Introduction

The American Diabetes Association has recommended that the levels of glycated hemoglobin (HbA1c) and low-density lipoprotein cholesterol (LDL-C) in patients with type 2 diabetes be maintained at <7% (53 mmol/mol) and <100 mg/dL, respectively [1]; however, a majority of patients with diabetes do not have these parameters under control, either individually or in combination [2].

HbA1c control has been shown to be associated with a reduced risk of microvascular complications in the general population of patients with diabetes [3–5] but the evidence is mixed with regards to cardiovascular benefits. Some studies have observed cardiovascular benefits of HbA1c control in relatively less severe patients with diabetes [6–8] or those newly diagnosed with or screened for diabetes [8–10]; however, a recent study on the impact of early use of insulin treatment to normalize glucose levels found no difference in cardiovascular benefits compared with standard treatment [11]. In addition, cardiovascular benefits of LDL-C control in diabetes have been well documented [12–16]. An intensive, multifactorial intervention approach in type 2 diabetes patients, designed to simultaneously target HbA1c, cholesterol levels, and other risk factors, has been shown to significantly reduce the risk of cardiovascular morbidity and mortality, microvascular complications, and rates of cardiovascular surgery [17, 18].

The benefits of achieving both HbA1c and LDL-C goals, compared with achieving just one, have not been quantified among newly diagnosed type 2 diabetes patients. Cardiovascular disease is the most frequent cause of death in patients with diabetes [19]. While some studies suggest that tight glycemic control reduces cardiovascular risk in patients who are newly diagnosed with diabetes [9], it is unclear whether achieving the HbA1c goal in addition to the LDL-C goal will have additional cardiovascular benefits. The primary objective of this study was to compare the clinical benefits of dual-goal achievement versus achievement of the LDL-C goal only in patients with newly diagnosed type 2 diabetes. Additional comparisons included dual-goal achievement versus HbA1c goal achievement and no-goal achievement, only LDL-C goal versus no-goal achievement, and only HbA1c goal versus no goal achievement. Furthermore, this study sought to better understand dual-goal achievement status and identify patient characteristics associated with the attainment of both goals.

Methods

Data Source

Electronic medical records from the South Central Veterans Affairs Health Care Network, Veterans Integrated Service Network (VISN) 16 (http://www.visn16.va.gov/) were used in this retrospective observational study. The VISN 16 data warehouse is an integrated, de-identified, individual-level database that includes records for more than 445,000 veterans from 10 medical centers and 40 outpatient clinics in the South Central Region of the United States (i.e., Oklahoma, Arkansas, Louisiana, Mississippi, and parts of Texas, Missouri, Alabama, and Florida). The database contains information regarding demographics, vital signs, laboratory results, diagnoses and procedures, inpatient and outpatient services (e.g., admission date, length of stay, emergency room visits), drug prescriptions, and database enrollment history. All data comply with the Health Insurance Portability and Accountability Act (HIPAA). This study was approved by the Institute Review Board including a waiver of informed consent and HIPPA authorization and Research and Development Committee of the Southeast Louisiana Veterans Heath Care System. This article does not contain any studies with human or animal subjects performed by any of the authors.

Sample Selection

To identify patients who were recently diagnosed with type 2 diabetes mellitus, patients were required to have ≥2 diagnoses of type 2 diabetes mellitus between January 1, 2004 and June 30, 2010, but no such diagnosis for at least 1 year from the start of their enrollment in the database (Fig. 1). Patients with more than 1 diagnosis of type 1 diabetes mellitus were excluded. During the 3-month period surrounding the first type 2 diabetes diagnosis, all patients were required to have at least one measurement of HbA1c and LDL-C within 30 days of each other; the date of the earliest HbA1c measurement in this period was defined as the index date. Patients were required to be ≥18 years old at the time of the index date, be continuously enrolled for at least 12 months, and have ≥1 measurement of both HbA1c and LDL-C in the 7- to 12-month period after the index date. Patients were excluded if they had taken antidiabetic drugs (oral antidiabetic drugs or insulin) or had a measurement of HbA1c ≥7.0% (53 mmol/mol) prior to the first diagnosis of type 2 diabetes mellitus or the index date, whichever was earlier.

