.57,
of $250 per annum for 13 one-hour Exercise Specialist sessions. QALYs = 8.77), resistance training (life-years = 11.51,
Sensitivity analysis included alternative estimates of this cost. QALYs = 8.73), and no program (life-years = 11.48, QALYs =
8.70).
Utilities The resistance, aerobic, and combined programs were more
Utility values were necessary for all possible health states within effective and more costly compared with no exercise program.
the model. For the time when patients experience no complica- The incremental cost per QALY was $206,985, $116,793, and
tions, the utility value was derived from the DARE clinical trial $37,872 for the resistance, aerobic, and combined programs,
(20). In the trial, all patients completed the SF-36 questionnaire respectively, as compared with no exercise program. The incre-
at baseline (25). The responses to this questionnaire were used mental cost per QALY for the combined program was $4,792
to estimate a mean baseline utility value for all patients based compared with the aerobic program and $8,570 compared with
on the SF6D scoring algorithm (3) (Table 1). the resistance program.

INTL. J. OF TECHNOLOGY ASSESSMENT IN HEALTH CARE 28:3, 2012 230

 Cost-effectiveness of exercise in diabetes

Table 2. Base Results: Cost-Effectiveness

Incremental cost per QALY Incremental cost per life-year

Exercise program Cost QALY Life-years versus aerobic versus resistance versus no program versus aerobic versus resistance versus no program

Combined $40,050 8.94 11.79 $4,792 $8, 570 $37, 782 $3,556 $6, 096 $28, 494
Aerobic $39,250 8.77 11.57 $26, 601 $116, 793 $15, 780 $90, 839
Resistance $38,319 8.73 11.51 $206, 985 $233, 693
No program $31,075 8.70 11.48

Note. QALY, quality-adjusted life-year.

Table 3. Sensitivity Analysis

Incremental cost per QALY gained

Assumption Combined versus no program Combined versus resistance Combined versus aerobic

Base Case $37, 872 $8, 570 $4, 792

Smokers $36, 985 $7, 929 $3, 358
African-American patients $45, 100 $8, 021 $3, 822
Asian-Indian patients $36, 702 $9, 003 $3, 645
Patients with PVD at baseline $45, 211 $8, 423 $5, 770
Patients with Atrial fibrillation at baseline $39, 177 $15, 269 $5, 259
12 months treatment effects from DARE RCT $39, 506 $8, 707 $3, 239
Double cost of gym membership and exercise sessions $39, 970 $10, 747 $6, 878
Lifetime gym membership with 1 year of exercise specialists $24, 429 $6, 042 $2, 369
1 year gym membership with exercise specialists $3, 982 $5, 411 $1, 764
Increase costs of complications by 50% $37, 443 $9, 612 $3, 661
Decrease costs of complications by 50% $36, 411 $5, 311 $3, 239
Healthcare system perspective (no exercise program costs) $3, 090 $6, 479 $2, 509
Discount rate 0% $32, 260 $10, 939 $1, 940
Discount rate 3% $35, 333 $9, 360 $3, 941
Discount rate 10% $50, 055 $14, 837 $15, 655

Note. # Caucasian, non-smokers without PVD or atrial fibrillation, lifetime gym membership with exercise specialists, from societal perspective.
PVD, peripheral vascular disease.

Sensitivity Analysis DISCUSSION

The results were consistent across each of the univariate sen-
sitivity analyses (Table 3). In all cases, the combined exercise
program resulted in the greatest increase in life expectancy
and quality-adjusted life expectancy. At a maximum value of
$50,000 per QALY (18) the combined exercise program re-
mained cost-effective as compared with the three alternatives
across all sensitivity analyses.
The cost-effectiveness acceptability curve depicts the prob-
ability that each of the interventions is cost-effective for values
of a QALY ranging from $0 to $100,000. (Figure 1) For a ceiling
of $50,000 per QALY (18), the probability that the combined
program is the most cost-effective is 55.5 percent.
Exercise has been shown to have a positive effect for individuals
with type 2 diabetes. Previous studies have shown that aerobic
exercise training and resistance exercise training significantly
affect HbA1c values. This was confirmed by the DARE clinical
trial which also found that a combined aerobic and resistance
exercise program is more effective than programs involving only
one form of exercise. In this study, we used data from the DARE
clinical trial combined with a computer simulation model and
Canadian specific cost data to examine the cost-effectiveness of
exercise programs for persons with type 2 diabetes.
Our study found that the combined exercise program was
the most cost-effective of the alternatives compared with the

