Abstract
Aims:
We compared risk factors for three CVD manifestations and a composite outcome over 25 years’ follow-up in the Pittsburgh Epidemiology of Diabetes Complications (EDC) prospective cohort study of childhood-onset (<17 years) type 1 diabetes (n=658).
Methods:
First CVD manifestations examined were: 1) major atherosclerotic cardiovascular event (MACE, i.e. CVD death, myocardial infarction, stroke), 2) coronary revascularization, 3) soft coronary artery disease (CAD, i.e. ischemia ECG, angina), and a 4) composite (MACE + revascularization) outcome. Baseline and time-varying mean and current risk factors, including medication use, were assessed, in diabetes duration-adjusted models.
Results:
MACE (n=107) was predicted by ln(albumin excretion rate) (AER, HR=1.3, p<.0001), systolic BP (SBP, HR=1.03, p<.0001), white blood cell count (WBC, HR=1.2, p<.0001), HbA1c (HR=1.2 p=0.03), LDLc (HR=1.01, p=0.03). Soft CAD (n=91) was predicted by ln(AER) (HR=1.2, p=0.004), SBP (HR=1.03, p=0.0002), WBC (HR=1.2, p=0.0003), HbA1c (HR=1.2, p=0.005). Revascularization (n=38) was predicted by LDLc (HR=1.03, p<.0001), eGFR (HR=0.98, p=0.002), HbA1c (HR=1.3, p=0.03). Adding revascularization to MACE enhanced the role of LDLc, while diminishing that of HbA1c, compared to MACE alone.
Conclusions:
Important risk factor associations may be affected by examining composite CVD outcomes. More research is needed to determine how to best incorporate revascularization into composite CVD definitions.
Keywords: Cardiovascular disease manifestations, Composite CVD Outcomes, Revascularization, CVD Risk Factors
1. Introduction
Type 1 diabetes continues to be associated with a dramatically higher risk of cardiovascular disease (CVD) compared to the background population (1,2). No single risk factor has been solely implicated in this increased CVD risk. Additionally, there is evidence that the relative impact of risk factors may vary by harder (i.e. myocardial infarction or stroke) and softer (i.e. angina or ischemic EGC) CVD manifestations (3). Thus, combining the various manifestations of CVD into a single composite endpoint definition may obscure associations between risk factors and specific outcomes in observational cohort studies. Furthermore, while concern has been raised regarding composite endpoints in cardiovascular outcomes trials (4,5) and CVD risk engines (6), investigations in observational studies are lacking. In particular, including revascularization procedures in composite endpoints is questionable because the procedure is the result of a medical decision-making process, introducing subjectivity into the CVD definition (4,7). Improved understanding of the differing risk factors for the various manifestations of CVD may help to identify patients at greatest risk of major morbidity and mortality in the type 1 diabetes population. Thus, our objective was to assess risk factors for three first CVD manifestations, comprising major atherosclerotic events (MACE, i.e. CVD death, myocardial infarction, or stroke), coronary revascularizations, and soft coronary artery disease (CAD, i.e. angina or ischemic ECG) in a well-characterized type 1 diabetes cohort.
2. Subjects, Materials and Methods
2.1. Study Population
The Pittsburgh Epidemiology of Diabetes Complications (EDC) Study is a prospective cohort study of childhood-onset (<17 years old) type 1 diabetes. All participants (n=658) were diagnosed, or seen within one year of diagnosis, at Children’s Hospital of Pittsburgh between 1950 and 1980. The cohort has been described in detail elsewhere (8,9). In brief, participants have been followed since 1986–1988, initially with biennial examinations for ten years and thereafter with biennial questionnaires and further examinations at 18- and 25-years post-baseline. Research protocols were approved by the University of Pittsburgh institutional review board and all participants provided written informed consent.
2.2. Ascertainment of CVD Outcomes
EDC participants were followed for 25 years (from 1986–88 to 2011–2013) to assess CVD incidence. Participants were classified by their first CVD event into the following three outcomes: 1) MACE, defined as the first instance of CVD death, nonfatal myocardial infarction (MI, including clinical events and subclinical myocardial infarction on ECG, i.e. Minnesota code 1.1 or 1.2), or nonfatal stroke, 2) coronary revascularization procedure or blockage ≥50%, and 3) soft CAD, defined as ischemic EGC (Minnesota code 1.3, 4.1–4.3, 5.1–5.3, or 7.1) at EDC study visit or EDC physician-diagnosed angina. Fatal events were ascertained using medical records, death certificates, autopsy reports, and/or interview with next of kin and classified according to the Diabetes Epidemiology Research International (DERI) system (10). Medical records were requested to confirm nonfatal MI, stroke, coronary revascularization, and blockage.
2.3. Risk Factor Assessment
Risk factors were assessed at baseline and repeated at 2-, 4-, 6-, 8-, 10-, and 18-years of follow-up. For the first 18 months of the study, HbA1c was measured in fasting blood samples using microcolumn cation exchange (Isolab, Akron, OH, USA). For the remainder of the first 10 years of follow-up, HbA1 was measured using automated high-performance liquid chromatography (Diamat; Biorad, Hercules, CA, USA). The two assays had high agreement (r=0.95; Diamat HbA1=-0.18+1.00[Isolab HbA1]). HbA1 values were converted to DCCT-aligned HbA1c values using a regression equation derived from duplicate assays (DCCT HbA1c=0.14 + 0.83[EDC HbA1]) (11). At the 18-year examination, HbA1c was measured using the DCA 2000 analyzer (Bayer Healthcare LLC. Elkhart, IN) and converted to DCCT-aligned HbA1c by the equation: DCCT HbA1c=(EDC HbA1c-1.13)/0.81. From baseline through the 10-year examination, serum total cholesterol and triglycerides were determined enzymatically (12,13) and HDL-cholesterol was determined using a modified precipitation technique (14) based on the lipid research clinics method (15). At the 18-year examination, serum lipids were measured using the Cholestech LDX (Cholestech Corp., Hayward, CA). Hyperlipidemia was defined as LDLc≥130 mg/dl or use of lipid-lowering medication. White blood cell count (WBC) was determined using a Coulter Counter (Coulter Electronics, Inc., Hialeah, FL). Blood pressure was measured according to the Hypertension Detection and Follow-Up protocol (16) with a random-zero sphygmomanometer, replaced by an aneroid sphygmomanometer at the 18-year exam. Hypertension was defined as blood pressure ≥140/90 mmHg or use of blood pressure-lowering medication. Pulse pressure was calculated as the difference between systolic and diastolic pressures. Pulse rate (beats/minute) was determined by palpitating the radial pulse for 30 seconds and multiplying by two. Serum creatinine was measured using an Ectachem 400 Analyzer (Eastman Kodak Co.). Serum and urinary albumin were measured by immunonephelometry (17). Albumin excretion rate (AER) was calculated for each of three timed urine samples (24-hour, overnight, and 4-hour collections obtained over a two-week period); the median of the three AERs was used in analyses. Overt nephropathy was defined as AER>200 μg/min in at least two of the three timed urine samples. Glomerular filtration rate (eGFR) was estimated by the CKD-EPI creatinine equation (18). Height and weight were measured using standard methods to calculate BMI. Waist and hip circumference were measured at least twice; the average of each was used to calculate the waist-hip ratio (WHR).