Fig. 1
figure 1

Sample selection for patients with newly diagnosed type 2 diabetes. Asterisks the date of the earliest HbA1c measurement was considered the index date. HbA1c glycated hemoglobin A1c, LDL-C low-density lipoprotein cholesterol

Full size image

Data Preparation

Longitudinal data were analyzed according to 6-month cycles, starting from the index date. Average HbA1c and LDL-C levels were estimated for each cycle using the area under the curve method [20, 21]. For each cycle, the following estimated averages were used to group patients into one of four goal achievement categories: dual-goal [HbA1c <7% (53 mmol/mol) and LDL-C <100 mg/dL], HbA1c only [HbA1c <7% (53 mmol/mol) and LDL-C ≥100 mg/dL], LDL-C only [LDL-C <100 mg/dL and HbA1c ≥7% (53 mmol/mol)], or neither goal [HbA1c ≥7% (53 mmol/mol) and LDL-C ≥100 mg/dL].

Patient Characteristics

Patient characteristics during the first 6-month cycle were summarized for the overall sample as well as stratified by goal achievement status. Demographic information included age at index date, gender, race, body mass index (BMI), and year of index date. The history of diabetes-related complications [microvascular (diabetic retinopathy, neuropathy, or nephropathy), macrovascular (coronary artery disease, cerebrovascular disease, peripheral vascular disease, or atherosclerosis), and other complications such as infection, ocular problems, etc.], comorbidities, and surgical procedures were identified and summarized as of the first cycle using International Classification of Diseases, Clinical Modification, Version 9 (ICD-9-CM) codes. Diabetic medications used by patients during the first cycle were categorized by drug therapeutic class. Healthcare resource utilization during the first cycle was categorized by inpatient, outpatient, and emergency room (ER) visits. Differences in characteristics between the four patient groups according to goal achievement status were assessed using the analysis of variance method for continuous variables and Chi-squared tests for categorical variables.

Clinical Outcomes

Clinical outcomes were selected a priori and included the following: (a) a composite cardiovascular-related endpoint [cerebrovascular disease (stroke), acute myocardial infarction (MI), or cardiovascular death (defined by a diagnosis of coronary artery disease or cerebrovascular disease on the day of death)], (b) acute coronary syndromes (acute MI or unstable angina), (c) a composite endpoint for microvascular complications, and (d) cardiovascular surgical procedures (coronary artery bypass graft or percutaneous coronary intervention) (Supplementary Appendix 1).

For each specific clinical event analyzed, patients were excluded from the analysis if the event occurred before the end of the first cycle, and patients were followed from the start of the second cycle until the first event, death, or end of the data. Cox proportional hazards regression models were used for both descriptive and multivariate analyses. In the descriptive analysis, Cox proportional hazards regression models with time-varying goal achievement status were used to estimate the proportion of patients who did not experience specific clinical events over time. Furthermore, Cox proportional hazards models were expanded in the multivariate analysis to adjust for potential confounding factors: demographics, BMI, and index year as of the index date, cumulative diabetes-related comorbidity/complication history, resource utilization, and medication use. Goal achievement status and confounding factors were measured during a given cycle while the clinical event was measured at the following cycle. The risk of each clinical event by goal achievement status was quantified using adjusted hazard ratios (HRs) with 95% confidence intervals (95% CIs).

Characteristics Associated with Dual-Goal Achievement

To identify characteristics associated with dual-goal achievement 7–12 months after the index date, a logistic regression model was used to assess the effects of individual factors. These include demographics, dual-goal achievement within 6 months following the index date, and complications and comorbidities, diabetic medication use, surgical procedures, resource utilization (i.e., outpatient, inpatient, and ER visits), the Charlson Comorbidity Index (CCI, a validated measure of the overall health status) [22] within the 1-year period surrounding the index date (i.e., 6 months before and 6 months after the index date). The likelihood of achieving both goals relative to not achieving both goals was quantified using adjusted odds ratios (ORs) with 95% CIs. SAS software version 9.2 (Cary, NC, USA) was used to conduct statistical analyses, and a two-tailed α level of 0.05 was used to determine statistical significance.

