UNDER EMBARGO UNTIL AUGUST 6, 2013, 12:01 AM ET

Lifetime Direct Medical Costs of Treating Type 2 Diabetes and Diabetic Complications
Xiaohui Zhuo, PhD, Ping Zhang, PhD, Thomas J. Hoerger, PhD
This activity is available for CME credit. See page A3 for information.

Background: Lifetime direct medical cost of treating type 2 diabetes and diabetic complications in the U.S. is unknown. Purpose: This study provides nationally representative estimates of lifetime direct medical costs of
treating type 2 diabetes and diabetic complications in people newly diagnosed with type 2 diabetes, by gender and by age at diagnosis.

Methods: A type 2 diabetes simulation model was used to simulate the disease progression and
direct medical costs among a cohort of newly diagnosed type 2 diabetes patients. The study sample used for the simulation was based on data from the 20092010 National Health and Nutritional Examination Survey. The costs of treating type 2 diabetes and diabetic complications were derived from published literature. Annual medical costs were accumulated over the life span of type 2 diabetes to determine the lifetime medical costs. All costs were calculated from a healthcare system perspective, and expressed in 2012 dollars.

Results: In men diagnosed with type 2 diabetes at ages 2544 years, 4554 years, 5564 years, and 65 years, the lifetime direct medical costs of treating type 2 diabetes and diabetic complications were $124,700, $106,200, $84,000, and $54,700, respectively. In women, the costs were $130,800, $110,400, $85,500, and $56,600, respectively. The agegender weighted average of the lifetime medical costs was $85,200, of which 53% was due to treating diabetic complications. The cost of managing macrovascular complications accounted for 57% of the total complication cost. Conclusions: Over the lifetime, type 2 diabetes imposes a substantial economic burden on healthcare systems. Effective interventions that prevent or delay type 2 diabetes and diabetic complications might result in substantial long-term savings in healthcare costs.
(Am J Prev Med 2013;45(3):253 261) Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine

Introduction

early 26 million Americans currently live with diabetes. The disease imposes large and growing economic burdens on the healthcare system and society. The estimated total economic cost of diabetes in 2012 was $245 billion, including $176 billion in direct medical costs and $69 billion in reduced productivity.1 The number of people in the U.S. with diabetes is

From the Division of Diabetes Translation (Zhuo, Zhang), National Center for Chronic Disease Prevention and Health Promotion, CDC, Atlanta, Georgia; and the Research Triangle International (Hoerger), Research Triangle Park, North Carolina Address correspondence to: Xiaohui Zhuo, PhD, Division of Diabetes Translation, CDC, 4770 Buford Highway, MS K-10, Atlanta GA. E-mail: xzhuo@cdc.gov. 0749-3797/$36.00 http://dx.doi.org/10.1016/j.amepre.2013.04.017

projected to nearly double by 2034, with medical costs nearly tripling, to $336 billion.2 The economic burden of diabetes can be measured on the basis of prevalent cases or incident cases. The prevalence-based approach, which is used in most existing cost-of-diabetes studies, assesses the costs associated with diabetes in a given year, regardless of the length of time since diabetes became evident.36 A recent prevalence-based study, conducted by the American Diabetes Association, estimated that people with diabetes incur an average medical cost of  $13,700 per year, of which  $7900 is attributed to diabetes.1 Although the prevalence-based cost studies provide useful measurement on the economic burden of the disease for a given period of time, they do not reveal the long-term economic consequences associated with a new incident case of type 2 diabetes.
Am J Prev Med 2013;45(3):253 261 253

Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine

254

Zhuo et al / Am J Prev Med 2013;45(3):253 261 complications are measured as the primary endpoint outcomes in the United Kingdom Prospective Diabetes Study (UKPDS) and most of the other major clinical trials.1519 The clinical progression of the ve disease pathways was primarily based on data from the UKPDS,15,2023 which is the largest and longest clinical study concerning the effect of intensive glycemic control on diabetes-related microvascular and macrovascular complications among newly diagnosed type 2 diabetic patients. Patients in the intensive intervention arm of the UKPDS received an intensive glycemic control with a treatment goal of a hemoglobin A1c level of 7.0%. Patients who had hypertension or hyperlipidemia also received intensive blood pressure control and intensive cholesterol control, respectively.24 Such treatment regimens have been part of the treatment standard recommended by the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD).25 In the model, the deaths of the people with diabetes could be attributed to end-stage renal disease; lower extremity amputation (LEA); coronary heart disease; stroke; and other causes. In addition to the increased mortalities due to type 2 diabetesrelated complications, higher mortalities were assumed for people with type 2 diabetes than for those without.26