Socioeconomic status (SES) factors, including highest level of education, professional employment status, and income, were assessed at the EDC cycle closest to age 28, using methodology which has been described previously (19,20). Briefly, age 28 was chosen as this time period permits reasonable establishment of educational, occupational, and financial status, while minimizing the effect of advanced diabetes complications on income potential. Highest level of education was ascertained through self-report, categorized as some high school, high school graduate, some college, college graduate (i.e. bachelor’s degree), and education beyond college. Occupation category was determined using the Hollingshead Index of Social Position (21). Occupation status was classified as professional (Hollingshead 1A–3C) or nonprofessional (Hollingshead 4A–7X). Household income was self-reported by the use of categories of annual pre-tax income in U.S. dollars earned by each household. Income was grouped into one of five income categories based on the Age 28 study cycle. For EDC study visits 1–3 (1986–1992), annual household income categories were: ≤$10,000, $10,001–20,000, $20,001–30,000, $30,001–40,000, and >$40,000. For EDC study visits 4–10 (1992–2006), income categories were: ≤$20,000, $20,001–30,000, $30,001–40,000, $40,001–50,000, and >$50,000. Income was dichotomized into two groups: member of the highest income level at Age 28 study cycle (i.e. >$40,000 or >$50,000) versus all others. Health insurance status and plan type were ascertained by questionnaire, beginning at the 1990–1992 (visit 3) follow-up. Insurance plans types were categorized as: 1) employer group plan (obtained through the participant’s or their spouse or parent’s employer), 2) individual plan, and 3) Medicare or Medicaid.
Insulin dose was calculated as total insulin units per day divided by body weight (kg). Past and current smoking status, alcohol consumption, lipid- and blood pressure-lowering medications, and first degree family history of hypertension, myocardial infarction, diabetes before age 30, and diabetes after age 30 were obtained by self-administered questionnaire. Physical activity was assessed using the Paffenbarger Questionnaire (22) and average total weekly energy expenditure was calculated (kcal/week). Hypoglycemia requiring assistance was defined as reporting at least one hypoglycemic episode in the past 2 years resulting in unconsciousness and/or hospitalization or a hypoglycemic episode in the past 12 months that was not recognized by the participant (i.e. someone else had to tell or help the participant).
2.4. Statistical Analyses
Participants with prevalent CVD at baseline (n=54) were excluded from analyses. In addition to the three individual CVD manifestations (i.e. MACE, soft CAD, and revascularization), a combined outcome of MACE + revascularization (whichever occurred first) was also examined to determine whether risk factors differed when the two hard CVD outcomes were combined, compared to each alone. Pairwise comparisons of continuous characteristics between each CVD manifestation group were compared using t-tests or Wilcoxon rank sum tests if non-normally distributed. Dichotomous variables were compared between groups using chi-square or Fisher’s exact test, where appropriate. To account for multiple comparisons, given the four CVD outcomes being examined, we adjusted the significance level to p=0.01 (i.e. 0.05 divided by four).
Kaplan-Meier curves were fit for each CVD manifestation (MACE, soft CAD, revascularization). In Cox models, three forms of each risk factor variable were assessed: 1) fixed baseline, 2) time-varying current/most recent, and 3) time-varying updated mean, with the exceptions of diabetes duration and age at baseline, sex, highest education, income and employment status at age 28, and health insurance status and plan type category at the time of event or last follow-up, which were assessed as fixed values only. Separately for each first CVD manifestation and for the combined MACE + revascularization outcome, Cox models were used to estimate the relative risk associated with a one-unit increment in the risk factors in both unadjusted univariable models for each form of each factor individually and in fully-adjusted multivariable models. The fully-adjusted models were selected using two approaches: 1) offering all variables with a univariable association with p<0.1 and performing backward selection with a p-value cut-off of 0.05 to be retained in the final model and 2) offering all variables, regardless of univariable associations, and performing backward selection with a p-value cut-off of 0.05 to be retained in the final model. For all outcomes, the comparison group comprised the participants who never developed CVD during the 25-year follow-up. Kaplan-Meier curves were fit using R. Cox analyses were performed using SAS v. 9.4 (SAS Institute, Inc., Cary, NC).
3. Results
Of the 604 participants free of CVD at baseline, 236 (39.1%) had a first incident CVD event during the 25-year follow-up period. Of these CVD events, 107 were MACE, 38 revascularization (including 8 individuals with blockage>50% but medically managed), and 91 were soft CAD (31 ischemic ECG, 60 angina). The frequencies of the specific outcomes within each category are presented in Supplementary Table 1. Detailed information on the extent of angina that may have precipitated revascularization was not available. However, 24 of the 38 revascularization cases granted access to medical records. A description of the review is in the Supplementary Material.
Kaplan-Meier curves for each first CVD manifestation category are shown in Figure 1. Key baseline characteristics by CVD manifestation are presented in Table 1. A complete comparison of baseline characteristics between each first CVD manifestation and no CVD can be found in Supplementary Tables 2–4. Those who went on to develop CVD were older and had longer diabetes duration than those who did not develop CVD, regardless of manifestation. LDL-c was higher in all manifestations compared to no CVD, but was highest in those whose first event was a revascularization, while triglycerides were highest in those whose first event was MACE. HbA1c at baseline was slightly lower in those whose first event was soft CAD, compared to the other manifestations and no CVD. Both systolic and diastolic blood pressures showed a gradient by severity of CVD manifestation, being highest for MACE, followed by revascularization, and soft CAD being only slightly elevated compared to no CVD. AER, white blood cell count, and smoking were all much higher at baseline in those who developed MACE as their first CVD event compared to any other manifestation. Those whose first CVD manifestation was revascularization were more likely to have attained a bachelor’s degree or higher education, hold professional employment, or have an income in the highest category (>$40,000 or $50,000), though these differences did not reach statistical significance.
Figure 1.
Kaplan-Meier survival curves for each first manifestation of cardiovascular disease
Table 1.
