United States Veterans are at excess risk for type 2 diabetes, but population differentials in risk have not been characterized. We determined risk of type 2 diabetes in relation to prediabetes and dyslipidemic profiles in Veterans at the VA New York Harbor (VA NYHHS) during 2004–2014. Prediabetes was based on American Diabetes Association hemoglobin A1c (HbA1c) testing cut-points, one of several possible criteria used to define prediabetes. We evaluated transition to type 2 diabetes in 4,297 normoglycemic Veterans and 7,060 Veterans with prediabetes. Cox proportional hazards regression was used to relate HbA1c levels, lipid profiles, demographic, anthropometric and comorbid cardiovascular factors to incident diabetes (Hazard Ratio [HR] and 95% confidence intervals). Compared to normoglycemic Veterans (HbA1c: 5.0–5.6%; 31–38 mmol/mol), risks for diabetes were >2-fold in the moderate prediabetes risk group (HbA1c: 5.7–5.9%; 39–41 mmol/mol) (HR 2.37 [1.98–2.85]) and >5-fold in the high risk prediabetes group (HbA1c: 6.0–6.4%; 42–46 mmol/mol) (HR 5.59 [4.75–6.58]). Risks for diabetes were increased with elevated VLDL (≥40mg/dl; HR 1.31 [1.09–1.58]) and TG/HDL (≥1.5mg/dl; HR 1.34 [1.12–1.59]), and decreased with elevated HDL (≥35mg/dl; HR 0.80 [0.67–0.96]). Transition to diabetes in Veterans was related in age-stratified risk score analyses to HbA1c, VLDL, HDL and TG/HDL, BMI, hypertension and race, with 5-year risk differentials of 62% for the lowest (5-year risk, 13.5%) vs. the highest quartile (5-year risk, 21.9%) of the risk score. This investigation identified substantial differentials in risk of diabetes in Veterans, based on a readily-derived risk score suitable for risk stratification for type 2 diabetes prevention.
Citation: Davis PJ, Liu M, Sherman S, Natarajan S, Alemi F, Jensen A, et al. (2018) HbA1c, lipid profiles and risk of incident type 2 Diabetes in United States Veterans. PLoS ONE 13(9): e0203484. https://doi.org/10.1371/journal.pone.0203484
Editor: Alberto G. Passi, University of Insubria, ITALY
Received: April 27, 2018; Accepted: August 21, 2018; Published: September 13, 2018
Copyright: © 2018 Davis et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data for this study were obtained via access to the national Corporate Data Warehouse (CDW) of the Department of Veterans Affairs (VA) Office of Information and Technology (Washington D.C.). VA patient medical and administrative data are accessible from the VINCI Workspace (Veterans Affairs Informatics and Computing Infrastructure), a VA-hosted computing environment, through a secure gateway (firewall) using a Remote Desktop Connection (RDC). VINCI provides a centralized, secure development and research platform for conducting VA studies and allows for data storage, extraction, processing and analysis. Access to linked patient level data was obtained through the Veterans Health Administration (VHA), National Data Services (NDS) and the VA Information Resource Center (VIReC), one of three Health Services Research and Development Service (HSR&D) resource centers that authorize and maintain research access to patient data and resources for research conducted within the VA.
Funding: The authors received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Type 2 diabetes is a chronic metabolic disorder defined by persistent hyperglycemia due to increased insulin resistance and/or impaired insulin secretion [1–2]. In the United States, more than 30 million people (9.4% of the population) had type 2 diabetes in 2015 . With aging and increasing rates of obesity, prevalence of type 2 diabetes is also rising and is currently the seventh leading cause of death . Furthermore, type 2 diabetes leads to greater risk for cardiovascular disease and other conditions, such as blindness, kidney failure, and lower limb amputations, and is associated with greatly reduced overall life expectancy [1–4]. Nearly a quarter of U.S. Veterans have type 2 diabetes, more than double the prevalence of the general population [3,5–6].