In contrast, the incidence-based costing method measures the cumulative costs of an incident case of diabetes from onset or diagnosis to death. Compared with the prevalence-based costs, such a method is more useful for understanding the long-term scal impact of type 2 diabetes on individual and healthcare systems. Incidencebased costs are also critical for public health ofcials and private insurers in evaluating the potential long-term nancial return and cost effectiveness of type 2 diabetes prevention programs. Despite the relevance, few studies have assessed the lifetime medical cost of type 2 diabetes. To our knowledge, no national-level estimates of the lifetime cost of diabetes are yet available in the U.S. Estimating incidence-based costs is challenging, because it usually requires a longitudinal follow-up of patients. Such longitudinal data are relatively scarce and expensive to collect. Without empirical data, incidencebased estimates can be calculated using a disease simulation model that simulates the natural progression of diseases and the associated costs. In the present study, a validated type 2 diabetes simulation model was used to estimate the lifetime direct medical costs of treating type 2 diabetes and diabetic complications in people with newly diagnosed type 2 diabetes in the U.S.

Direct Medical Costs

Direct medical costs were dened as the medical cost of managing type 2 diabetes and medical costs of treating diabetic complications. Lifetime direct medical costs were calculated in two steps. First, annual direct medical costs of diabetes management and treating diabetes-related complications were estimated based on published literature. Second, these annual estimates were assigned to the life-years in which an individual had diabetes, and life-years in which patients had complications, respectively. Annual medical costs were accumulated from the diagnosis of type 2 diabetes in 2010 to death to determine the per capita lifetime direct medical costs for each of the eight age- and gender-specic cohorts. After calculating the per capita costs of the people in the eight cohorts, the national weighted average lifetime cost per person was calculated by the sum of the products of the per capita costs of each cohort by the proportions of the populations they represented. The weighted average represents the lifetime direct medical cost for an average person who was diagnosed with type 2 diabetes in 2010. All costs were rounded to the nearest $100 and expressed in 2012 U.S. dollars. Future costs were discounted at a 3% annual discount rate.

Methods
Type 2 Diabetes Simulation Model
The current simulation model is a Markov-based lifetime model that follows a cohort of people from the time of type 2 diabetes diagnosis to death.7,8 The model includes the annual transition probabilities between disease stages and death. The basic model structure has been described previously,7,9,10 and the technical details of the model have also been presented.9 The model has been validated against major clinical trials and been found to accurately predict the progression of the disease and development of complications.8 This model has previously been used to evaluate the cost effectiveness of interventions designed to prevent or delay type 2 diabetes or prevent diabetic complications.7,9,11 For this analysis, the model was used to simulate the disease progression among eight cohorts of newly diagnosed type 2 diabetes patients. They are women and men diagnosed with type 2 diabetes at ages 2544 years, 4554 years, 5564 years, and 65 years in 20092010. Demographic characteristics and prevalence of hypertension, cholesterol level, and smoking status of the simulated cohorts are based on data of year 20092010 from the National Health and Nutritional Examination Survey.12 The most common diabetic complications are simulated by ve separate modules, including nephropathy; neuropathy; and retinopathy (microvascular) and stroke and coronary heart diseases (macrovascular; Figure 1).13 Approximately 58% of patients with diabetes have at least one microvascular or macrovascular complication.14 These ve diabetic complications were included in the model primarily because, rst, the injurious effects of hyperglycemia on microvascular and macrovascular systems are well established by a large body of clinical evidence13; second, those

Medical Costs of Diabetes Management

The cost of managing type 2 diabetes included the cost of medications, physician visits, and self-testing devices (Table 1). Based on a treatment algorithm recommended by the ADA and the EASD,25 newly diagnosed diabetic patients received, at diagnosis, both metformin and a lifestyle intervention. Other drugs (e.g., sulfonylureas, glitazones, and insulin) would be added if needed to achieve/maintain a hemoglobin A1c level of o7.0%. Metformin was used in 54% of treatments, sulfonylureas in 34%, glitazones in 28%, and insulin in 28%.27 Total medication costs were calculated by multiplying the total quantity of each drug used by the price for that drug. Price data were obtained from Red Book 2010: Pharmacys Fundamental Reference (hereafter simply Red Book 2010).28 www.ajpmonline.org

