Decreasing incidence of type 2 diabetes mellitus in the United States, 2007–2012: Epidemiologic findings from a large US claims database

doi:10.1016/j.diabres.2016.04.043

Diabetes Research and Clinical Practice

Volume 117, July 2016, Pages 111-118

https://doi.org/10.1016/j.diabres.2016.04.043 Get rights and content

Abstract

Aims

To explore epidemiological trends in type 2 diabetes mellitus (T2D) in the US between 2007 and 2012 using a large US claims database, with a particular focus on demographics, prevalence, newly-diagnosed cases, and comorbidities.

Methods

Truven Health MarketScan® Databases were used to identify patients with claims evidence of T2D in the years 2007 and 2012. Newly-diagnosed T2D was characterized by an absence of any T2D claims or related drug claims for 6 months preceding the index claim. Demographic and comorbidity characteristics of the prevalent and new-onset T2D groups were compared and analyzed descriptively for trends over time.

Results

The overall prevalence of T2D remained stable from 2007 (1.24 million cases/15.07 million enrolled; 8.2%) to 2012 (2.04 million cases/24.52 million enrolled; 8.3%), while the percentage of newly-diagnosed cases fell dramatically from 2007 (152,252 cases; 1.1%) to 2012 (147,011 cases; 0.65%). The mean age of patients with prevalent T2D was similar in 2007 (60.6 y) and 2012 (60.0 y), while the mean age of newly-diagnosed T2D patients decreased by 3 years from 2007 (57.7 y) to 2012 (54.8 y). Hypertension and hyperlipidemia were the most common comorbidities, evident in 50–75% of T2D patients, and increased markedly from 2007 to 2012 in both prevalent and new-onset T2D populations. Cardiovascular disease decreased slightly in prevalent (−0.9%) and new-onset (−2.8%) cases.

Conclusions

This large US health claims database analysis suggests stabilization in prevalence and declining incidence of T2D over a recent 5-year period, a downward shift in age at T2D diagnosis, but increases in several comorbidities.

Introduction

Diabetes continues to pose a substantial global healthcare challenge with marked medical and socioeconomic ramifications. Recent estimates from the International Diabetes Federation suggest that 415 million adults worldwide were living with diabetes in 2015 (1 in 11 adults), and that as many as 642 million people (1 in 10 adults) may be affected by the year 2040 [1]. In the US alone, 29.3 million adults had a diagnosis of diabetes in 2015 [1]. Worldwide, most countries spend between 5% and 20% of all healthcare dollars on diabetes [1]. In the US, the estimated national cost of diabetes in 2015 was $320 billion [1]. Data from 2012 indicated that costs associated with diabetes accrue from a combination of direct health expenditures (72%) and indirect costs from productivity losses, job absenteeism, unemployment, and premature mortality (28%) [2].

Given the large number of people affected by diabetes, the majority of whom have type 2 diabetes mellitus (T2D), trends in prevalence and incidence are of interest for gauging the impact of disease prevention efforts and forecasting future healthcare resource needs. The purpose of this study was to explore epidemiological trends in T2D in the US between 2007 and 2012. This was accomplished using a large US claims database from which objective data pertaining to demographics, medical services for T2D (according to ICD-9 codes), and prescriptions for T2D drugs were analyzed for 1.2 million and 2.0 million T2D patients in 2007 and 2012, respectively. The objectives of the study were to ascertain changes between the two study years in T2D incidence and prevalence, demographic characteristics of T2D patients, and patterns of comorbidities related to T2D. Trends derived from the analyses of these real-world data could be useful for allocating healthcare resources.

Section snippets

Subjects

The Truven Health MarketScan® Databases were used for cross-sectional analyses. These databases include claims from persons residing in all states of the US, include only de-identified data, and are fully compliant with the Health Insurance Portability and Accountability Act of 1996. The Commercial Claims and Encounters (CCAE) database is a collection of inpatient, outpatient, and pharmaceutical claims from over 200 US insurance carriers and large, self-insuring companies. The Medicare

Variables

The variables examined included age, sex, and geographical location, as well as type of insurance (commercial vs Medicare). Geographic regions of the US were defined as Central, Northeast, South, and West. Comorbidities were identified using ICD-9 codes in claims during the analysis years (2007 and 2012). The comorbidities examined included those known to be commonly associated with diabetes and included hypertension, hyperlipidemia, cardiovascular diseases, cerebrovascular diseases, peripheral

Prevalence population

The total numbers of enrolled persons in the databases were 15.07 million in 2007 and 24.52 million in 2012 (Table 1). Of these, 1.2 million were diagnosed with T2D in 2007, and 2.04 million were diagnosed in 2012, for a T2D prevalence of 8.2% in 2007 and 8.3% in 2012. Slightly more men than women had T2D, with men accounting for 52.5% of the diabetes population in 2007 and 52.1% in 2012.

In both years, T2D prevalence was markedly greatest among patients 60 years and over compared with younger

Discussion

This analysis of a very large claims database analysis, including more than 15 million patients in 2007 and more than 24 million patients in 2012, highlights a number of important results and trends in the epidemiology of T2D and its associated comorbidities in the US.