Prediabetes describes individuals at increased risk for future development of type 2 diabetes [5,7]. Those with prediabetes have higher than normal blood glucose concentrations, but lower than diabetic diagnostic criteria [1,5,7–8]. Since the early 2000s, the American Diabetes Association has recommended assessment of glycated hemoglobin (HbA1c), a type of hemoglobin that indicates an individual’s three-month average plasma glucose concentration , as an alternative to glucose tolerance testing for diagnostic surveillance of diabetes and prediabetes. Despite the higher prevalence of type 2 diabetes among Veterans, little is known about the prevalence of prediabetes among Veterans and its impact on their development of type 2 diabetes .
Epidemiologic studies have shown that dyslipidemic profiles are independent risk factors for type 2 diabetes and can manifest prior to development of the disease [11–12]. Furthermore, dyslipidemia can cause decreased pancreatic beta cell function and survival particularly among diabetic patients [11,13–14]. Additional research is needed to examine the effects of lipid patterns in prediabetic and nondiabetic individuals on risk for incident type 2 diabetes.
The relationship of HbA1c, dyslipidemias and other risk factors to type 2 diabetes occurrence in Veterans is incompletely understood. Therefore, we investigated the association of HbA1c levels and associated lipid profiles with development of incident type 2 diabetes among Veterans at the Veterans Affairs New York Harbor Healthcare System (VA NYHHS).
Materials and methods
Data for this study were obtained via access to the national Corporate Data Warehouse (CDW) of the Department of Veterans Affairs (VA) Office of Information and Technology (Washington D.C.). VA patient medical and administrative data are accessible from the VINCI Workspace (Veterans Affairs Informatics and Computing Infrastructure), a VA-hosted computing environment, through a secure gateway (firewall) using a Remote Desktop Connection (RDC) [15–16]. VINCI provides a centralized, secure development and research platform for conducting VA studies and allows for data storage, extraction, processing and analysis . Access to linked patient level data was obtained through the Veterans Health Administration (VHA), National Data Services (NDS) and the VA Information Resource Center (VIReC), one of three Health Services Research and Development Service (HSR&D) resource centers that authorize and maintain research access to patient data and resources for research conducted within the VA [15–16].
Data extracted for use in this study in the VINCI workspace included information on patient demographics, primary care visits, laboratory data, pharmacy claims data, medical diagnostic and procedures data as ICD-9 codes, vital status and mortality records. Demographic information was based on self-report by patients and/or VA employee recorded at baseline and grouped based on existing standardized categories used in VA medical records [10,15–17]. Covariates included race (white, or other), ethnicity (not Hispanic or Latino/ other [declined/unknown], Hispanic or Latino), marital status (never married, married, separated, divorced, widowed), smoking status (never or ever smoker), body mass index ([BMI] calculated as the average weight (kg) during the baseline year or year prior to entry, divided by height (m2) and categorized [<18.5, 18.5–24.9, 25–29.9, ≥30]). Medical record information was also extracted for history of cardiovascular disease at baseline, defined as having ischemic heart disease (IHD; ICD-9: 410–414), cerebrovascular accident (CVA; ICD-9: 430–434), congestive heart failure (CHF; ICD-9: 428), or peripheral vascular disease (PVD; ICD-9: 443), and for history of hypertension (baseline BP ≥140/90; HTN; ICD-9:401). Systolic blood pressure (mmHg) and diastolic blood pressure (mmHg) were included as the average measurement taken during the baseline year or year prior. Data for lipid profiles were average annual fasting measurements taken during the baseline year or the year prior. This included low-density lipoprotein-cholesterol (LDL-C) (≥130, <130 mg/dl), high-density lipoprotein-cholesterol (HDL-C) (≥35, <35 mg/dl), total cholesterol (TC) (≥200, <200 mg/dl), triglyceride (TG) (≥150, <150 mg/dl), very low-density lipoprotein–cholesterol (VLDL) (≥40, <40 mg/dl) calculated as triglycerides/5 according to the Friedewald equation , TG/HDL ratio (≤1.5, >1.5 mg/dl), TC/HDL ratio (≥5.0, <5.0 mg/dl), and LDL/HDL ratio (≥4.0, <4.0 mg/dl).