Zhuo et al / Am J Prev Med 2013;45(3):253 261

Coronary heart disease
Angina

255

Normal

CHD

CA/MI

Death

History of CA/MI

Stroke

Within-year events

Normal

Stroke

History of stroke

Death

Nephropathy
Normal Low-micro/ high-micro Clinical nephropathy ESRD ESRD death

Neuropathy
Normal Peripheral neuropathy LEA History of LEA
Subsequent LEA

LEA death

Retinopathy
Normal Photocoagulation Blind

Figure 1. Simulations of the development of diabetic complication

Note: The ve diabetic complications are the most common complications among type 2 diabetic patients. It was estimated that 35% of diabetic patients had heart disease or stroke (www.cdc.gov/diabetes/statistics/cvd/g3.htm); 29% of people with diabetes aged 40 years had diabetic retinopathy (diabetes.niddk.nih.gov/dm/pubs/statistics/#Blindness); about 60%70% of people with diabetes had mild to severe forms of neuropathy. About 40% of diabetic patients have chronic renal disease (www.ncbi.nlm.nih.gov/pubmed/20338960). CA, cardiac arrest; CHD, coronary heart disease; ESRD, end-stage renal disease; LEA, lower-extremity amputation; MI, myocardial infarction

The frequencies of physician ofce visits were derived from data of the National Health Interview Survey Diabetes Supplement.29 The data suggest that people with type 2 diabetes who were not using insulin made an average of seven physician ofce visits per year; those using insulin made an average of eight ofce visits per year. During a regular ofce visit, patients were assumed to receive a hemoglobin A1c test and 25 minutes of counseling. In addition to regular visits, patients received a more comprehensive visit per year, in which patients would receive 35 minutes of counseling, a September 2013

hemoglobin A1c test, a urinalysis, an albumin and creatinine test, and a lipid test. Based on the 2012 Medicare fee schedule, the cost of a regular visit was estimated at $50 and the cost of an annual visit at $110. The number of visits was multiplied by their respective unit cost by visit type to determine the annual total cost for physician ofce visits. To calculate the annual cost of selftesting, the level of use of self-testing was rst estimated based on data from the National Health Interview Survey Diabetes

256

Zhuo et al / Am J Prev Med 2013;45(3):253 261 Supplement.29 On average, insulin users conducted self-testing three times daily; those not using insulin conducted self-testing once daily. The self-testing devices included glucose test strips, lancets, glucose meters, and glucagon. The total number of each device used per year was multiplied by the unit cost of that device (based on Red Book 2010) to determine the annual cost of self-testing.

Table 1. Direct medical costs of type 2 diabetic care and diabetic complications
Cost of one-time incidence ($)
TYPE 2 DIABETES CARE Medications 148597a Nathan (2009)25 Alexander (2008)27 Red Book 201028

Annual medical cost ($)

Study/source

Physician visits Noninsulin users 477 National Health Interview Survey data29 Medicare Fee Schedule 2012

Medical Costs of Diabetic Complications

Annual costs of diabetic complications were dened as the cost of a one-time event in the year it rst occurred and the annual treatment cost in subsequent years. A comprehensive literature review was conducted to obtain the up-to-date estimates on the medical costs of the disease states simulated in the diabetic complication model (Table 1). Medical costs of nephropathy include the costs of microalbuminuria, macroalbuminuria/clinical nephropathy, and end-stage renal disease. The costs came from a study of Kaiser Permanente patients29 and the U.S. Renal Data System.35 The costs of diabetic neuropathy, including peripheral neuropathy, diabetic foot ulcer and LEA, are based on a study by OBrien et al.32 Fatal and nonfatal LEA were reported separately. Medical costs of diabetic retinopathy were separated into the cost of photocoagulation and blindness, taken from Rein and colleagues study33 based on private insurance and Medicare claims data. The medical costs of coronary heart diseases and stroke were primarily based on a study by OBrien et al.32 and a study by Taylor et al.34 The medical costs of death caused by complications other than neuropathy and coronary heart disease were based on a study of Medicare data.35

Insulin users

522

Self-testing device Noninsulin users 90 National Health Interview Survey data29 Red Book 201028

Insulin users DIABETIC COMPLICATIONS Nephropathy Microalbuminuria Macroalbuminuria/Clinical nephropathy End-stage renal disease Neuropathy Peripheral neuropathy Diabetic foot ulcer Initial lower-extremity amputation Subsequent lower-extremity amputation Retinopathy Photocoagulation Aged o65 years Aged 65 years Glaucoma/blindness Aged o64 years Aged 65 years Coronary heart disease Angina 8,464 1850 1850 Fatal Nonfatal Fatal Nonfatal 78,480 57,032 78,480 57,032 523