231 INTL. J. OF TECHNOLOGY ASSESSMENT IN HEALTH CARE 28:3, 2012

Coyle et al.
100%
90%
Probabiity Strategy is Most Cost Effective
80%
70%
60%
50%
40%
30%
20%
10%
0%
$0
Figure 1. Cost-effectiveness acceptability curve, depicting the probability of each of the four comparators which is the most cost-effective given different thresholds relating to a decision maker’s willingness to pay for a
quality-adjusted life-year (QALY) gained. The probability for resistance and aerobic was 0% for all ratios.
cost per QALY gained being less than $50,000 against all com-
parators. Sensitivity analysis confirmed that this finding was
robust to all changes in assumptions.
There are limitations when conducting economic analysis in
chronic diseases when based on clinical trials of short duration.
Analysis incorporated the 6-month outcomes which may not be
sustained over the long-term. However, the sensitivity analysis
incorporated the 12-month outcomes which allows for non-
compliance with exercise program beyond the initial 6-month
period. It should be noted that examining 12-month outcomes
may involve dilution of the original differences between ex-
ercise programs and controls because subjects (including con-
trols) were allowed to follow any exercise program they chose
during months 7–12 and there were no restrictions on medica-
tion changes during this period. The results of the sensitivity
analysis found that analysis based on the 12-month outcomes
led to the same interpretation as the analysis based on 6-month
outcomes. There may still be concerns that the benefits obtained
within the first 12 months might not be sustained over a longer
period of time, although the base analysis assumes that contin-
ued gym membership and exercise specialist sessions would be
funded for the patient’s lifetime.
A limitation of the clinical trial is that the duration may not
have been long enough to determine any quality of life benefits
INTL. J. OF TECHNOLOGY ASSESSMENT IN HEALTH CARE 28:3, 2012
$10,000 $20,000 $30,000 $40,000 $50,000 $60,000 $70,000 $80,000 $90,000 $100,000
Ceiling Ratio
No program Resistance Combined Aerobic
from exercise. In a previous study from the DARE trial, the
assumption that combined aerobic and resistance exercise train-
ing or resistance exercise alone would show improvements in
well-being and health status in the short-term was only partially
supported (16). Thus, analysis does not include consideration of
any additional clinical benefits to be obtained from encouraging
exercise within persons with type 2 diabetes nor does it consider
whether there are any additional costs to participants—that is,
costs of exercise clothing and lost time to pursue other activ-
ities. In the long-term, individuals will decide on whether to
continue their exercise program based on an evaluation of the
relative benefits versus costs.
Analysis assumes that exercise training during the first year
will have a one-time effect on ameliorating risk factors for type
2 diabetes. Thus factors such as HbA1c will be assumed to
increase after the first year based on the standard assumptions
within the UKPDS model. This is likely to bias against the effec-
tiveness of exercise programs if continued benefits in modifying
risk factors are obtained.
The costs of gym membership were obtained from dif-
ferent centers in Ottawa, Calgary, and Toronto. It is possi-
ble that these estimates may not be generalizable to other
communities. A sensitivity analysis was conducted doubling
the costs of membership and exercise sessions. This had
232