Baseline characteristics by first manifestation of cardiovascular disease
| No CVD (n=368) | MACE* (n=107) | Soft CAD⍰(n=91) | REVǂ (n=38) | MACE + REV (n=145) | |
|---|---|---|---|---|---|
| Age (years) | 24 (7.1) | 31 (6.6)N | 32 (7.0)N | 30 (6.9)N | 31 (6.7)N |
| T1D Duration (years) | 16 (6.4) | 23 (7.0)N | 24 (7.1)N | 22 (7.7)N | 23 (7.1)N |
| Sex, % men (n) | 49.5% (182) | 54.2% (58) | 45.1% (41) | 60.5% (23) | 55.9% (81) |
| Race, non-white, % (n) | 2.5% (9) | 3.7% (4) | 1.1% (1) | 0% (0) | 2.8% (4) |
| Total Cholesterol (mg/dl) | 181 (37.6) | 208 (41.3)N | 195 (42.3)N | 213 (46.6)N | 208.9 (42.7)N |
| LDL-c (mg/dl) | 107 (30.0) | 129 (34.3)N | 119 (35.5) N,R,M+R | 141 (40.6) N,s | 132.2 (36.3)N,S |
| HDL-c (mg/dl) | 55 (12.3) | 53 (12.0) | 52 (13.1) | 51 (9.2) | 52.5 (11.3)N |
| Triglycerides (mg/dl)§ | 75 (56, 102) | 104 (75, 172)N | 89 (63, 113) | 88 (69, 131) | 96 (73, 161)N |
| HbA1c (%) | 8.8 (1.5) | 9.0 (1.5)s | 8.4 (1.3)M,M+R | 9.0 (1.8) | 9.0 (1.6)s |
| (mmol/mol) | 72 (17) | 74 (17)s | 68 (14)M,M+R | 75 (19) | 75 (17)s |
| Frequency of checking blood sugar/week§ | 2 (0–14) | 1 (0–10) | 1 (0–10) | 0 (0–7) | 1 (0–8) |
| BMI (kg/m2) | 23 (3.1) | 24 (3.5) | 24 (3.4) | 24 (2.6) | 24 (3.3) |
| Systolic BP (mmHg) | 109 (11.8) | 122 (18.8)N,S | 114 (13.5) N,M,M+R | 117(16.1)N | 120 (18.1)N,s |
| Diastolic BP (mmHg) | 70 (9.3) | 78 (13.8)N,S | 73 (10.1)M,M+R | 76 (11.5)N | 77 (13.2)N,S |
| Pulse Rate (bpm) | 78 (9.7) | 80 (10.6) | 79 (9.7) | 78 (8.6) | 79 (10.1) |
| White Blood Cell Count (×109 cells/L) | 6.2 (1.7) | 7.5 (2.4)N,R | 6.9 (1.9) | 6.3 (1.7)M | 7.2 (2.3)N |
| Albumin Excretion Rate (μg/min)§ | 12 (7, 33) | 105 (13, 712)N,S | 23 (9, 242) N,M | 25(11, 528)N | 67 (12, 578)N |
| Estimated Glomerular Filtration Rate (eGFR) (mL/min/1.73 m2) | 109 (29.6) | 94 (32.5)N | 100 (27.7)N | 93 (30.7)N | 93.8 (32.0)N |
| Current smoker, % (n) | 17.9% (66) | 36.5% (39)N | 27.5% (25) | 29.0% (11) | 34.5% (50)N |
| Bachelor’s Degree or greater, % (n)ǁ | 35.6% (131) | 18.7% (20)N | 22.0% (20)N | 36.8% (14) | 23.5% (34)N |
| Professional employment, % (n)ǁ,# | 37.8% (128) | 22.1% (23)N | 24.1% (21) | 42.9% (15) | 27.3% (38)N |
| Highest Income category, % (n)f | 15.3% (42) | 9.8% (9) | 11.9% (8) | 23.3% (7) | 13.1% (16) |
MACE=major atherosclerotic cardiovascular event (CVD death, myocardial infarction, or stroke);
SOFT=soft cardiovascular disease event (ischemic ECG or angina);
REV=revascularization procedure or blockage ≥50%;
Values are median (interquartile range);
Using socioeconomic status closest to age 28 as described in Secrest et al. 2011;
declined to answer: employment n=38 (6%), income n=141 (23%).
Superscripts on the values represent differences between first manifestations or no CVD at p<0.01 significance level, N=None, M=MACE, S=SOFT, R=REV
A comparison of most recent risk factors across each CVD manifestation is presented in Table 2. Those who first developed MACE or soft CAD had similarly high most recent levels of triglycerides compared to those with revascularization/blockage. Those developing MACE had much higher most recent AER than any other group. Lipid-lowering medication use was twice as common in those with revascularization as a first event than those with MACE or soft CAD. Greater than median leisure-time physical activity was also much more frequent among those with revascularization compared to MACE and soft CAD. The proportion with health insurance coverage was similarly high in all groups (>90%), however, the distribution of plan type somewhat differed, with a higher proportion of participants with MACE covered by Medicare or Medicaid at the time of their event compared to the other groups.
Table 2.
Most recent characteristics prior to first manifestation of cardiovascular disease
| MACE* (n=107) | Soft CAD⍰(n=91) | REVǂ (n=38) | MACE + REV (n=145) | |
|---|---|---|---|---|
| Total Cholesterol (mg/dl) | 211 (38.1) | 205 (43.7) | 204 (52.9) | 209 (42.3) |
| LDL-c (mg/dl) | 131 (35.6) | 128 (40.9) | 129 (45.2) | 131 (38.2) |
| HDL-c (mg/dl) | 54 (15.9) | 53 (14.4) | 54 (14.6) | 54 (15.5) |
| Triglycerides (mg/dl)d | 110 (73.7–148.4) | 106 (68.7–148.0) | 88 (66.0–114.4) | 102 (75.2, 142.6) |
| Lipid-lowering medications, % (n) | 12.2% (13) | 14.3% (13) | 26.3% (10) | 15.9% (23) |
| HbA1c (%) | 9.0 (1.86) | 9.2 (1.69) | 8.8 (1.55) | 9.0 (1.78) |
| (mmol/mol) | 75 (20) | 77 (18) | 72 (17) | 75 (19) |
| Insulin Dose (units/kg body weight) | 0.70 (0.33) | 0.69 (0.19) | 0.63 (0.27) | 0.68 (0.31) |
| Frequency of checking blood sugar/week | 7 (0–21) | 8.5 (1–21) | 15 (6–26.5) | 10 (0–25) |
| BMI (kg/m2) | 25 (4.9) | 26 (4.8) | 26 (3.7) | 26 (4.6) |
| Systolic BP (mmHg) | 126 (20.1) | 125 (18.4) | 122 (17.8) | 125 (19.6) |
| Diastolic BP (mmHg) | 78 (14.4) | 74 (12.8) | 73 (13.6) | 75 (14.2) |
| White Blood Cell Count (×109 cells/L) | 7.8 (2.5) | 7.6 (2.2) | 7.3 (2.6) | 7.65 (2.49) |
| Albumin Excretion Rate (μg/min)§ | 156 (21.3–944)s | 45 (9.9–2208)M | 45 (7.4–492.7) R | 106 (14.4, 880.1) |
| Estimated Glomerular Filtration Rate (eGFR) (mL/min/1.73 m2) | 79 (35.5) | 88 (27.2) | 76 (32.1) | 78 (34.6) |
| Current smoker, % (n) | 24.3% (26) | 23.1% (21) | 21.1% (8) | 23.5% (34) |
| Physical Activity, % >1512 kcal/wk (n) | 26.4% (28) | 27.5% (25) | 47.4% (18) | 31.9% (46) |
| Has health Insurance, % (n)ǁ | 94.3% (50) | 91.2% (52) | 91.7% (22) | 93.5% (72) |
| Type of health insurance, % (n)# | ||||
| Employer group plan | 55.8% (29) | 70.2% (40) | 70.8% (17) | 60.5% (46) |
| Individual plan | 11.5% (6) | 8.8% (5) | 12.5% (2) | 11.8% (9) |
| Medicare/Medicaid | 26.9% (14) | 12.3% (7) | 8.3% (2) | 21.1% (16) |
MACE=major atherosclerotic cardiovascular event (CVD death, myocardial infarction, or stroke);
SOFT=soft cardiovascular disease event (ischemic ECG or angina);
REV=revascularization procedure or blockage ≥50%;
Values are median (interquartile range);
Due to missing data for most recent health insurance status, total n=341 No CVD, 53 MACE, 24 revascularization/ blockage, 57 soft CAD;
Due to missing data for insurance type, total n=332 No CVD, 52 MACE, 24 revascularization/ blockage, 57 soft CAD.