278

430 3,936 78,795

Nichols (2011)30 Nichols (2011)30 USRDS 201231

17,959 OBrien (2003)32

686 649 Rein (2006)33

3654 6101

2187
(continued on next page)

Sensitivity Analyses
The uncertainties of the lifetime cost estimates were examined by www.ajpmonline.org

Zhuo et al / Am J Prev Med 2013;45(3):253 261

257

2 diabetes in 2010. The lifetime medical cost of treating Cost of diabetes and diabetic comone-time Annual plications in men who were incidence medical diagnosed at ages 2544 ($) cost ($) Study/source years was $124,700. The lifeDeath from cardiac arrest or 1,186 O'Brien (2003)32 time direct medical cost was myocardial infarction lower in patients who were without hospitalization diagnosed when older: Death from cardiac arrest or 24,414 $106,200 for men diagnosed myocardial infarction with type 2 diabetes at ages within 30 days of hospitalization 4554 years, $84,000 when diagnosed at 5564 years, Nonfatal cardiac arrest or 42,662 myocardial infarction and $54,700 when diagnosed at aged 65 years. Women History of cardiac arrest or 2,358 myocardial infarction with type 2 diabetes had slightly higher lifetime medStroke, by age in years ical cost. The costs were 44 81,850 15,682 $130,800 when diagnosed at 4554 81,850 8,952 ages 2544 years, $110,400 when diagnosed at ages 45 5564 81,850 28,943 OBrien (2003)32 54 years, $85,500 when diag34 6574 63,374 23,139 Taylor (1996) nosed at ages 5564 years, 7584 60,199 20,084 and $56,600 when diagnosed at ages 65 years. The over85 46,840 8,788 all gender- and age-weighted Cost of death, by age in years average lifetime medical cost 74 19,870 of treating diabetes and its 7584 17,307 Hogan (2001)35 related complications was $85,200. 85 11,747 Diabetic complications Note: All costs are in 2012 dollars. account for 48%64% of a The estimated annual cost increases from $148 in the year of diagnosis to $597 in the year of death or up to 9 the lifetime medical cost. age 94 years. A detailed calculation has been presented previously. USRDS, U.S. Renal Data System Of the medical costs of diabetic complications, 57% varying each of the cost parameters used in the model. The baseline were spent on treating stroke and coronary heart disease. value of the cost parameters was varied from the lower value to the Patients diagnosed at older ages spent a greater proporupper value of the 95% CIs. If CIs were not available, the parameter tion of their expenses on diabetic complications than was varied from 50% and 150% of the baseline values. In addition, an younger patients. For example, men diagnosed with type alternative method for calculating costs from diabetic complications 2 diabetes when aged 65 years spent about 64% on was used. In the base-case analysis, the cost of diabetic complications treating diabetic complications, compared with 50% in was computed using an additive costing method. Therefore, medical men diagnosed when aged 2544 years. costs other than the modeled diabetic complications (e.g., hypoglyTable 1. (continued )
cemia and stomach nerve damage) were not included in the cost calculation. In the alternative costing method, complication costs were estimated using a regression-based multiplicative model based on data from an employer-based insurance plan.36 The annual medical cost of complications was calculated as the product of the estimated baseline cost and the multipliers corresponding to cohort characteristics and all the possible diabetic complications.

Sensitivity Analyses
Figure 2 shows the results of the sensitivity analyses in men who were diagnosed at ages 4554 years. Results from other cohorts were similar. The annual treatment cost of stroke and the cost of incident cases of nonfatal cardiac arrest or myocardial infarction were most inuential on the lifetime cost estimates. The lifetime medical cost would be reduced by up to $10,000 or increased by up to $8100 if either of the two parameters changed.