little effect on the estimate of the incremental cost per QALY
gained.
The application of economic analysis is often limited due to
lack of recognition of an acceptable threshold value for assess-
ing cost-effectiveness. Analysis adopted a threshold of $50,000
based on previous reports that a range of $40,000 to $60,000 is
used as an informal threshold within the province of Ontario in
decision making regarding the funding of pharmaceuticals (18).
Given this, the results of the analysis suggest that a combined
aerobic and resistance exercise program is more cost-effective
than a resistance exercise program, an aerobic exercise pro-
gram, or no program. However, this requires the assumption
that the informal threshold used for pharmaceuticals can be
interpolated to other nonpharmacological interventions. Thus,
our conclusion that the program is cost-effective is reliant on a
decision maker’s willingness to pay for a QALY being at least
$37,782.
We could not find any similar studies which have formally
evaluated cost-effectiveness of exercise programs for persons
with type 2 diabetes. There have been several studies that have
demonstrated that lifestyle modification (including exercise)
are cost-effective interventions for patients with prediabetes or
are at high risk of developing diabetes (e.g., 1;14). A study of
twenty-five patients did demonstrate a significant reduction in
costs of complications with exercise although the small scale
nature of this study and the lack of a formal economic analysis
precludes further conclusions to be drawn (6).
The conclusions of the analysis were insensitive to changes
in several parameters. Results were most sensitive to assump-
tions relating to the cost of gym membership and visits with
exercise specialists. If only the first years costs of gym member-
ship are included, the incremental costs per QALY gained was
only $3,982.
The base analysis can be considered highly conservative
as it assumes both no additional benefits on risk factor modi-
fication beyond the first year but included the continued costs
of exercise programs for lifetime. Despite these assumptions,
analysis suggests that it would be at least as cost-effective
for a healthcare ministry to cover the costs of a combined
aerobic and resistance exercise program for individuals with
type 2 diabetes; including both gym memberships and sessions
with an exercise specialist; as covering the costs of diabetic
medications
CONCLUSION
A combined program providing training in both resistance and
aerobic exercise was the most cost-effective of the alternatives
compared based on previous funding decisions. Within a public
healthcare system, the funding of exercise training for individ-
uals with type 2 diabetes can be considered an efficient use of
resources.
233
Cost-effectiveness of exercise in diabetes
CONTACT INFORMATION
Doug Coyle, PhD, Professor, Department of Epidemiology and
Community Medicine, University of Ottawa, Ottawa, Ontario,
Canada
Kathryn Coyle, MSc, Senior Associate, Applied Health Eco-
nomic Research Unit, Ottawa, Ontario, Canada
Glen P. Kenny, PhD, University of Ottawa Research Chair in En-
vironmental Physiology, School of Human Kinetics, Faculty of
HealthSciences, University of Ottawa, Ottawa, Ontario, Canada
Normand G. Boulé, PhD, Professor, School of Human Kinet-
ics, Faculty of Health Sciences, University of Ottawa, Faculty
of Physical Education and Recreation, University of Alberta,
Edmonton, Canada
George A. Wells, PhD, Professor, Department of Epidemiol-
ogy and Community Medicine, Faculty of Medicine, University
of Ottawa; Director, Cardiovascular Research Methods Centre,
University of Ottawa Heart Institute, Ottawa, Ontario, Canada
Michelle Fortier, PhD, Professor, School of Human Kinetics,
Faculty of Health Sciences, University of Ottawa, Department
of Psychology, Faculty of Medicine, University of Ottawa,
Ottawa, Ontario, Canada
Robert D. Reid, PhD, Associate Professor, Medicine, University
of Ottawa; Associate Director, Minto Prevention and Rehabil-
itation Centre, University of Ottawa Heart Institute, Ottawa,
Ontario, Canada
Penny Phillips, MA, Research Coordinator, Clinical Epidemi-
ology Program, Ottawa Health Research Institute, Ottawa, On-
tario, Canada
Ronald J. Sigal, Professor, School of Human Kinetics, Faculty of
Health Sciences, Department of Medicine, Faculty of Medicine,
University of Ottawa, University of Ottawa, Ottawa, Ontario,
Canada; Professor, Departments of Medicine, Cardiac Sciences
and Community Health Sciences, Faculties of Medicine and
Kinesiology, University of Calgary, Calgary, Alberta, Canada
CONFLICTS OF INTEREST
Glen Kenny, Penny Phillips and Ronald Sigal report having
received a grant to their institution from Canadian Institutes of