Superscripts on the values represent differences between first manifestations or no CVD at p<0.01 significance level, N=None, M=MACE, S=SOFT, R=REV
The unadjusted hazard ratios and 95% confidence intervals for a one-unit increment in selected fixed baseline and time-varying most recent or updated mean risk factors are presented in Table 3. The remaining risk factors are presented in Supplementary Table 5. Most variables were similarly univariately associated with all CVD manifestations. Exceptions included current physical activity expenditure of ≥1512 kcal per week, which was protective against MACE and soft CAD, but not associated with revascularization. Current hypertension was more strongly associated with MACE and soft CAD than revascularization. Triglycerides, including baseline, current, and updated mean, were most strongly associated with MACE, while current hyperlipidemia (LDLc≥130 mg/dl or use of lipid-lowering medication) was most strongly associated with revascularization. Mean insulin dose was protective of both revascularization and soft CAD, but not MACE. Hypoglycemia was marginally associated with an increased risk of soft CAD, but not MACE or REV. Baseline, updated mean, and current HbA1c were each similarly associated with all manifestations, though the relationships were slightly weaker for soft CAD than MACE or revascularization.
Table 3.
Univariable associations between risk factors and incidence of each manifestation of cardiovascular disease
| MACE (107 events) | Soft CAD (91 events) | REV (38 events) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Typea | HRb | 95% Cl | p-value | HRb | 95% Cl | p-value | HRb | 95% Cl | p-value | ||
| Demographic | |||||||||||
| Physical | |||||||||||
| Age (years) | B | 1.13 | 1.10, 1.16 | <.0001 | 1.13 | 1.10, 1.17 | <.0001 | 1.12 | 1.07, 1.16 | <.0001 | |
| Mean BMI (kg/m2) | M | 1.05 | 0.99, 1.12 | 0.11 | 1.09 | 1.02, 1.16 | 0.009 | 1.06 | 0.96, 1.18 | 0.25 | |
| Behavioral | |||||||||||
| Smoking (yes v. no) | B | 2.67 | 1.80, 3.97 | <.0001 | 1.83 | 1.15, 2.90 | 0.01 | 2.18 | 1.08, 4.40 | 0.03 | |
| Smoking (yes v. no) | C | 1.69 | 1.08, 2.64 | 0.02 | 1.55 | 0.95, 2.54 | 0.08 | 1.73 | 0.79, 3.79 | 0.17 | |
| Alcohol (occasional/regular v. none) | C | 0.73 | 0.49, 1.08 | 0.11 | 1.06 | 0.68, 1.64 | 0.81 | 0.95 | 0.48, 1.86 | 0.88 | |
| Activity (≥1512 kcal/wk) | C | 0.44 | 0.29, 0.68 | 0.0002 | 0.44 | 0.28, 0.69 | 0.0005 | 1.16 | 0.61, 2.20 | 0.64 | |
| Family History | |||||||||||
| Hypertension | B | 1.03 | 0.86, 1.25 | 0.74 | 1.10 | 0.93, 1.30 | 0.26 | 1.15 | 0.93, 1.42 | 0.21 | |
| Myocardial Infarction | B | 2.60 | 1.69, 4.02 | <.0001 | 2.20 | 1.35, 3.60 | 0.002 | 3.48 | 1.66, 7.28 | 0.001 | |
| Diabetes Before Age 30 | B | 1.09 | 0.44, 2.68 | 0.85 | 2.29 | 1.19, 4.44 | 0.01 | 1.75 | 0.53, 5.74 | 0.36 | |
| Diabetes After Age 30 | B | 2.04 | 1.18, 3.53 | 0.01 | 1.50 | 0.73, 2.91 | 0.28 | 1.83 | 0.64, 5.21 | 0.26 | |
| Socioeconomic Factors | |||||||||||
| Highest Education (≥bachelor’s degree) | Bc | 0.41 | 0.25, 0.66 | 0.0003 | 0.51 | 0.31, 0.83 | 0.007 | 0.87 | 0.45, 1.68 | 0.67 | |
| Professional Employment (yes v. no) | Bc | 0.49 | 0.31, 0.78 | 0.003 | 0.54 | 0.33, 0.88 | 0.01 | 1.09 | 0.56, 2.14 | 0.79 | |
| Income (highest level v. all others) | Bc | 0.58 | 0.29, 1.16 | 0.13 | 0.68 | 0.32, 1.42 | 0.30 | 1.46 | 0.63, 3.41 | 0.38 | |
| Health Insurance | C | ||||||||||
| None | Reference | Reference | Reference | ||||||||
| Employer group plan | 0.66 | 0.20, 2.16 | 0.49 | 0.57 | 0.23,1.45 | 0.24 | 0.55 | 0.13, 2.37 | 0.42 | ||
| Individual plan | 1.49 | 0.37, 5.94 | 0.58 | 0.83 | 0.24, 2.89 | 0.77 | 1.14 | 0.19, 6.81 | 0.89 | ||
| Medicare/Medicaid | 1.17 | 0.34, 4.10 | 0.80 | 0.41 | 0.13, 1.29 | 0.13 | 0.27 | 0.04, 1.91 | 0.19 | ||