Results
Table 2 presents the lifetime direct medical costs (in 2012 dollars) in women and men who were diagnosed with type
September 2013

258

Zhuo et al / Am J Prev Med 2013;45(3):253 261

Table 2. Lifetime direct medical costs in patients with type 2 diabetes

Age at diagnosis, years 2544
MEN Total, $ (%) Diabetes management Nephropathy Neuropathy Retinopathy Coronary heart disease Stroke WOMEN Total, $ (%) Diabetes management Nephropathy Neuropathy Retinopathy Coronary heart disease Stroke 130,800 67,600 (52) 16,200 (12) 12,200 (9) 9,400 (7) 6,600 (5) 18,800 (14) 110,400 52,800 (48) 9,100 (8) 8,100 (7) 10,300 (9) 10,100 (9) 20,000 (18) 85,500 39,700 (46) 5,200 (6) 5,400 (6) 6,900 (8) 9,700 (11) 18,600 (22) 56,600 25,300 (45) 2,400 (4) 2,800 (5) 3,800 (7) 7,600 (13) 14,700 (26) 124,700 61,800 (50) 10,600 (9) 13,400 (11) 5,900 (5) 19,200 (15) 13,800 (11) 106,200 49,100 (46) 7,100 (7) 6,600 (6) 8,400 (8) 15,400 (15) 19,600 (18) 84,000 35,600 (42) 4,000 (5) 4,200 (5) 5,400 (6) 14,800 (18) 20,000 (24) 54,700 19,700 (36) 1,600 (3) 2,000 (4) 2,700 (5) 13,000 (24) 15,700 (29)

4554

5564

65

Note: All costs are in 2012 U.S. dollars (discounted at 3% per year for any costs beyond 2012). Percentage values in parentheses indicate percentage of total.

Varying any of the other individual cost components resulted in a change of no more than $3000 from the base-case estimates. Adopting an alternative multiplicative costing method for the costs of diabetic complications increased the total lifetime cost by up to $3200.

Discussion
Type 2 diabetes imposes a persistent economic burden over the life span. However, no estimates of lifetime medical costs of type 2 diabetes are available in the U.S. The present study provides estimates of the lifetime costs of treating type 2 diabetes and diabetic complications, stratied by gender and the age of diagnosis. The current estimates indicate that the nancial burden of a new case of type 2 diabetes imposed on the healthcare system is substantial, and this nancial burden is particularly high in people diagnosed with type 2 diabetes at younger ages, primarily because of the longer cumulative exposure to diabetes. Women were found to have greater lifetime medical costs than men. This is primarily because even though women have fewer complications, on average, they live longer than men.20,25

The current estimated lifetime medical costs of diabetic complications were in line with the estimates previously reported by Caro and colleagues (an estimated lifetime medical cost of diabetic complications to be $47,000 over 30 years).37 Among the few lifetime cost studies, Birnbaum and colleagues estimated the lifetime incremental medical cost among diabetic women using data from an employer-based insurance claims.38 However, the Birnbaum estimates are different than those reported in the present study, because the Birnbaum study estimated the incremental cost by comparing the costs of people with and without diabetes. In addition, the authors also included both direct and indirect medical costs among women with diabetes (type 1 and type 2). In contrast, the present study examined direct medical costs among type 2 diabetes patients only. No comparison was made of the cost with that of the people without diabetes. A study by Goldman and colleagues assessed the lifetime medical spending among the elderly under hypothetic scenarios of diabetes prevention.39 However, their estimate represents the medical cost of average people (i.e., people with or without diabetes) and is therefore not comparable.
www.ajpmonline.org

Zhuo et al / Am J Prev Med 2013;45(3):253 261

Multiplicative costing to calculate complication cost Annual cost of stroke Cost of incident stroke Annual cost of CA/MI Cost of incident nonfatal CA/MI Annual cost of angina Annual cost of ESRD Cost of death from other causes for ages <65 years Cost of CA/MI-caused death (with hospitalization) One-time incidence cost of angina Cost of stroke-caused death One-time incidence cost of ESRD Cost of LEA-caused death Annual cost of blindness Drug costs reduced by 30% 10000 8000 6000 4000 2000 0 2000 4000 6000 Deviation from base-case estimate, $ 8000 10000

259

Figure 2. Univariate sensitivity analyses in diabetic men diagnosed at age 4554 years
Note: The horizontal axis is the difference of the lifetime cost from the estimate in base-case analysis. For example, varying the cost of new incident stroke to the upper limit of its estimate increased the lifetime cost estimate by more than $4000. CA, cardiac arrest; ESRD, end-stage renal disease; LEA, lower-extremity amputation; MI, myocardial infarction

