Trends in Cardiovascular Disease Risk Factors in Individuals With and Without Diabetes Mellitus in the Framingham Heart Study
Background— Individuals with diabetes mellitus are at 2- to 3-fold increased risk for cardiovascular disease (CVD) relative to those without diabetes. Our objective was to examine CVD risk factor level changes among individuals with and without type 2 diabetes mellitus from 1970 to 2005 in the Framingham Heart Study.
Methods and Results— We included 4195 participants (3990 with no diabetes and 205 with diabetes) 50 years of age and 3495 participants (3178 with no diabetes and 317 with diabetes) 60 years of age. Contemporaneous CVD risk factor levels were measured; linear regression models were used to assess the interaction between diabetes status and calendar year on CVD risk factor levels. Among 50-year-olds without diabetes mellitus, there was an increase in body mass index of 0.39 kg/m2 per 10 years, whereas for those with diabetes, there was an increase of 2.52 kg/m2 (P value for the diabetes-by–calendar year interaction [P for interaction] <0.001). For low-density lipoprotein cholesterol, the mean decrease was −7.43 mg/dL per decade (nondiabetes) and −15.5 mg/dL for diabetes (P for interaction=0.002). For systolic blood pressure, the mean decrease was −3.35 mm Hg per decade (nondiabetes) and −3.50 mm Hg for diabetes (P for interaction=0.97). The direction of the trends for those with diabetes relative to those without diabetes was similar for 60-year-olds.
Conclusions— Compared with individuals without diabetes mellitus, individuals with diabetes experienced a greater increase in body mass index, a greater decrease in low-density lipoprotein cholesterol, and a similar magnitude of decline in systolic blood pressure. Individuals with diabetes mellitus have not experienced the necessary declines in CVD risk factors to overcome their increased risk of CVD. Further efforts are needed to aggressively control CVD risk factors among individuals with diabetes mellitus.
Received December 21, 2008; accepted May 26, 2009.
Cardiovascular disease (CVD) risk factor control is critical to reducing the CVD risk of those with diabetes mellitus. Recent clinical trials have demonstrated that the use of more aggressive targets for blood pressure and cholesterol control among individuals with diabetes results in a reduced incidence of CVD events.1–3 Despite this, data from multiple sources suggest that those with diabetes do not experience optimal risk factor control compared with those without diabetes mellitus.4–6 Although individuals with diabetes have experienced substantial declines in dyslipidemia, blood pressure, and cigarette smoking,7 the experience relative to individuals without diabetes has not been well described. Therefore, individuals with diabetes may not have achieved the level of risk factor control needed to reduce their burden of CVD morbidity and mortality compared with that of their nondiabetic counterparts.
Clinical Perspective on p 220
The primary goal of the present analysis was to characterize the trends in body mass index (BMI), blood pressure, cholesterol, and smoking among individuals with and without type 2 diabetes mellitus from 1970 through 2005 in the Framingham Heart Study. We also assessed the prevalence of treatment and control of CVD risk factors among individuals with and without type 2 diabetes mellitus.
The design of the Framingham Heart Study Original, Offspring, and Third Generation cohorts has been described previously.8–11 Briefly, the Framingham Heart Study Original cohort began in 1948 with the enrollment of 5209 individuals 28 to 62 years of age who underwent biennial medical examinations.8,9 In 1971, 5124 offspring of the original participants and their spouses were enrolled into the Offspring cohort and underwent examinations approximately every 4 years.10 In 2002, 4095 children of the offspring participants and their spouses were enrolled in the Third Generation Cohort.11 As of the present analysis, the Third Generation participants have only attended their first examination.
Our aim was to examine the experience of 50- and 60-year-old individuals with and without diabetes over the time period from 1970 to 2005. We chose this design to describe the experience of a population at a given age over time rather than the aging process of the population, with the goal of examining the changes in risk factor levels cross-sectionally rather than examining changes within a given person. Individuals 50 and 60 years of age were chosen because those ages provided sufficient sample sizes across each decade to allow for meaningful comparisons. The study was approved by the Institutional Review Board at Boston Medical Center, and written informed consent was provided by all participants. For information on diabetes status assessment and risk factor assessment and definition, please see the online-only Data Supplement.