Health Research. The other authors have no potential conflicts
of interest.
REFERENCES
1. Bertram MY, Lim SS, Barendregt JJ, Vos T. Assessing the cost-
effectiveness of drug and lifestyle intervention following opportunistic
screening for pre-diabetes in primary care. Diabetologia. 2010;53:875-
881.
2. Boulé NG, Haddad E, Kenny GP, Wells GA, Sigal RJ. Effects of exercise
on glycemic control and body mass in type 2 diabetes mellitus: A meta-
analysis of controlled clinical trials. JAMA. 2001;286:1218-1227.
3. Brazier J, Roberts, Deverill M. The estimation of a preference based
measure of health from the SF36. J Health Econ. 2002;21:271-292.
4. Briggs A, Sculpher M, Buxton M. Uncertainty in the economic evaluation
of health care technologies: The role of sensitivity analysis. Health Econ.
1994;3:95-104.
INTL. J. OF TECHNOLOGY ASSESSMENT IN HEALTH CARE 28:3, 2012
Coyle et al.
5. Briggs AH. Handling uncertainty in cost-effectiveness models. Pharma-
coeconomics. 2000;17:479-500.
6. Brun JF, Bordenave S, Mercier J, Jaussent A, Picot MC, Préfaut C.
Cost-sparing effect of twice-weekly targeted endurance training in type
2 diabetics: A one-year controlled randomized trial. Diabetes Metab.
2008;34:258-265.
7. CADTH. Guidelines for the economic evaluation of health technologies:
Canada [3rd Edition]. Ottawa: CADTH; 2006.
8. Clarke P, Gray A, Holman R. Estimating utility values for health states
of type 2 diabetic patients using the EQ-5D (UKPDS 62). Med Decis
Making. 2002;22:340-349.
9. Clarke PM, Gray AM, Briggs A, et al. A model to estimate the lifetime
health outcomes of patients with Type 2 diabetes: The United Kingdom
Prospective Diabetes Study (UKPDS 68) Outcomes Model. Diabetologia.
2004;47:1747-1759.
10. Coyle D, Palmer AJ, Tam R. Economic evaluation of PioglitazoneHCl in
the management of Type 2 Diabetes Mellitus in Canada. Pharmacoeco-
nomics. 2002;20:31-42.
11. Doubilet P, Begg CB, Weinstein MC, et al. Probabilistic sensitivity anal-
ysis using Monte Carlo simulation: A practical approach. Med Decis
Making. 1985;5:157-177.
12. Eves ND, Plotnikoff RC. Resistance Training and Type 2 Diabetes. Con-
siderations for implementation at the population level. Diabetes Care.
2006;29:1933-1941.
13. Goeree R, Lim ME, Hopkins R, et al. Prevalence, total and excess costs of
diabetes and related complications in Ontario, Canada. Can J Diabetes.
2009;33:35-45.
14. Herman WH, Hoerger TJ, Brandle M, et al. The cost-effectiveness of
lifestyle modification or metformin in preventing type 2 diabetes in adults
with impaired glucose tolerance. Ann Intern Med. 2005;142:323-332.
15. Lamy A, Yusuf S, Pogue J, Gafni A. Cost implications of the use of
INTL. J. OF TECHNOLOGY ASSESSMENT IN HEALTH CARE 28:3, 2012
ramipril in high-risk patients based on the Heart Outcomes Prevention
Evaluation (HOPE) study. Circulation. 2003;107:960-965.
16. Reid RD, Tulloch HE, Sigal RJ, et al. Effects of aerobic exercise, resis-
tance exercise or both, on patient-reported health status and well-being in
type 2 diabetes mellitus: A randomised trial. Diabetologia. 2010;53:632-
640.
17. Report of the Expert Committee on the Diagnosis and Classification of
Diabetes Mellitus. Diabetes Care. 1997;20:1183-1197.
18. Sapsford R. Deputy Minister’s response to the investigation into the Min-
istry of Health and Long-Term Care’s decision-making concerning the
funding of Avastin for colorectal cancer patients. Letter dated August
26th 2009.
19. Selvin E, Marinopoulos S, Berkenblit G, et al. Meta-analysis: Glycosy-
lated hemoglobin and cardiovascular disease in diabetes mellitus. Ann
Intern Med. 2004;141:421-431.
20. Sigal RJ, Kenny GP, Boulé NG, et al. Effects of aerobic exercise, resis-
tance exercise, or both on glycemic control in type 2 diabetes: A random-
ized trial. Ann Intern Med. 2007;147:357-369.
21. Snowling NJ, Hopkins WG. Effects of different modes of exercise training
on glucose control and risk factors for complications in type 2 diabetic
patients: A meta-analysis. Diabetes Care. 2006;29:2518-2527.
22. Stratton IM, Adler AI, Neil HA, et al. Association of glycaemia
with macrovascular and microvascular complications of type 2 diabetes
(UKPDS 35): Prospective observational study. BMJ. 2000;321:405-412.
23. Thomas DE, Elliott EJ, Naughton GA. Exercise for type 2 diabetes mel-
litus. Cochrane Database Syst Rev. 2006;3:CD002968.
24. van Hout BA, Al MJ, Gordon GS, et al. Costs, effects and C/E ratios
alongside a clinical trial. Health Econ. 1994;3:309-319.
25. Ware JE, Snow KK, Kolinski M, Gandeck B. SF-36 Health survey manual
and interpretation guide. Boston, MA: The Health Institute, New England
Medical Centre; 1993.
234
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.