Results

Patient Characteristics

A total of 16,829 newly diagnosed patients were included in the analysis (Fig. 1). Patient characteristics for the overall population are shown in Table 1. As of the index date, most patients were over 55 years (80.8%; mean age 63.3 years), male (96.8%), and white (67.0%), with a mean BMI of 31.8 kg/m2. Patients had a history of several diabetes-related complications and comorbidities including hypertension (79.3%), hyperlipidemia (72.9%), depression (39.5%), microvascular complications (19.1%), macrovascular complications (38.4%), infection (13.0%), and ocular problems (30.1%). The percentages of patients who had undergone lower extremity amputation, coronary artery bypass graft, or percutaneous coronary intervention were 0.2, 0.5, and 1.4%, respectively. The majority of patients were taking oral antidiabetic (52.8%), antihypertensive (67.9%), and lipid-lowering drugs (69.4%), while 10.2% of the patients received insulin during the first 6 months after index date. The majority of the patients had an outpatient visit (98.6%) during the first 6 months following the index date, but only 3.8% of the patients had an ER visit; 10.1% had an inpatient visit during the same period.

Table 1 Patient baseline characteristics, demographics, comorbidities, complications, medications and resource use
Full size table

During the first 6-month cycle, 39.0% of patients achieved both goals (dual-goal achievers), while 14.5% achieved only the LDL-C goal (LDL-C achievers), 32.2% achieved only the HbA1c goal (HbA1c achievers), and 14.4% did not achieve either goal (no-goal achievers) (Table 1). Compared with LDL-C achievers, HbA1c achievers, and no-goal achievers, dual-goal achievers were significantly older (66.2 years vs. 61.8, 62.5, and 58.7 for LDL-C, HbA1c, and no-goal achievers, respectively; all P < .001) and more likely to be white (70.5% vs. 66.7%, 65.8%, and 60.1%; all P < .001). History of depression was the lowest (34.7% vs. 39.4%, 43.4%, and 43.5% for LDL-C, HbA1c, and no-goal achievers, respectively; all P < .001) but history of macrovascular complications (46.6% vs. 42.9%, 31.0%, and 27.9%; all P < .001) and hypertension (82.5% vs. 79.0%, 77.1%, and 75.9%; all P < .001) was the highest for dual-goal achievers compared with other patient groups. History of microvascular complications was lower for dual-goal achievers than LDL-C goal achievers (18.1% vs. 23.1%; P < .001), but similar between dual-goal achievers and HbA1c goal achievers (18.1% vs. 17.3%; P = .245).

Clinical Outcomes

The median duration of follow-up time was 3.7 years. Most patients were free of the composite cardiovascular endpoint (96.4, 96.3, 96.6, and 95.9% for dual-goal, LDL-C, HbA1c, and no-goal achievers, respectively), Acute Coronary Syndrome (ACS) (98.7, 98.2, 98.8, and 98.0%), or microvascular complications (89.8, 83.5, 89.7, and 84.6%) during the first year. After 3 years of follow-up, a large majority of patients remained free of the composite cardiovascular endpoint (91.2, 91.0, 91.6, and 90.0% for dual-goal, LDL-C, HbA1c, and no-goal achievers, respectively), ACS (96.5, 95.2, 96.9, and 94.7%) or microvascular complications (77.0, 64.6, 76.9, and 66.6%).

After adjusting for potential confounding factors, the Cox proportional hazards regressions revealed that newly diagnosed diabetes patients who achieved both goals were associated with a lower risk of microvascular complications than patients who achieved only the LDL-C goal or those who did not achieve either goal. In particular, dual-goal achievement was associated with a lower risk of microvascular complications compared with only LDL-C goal achievers [adjusted HR (95% CI): 0.69 (0.63, 0.76)]. Analyses of all other diabetes-related complications and surgical procedures revealed no other statistically significant benefits for dual-goal achievers relative to single-goal achievers. Compared with no-goal achievers, dual-goal achievers had a significantly decreased risk of experiencing the composite cardiovascular-related endpoint [0.74 (0.62, 0.87)], acute coronary syndromes [0.69 (0.54, 0.88)], microvascular complications [0.70 (0.63, 0.78)], and coronary artery bypass grafting [0.40 (0.23, 0.68)] (Fig. 2).