Prediabetes was defined as hemoglobin A1c (HbA1c) of 5.7–6.4% (39–46 mmol/mol), based upon American Diabetes Association criteria, one of several possible definitions for prediabetes . We further defined HbA1c of 5.0–5.6% (31–38 mmol/mol) as normoglycemic and divided prediabetes as prediabetic “moderate” risk (HbA1c: 5.7–5.9% or 39–41 mmol/mol) and prediabetic “high” risk groups (HbA1c: 6.0–6.4% or 42–46 mmol/mol), based on other research to assess predictiveness for transition from prediabetes to diabetes [1–2,5,7,19]. A diagnosis of type 2 diabetes was defined as having ≥2 ICD-9 diagnosis codes (250, 357.2, 362.0, 366.4) from clinical encounters and/or ≥2 prescriptions of DM medications other than metformin, based on the VA Primary Care Almanac approach [17,20]. This method of type 2 diabetes status ascertainment has been previously validated at the VA .
Among Veterans enrolled in primary care from January 1, 2004 to December 31, 2014, at the Veterans Affairs New York Harbor Healthcare System (VA NYHHS), we identified 14,361 Veterans from the CDW, ≥18 years of age, who received at least 2 HbA1c tests and had no prior history of diabetes. After exclusion of 1,549 subjects (10.8%) with missing data, there remained 12,812 subjects for study.
Individuals were classified at baseline according to the average of their first two HbA1c results: low glycemia <5.0% (<31 mmol/mol), normoglycemia 5.0–5.6% (31–38 mmol/mol), prediabetes moderate risk 5.7–5.9% (39–41 mmol/mol), prediabetes high risk 6.0–6.4% (42–46 mmol/mol) or type 2 diabetes risk range ≥6.5% (≥48 mmol/mol). Frequencies and proportions within these groups were calculated and the χ2 test and one-way ANOVA test was used to compare proportions for categorical and continuous variables across all groups. Crude type 2 diabetes incidence rates (IR) were calculated based on the number of type 2 diabetes cases and respective person-time at risk for study participants. Kaplan-Meier method and survival curves were used to calculate survival probability over time, with respect to type 2 diabetes incidence stratified by HbA1c category .
Cox proportional hazards models were used to assess the risk of developing incident type 2 diabetes, after adjustment for covariates. For assessment of type 2 diabetes risk, subjects were categorized by HbA1c level as <5.0% (<31 mmol/mol), 5.7–5.9% (39–41 mmol/mol), and 6.0–6.4% (42–46 mmol/mol), and were compared to those categorized as 5.0–5.6% (31–38 mmol/mol), as the reference category (Subjects in the diabetic risk range for HbA1c ≥6.5% (≥48 mmol/mol) at study entry were excluded from the analyses of type 2 diabetes incidence). Time to incident type 2 diabetes was based on the interval in days between the second HbA1c test (date of entry), subsequent to January 1, 2004, and date of the occurrence of type 2 diabetes. Subject follow-up was censored at the date of death or the end of the follow-up period, December 31, 2014, whichever came first. Estimates of HRs for all models were based on Sandwich variance estimates and ties were handled using the Efron method . All models controlled for sex, race, ethnicity, marital status, BMI, smoking status, HTN, CHF, CVA, IHD, PVD, LDL, HDL, TC, TG, VLDL, TG/HDL, TC/HDL, and LDL/HDL. The proportional hazards assumption was verified by graphs of the Kaplan-Meier estimates of the survival function, log-(-)log plots, including a product term of the variables with log(time) in the Cox model and by using the proportionality test statement to assess all variables. All models were stratified by age (18–54, 55–64, ≥65) .
To develop estimates of the 5-year risk for incident type 2 diabetes, we constructed a 5-year risk equation [22–23]. We carried out backward elimination in the dataset with all predictors included in the stratified Cox regression analysis (p<0.10 for stay) to identify the subset of variables most strongly predictive of risk . Use of forward elimination and stepwise selection methods yielded similar results. Variables selected in the final model included, HbA1c (%), body mass index (BMI), hypertension (HTN), race, VLDL, HDL and TG/HDL ratio denoted by X1-X7. HbA1c and BMI were treated as continuous variables based on the average of two HbA1c test results and BMI values at baseline. The overall C index is used as an extension of the area under the receiver operating characteristic curve (aROC) and therefore was calculated for overall ability of the risk equation to discriminate between those who develop type 2 diabetes from those who did not .