The mean age of patients with prevalent T2D was similar in 2007 and 2012. However, the mean age of newly-diagnosed T2D patients decreased by 3 years from 2007 to 2012, while the mean ages for the overall database populations in the

Conflicts of interest

WW, EK, TH, SK, BS and JB are employees of, and stock/shareholders in, Novo Nordisk Inc. YL was a contract employee of Novo Nordisk, Inc. at the time the study was performed (currently employed at Truven Health Analytics, Cambridge, MA, USA).

Acknowledgements

This study was funded by Novo Nordisk Inc. (Plainsboro, NJ); editorial assistance for this paper was provided by Churchill Communications (Maplewood, NJ) with funding from Novo Nordisk Inc.

References (16)

L.L. Lipscombe et al.
Trends in diabetes prevalence, incidence, and mortality in Ontario, Canada 1995–2005: a population-based study
Lancet
(2007)
M. Virally et al.
Type 2 diabetes mellitus: epidemiology, pathophysiology, unmet needs and therapeutical perspectives
Diabetes Metab
(2007)
E. Eschwege et al.
Type 2 diabetes mellitus in France in 2012: results from the ObEpi survey
Diabetes Metab
(2015)
International Diabetes Federation. IDF Diabetes Atlas 7th ed., Brussels, Belgium; International Diabetes Federation,...
American Diabetes Association
Economic costs of diabetes in the US in 2012
Diabetes Care
(2013)
Centers for Disease Control and Prevention. National Diabetes Statistic Report, 2014. Available at:...
A. Menke et al.
Prevalence of and trends in diabetes among adults in the United States, 1988–2012
JAMA
(2015)
American Diabetes Association. Statistics About Diabetes: Data from the National Diabetes Statistics Report, 2014...

There are more references available in the full text version of this article.

Cited by (31)