Fig. 2
figure 2

Risk of experiencing diabetes-related complications and surgeries in patients with newly diagnosed type 2 diabetes. Dual: patients achieving both LDL-C and HbA1c goals; HbA1c: patients achieving only the HbA1c goal; LDL-C: patients achieving only the LDL-C goal; none: patients achieving neither goal (please see text for details). CI confidence interval, HbA1c glycated hemoglobin A1c, LDL-C low-density lipoprotein cholesterol

Full size image

Compared with no-goal achievers, both groups of single-goal achievers had a significantly lower risk of experiencing composite cardiovascular endpoint [LDL goal achievers: 0.76 (0.71, 0.81); HbA1c goal achievers: 0.80 (0.74, 0.87)], acute coronary syndrome [LDL goal achievers: 0.76 (0.67, 0.86); HbA1c goal achievers: 0.66 (0.60, 0.72)], and coronary artery bypass grafting [LDL goal achievers: 0.63 (0.45, 0.88); HbA1c goal achievers: 0.59 (0.44, 0.80)]. HbA1c achievers had a lower risk of experiencing microvascular complications [0.73 (0.71, 0.74)], and undergoing percutaneous coronary intervention [0.66 (0.53, 0.82)] relative to no-goal achievers.

Characteristics Associated with Dual-Goal Achievement

Out of 16,829 patients, 7432 (44.2%) achieved dual-goal status while 9397 (55.8%) were non-dual-goal achievers in the 7–12 months following the index date. The descriptive statistics of characteristics of these patients are summarized in Supplementary Appendix 2. The logistic regression reveals that patient characteristics associated with an increased likelihood of dual-goal achievement included dual-goal achievement during 0–6 months following the index date [adjusted OR (95% CI): 16.19 (14.92, 17.58)], a diagnosis of ketoacidosis (without coma) and hyperosmolarity [2.20 (1.18, 3.94)], the use of lipid-lowering drugs [1.53 (1.39, 1.69)], CCI scores between 1–3 and ≥3 [compared with CCI = 0; 1.15 (1.03, 1.29) and 1.40 (1.08, 1.81), respectively], index years of 2007 [1.16 (1.01, 1.33)], 2008 [1.19 (1.04, 1.37)], or 2009 [1.21 (1.03, 1.42)] relative to an index year of 2004, and age [1.02 (1.01, 1.02)]. Conversely, insulin use [0. 70 (0.61, 0.81)] and a diagnosis of retinopathy [0.67 (0.55, 0.82)], neuropathy [0.85 (0.76, 0.97)], or congestive heart failure [0.84 (0.70, 0.99)] were associated with a significantly decreased likelihood of dual-goal achievement (Table 2).

Table 2 Characteristics associated with dual-goal achievement
Full size table

Discussion

The results from this study show that the achievement of both HbA1c and LDL-C goals in patients with newly diagnosed type 2 diabetes is associated with an additional reduction of microvascular complication rates, compared with achievement of the LDL-C goal alone. The benefit of HbA1c goal has a so-called “glycemic legacy” beyond a finite period of intensive management, as shown in 10-year follow-up of the UKPDS study (ClinicalTrials.gov # NCT01099865) [9, 23]. In addition, this study found that dual- and single-goal achievements are both generally superior to not achieving either goal, and result in a lower risk of cardiovascular events, acute coronary syndrome, microvascular complications, and cardiovascular surgery. Based on our results, 44.2% of patients achieved both goals in the 7–12 months following the index date. Prior dual-goal achievement, older age, use of lipid-lowering medication, and prior diagnosis of ketoacidosis (without coma) and hyperosmolarity are associated with an increased likelihood of dual-goal achievement. Conversely, factors associated with a decreased likelihood of dual-goal achievement include retinopathy, neuropathy, congestive heart failure and prior use of insulin.

Benefits of Dual- and Single-Goal Achievements

The absence of cardiovascular benefits in patients with early-stage type 2 diabetes who achieved dual HbA1c–LDL-C control is consistent with observations available in the literature. In a previous study with a mean follow-up of 5.3 years, an early intensive multifactorial intervention (aimed at reducing HbA1c, LDL-C, and blood pressure) in patients with diabetes mellitus detected by screening was not associated with any statistically significant difference in the rate of cardiovascular events and all-cause mortality, compared with routine care, despite significantly better improvements in HbA1c and LDL-C levels in the intensive treatment group [24]. However, in Steno-2, a randomized study in patients with established diabetes (mean disease duration: 6 years), multifactorial treatment was associated with significantly lower rates of cardiovascular disease and microvascular complications after 7.8 years of follow-up, compared with standard care [17], and further 5.5 years of follow-up demonstrated significant benefits on cardiovascular mortality [18]. In both periods of the Steno-2 study, multifactorial treatment resulted in a higher proportion of patients who achieved HbA1c levels <6.5% (48 mmol/mol) and in mean LDL-C levels <100 mg/dL [17, 18]. Taken together, data from our analysis and the studies that assessed the effects of multifactorial intervention [17, 18, 24] suggest that cardiovascular benefits of dual-goal over single-goal achievements in patients with type 2 diabetes may be observed over a long term. Possibly, a follow-up time of 3.7 years in our study was not sufficient to reveal such benefits. The 1- and 3-year cumulative cardiovascular event-free rates (96.4%, and 91.2%, respectively) among patients with dual-goal achievement in our study were similar to patients with intensive treatment in the Steno-2 study (approximately 97.8% and 87.5% by year 1 and 3, respectively) [18].