The substantial lifetime medical cost highlights the potential economic return of diabetes prevention. Policymakers have called for additional efforts to prevent type 2 diabetes, particularly among people who are at high risk for developing type 2 diabetes.40 Major clinical trials from the U.S. and other countries have established that, among those with elevated glucose levels, structured lifestyle modications can reduce the risk of developing diabetes by 40%60%.4143 A recent study further suggests that the reduction in the risk of diabetes can be achieved at a relatively low cost.44 However, the cost effectiveness of such prevention efforts depends on the magnitude of the downstream costs that ultimately would be saved. The current lifetime cost estimates clearly show that, if type 2 diabetes and diabetic complications could be prevented, a substantial downstream cost could be potentially saved. This nding reinforces earlier evidence that clinical and communitybased interventions to prevent type 2 diabetes could be highly cost effective.45 The present study also indicates the economic importance of glycemic management to prevent diabetic complications in type 2 diabetes patients. Diabetic complications accounted for more than half of the lifetime direct medical costs among patients with type 2 diabetes, and the majority of those costs were associated with macrovascular diseases. Poor adherence to recognized standards of management care in glycemic, blood pressure, and cholesterol control is a primary cause of
September 2013

increased risk for diabetic complications.46 Among people with diabetes, only 3% of insulin users and 1% of nonusers met all ve of the ADAs recommended standards for risk factor management and complication prevention.47 The present study suggests that decreasing the incidence of complications could have a major cost impact. Therefore, from a cost perspective, it is important that people with diabetes achieve appropriate treatment targets of diabetes management care to reduce the risk of complications. The current lifetime cost estimates provide a basis for public or private healthcare systems to assess the longterm scal impact of type 2 diabetes. Currently, there are about 26 million American adults who have diagnosed diabetes; 90%95% of those have type 2 diabetes,48 and given the lifetime cost of medical treatment, the total economic resources used for caring for this population in the next several decades would be enormous. For example, the current population aged 65 years would be expected to create about $170 billion in lifetime medical spending. Further, given that prevalence of type 2 diabetes increases in younger people,49 lifetime medical costs in future cohorts of patients would likely increase. Because of the large degree of uncertainty with extant estimates of the indirect costs of type 2 diabetes, the lifetime indirect costs were not considered in the present study. A large body of literature has suggested that type 2 diabetes is associated with reduced workforce participation, workday absence, and reduced productivity at work.

260

Zhuo et al / Am J Prev Med 2013;45(3):253 261

In fact, a recent prevalence-based study estimated that the annual productivity loss per person with diabetes (including type 1 and type 2) was nearly half that of direct medical costs.1 If this estimate is similar for incidence-based costs, people with diabetes would also incur considerable indirect costs over their lifetime.

Limitations
The study has several limitations. First, as in other modeling studies, it was assumed that current treatment costs for type 2 diabetes and its complications would remain relatively stable. If any signicant changes occur because of advances in medical technologies or other unforeseen factors, the current lifetime estimates would need revision. Second, the simulation model was developed based on data collected from multiple sources and thus was inevitably subject to the accuracy of the data sources and the assumptions employed by the researchers in source studies. In particular, because no U.S.-based data were available at the time of the analysis, the simulation of the progression of diabetes was primarily based on data from a UK-based clinical trial. However, because diabetic pathology is generally believed to be similar, UKPDS has been used as the clinical base for the ADAs recommendation for the standard of diabetes care in the U.S.50,51 Third, because of constraints on computation power, a full probabilistic sensitivity analysis could not be performed. Ideally, the model parameters would be sampled from the distributions simultaneously to determine the change of the lifetime cost estimate. Fourth, without a nondiabetic cohort as a comparison group, it was not possible to compare the costs for people with/without diabetes; thus, the incremental cost attributable to type 2 diabetes could not be calculated. Future models could examine the incremental cost if longitudinal data are available in both groups. Finally, the current model only simulated ve of the most common diabetic complications. Because of lack of data, other complications such as hypoglycemia were not considered. Had those costs been included, the lifetime medical cost of diabetic complications would have been greater.

The authors thank Lawrence Barker, PhD, and Barbara Bardenheier, PhD, of the CDC for helpful comments, and Tony Pearson-Clarke of the CDC for the excellent editing. The ndings and conclusions in this article are those of the authors and do not necessarily represent the ofcial position of the CDC. No nancial disclosures were reported by the authors of this paper.