The risk equation developed from these findings includes X1-X7 as the baseline predictors with estimated coefficients expressed as β1-β7 for modeling the 5-year risk equation [22–23]. To account for potential violations of the proportional hazards assumptions due to age, age-stratified Cox proportional hazards models were used to derive the estimates of hazard ratios and baseline hazard functions. The mean risk score and baseline survival S0(5) was assessed at 5-years for each age stratified category when all factors were equal to their mean.
All data analyses were conducted using SAS 9.1 (SAS Institute, Cary, NC), SAS Enterprise Guide 6.1 (Cary, NC), and SQL Server Management Studio (Microsoft, 2014). Significance level was set at α = 0.05.
HbA1c and type 2 diabetes risk
The study population of 12,812 adult Veterans not known to have type 2 diabetes was predominantly white (54.5%), male (95.8%), not Hispanic or Latino (84.9%), and greater than 55 years of age (66.6%). Many of these individuals were overweight (BMI: 25–29.9) (39.6%) or obese (BMI: ≥30) (38.8%) and had hypertension (77.4%) where those with a HbA1c ≥5.7% had higher percentages of these conditions (Table 1). Based on HbA1c, 33.5% of these subjects were normoglycemic, 28% were in the moderate prediabetic risk range, 27.1% were in the high prediabetic risk range, 5.9% had low glycemia and 5.4% were in the diabetic risk range.
Excluding 695 subjects with HbA1c in the diabetic range at study entry, we identified 1,270 incident cases of type 2 diabetes in 12,117 nondiabetic study participants followed on average for 3.6 years (Table 2). Type 2 diabetes incidence rates increased across the range of HbA1c, from 1.08 (per 100 person-years [0.76–1.53]) for those with low glycemia to 6.41 (per 100 person-years [5.96–6.89]) among those in the high prediabetic range. Both the prediabetic moderate risk (HR 2.37 [1.98–2.85]) and the prediabetic high risk groups (HR 5.59 [4.75–6.58]) showed significantly increased type 2 diabetes risk compared to the normoglycemic reference group, after adjustment for other factors (Table 2, Model 2). Type 2 diabetes risks were nearly the same for those in the low glycemia and normoglycemic categories of HbA1c.
Dyslipidemias and type 2 diabetes risk
Incident type 2 diabetes was also assessed by lipid categories. A majority of study subjects had levels of LDL <130 mg/dl (82.4%), HDL ≥35 mg/dl (90.7%), TC<200 mg/dl (74.8%), TG<150 mg/dl (73.1%), VLDL<40 mg/dl (87.8%), TG/HDL >1.5 mg/dl (74.7%), TC/HDL<5.0 mg/dl (88.6%) and LDL/HDL <4.0 mg/dl (97.7%). Cox proportional hazards models showed significantly increased risks for type 2 diabetes with respect to VLDL ≥40 mg/dl (HR 1.31 [1.09–1.58]), TG/HDL ≥1.5 mg/dl (HR 1.34 [1.12–1.59]), while a decreased risk was observed for HDL ≥35 mg/dl (HR 0.80 [0.67–0.96]) after adjustment for all other factors. Increased risks were observed for other lipid factors, but were not statistically significant (Table 3).
To investigate possible effects on study results from the implementation of new clinical practice guidelines for HbA1c use at the VA, we compared analyses of data from 2011–2014 (after implementation of the new HbA1c practice guidelines) to results from 2004–2008 (pre-implementation). We also assessed impact of more restrictive eligibility criteria of having had 5 or more HbA1c tests vs. our criteria of 2 or more tests. Associations observed overall, were similar to those found in subset analyses (S1 Appendix).
Type 2 diabetes risk prediction
Based on the backward elimination method from a fully adjusted model (Table 2, Model 2), we developed an age-stratified 5-year risk prediction model for transition to type 2 diabetes (Table 4). For example, a 0.1-unit increase in HbA1c is associated with a 1.33 increase in the HR and a one unit increase in BMI is associated with a more modest 1.03 increase in the HR, each after adjusting for other factors. The C statistic calculated for overall discrimination for the model was good at 0.70 (95% CI: 0.65–0.74) . Based on an age-stratified risk prediction model including HbA1c, VLDL, HDL and TG/HDL, BMI, hypertension and race (S2 Appendix), there is an 18.6% average potential 5-yr risk for incident type 2 diabetes among prediabetic Veterans. For individuals in the population, there is a 5-year risk differentials of 62% for the lowest (5-year risk, 13.5%) vs. the highest quartile (5-year risk, 21.9%) of the risk score. Global models including all interaction terms demonstrated similar model fit (AIC = 22749.2, SBC = 22983.579) compared to models without interaction terms (AIC = 22746.05, SBC = 22909.57), suggesting that the more conservative model without interaction terms is suitable.