The observed cardiovascular benefits of single-goal achievement (HbA1c or LDL-C) versus no-goal achievement are also consistent with literature [25–27].

Status and Characteristics Associated with Dual-Goal Achievement

Given that, only 44.2% of patients achieved both the HbA1c and LDL-C goals in the 7–12 month period following the index date, our results suggest an unmet need in controlling major risk factors for patients with newly diagnosed diabetes. This is consistent with available literature [2].

Our results also show that prior dual-goal achievement, older age, and the use of lipid-lowering drugs are associated with dual-goal achievement. On the contrary, insulin use and diabetes-related complications such as neuropathy, retinopathy, and congestive heart failure, which indicate a more advanced phase of the disease, are identified as characteristics associated with a decreased likelihood of dual-goal achievement. These results highlight the importance of timely intervention and regular medical care in achieving HbA1c and LDL-C goals.

These results are consistent with findings from several other studies. Older age was associated with the achievement of HbA1c goal in a recent retrospective cohort study of individuals with type 2 diabetes [28] and with the achievement of simultaneous glycemic, lipid, and blood pressure control in patients with diabetes, hypercholesterolemia, and hypertension [29]. In addition, a recently presented conference proceeding of a retrospective cohort study found that patients with diabetes who attained HbA1c levels <7% (53 mmol/mol) were generally older and less likely to receive insulin than patients who had not achieved HbA1c goal [30]. The Look AHEAD study (ClinicalTrials.gov # NCT00017953) suggested that insulin use and non-utilization of lipid-lowering drugs was associated with a failure to achieve all three goals (HbA1c, LDL-C, and blood pressure) among overweight and obese patients with diabetes [31]. Finally, achievement of the LDL-C goal has been directly associated with older age, and inversely associated with baseline LDL-C [32].

Limitations and Strengths

Due to the retrospective observational design, the analysis may have been affected by unobserved differences that were not taken into account in the model. Although we used strict selection criteria, there is a possibility we included some patients who were not truly newly diagnosed with T2DM. Specifically, we observed an unexpectedly high rate of insulin use during the 6-month post-index period. Patients who did not achieve either goal were on average about 8 years younger and used 4 times more insulin than those who achieved both goals (20.9% vs. 5.0%). It is possible that some of these younger patients who were using insulin were suffering from latent autoimmune diabetes of adults instead of type 2 diabetes [33]. In addition, some important information was not captured in the electronic medical records, including disease severity, disease duration, lifestyle modifications, and any potential (but unlikely) health care services that were provided outside of the Veterans Affairs (VA) health system. We did not look into the specific information about alternative cholesterol-lowering drugs and antihypertensive drugs. Furthermore, the VA database predominantly consists of male patients, which may limit generalization of findings. One of the major advantages of using electronic records from VA health system is that the lab values are recorded over time, allowing for a longitudinal study design with a median follow-up period of 3.7 years.

Similar studies in the general population, designed to assess the effects of medication, lifestyle changes, or triple-goal achievement (HbA1c, LDL-C, and blood pressure) may provide additional information. In addition, analyses with longer follow-up times may reveal benefits of dual- or triple-goal achievements on cardiovascular outcomes that were not observed in this study.

Conclusion

In US veterans with a newly recorded diagnosis of type 2 diabetes, the main benefit of achieving both HbA1c and LDL-C goals over achieving only LDL-C goal appears to be a reduced rate of microvascular complications. However, current rates of dual-goal achievement are suboptimal. Prior dual-goal achievement and use of lipid-lowering drugs were both associated with a higher rate of dual-goal achievement, which highlights the importance of early intervention and regular medical care.