Evolution of the burden of diabetes among adults and children in Québec, Canada, from 2001 to 2019: A population-based longitudinal surveillance study
2024, Diabetes Epidemiology and Management
Many developed countries, including Canada, have observed reductions in incidence of diabetes. Given the latest improvements in the case definition of diabetes for the younger population in Quebec, Canada, we sought to examine the evolution of diabetes among adults and children in Quebec, between 2001 and 2019.
Crude and age-standardized incidence and prevalence of diabetes among individuals ≥1 year were calculated using data from the Quebec Integrated Chronic Disease Surveillance System (n≈8,351,500 in 2019), using two case definitions for adults and the youth respectively. Age-standardized all-cause hospitalizations and mortality proportions were calculated among the population ≥20 years.
Between 2001 and 2019, age-standardized incidence decreased by 30%, with a crude incidence of 4.6 per 1,000 in 2019. Incidence rates decreased from age group ≥50 years but increased by 25% for the group of 1-19 years. Age-standardized prevalence increased by 42% (crude prevalence in 2019: 8.1%). Males had higher incidence and prevalence of diabetes, with an incremental gap between sexes increasing with age. All-cause hospitalization and mortality proportions among individuals with diabetes declined by 21% and 29% respectively between 2001 and 2019. Age-standardized hospitalizations and mortality ratios for individuals with/without diabetes remained stable and were 2.7 (99% Confidence Intervals [CI]: 2.7-2.8) and 2.2 (99% CI: 2.1-2.3) in 2019, respectively.
Despite the reduction of incidence among adults, diabetes incidence increased among the youth and remained high among adults, especially for males. These results highlight the importance of improving earlier preventive care and initiatives for reducing the diabetes burden in Quebec.
Discovery of lipid profiles of type 2 diabetes associated with hyperlipidemia using untargeted UPLC Q-TOF/MS-based lipidomics approach
2021, Clinica Chimica Acta
The incidence of type 2 diabetes (T2D) is rising rapidly and has become an important public health problem. According to reports, people with T2D often have hyperlipidemia. Hence, in the current study, a plasma non-targeted lipidomics method was used to study the differences in lipid profile between 36 T2D-associated hyperlipidemia patients and 43 healthy controls by ultra-performance liquid chromatography coupled with quadrupole time-of-flight high-definition mass spectrometry (UPLC Q-TOF/MS). Furthermore, we studied the differences in lipid profile between 36 T2D-associated hyperlipidemia patients and 41 T2D patients. Principal component analysis (PCA), orthogonal partial least squares-discriminant analysis (OPLS-DA), S-plot and heatmap were used to analyze the lipid changes between the groups. Compared with the healthy control group, 37 lipids were significantly altered in the T2D-associated hyperlipidemia group, and when compared with the T2D group, 22 lipids were significantly altered in the T2D-associated hyperlipidemia group. Of all the detected lipids categories which included sphingolipids, glycerolipids, glycerophospholipids, prenol lipids and saccharolipids, glycerophospholipids accounted for the largest proportion in the two groups. Also, this study found that glycerophospholipid metabolism pathway was the most relevant pathway for these lipid metabolisms. The identified lipids may enhance the disease prediction and provide a new tool to monitor the progression of T2D-associated hyperlipidemia.
Incidence of diabetes mellitus in Italy in year 2018. A nationwide population-based study of the ARNO Diabetes Observatory
2021, Nutrition, Metabolism and Cardiovascular Diseases
To assess incidence of diabetes in Italy in 2018 by the use of administrative claims from several million residents. Differences in rates in men and women across decades of age were investigated. Incident rates of insulin or noninsulin treated subjects were also examined.
We analyzed administrative healthcare claims of 11,300,750 subjects monitored by the ARNO Diabetes Observatory. Incident cases of diabetes were identified by glucose lowering drug prescriptions, disease-specific co-payment exemptions and hospital discharge codes related to diabetes occurring in 2018 but not in 2017.
We identified 697,208 subjects with ascertained diabetes. Incident cases were 65,932, with a rate of 5.83 per 1000 person-years (p-y). Incidence of drug-treated diabetes (n = 60,271) was 5.33 per 1000 p-y. Subjects receiving only insulin prescriptions were 5652 (rate 0.50 per 1000 p-y) and those receiving only prescriptions of noninsulin medications were 51,085 (rate 4.52 per 1000 p-y). Incidence rates progressively increased across decades until age 80 and then dropped by 25–30%. Overall, incident rates were generally higher in women aged 11–40 and in men aged ≥51.
Recent cases represented ~10% of the population of diabetic subjects. Incidence of noninsulin-treated diabetes was almost 10-fold higher than incidence of insulin-treated diabetes. Substantial differences in incidence rates were observed in men and women of several decades of age: women more affected in adolescence and young adult age, men more affected in mature and advanced age. These data provide further understanding on the epidemiological burden of the disease in Italy.
The Lancet Commission on diabetes: using data to transform diabetes care and patient lives
2020, The Lancet
Risk factor control among Black and White adults with diabetes onset in older adulthood: The Reasons for Geographic and Racial Differences in Stroke (REGARDS) study
2020, Preventive Medicine
The objective of this study was to determine whether attainment of clinical and lifestyle targets varied by race and sex among adults with diabetes onset in older adulthood.
This study included 1420 black and white adults from the REGARDS study without diabetes at baseline (2003–07) but with diabetes onset at the follow-up exam (2013–16). Attainment of clinical targets (A1c <8%; blood pressure < 140/90 mmHg; and statin use) and lifestyle targets (not smoking; physical activity≥ 4 times/week; and moderate/no alcohol use) was assessed at the follow-up exam. Modified Poisson regression was used to obtain prevalence ratios (PR) for meeting clinical and lifestyle targets stratified by race and sex, separately.
The mean age was 71.5 years, 53.6% were female, and 46.1% were black. The majority were aware of their diabetes status (85.7%) and used oral or injectable hypoglycemic medications (64.8%). Overall, 39.4% met all 3 clinical targets and 18.8% met all 3 lifestyle targets. Meeting A1c and blood pressure targets were similar by race and sex. Statin use was more prevalent for men than women among white adults (PR = 1.13; 95% CI = 0.99–1.29) and black adults (PR = 1.23; 95% CI = 1.06–1.43). For lifestyle factors, the non-smoking prevalence was similar by race and sex, while white men were more likely than white women to be physically active.
Although the attainment of each clinical and lifestyle target separately was generally high among adults with diabetes onset in older adulthood, race and sex differences were apparent. Comprehensive management of clinical and lifestyle factors in people with diabetes remains suboptimal.
Segmentation of retinal layers from OCT scans
2020, Diabetes and Retinopathy: Volume 2: Computer-Assisted Diagnosis
The accurate segmentation of retinal layers of the eye in a 3D optical coherence tomography (OCT) data provides relevant clinical information. This chapter introduces a three-dimensional (3D) segmentation approach that uses an adaptive patient-specific retinal atlas as well as an appearance model for 3D OCT data. To reconstruct that atlas of 3D retinal scan, we first segment the central area of the macula at which we can clearly identify the fovea. Markov Gibbs random field (MGRF) including intensity, shape, and spatial information of 12 layers of retina was all used to segment the selected area of retinal fovea. A set of coregistered OCT scans that were gathered from 200 different individuals were used to build a 2D shape prior. This shape prior was adapted in a following step to the first-order appearance and second-order spatial interaction MGRF model. After segmenting the center of the macula “foveal area,” the labels and appearances of the layers that have been segmented were used to have the adjacent slices segmented as well. The last step was then repeated recursively until the 3D OCT scan of the patient is segmented. This approach was tested on 30 individuals while some of them were normal and others were pathological, and then compared to a manually segmented ground truth and finally these results were verified by medical retina experts. Metrics such as dice similarity coefficient, agreement coefficient, and average deviation metrics were used to measure the performance of the proposed approach. Accomplished accuracy by the proposed approach shows promising results with noticeable advantages over the state-of-the-art 3D OCT approach.

View all citing articles on Scopus

View full text