Our research in a large cohort of Veterans in New York City shows that prediabetic Veterans classified as having a moderate risk for type 2 diabetes have a more than 2-fold increase in risk of incident type 2 diabetes, and that Veterans classified at high risk have a more than 5-fold increase in risk for this disease. Our study also identified associations of selected cardiovascular risk factors with modest changes in type 2 diabetes risk. Furthermore, our research provides a useful prediabetes specific risk equation for the estimation of 5-year risk of incident type 2 diabetes in this cohort, demonstrating stratification of individuals over a broad risk range, based on selected demographic and clinical factors. The HbA1c test used in conjunction with lipid profiles and other factors in the proposed risk equation, serves as a simple tool to predict risk of incident type 2 diabetes in this population.
Our research is the first systematic evaluation in Veterans of type 2 diabetes risk associated with HbA1c, as the introduction of screening with HbA1c for prediabetes and type 2 diabetes at the VA is relatively recent . Our study showed the importance of examining moderate versus high prediabetes risk groups and highlights the high prevalence of prediabetes at the VA where more than half of all Veterans tested have this condition. Efforts to mitigate transition of this high risk state to incident type 2 diabetes are of critical importance within the older and sicker Veteran population.
Dyslipidemia is a major risk factor for cardiovascular disease, particularly among diabetics, which remains the number one cause of death in the United States [25–26]. The characteristic features of dyslipidemia, such as increased triglycerides and LDL levels combined with low HDL, are metabolically interrelated and begin occurring several years before the onset of type 2 diabetes and, as we and others show, may serve as independent risk factors for the condition [11–12,27]. Additionally, increased production of VLDL plasma levels by hepatocytes significantly contribute and are thought to initiate a cascade of progressively abnormal levels of atherogenic lipids .
Hypercholesterolemia also contributes to decreased insulin production and pancreatic beta cell dysfunction [28–29]. Increasing cholesterol and LDL levels related to beta cell functional decline were found to be higher in prediabetic and diabetic individuals [30–31]. Cholesterol homeostasis is therefore an important factor in delaying or preventing insulin secretory defects [13–14,28,31]. A recent study using mouse models has also demonstrated the pathogenesis of beta cell dysfunction and insulin resistance by presence of lipoproteins . Increased triglyceride levels were shown to alter insulin secretion along inflammatory free fatty acid metabolic pathways and influence pancreatic beta cell function and survival [14,31–32].
High-density lipoproteins, on the other hand, stimulate insulin secretion, inhibit beta cell apoptosis [11,33] and may also enhance skeletal muscle glucose uptake . Conversely, increased BMI and atherogenic lipids in prediabetes and type 2 diabetes adds metabolic stress that interferes with HDL production, thereby exacerbating pancreatic cell dysfunction and insulin resistance . Decreased HDL levels may therefore worsen glycemic control and accelerate progression to type 2 diabetes in prediabetic individuals , although not all studies agree .
Our study found increased type 2 diabetes risk with increasing VLDL and TG/HDL, while a decreased risk was seen for elevated HDL. While TG/HDL ratios predict for cardiovascular disease, additional data is needed regarding prediction of type 2 diabetes risk among prediabetics and normoglycemic individuals [37–39]. TG/HDL has been related to insulin resistance  and type 2 diabetes  consistent with our findings.
Strengths of this study include a large population at the VA NYHHS, use of robust electronic medical record information and survival analysis to model relevant outcomes and a validated methodology for case ascertainment [10,15–17,20]. The move toward use of HbA1c in recent years, has gained traction due to several advantages as a diagnostic test [1–3,7]. HbA1c testing is now better standardized as a diagnostic test in the U.S. and has less intraindividual biologic variability compared to fasting plasma glucose (FPG) [1–2,7]. It also provides a better measure of overall levels of glycemia by reflecting longer-term exposure intervals (~3 months, the average half-life of a red blood cell), has the practical advantage of not requiring fasting at collection, higher repeatability and serves as a better guide to clinical management [1,3,7,9–10,17].