References
1. American Diabetes Association. Economic costs of diabetes in the U.S. in 2012. Diabetes Care 2013;36(4):103346. 2. Huang ES, Basu A, O'Grady MJ, Capretta JC. Using clinical information to project federal health care spending. Health Aff (Millwood) 2009;28(5):w978990. 3. American Diabetes Association. Economic costs of diabetes in the U.S. in 2007. Diabetes Care 2008;31(3):596615. 4. Chen Y, Quick WW, Yang W, et al. Cost of gestational diabetes mellitus in the U.S. in 2007. Popul Health Manag 2009;12(3):16574. 5. Ettaro L, Songer TJ, Zhang P, Engelgau MM. Cost-of-illness studies in diabetes mellitus. Pharmacoeconomics 2004;22(3):14964. 6. Songer TJ. Studies on the cost of diabetes. Diabetes public health resource 1998. www.cdc.gov/diabetes/pubs/costs/intro.htm#references. 7. CDC Diabetes Cost-effectiveness Group. Cost-effectiveness of intensive glycemic control, intensied hypertension control, and serum cholesterol level reduction for type 2 diabetes. JAMA 2002;287(19): 254251. 8. Hoerger TJ, Segel JE, Zhang P, Sorensen SW. Validation of the CDCRTI diabetes cost-effectiveness model. Research Triangle Park NC: RTI International, 2009. 9. Herman WH, Hoerger TJ, Brandle M, et al. The cost-effectiveness of lifestyle modication or metformin in preventing type 2 diabetes in adults with impaired glucose tolerance. Ann Intern Med 2005;142(5): 32332. 10. Hoerger TJ, Harris R, Hicks KA, Donahue K, Sorensen S, Engelgau M. Screening for type 2 diabetes mellitus: a cost-effectiveness analysis. Ann Intern Med 2004;140(9):68999. 11. Zhuo X, Zhang P, Gregg EW, et al. A nationwide community-based lifestyle program could delay or prevent type 2 diabetes cases and save $5.7 billion in 25 years. Health Aff (Millwood) 2012;31(1):5060. 12. National Center for Health Statistics. National Health and Nutrition Examination Survey Data. Hyattsville MD: DHHS, CDC, 2010. 13. Fowler M. Microvascular and macrovascular complications of diabetes. Clin Diabetes 2011;29(3):6. 14. Mitka M. Report quanties diabetes complications. JAMA 2007;297 (21):23378. 15. UK Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. BMJ 1998;317(7160):70313. 16. Action to Control Cardiovascular Risk in Diabetes Study Group, Gerstein HC, Miller ME, Byington RP, et al. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med 2008;358(24):254559. 17. Gazi IF, Mikhailidis DP. Efcacy and safety of atorvastatin in the prevention of cardiovascular end points in subjects with type 2 diabetes: the Atorvastatin Study for Prevention of Coronary Heart Disease Endpoints in Non-Insulin-Dependent Diabetes Mellitus (ASPEN): response to Knopp. Diabetes Care 2006;29(11):2561; author reply 25612. 18. ADVANCE Collaborative Group. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med 2008;358(24):256072.

Conclusion
Type 2 diabetes imposes a substantial lifetime economic burden on healthcare systems. This high lifetime cost highlights the potential long-term scal return on investment of interventions to prevent or delay type 2 diabetes or diabetic complications. The current lifetime cost estimates can be used as one of the benet measures to evaluate the economic efciency of various type 2 diabetes prevention and control policies or programs.