| Traditional | |||||||||||
| Blood Pressure | |||||||||||
| Systolic (mmHg) | B | 1.05 | 1.04, 1.06 | <.0001 | 1.04 | 1.02, 1.05 | <.0001 | 1.06 | 1.03, 1.08 | <.0001 | |
| Systolic (mmHg) | C | 1.04 | 1.03, 1.05 | <.0001 | 1.05 | 1.04, 1.06 | <.0001 | 1.04 | 1.02, 1.06 | <.0001 | |
| Systolic (mmHg) | M | 1.06 | 1.05, 1.08 | <.0001 | 1.05 | 1.04, 1.07 | <.0001 | 1.07 | 1.04, 1.09 | <.0001 | |
| Diastolic (mmHg) | B | 1.07 | 1.05, 1.09 | <.0001 | 1.03 | 1.01, 1.06 | 0.007 | 1.07 | 1.03, 1.11 | 0.0001 | |
| Diastolic (mmHg) | C | 1.05 | 1.03, 1.070 | <.0001 | 1.03 | 1.01, 1.05 | 0.004 | 1.04 | 1.01, 1.08 | 0.012 | |
| Diastolic (mmHg) | M | 1.08 | 1.06, 1.099 | <.0001 | 1.04 | 1.01, 1.07 | 0.004 | 1.09 | 1.04, 1.13 | <.0001 | |
| Hypertension (yes v. no) | C | 4.58 | 3.12, 6.724 | <.0001 | 4.06 | 2.65, 6.20 | <.0001 | 2.62 | 1.36, 5.06 | 0.004 | |
| Any hypertension | E | 4.31 | 2.91, 6.364 | <.0001 | 3.62 | 2.37, 5.53 | <.0001 | 2.93 | 1.53, 5.60 | 0.001 | |
| Pulse rate (bpm) | B | 1.03 | 1.01, 1.05 | 0.006 | 1.02 | 0.998, 1.04 | 0.076 | 1.01 | 0.98, 1.04 | 0.575 | |
| Lipids | |||||||||||
| Total Cholesterol (mg/dl) | B | 1.01 | 1.01, 1.018 | <.0001 | 1.01 | 1.004, 1.013 | 0.0002 | 1.02 | 1.01, 1.023 | <.0001 | |
| In(Triglycerides) (mg/dl) | B | 2.58 | 1.94, 3.44 | <.0001 | 1.80 | 1.28, 2.51 | 0.0007 | 1.82 | 1.07, 3.09 | 0.03 | |
| HDLc (mg/dl) | B | 0.98 | 0.97, 1.001 | 0.081 | 0.98 | 0.96, 1.00 | 0.048 | 0.97 | 0.940, 1.00 | 0.053 | |
| LDLc (mg/dl) | B | 1.02 | 1.01, 1.03 | <.0001 | 1.01 | 1.006, 1.017 | <.0001 | 1.03 | 1.02, 1.032 | <.0001 | |
| Inflammation | |||||||||||
| White blood cell count | B | 1.32 | 1.23, 1.43 | <.0001 | 1.23 | 1.11, 1.36 | <.0001 | 1.09 | 0.91, 1.31 | 0.35 | |
| White blood cell count | C | 1.24 | 1.15, 1.34 | <.0001 | 1.20 | 1.11, 1.31 | <.0001 | 1.19 | 1.03, 1.37 | 0.02 | |
| Diabetes-Related | |||||||||||
| History | |||||||||||
| Duration of Diabetes (years) | B | 1.12 | 1.10, 1.15 | <.0001 | 1.13 | 1.10, 1.16 | <.0001 | 1.11 | 1.07, 1.16 | <.0001 | |
| Insulin Dose (units/kg/day) | B | 0.57 | 0.24, 1.32 | 0.19 | 0.20 | 0.07, 0.50 | 0.0009 | 0.07 | 0.01, 0.03 | 0.001 | |
| Nephropathy | |||||||||||
| Estimated GFR (ml/min/1.73m2) | B | 0.98 | 0.97, 0.99 | <.0001 | 0.99 | 0.98, 0.99 | 0.0002 | 0.98 | 0.97, 0.99 | <.0001 | |
| AER (μg/min) | C | 1.54 | 1.42, 1.67 | <.0001 | 1.37 | 1.25, 1.50 | <.0001 | 1.41 | 1.21, 1.63 | <.0001 | |
| Overt Nephropathy (yes v. no) | E | 6.22 | 4.22, 9.18 | <.0001 | 2.97 | 1.94, 4.534 | <.0001 | 4.14 | 2.1, 7.84 | <.0001 | |
| Hypoglycemia | |||||||||||
| Requiring Assistance (yes v. no) | C | 0.87 | 0.59, 1.28 | 0.49 | 1.45 | 0.95, 2.19 | 0.08 | 0.90 | 0.47, 1.72 | 0.75 | |
| Glycemia | |||||||||||
| HbA1c (%) | B | 1.14 | 1.01, 1.28 | 0.04 | 0.88 | 0.76, 1.02 | 0.10 | 1.17 | 0.96, 1.42 | 0.12 | |
| HbA1c (%) | C | 1.15 | 1.03, 1.29 | 0.01 | 1.18 | 1.04, 1.33 | 0.01 | 1.16 | 0.96, 1.38 | 0.12 | |
| HbA1c (%) | M | 1.27 | 1.11, 1.47 | 0.0007 | 1.06 | 0.90, 1.24 | 0.50 | 1.32 | 1.00, 1.72 | 0.05 | |
The final multivariable models for each manifestation are shown in Table 4. For MACE, diabetes duration, current ln(AER), mean systolic BP, baseline white blood cell count, baseline LDLc, and mean HbA1c remained significant predictors after backward selection. For revascularization, baseline LDLc, diabetes duration, baseline eGFR, and current HbA1c were significant predictors. Finally, for soft CAD, diabetes duration, current systolic BP, current white blood cell count, current ln(AER), and current HbA1c were significant predictors. When MACE and revascularization are combined into a composite outcome, the final model is similar to that seen for MACE, with the notable exception that the association with HbA1c is obscured.
Table 4.