A limitation in HbA1c testing to distinguish normoglycemic from prediabetic subjects (or moderate from high risk prediabetics) is that the cutpoints for these groups are somewhat arbitrary, with different research groups using slight modifications of the definitions we employed [1–2,5,7,19]. In fact, prediabetes and associated HbA1c (or other measures of glycemic state, such as fasting glucose) are on a continuum from health to disease state and may also have different implications for people with varying profiles of other diabetes risk factors. Also, we used cut-points for HDL (≥35 versus <35 mg/dl), as recommended by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) and ADA for diabetes risk evaluation in overweight adults [41–43], recognizing that optimum targets of HDL levels >40 mg/dl for men and >50 mg/dl for women have also been suggested [44–45]. Although the Veterans Administration provides a rich resource for research, there are also certain limitations to use of data from this source. We did not have information on diet or physical activity of study participants, nor did we evaluate the impact of medication use, which is a topic of ongoing research. We did not consider the potential influence of medication use on diabetes risk, although this could be dealt with in detail in subsequent research on this population. Other limitations of our study include the predominately white male and older study population, and possible selection bias of patients given HbA1c testing at the VA, although our sensitivity analysis by time period and frequency of pre-entry HbA1c tests showed similar associations when more restrictive eligibility criteria were used. Additionally, intraindividual variation in fasting lipid levels may contribute to potential misclassification of individuals on dyslipidemia profiles.
This study demonstrates the importance of HbA1c screening at the VA to identify those at greater risk for incident type 2 diabetes. Results also add evidence on the importance of lipid components as independent risk factors for type 2 diabetes. Findings suggest that prevention strategies should target prediabetic Veterans with HbA1c levels ≥5.7% (42 mmol/mol) and prediabetic dyslipidemia to reduce incident type 2 diabetes among Veterans.
This work was made possible by cooperative agreement between New York University and the United States Department of Veterans Affairs. We are grateful for the assistance provided by Ashley Jensen and Dr. Sanja Avramovic in accessing VA research resources.
- 1. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2014;37(Suppl 1):S81–S90.
- 2. The International Expert Committee. International Expert Committee report on the role of the A1c assay in the diagnosis of diabetes. Diabetes Care. 2009;32(7):1327–34.
- 3. Centers for Disease Control and Prevention. National Diabetes Statistics Report, 2017. Atlanta, GA: Centers for Disease Control and Prevention, U.S. Department of Health and Human Services; 2017.
- 4. Deshpande AD, Harris-Hayes M, Schootman M. Epidemiology of diabetes and diabetes-related complications. Phys Ther. 2008;88(11):1254–64. pmid:18801858
- 5. Tabák AG, Herder C, Rathmann W, Brunner EJ, Kivimäki M. Prediabetes: a high-risk state for developing diabetes. Lancet. 2012;379(9833):2279–90. pmid:22683128
- 6. U.S. Department of Veterans Affairs. Health Services Research & Development [Internet]. Health Services Research & Development; 2014 [cited 2016 Jan 15]. Improving health and care for Veterans with diabetes. Available from: https://www.hsrd.research.va.gov/news/feature/diabetes.cfm#ref.
- 7. Grundy SM. Pre-diabetes, metabolic syndrome, and cardiovascular risk. J Am Coll Cardiol. 2012;59(7):635–43. pmid:22322078
- 8. Bansal N. Prediabetes diagnosis and treatment: A review. World J Diabetes. 2015;6(2):296–303. pmid:25789110
- 9. Goldstein DE, Little RR, Lorenz RA, Malone JI, Nathan D, Peterson CM, et al. Tests of glycemia in diabetes. Diabetes Care. 2004;27(7):1761–73. pmid:15220264
- 10. Miller DR, Safford MM, Pogach LM. Who has diabetes? Best estimates of diabetes prevalence in the Department of Veterans Affairs based on computerized patient data. Diabetes Care. 2004;27(Suppl 2):B10–B21.