www.ajpmonline.org

Zhuo et al / Am J Prev Med 2013;45(3):253 261

19. Hemmingsen B, Lund SS, Gluud C, et al. Intensive glycaemic control for patients with type 2 diabetes: systematic review with meta-analysis and trial sequential analysis of randomised clinical trials. BMJ 2011;343:d6898. 20. Adler AI, Stevens RJ, Manley SE, Bilous RW, Cull CA, Holman RR. Development and progression of nephropathy in type 2 diabetes: the United Kingdom Prospective Diabetes Study (UKPDS 64). Kidney Int 2003;63(1):22532. 21. Haak T. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet 1998;352(9131):83753. 22. UK Prospective Diabetes Study (UKPDS) Group. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). Lancet 1998;352(9131):85465. 23. Stevens RJ, Kothari V, Adler AI, Stratton IM. The UKPDS risk engine: a model for the risk of coronary heart disease in Type II diabetes (UKPDS 56). Clin Sci (Lond) 2001;101(6):6719. 24. UK Prospective Diabetes Study (UKPDS) Group. Intensive bloodglucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 1998;352(9131):83753. 25. Nathan DM, Buse JB, Davidson MB, et al. Medical management of hyperglycemia in type 2 diabetes: a consensus algorithm for the initiation and adjustment of therapy: a consensus statement of the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care 2009;32(1):193203. 26. Gregg EW, Gu Q, Cheng YJ, Narayan KM, Cowie CC. Mortality trends in men and women with diabetes, 1971 to 2000. Ann Intern Med 2007;147(3):14955. 27. Alexander GC, Sehgal NL, Moloney RM, Stafford RS. National trends in treatment of type 2 diabetes mellitus, 1994-2007. Arch Intern Med 2008;168(19):208894. 28. Red book: pharmacys fundamental reference, 2010 edition. New York: Thomson Reuters, 2010. 29. National Center for Health Statistics. National Health Interview Survey diabetes supplement. Hyattsville MD: NCHS, 2006. 30. Nichols GA, Vupputuri S, Lau H. Medical care costs associated with progression of diabetic nephropathy. Diabetes Care 2011;34(11):23748. 31. NIH, National Institute of Diabetes and Digestive and Kidney Diseases. USRDS 2012 annual data report: atlas of chronic kidney disease and end-stage renal disease in the U.S. Bethesda MD: US Renal Data System, 2012. 32. O'Brien JA, Patrick AR, Caro JJ. Cost of managing complications resulting from type 2 diabetes mellitus in Canada. BMC Health Serv Res 2003;3(1):7. 33. Rein DB, Zhang P, Wirth KE, et al. The economic burden of major adult visual disorders in the U.S. Arch Ophthalmol 2006;124(12): 175460.

261

34. Taylor TN, Davis PH, Torner JC, Holmes J, Meyer JW, Jacobson MF. Lifetime cost of stroke in the U.S. Stroke 1996;27(9):145966. 35. Hogan C, Lunney J, Gabel J, Lynn J. Medicare beneciaries costs of care in the last year of life. Health Aff (Millwood) 2001;20(4): 18895. 36. Brandle M, Zhou H, Smith BR, et al. The direct medical cost of type 2 diabetes. Diabetes Care 2003;26(8):23004. 37. Caro JJ, Ward AJ, O'Brien JA. Lifetime costs of complications resulting from type 2 diabetes in the U.S. Diabetes Care 2002;25(3):47681. 38. Birnbaum H, Leong S, Kabra A. Lifetime medical costs for women: cardiovascular disease, diabetes, and stress urinary incontinence. Womens Health Issues 2003;13(6):20413. 39. Goldman DP, Zheng Y, Girosi F, et al. The benets of risk factor prevention in Americans aged 51 years and older. Am J Public Health 2009;99(11):2096101. 40. CDC. National Diabetes Prevention Program. 2011 11/04/2011 [cited 2012 06/01/2012]. www.cdc.gov/diabetes/prevention/index.htm. 41. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002;346(6):393403. 42. Pan XR, Li GW, Hu YH, et al. Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance. The Da Qing IGT and Diabetes Study. Diabetes Care 1997;20(4):53744. 43. Tuomilehto J, Lindstrom J, Eriksson JG, 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):134350. 44. Ali MK, Echouffo-Tcheugui J, Williamson DF. How effective were lifestyle interventions in real-world settings that were modeled on the Diabetes Prevention Program? Health Aff (Millwood) 2012;31(1): 6775. 45. Li R, Zhang P, Barker LE, Chowdhury FM, Zhang X. Cost-effectiveness of interventions to prevent and control diabetes mellitus: a systematic review. Diabetes Care 2010;33(8):187294. 46. De Geest S, Sabate E. Adherence to long-term therapies: evidence for action. Eur J Cardiovasc Nurs 2003;2(4):323. 47. Beckles GL, Engelgau MM, Narayan KM, Herman WH, Aubert RE, Williamson DF. Population-based assessment of the level of care among adults with diabetes in the U.S. Diabetes Care 1998;21(9): 14328. 48. CDC. National diabetes fact sheet. 2011. www.cdc.gov/diabetes/pubs/ pdf/ndfs_2011.pdf. 49. Koopman RJ, Mainous AG 3rd, Diaz VA, Geesey ME. Changes in age at diagnosis of type 2 diabetes mellitus in the United States, 1988 to 2000. Ann Fam Med 2005;3(1):603. 50. American Diabetes Association. Implications of the United Kingdom Prospective Diabetes Study. Diabetes Care 1998;21(12):21804. 51. American Diabetes Association. Standards of medical care in diabetes 2010. Diabetes Care 2010;33(S1):S1161.

September 2013