Final multivariable models for each manifestation of cardiovascular disease
| Type* | HR⍰ | 95% Cl | p-value | |
|---|---|---|---|---|
| MACE | ||||
| Diabetes Duration | B | 1.11 | 1.08, 1.15 | <.0001 |
| In(AER) | C | 1.32 | 1.19, 1.48 | <.0001 |
| Systolic BP | M | 1.03 | 1.01, 1.05 | <.0001 |
| White Blood Cell Count | B | 1.23 | 1.12, 1.36 | <.0001 |
| HbA1c | M | 1.20 | 1.02, 1.41 | 0.03 |
| LDLc | B | 1.01 | 1.001, 1.013 | 0.04 |
| Soft CAD | ||||
| Diabetes Duration | B | 1.13 | 1.10, 1.16 | <.0001 |
| Systolic BP | C | 1.03 | 1.01, 1.04 | 0.0002 |
| White Blood Cell Count | C | 1.17 | 1.08, 1.28 | 0.0003 |
| In(AER) | C | 1.18 | 1.05, 1.32 | 0.004 |
| HbA1c | C | 1.21 | 1.06, 1.39 | 0.005 |
| REV | ||||
| LDLc | B | 1.03 | 1.02, 1.03 | <.0001 |
| Diabetes Duration | B | 1.10 | 1.05, 1.16 | 0.0001 |
| eGFR | B | 0.98 | 0.97, 0.99 | 0.002 |
| HbA1c | C | 1.25 | 1.02, 1.23 | 0.03 |
| MACE+REV | ||||
| Diabetes Duration | B | 1.09 | 1.07, 1.12 | <.0001 |
| In(AER) | C | 1.32 | 1.21, 1.44 | <.0001 |
| Systolic BP | B | 1.03 | 1.01, 1.04 | <.0001 |
| LDLc | B | 1.01 | 1.005, 1.014 | <.0001 |
| White Blood Cell Count | B | 1.19 | 1.09, 1.30 | 0.0001 |
Variable type: B=baseline (fixed), C=current/most recent (time-varying), M=updated mean (time-varying)
Hazard Ratio
4. Discussion
Our findings demonstrate that there may be important differences in risk factors between various CVD manifestations in type 1 diabetes. In general, MACE and soft CAD had similar risk factors, but there was a shift toward a greater contribution of renal and blood pressure factors in MACE, consistent with past findings (24). LDLc was the strongest independent predictor of revascularization, while it was a more modest predictor of MACE. For all first manifestations of CVD, HbA1c was consistently, though modestly, associated with increased risk. However, when revascularization was added to MACE in a composite outcome, the association with LDLc was enhanced while the already relatively weak association with HbA1c was diminished, compared to MACE alone. These results suggest that examining composite CVD outcomes may obscure important risk factor associations in type 1 diabetes.
While sex was not significantly associated with risk of any CVD manifestation, a greater proportion of men first presented with MACE or revascularization, while more women first presented with ischemic ECG or angina, consistent with findings in the general population (25–27). AER was associated with both MACE and, less strongly, soft CAD events, while low eGFR was associated with revascularization. This stronger relationship between AER and MACE is consistent with prior results demonstrating that nephropathy seems to be associated with “harder” clinical CVD events, especially CVD mortality (24,28). Inflammation, as measured by white blood cell count, was associated with increased risk of MACE and soft CAD. For MACE, a model with smoking in place of white blood cell count was nearly identical, though worse in fit, to the final model shown in Table 4, suggesting that the higher rate of smoking is a strong contributor to the association between white blood cell count and MACE. This stronger association between smoking and myocardial infarction compared to angina has previously been observed in the general population (29–31) and reiterates the need for smoking cessation to prevent greater CVD morbidity and mortality. Smoking was also a strong risk factor for MACE in the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) (32) and for both CAD and stroke in the Finnish Diabetic Nephropathy Study (FInnDiane) (33). Notably, use of lipid- or blood pressure-lowering medications were not independently associated with any of the CVD manifestations in this cohort. Compared to the risk factors for MACE reported in DCCT/EDIC, there are similarities in the results from EDC reported here, including systolic blood pressure, smoking, and HbA1c, though the association between HbA1c and MACE was much stronger in DCCT/EDIC. A key difference is the lack of an association between AER and MACE in DCCT/EDIC, likely due to the delay in developing microvascular disease in DCCT/EDIC as a result of intensive insulin therapy. On the other hand, nephropathy was a strong risk factor for both CAD and stroke in the FinnDiane study (34), similar to EDC. Risk factors associated with revascularization have not been reported in DCCT/EDIC or FinnDiane.
Notable trends differentiated those with revascularization as a first event. First, though not independently associated with outcomes, lipid-lowering medication use prior to first event was more common in revascularization (26%) compared to MACE (12%) or soft CAD (14%). Second, those with revascularization were more likely to have attained at least a bachelor’s degree, be professionally employed, and have higher household income. Professional employment has previously been associated with shorter time to angiography and revascularization in patients with angina (35). Third, those first presenting with revascularization were more likely to engage in leisure-time physical activity, even compared to participants who did not develop CVD, raising the possibility that exercise-induced angina led some to seek treatment. Finally, while the proportion with health insurance coverage was similarly high for all manifestations (>90%), the distribution of insurance plan type differed, with a higher proportion of participants with MACE (27%) covered by Medicare or Medicaid at time of event versus just 8% of those with revascularization and 12% with soft CAD. Altogether, these differences underlie the difficulty of composite CVD endpoints which include revascularization: revascularization is as much a medical decision resulting from a cluster of exogenous factors as it is the result of a pathophysiological process. In addition to the aforementioned factors, the guidelines for recommending coronary revascularization in patients with stable manifestations of coronary artery disease remain a source of debate, as the benefits of revascularization over medical therapy alone are unclear in these patients (36,37). As a recent example, an observational study examining predictors of revascularization versus medical therapy alone in non—ST-segment—elevation acute coronary syndromes, found that patients with high physician-assessed risk were more likely to undergo revascularization, while those with objectively greater comorbidity were less likely to undergo revascularization (38). In a systematic review of international studies in the general population, lower SES was consistently associated with lower access to coronary procedures, including revascularization (39). This association was observed even in countries with universal health coverage, though the disparity was greater in countries without universal coverage, including the United States. The authors also found that reduced access to coronary angiography in people with lower SES seemed to be a major factor contributing to lower rates of revascularization in this group. There are several potential underlying causes for this reduced access in people with lower SES, even in those with health care coverage, including financial burden associated with copays, greater distance to care providers and other transportation barriers, and lower health literacy and/or lack of knowledge of treatment options.
Our study has many strengths, including the long, prospective follow-up. The EDC cohort is well-characterized, with data available for many clinical risk factors, and has been shown to be epidemiologically representative of childhood onset type 1 diabetes (40,41). The CVD events were adjudicated using death certificates and medical records, when possible, by reviewers who were masked to risk factors status. The main limitations of the study include the long period of follow-up may not represent the type 1 diabetes experience of people who have been diagnosed more recently. However, as we have recently reported in a contemporary subgroup, high CVD risk persists in this cohort (1). Similarly, the availability of therapies for lipid- and blood pressure-lowering changed dramatically during the 25-year follow-up period. We incorporated time-varying covariates for lipid and blood pressure medication use in our models to account for changing treatment status during the course of the study, though these variables did not remain independently associated with any CVD manifestations in the final models.