- 11. von Eckardstein A, Sibler RA. Possible contributions of lipoproteins and cholesterol to the pathogenesis of diabetes mellitus type 2. Curr Opin Lipidol. 2011;22(1):26–32. pmid:21102330
- 12. von Eckardstein A, Schulte H, Assmann G. Risk for diabetes mellitus in middle-aged Caucasian male participants of the PROCAM study: Implications for the definition of impaired fasting glucose by the American Diabetes Association. Prospective Cardiovascular Münster. J Clin Endocrinol Metab. 2000;85(9):3101–8. pmid:10999793
- 13. Bardini G, Rotella CM, Giannini S. Dyslipidemia and diabetes: Reciprocal impact of impaired lipid metabolism and beta-cell dysfunction on micro- and macrovascular complications. Rev Diabet Stud. 2012;9(2–3):82–93. pmid:23403704
- 14. Fryirs M, Barter PJ, Rye KA. Cholesterol metabolism and pancreatic beta-cell function. Curr Opin Lipidol. 2009;20(3):159–64. pmid:19417651
- 15. virec.research.va.gov [Internet]. U.S. Department of Veterans Affairs. VA Information Resource Center (VIReC): The Researcher's Guide to VA Data; [cited 2016 Jun 20]. Available from: http://www.virec.research.va.gov.
- 16. hsrd.research.va.gov [Internet]. U.S. Department of Veterans Affairs. Health Services Research & Development. VA Informatics and Computing Infrastructure (VINCI); [cited 2016 Jun 20]. Available from: http://www.hsrd.research.va.gov/for_researchers/vinci/.
- 17. Management of Diabetes Mellitus Update Working Group. VA/DoD Clinical Practice Guideline for the Management of Diabetes Mellitus. Version 4.0. Washington, DC: Veterans Health Administration and Department of Defense; 2010.
- 18. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18(6):499–502. pmid:4337382
- 19. Paprott R, Rosario AS, Busch MA, Du Y, Thiele S, Scheidt-Nave C, et al. Association between hemoglobin A1c and all-cause mortality: results of the mortality follow-up of the German National Health Interview and Examination Survey 1998. Diabetes Care. 2015;38(2):249–56. pmid:25414153
- 20. Primary Care Almanac [database on the VHA internet]. Washington (DC): VHA Support Service Center, Veterans Health Administration; 2010.
- 21. Therneau TM, Grambsch PM. Modeling Survival Data: Extending the Cox Model. Statistics for Biology and Health. Springer, New York; 2000.
- 22. Cederholm J, Eeg-Olofsson K, Eliasson B, Zethelius B, Gudbjörnsdottir S; Swedish National Diabetes Register. A new model for 5-year risk of cardiovascular disease in type 1 diabetes; from the Swedish National Diabetes Register (NDR). Diabet Med. 2011;28(10):1213–20. pmid:21627687
- 23. Yang X, So WY, Kong AP, Ho CS, Lam CW, Stevens RJ, et al. Development and validation of stroke risk equation for Hong Kong Chinese patients with type 2 diabetes: the Hong Kong Diabetes Registry. Diabetes Care. 2007;30(1):65–70. pmid:17192335
- 24. Pencina MJ, D'Agostino RB. Overall C as a measure of discrimination in survival analysis: model specific population value and confidence interval estimation. Stat Med. 2004;23(13):2109–23. pmid:15211606
- 25. Nelson RH. Hyperlipidemia as a risk factor for cardiovascular disease. Prim Care. 2013;40(1):195–211. pmid:23402469
- 26. Kochanek KD, Murphy SL, Xu J, Tejada-Vera B. Deaths: Final Data for 2014. Vital Stat Rep. 2016;65(4):1–121.