For all continuous risk factors (BMI, total cholesterol, low-density lipoprotein [LDL] cholesterol, systolic blood pressure, and diastolic blood pressure), linear regression models were fit to determine the association between calendar year and the risk factor level. The β-coefficients from this model represent the interpretation of the change in risk factor level per calendar year. For ease of interpretation, we multiplied the β-coefficients from this model by 10 so that they represent the interpretation of the change in risk factor level per 10 years. Models were fit separately for those with and without diabetes and by age group (50 and 60 years). The interaction between diabetes status and calendar year was assessed in the total data set by the inclusion of a cross-product term in the model, and its significance was assessed with a likelihood ratio test. All models were adjusted for sex. For dichotomous characteristics, the prevalence of each risk factor was calculated for each of the 3.5 decades (35 years) of study follow-up (1970 to 1979, 1980 to 1989, 1990 to 1999, and 2000 to 2005).
In secondary analyses, we further adjusted the linear regression models for BMI and repeated the analysis among those free of prevalent CVD (myocardial infarction, stroke, or coronary heart failure). Because we were also interested in the characteristics of individuals with diabetes over time, we secondarily examined mean fasting blood glucose and prevalence of diabetes treatment over time. Because the original cohort did not have fasting blood glucose measurements available, this analysis was restricted to participants in the Offspring and Third Generation cohorts. All variables were assessed for normality with box plots, and no violations were found. A P value <0.05 was considered statistically significant. All analyses were performed with SAS version 9.1 (Cary, NC).
Drs Preis and Fox had full access to the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study Sample Characteristics
The mean CVD risk factor levels by decade are presented in Table 1.
Change in CVD Risk Factor Levels Over Time
Table 2 shows the difference in CVD risk factor levels per decade in individuals with and without diabetes. Among individuals 50 years of age, those without diabetes had a BMI of 26.4 kg/m2 in the 1970s and a mean BMI of 27.4 kg/m2 in the 2000s, which represents a mean adjusted increase in BMI per decade of 0.39 kg/m2. In comparison, those with diabetes had a mean BMI of 28.3 kg/m2 in the 1970s and a mean BMI of 35.7 kg/m2 in the 2000s, which represents a mean increase of 2.52 kg/m2 per decade. The interaction between calendar year and diabetes status on BMI was statistically significant (age 50 years: P for interaction <0.001; Table 2; Figure, A).
Those with diabetes had a larger decline in both total and LDL cholesterol than those without diabetes (Table 2; Figure, B). For example, among those without diabetes, mean LDL cholesterol declined from 141 to 119 mg/dL over the 3.5 decades of observation, which represents a mean adjusted decrease of 7.43 mg/dL per decade. Among those with diabetes, LDL cholesterol declined from 161 to 111 mg/dL, which corresponds to a mean adjusted rate of decline of 15.5 mg/dL per decade. The interaction between diabetes status and calendar time on LDL levels was statistically significant (age 50 years: P for interaction=0.002).
The magnitude of decline in systolic and diastolic blood pressures was similar between those with and without diabetes (Table 2; Figure, C). For example, mean systolic blood pressure declined from 130 to 121 mm Hg (adjusted mean decline of 3.35 mm Hg per decade) among those without diabetes and from 141 to 129 mm Hg (adjusted mean decline of 3.50 mm Hg per decade) among those with diabetes. The corresponding decline per decade for diastolic blood pressure was 2.28 mm Hg among individuals without diabetes and 2.51 mm Hg among individuals with diabetes. The interaction between diabetes status and calendar time was not statistically significant for systolic blood pressure (age 50 years: P for interaction=0.97) or for diastolic blood pressure (age 50 years: P for interaction=0.78).
The direction of the trends for those with diabetes relative to those without diabetes was similar for 60-year-olds. All analyses presented in Table 2 were repeated with further adjustment for BMI, and the results were not materially different (online-only Data Supplement Table I).
Prevalence, Treatment, and Control of CVD Risk Factors
Overall, for those individuals who were 50 years old, the prevalence of hypertension declined among those without diabetes and did not change among those with diabetes (Table 3); however, the prevalence of hypertension treatment increased from 21.2% to 53.1% among those without diabetes and from 25.9% to 63.0% among those with diabetes over the period from 1970 to 1979 to the period from 2000 to 2005. Similarly, the prevalence of hypertension control increased among those without diabetes from 7.8% to 36.0% (Ptrend<0.001). Similar trends were observed in those with diabetes as well, although the probability value for trend was not significant (P=0.06). Surprisingly, only about one third of those without diabetes and 15% of those with diabetes experienced hypertension control in the 2000s.