Another limitation is the small number of revascularization/blockage as first CVD events. Additionally, our knowledge of what precipitated the revascularization is limited to what is noted in the hospital records. Thus we do not have detailed information on the severity of angina that may have led to the revascularization. However, we are able to provide some characterization of what led to the detection of the blockage and/or revascularization and, importantly, only one participant showed evidence that detection of coronary blockage was directly related to EDC study participation.
An additional limitation of our study is the lack of racial diversity (98% Caucasian), due to the demographics of Allegheny County, Pennsylvania (<15% African American), and lower type 1 diabetes incidence among African Americans during the period when the cohort was diagnosed. There is also a potential for ‘survivor bias’, particularly in the earlier years of the cohort (those diagnosed before 1965), because prevalent cases of CVD at baseline were excluded from these analyses of incident events. A further limitation is that assessment of HbA1c and lipids changed during the course of the study. While agreement between the methods was high, there remains the possibility that these changes could influence the time-varying analyses. Finally, socioeconomic status was assessed at a single time point, closest to age 28, to permit reasonable establishment of educational, occupational, and financial status, while minimizing the effect of advanced diabetes complications on income potential. In this cohort, the average diabetes duration was 20 years by age 28, underscoring the difficulty of assessing the effect of socioeconomic factors at older ages on CVD risk independently of other diabetes complications. However, it is possible that changes in socioeconomic status after this time point but prior to complication development could influence access to health care and thus CVD risk.
In conclusion, these results suggest that risk factors differ for the various first manifestations of CVD in this type 1 diabetes cohort. Additionally, when revascularization was added to MACE in a combined outcome, the role of LDLc was enhanced, while that of HbA1c was diminished, compared to examining MACE alone. These findings underscore the complexity associated with including revascularization in composite CVD endpoints. More research is needed to determine how to best incorporate coronary revascularization into composite CVD definitions for risk prediction.
Supplementary Material
Acknowledgements
Source of Funding
This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases at the National Institutes of Health (Grant No. R01-DK034818) and the Rossi Memorial Fund. R.G.M. is supported by American Diabetes Association Grant number 1-19-JDF-109.
Role of the Funding Sources
No funding source played a role in study design, collection, analysis, and interpretation of data, writing the report, or in the decision to submit the report for publication.
Abbreviations:
- CVD
cardiovascular disease
- MACE
major atherosclerotic coronary event
- CAD
Coronary artery disease
- EDC
Epidemiology of Diabetes Complications Study
- MI
myocardial infarction
- AER
albumin excretion rate
- eGFR
estimated glomerular filtration rate
- WBC
white blood cells
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Declaration of Competing Interest
Dr. Orchard is a consultant for Boehringer Ingelheim. Drs. Miller and Costacou have no relevant conflicts to report.
Data Statement
The data sets used in these analyses may be made available by the corresponding author upon reasonable request.
References
- 1.Miller RG, Mahajan HD, Costacou T, et al. A Contemporary Estimate of Total Mortality and Cardiovascular Disease Risk in Young Adults With Type 1 Diabetes: The Pittsburgh Epidemiology of Diabetes Complications Study. Diabetes Care. 2016; 39(12): 2296–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Livingstone SJ, Looker HC, Hothersall EJ, et al. Risk of cardiovascular disease and total mortality in adults with type 1 diabetes: Scottish registry linkage study. PLoS Med. 2012; 9(10): e1001321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Orchard TJ, Costacou T. When are type 1 diabetic patients at risk for cardiovascular disease? Curr Diab Rep 2010; 10(1): 48–54. [DOI] [PubMed] [Google Scholar]
- 4.Marx N, McGuire DK, Perkovic V, et al. Composite primary end points in cardiovascular outcomes trials involving type 2 diabetes patients: Should unstable angina be included in the primary end point? Diabetes Care. 2017; 40(9): 1144–51. [DOI] [PubMed] [Google Scholar]
- 5.Mannucci E, Mosenzon O, Avogaro A. Analyses of Results From Cardiovascular Safety Trials With DPP-4 Inhibitors: Cardiovascular Outcomes, Predefined Safety Outcomes, and Pooled Analysis and Meta-analysis. Diabetes care. 2016; 39(Suppl 2): S196–204. [DOI] [PubMed] [Google Scholar]
- 6.Allan GM, Nouri F, Korownyk C, et al. Agreement among cardiovascular disease risk calculators. Circulation. 2013; 127(19): 1948–56. [DOI] [PubMed] [Google Scholar]
- 7.Joffe HV, Parks MH, Temple R. Impact of cardiovascular outcomes on the development and approval of medications for the treatment of diabetes mellitus. Rev Endocr Metab Disord. 2010; 11(1): 21–30. [DOI] [PubMed] [Google Scholar]
- 8.Orchard TJ, Dorman JS, Maser RE, et al. Factors associated with avoidance of severe complications after 25 yr of IDDM. Pittsburgh Epidemiology of Diabetes Complications Study I. Diabetes Care. 1990; 13(7): 741–7. [DOI] [PubMed] [Google Scholar]
- 9.Orchard TJ, Dorman JS, Maser RE, et al. Prevalence of complications in IDDM by sex and duration. Pittsburgh Epidemiology of Diabetes Complications Study II. Diabetes. 1990; 39(9): 1116–24. [DOI] [PubMed] [Google Scholar]
- 10.Diabetes Epidemiology Research International Mortality Study Group. Major cross-country differences in risk of dying for people with IDDM. Diabetes Care. 1991; 14(1): 49–54. [DOI] [PubMed] [Google Scholar]
- 11.Prince CT, Becker DJ, Costacou T, et al. Changes in glycaemic control and risk of coronary artery disease in type 1 diabetes mellitus: findings from the Pittsburgh Epidemiology of Diabetes Complications Study (EDC). Diabetologia. 2007; 50(11): 2280–8. [DOI] [PubMed] [Google Scholar]
- 12.Bucolo G, David H. Quantitative determination of serum triglycerides by the use of enzymes. Clin Chem. 1973; 19(5): 476–82. [PubMed] [Google Scholar]
- 13.Allain CC, Poon LS, Chan CSG, et al. Enzymatic determination of total serum cholesterol. Lipids. 1974; 20(4): 470–5. [PubMed] [Google Scholar]