- 27. Adiels M, Olofsson SO, Taskinen MR, Borén J. Overproduction of very low-density lipoproteins is the hallmark of the dyslipidemia in the metabolic syndrome. Arterioscler Thromb Vasc Biol. 2008;28(7):1225–36. pmid:18565848
- 28. Hao M, Head WS, Gunawardana SC, Hasty AH, Piston DW. Direct effect of cholesterol on insulin secretion: a novel mechanism for pancreatic beta-cell dysfunction. Diabetes. 2007;56(9):2328–38. pmid:17575085
- 29. Brunham LR, Kruit JK, Pape TD, Timmins JM, Reuwer AQ, Vasanji Z, et al. Beta-cell ABCA1 influences insulin secretion, glucose homeostasis and response to thiazolidinedione treatment. Nat Med. 2007;13(3):340–7. pmid:17322896
- 30. Zhang L, Qiao Q, Tuomilehto J, Hammar N, Janus ED, Söderberg S, et al. Blood lipid levels in relation to glucose status in seven populations of Asian origin without a prior history of diabetes: the DECODA study. Diabetes Metab Res Rev. 2009;25(6):549–57. pmid:19585489
- 31. Zheng T, Gao Y, Tian H. Relationship between blood lipid profiles and pancreatic islet β cell function in Chinese men and women with normal glucose tolerance: a cross-sectional study. BMC Public Health. 2012;12:634. pmid:22876749
- 32. Sharma RB, Alonso LC. Lipotoxicity in the pancreatic beta cell: not just survival and function, but proliferation as well? Curr Diab Rep. 2014;14(6):492. pmid:24740729
- 33. Fryirs MA, Barter PJ, Appavoo M, Tuch BE, Tabet F, Heather AK, et al. Effects of high-density lipoproteins on pancreatic beta-cell insulin secretion. Arterioscler Thromb Vasc Biol. 2010;30(8):1642–8. pmid:20466975
- 34. Drew BG, Duffy SJ, Formosa MF, Natoli AK, Henstridge DC, Penfold SA, et al. High-density lipoprotein modulates glucose metabolism in patients with type 2 diabetes mellitus. Circulation. 2009;119(15):2103–11. pmid:19349317
- 35. Siebel AL, Heywood SE, Kingwell BA. HDL and glucose metabolism: current evidence and therapeutic potential. Front Pharmacol. 2015;6:258. pmid:26582989
- 36. Haase CL, Tybjaerg-Hansen A, Nordestgaard BG, Frikke-Schmidt R. High-density lipoprotein cholesterol and risk of type 2 diabetes: a Mendelian randomization study. Diabetes. 2015;64(9):3328–33. pmid:25972569
- 37. Hadaegh F, Hatami M, Tohidi M, Sarbakhsh P, Saadat N, Azizi F. Lipid ratios and appropriate cut off values for prediction of diabetes: a cohort of Iranian men and women. Lipids Health Dis. 2010;9:85. pmid:20712907
- 38. Giannini C, Santoro N, Caprio S, Kim G, Lartaud D, Shaw M, et al. The triglyceride-to-HDL cholesterol ratio: association with insulin resistance in obese youths of different ethnic backgrounds. Diabetes Care. 2011;34(8):1869–74. pmid:21730284
- 39. Quispe R, Martin SS, Jones SR. Triglycerides to high-density lipoprotein-cholesterol ratio, glycemic control and cardiovascular risk in obese patients with type 2 diabetes. Curr Opin Endocrinol Diabetes Obes. 2016;23(2):150–6. pmid:26863278
- 40. Vega GL, Barlow CE, Grundy SM, Leonard D, DeFina LF. Triglyceride-to-high-density-lipoprotein-cholesterol ratio is an index of heart disease mortality and of incidence of type 2 diabetes mellitus in men. J Investig Med. 2014;62(2):345–9. pmid:24402298
- 41. American Diabetes Association. Classification and diagnosis of diabetes. Diabetes Care. 2015;38(Suppl 1):S8–S16.
- 42. American Diabetes Association. Standards of Medical Care in Diabetes—2016. Diabetes Care. 2016;39(Suppl 1):S1–S112.
- 43. National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) [Internet]. National Institutes of Health; [cited 2018 July 30]. Risk Factors for Diabetes. Available from: https://www.niddk.nih.gov/health-information/communication-programs/ndep/health-professionals/game-plan-preventing-type-2-diabetes/prediabetes-screening-how-why/risk-factors-diabetes.
- 44. American Diabetes Association. Executive Summary: Standards of Medical Care in Diabetes—2014. Diabetes Care. 2014;37(Suppl 1):S5–S13.
- 45. American Diabetes Association. Dyslipidemia management in adults with diabetes. Diabetes Care. 2004;27(Suppl 1):S68–S71.