- 14.Warnick GR, Albers JJ. Heparin--Mn2+ quantitation of high-density-lipoprotein cholesterol: an ultrafiltration procedure for lipemic samples. Clin Chem. 1978; 24(6): 900–4. [PubMed] [Google Scholar]
- 15.Lipid Research Clinics Program: Manual of laboratory operations, Vol. 1: Lipid and liproprotein analysis. Washington, DC: National Institutes of Health, Department of Health, US Govt. Printing Office; 1974. [Google Scholar]
- 16.The hypertension detection and follow-up program. Hypertension detetection and follow-up program cooperative group. Prev Med (Baltim). 1976; 5: 207–15. [DOI] [PubMed] [Google Scholar]
- 17.Ellis D, Coonrod BA, Dorman JS, et al. Choice of urine sample predictive of microalbuminuria in patients with insulin-dependent diabetes mellitus. Am J kidney Dis Off J Natl Kidney Found. 1989; 13(4): 321–8. [DOI] [PubMed] [Google Scholar]
- 18.Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009; 150(9): 604–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Secrest AM, Costacou T, Gutelius B, et al. Associations Between Socioeconomic Status and Major Complications in Type 1 Diabetes: The Pittsburgh Epidemiology of Diabetes Complication (EDC) Study. Ann Epidemiol. 2011; 21(5). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kannel WB, Feinleib M, McNamara PM, et al. An investigation of coronary heart disease in families. The Framingham offspring study. Am J Epidemiol. 1979; 110(3): 281–90. [DOI] [PubMed] [Google Scholar]
- 21.Cirino P, Chin C, Sevcik R, et al. Measuring socioeconomic status: Reliability and preliminary validity for different approaches. Assessment. 2002; 9(2): 145–55. [DOI] [PubMed] [Google Scholar]
- 22.Paffenbarger R, Wing A, Hyde R. Physical activity as an index of heart attack risk in college alumni. Am J Epidemiol. 1978; 108: 161–75. [DOI] [PubMed] [Google Scholar]
- 23.Guo J, Miller RG, Costacou T, et al. Left ventricular systolic dysfunction predicts long-term major microvascular complication outcomes in type 1 diabetes. The Pittsburgh Epidemiology of Diabetes Complications (EDC) study of childhood onset diabetes. J Diabetes Complications. 2018; 32(3): 298–304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Orchard TJ, Olsen JC, Erbey JR, et al. Insulin Resistance – Related Factors , but not Glycemia , Predict Coronary Artery Disease in Type 1 Diabetes. Diabetes Care 2003; 26(5): 1374–9. [DOI] [PubMed] [Google Scholar]
- 25.Kannel WB, Feinleib M. Natural history of angina pectoris in the Framingham study. Prognosis and survival. Am J Cardiol. 1972; 29(2): 154–63. [DOI] [PubMed] [Google Scholar]
- 26.Lerner DJ, Kannel WB. Patterns of coronary heart disease morbidity and mortality in the sexes: a 26-year follow-up of the Framingham population. Am Heart J. 1986; 111(2): 383–90. [DOI] [PubMed] [Google Scholar]
- 27.Hemingway H, Langenberg C, Damant J, et al. Prevalence of Angina in Women Versus Men. Circulation. 2008; 117(12): 1526–36. [DOI] [PubMed] [Google Scholar]
- 28.Miller RG, Anderson SJ, Costacou T, et al. Risk stratification for 25-year cardiovascular disease incidence in type 1 diabetes: Tree-structured survival analysis of the Pittsburgh Epidemiology of Diabetes Complications study. Diabetes Vasc Dis Res. 2016; 13(4): 250–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hsia J, Aragaki A, Bloch M, et al. Predictors of angina pectoris versus myocardial infarction from the Women’s Health Initiative Observational Study. Am J Cardiol. 2004; 93(6): 673–78. [DOI] [PubMed] [Google Scholar]
- 30.Dudas K, Björck L, Jernberg T, et al. Differences between acute myocardial infarction and unstable angina: a longitudinal cohort study reporting findings from the Register of Information and Knowledge about Swedish Heart Intensive Care Admissions (RIKS-HIA). BMJ Open. 2013; 3(1): e002155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Pujades-Rodriguez M, George J, Shah AD, et al. Heterogeneous associations between smoking and a wide range of initial presentations of cardiovascular disease in 1 937 360 people in England: lifetime risks and implications for risk prediction. Int J Epidemiol. 2015; 44(1): 129–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Research Group. Risk factors for cardiovascular disease in type 1 diabetes. Diabetes. 2016; 65(5): 1370–1379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Feodoroff M, Harjutsalo V, Forsblom C, Groop PH, FinnDiane Study Group. Dose-dependent effect of smoking on risk of coronary heart disease, heart failure and stroke in individuals with type 1 diabetes. Diabetologia. 2018; 61(12): 2580–2589. [DOI] [PubMed] [Google Scholar]
- 34.Harjutsalo V, Thomas MC, Forsblom C, Groop PH, FinnDiane Study Group. Risk of coronary artery disease and stroke according to sex and presence of diabetic nephropathy in type 1 diabetes. Diabetes Obes Metab. 2018; 20(12): 2759–2767. [DOI] [PubMed] [Google Scholar]
- 35.Gaffney B, Kee F. Are the economically active more deserving? Br Heart J. 1995; 73(4): 385–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Stone GW, Hochman JS, Williams DO, et al. Medical Therapy With Versus Without Revascularization in Stable Patients With Moderate and Severe Ischemia. J Am Coll Cardiol. 2016; 67(1): 81–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Hamon M, Lemesle G, Meurice T, et al. Elective Coronary Revascularization Procedures in Patients With Stable Coronary Artery Disease: Incidence, Determinants, and Outcome (From the CORONOR Study). JACC Cardiovasc Interv. 2018; 11(9): 868–75. [DOI] [PubMed] [Google Scholar]
- 38.Wijeysundera HC, Sidhu MS, Bennell MC, et al. Predictors of initial revascularization versus medical therapy alone in patients with Non-ST-segment-elevation acute coronary syndrome undergoing an invasive strategy. Circ Cardiovasc Interv. 2016; 9(7): e003592. [DOI] [PubMed] [Google Scholar]
- 39.Schrӧder S, Richter M, Schrӧder J, Frantz S, Fink A. Socioeconomic inequalities in access to treatment for coronary heart disease: A systematic review. Int J Cardiol. 2016; 219: 70–78. [DOI] [PubMed] [Google Scholar]
- 40.Wagener DK, Sacks JM, LaPorte RE, Macgregor JM. The Pittsburgh study of insulin-dependent diabetes mellitus. Risk for diabetes among relatives of IDDM. Diabetes. 1982; 31(2): 136–44. [DOI] [PubMed] [Google Scholar]
- 41.Miller RG, Secrest AM, Sharma RK, et al. Improvements in the life expectancy of type 1 diabetes: the Pittsburgh Epidemiology of Diabetes Complications study cohort. Diabetes. 2012; 61(11): 2987–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.