The prevalence of high LDL cholesterol declined among those without diabetes and remained stable among those with diabetes. The proportion of individuals treated and controlled for high LDL cholesterol increased among both those with and without diabetes, although only about 20% of either those with or without diabetes experienced LDL cholesterol control in 2000 to 2005.
Among the 50-year-olds, the prevalence of obesity increased markedly, from 16.7% to 23.5% among those without diabetes and from 36.4% to 61.8% among those with diabetes (Table 3). The prevalence of cigarette smoking decreased from 43.8% to 19.2% among those without diabetes and from 58.1% to 17.1% among those with diabetes.
Overall, the results were similar for the 60-year-old group compared with the 50-year-old group. From the period 1970 to 1979 to the period 2000 to 2005, hypertension prevalence declined among those without diabetes and did not change among those with diabetes. Treatment and control of hypertension increased among both those with and without diabetes. The prevalence of high LDL cholesterol decreased among those without diabetes and did not change among those with diabetes. The proportion of individuals treated and achieving LDL control increased in both groups. Cigarette smoking declined among those with and without diabetes, and obesity increased in both groups.
Table 4 shows the glycemia-related characteristics of individuals with diabetes from the Offspring and Third Generation cohorts. Among individuals 50 years of age, mean fasting blood glucose did not decline appreciably, and diabetes treatment increased from 40.9% to 71.4% (Ptrend=0.004) from 1970–1979 to 2000–2005. Among the 60-year-olds, mean fasting blood glucose decreased (Ptrend<0.001). The prevalence of treatment increased, although the difference was not statistically significant.
We repeated our main analysis excluding individuals with prevalent CVD (online-only Data Supplement Table II). Overall, the results were very similar to those from the main analysis presented in Table 2. Additionally, we repeated our main analysis using the generalized estimating equations procedure to account for familial correlations, and found that the results were essentially unchanged (data not shown).
Mean BMI levels have increased and smoking prevalence and levels of cholesterol and blood pressure have decreased among individuals with and without diabetes. Individuals with diabetes had a greater increase in BMI and a greater decrease in cholesterol than those without diabetes; however, individuals with diabetes had a similar magnitude of decline for blood pressure as those without diabetes. The prevalence of treatment and control of hypertension and high LDL cholesterol increased from 1970 to 2005 among both those with and without diabetes; glucose-lowering treatment among individuals with diabetes has increased markedly, with concomitant declines in mean glucose. Finally, the prevalence of control of both hypertension and LDL cholesterol remains suboptimal for both individuals with and without diabetes.
In the Context of the Current Literature
To further reduce the differences in CVD rates between those with and without diabetes, efforts must be intensified to achieve guideline targets for systolic blood pressure and cholesterol levels among those with diabetes. In this context, the present results reveal parallel declines of blood pressure among those with and without diabetes. A study using data from the National Health and Nutrition Examination Survey showed a 27% decline in hypertension prevalence among those with diabetes over the period from 1971 to 2000.7 In the present study, the proportion of individuals with diabetes who achieved hypertension control was low, even in more recent years, and individuals with diabetes were less likely to achieve hypertension control than those without diabetes. This may be explained in part by the fact that we used a lower threshold (<130/80 mm Hg) for hypertension control among individuals with diabetes, which is harder to achieve. In addition, a prior study indicated that only one third of individuals with diabetes and hypertension were able to meet their blood pressure goals, even after behavioral management and pharmacotherapy.13 Furthermore, a recent study of hypertension trends in the US population from 1988 to 2000 showed that among individuals with diabetes, the proportion of individuals who achieved hypertension control did not change significantly between 1988 and 2000.14
In contrast to trends in hypertension, we observed steeper declines for cholesterol levels among those with diabetes compared with those without diabetes. This is likely owing to the increased use of lipid-lowering drugs, as evidenced by the increase in treatment and control of LDL cholesterol. Before the existence of statins, there were few effective lipid-lowering therapies available; however, the results from the present study show that the proportion of individuals achieving control, particularly among those with diabetes, is still relatively low. The present results are similar to data from the National Health and Nutrition Examination Survey, which showed a prevalence of hypercholesterolemia control in 1999 to 2000 of 22% for individuals with diabetes and 6% for individuals without diabetes.15 A study of health maintenance organization patients showed that 61% of individuals with diabetes did not reach their LDL cholesterol goal of <100 mg/dL after 6 months of lipid treatment.16
The present study demonstrated a dramatic increase in the prevalence of obesity among individuals with diabetes relative to those without diabetes, which has also been observed in a prior study of Spanish individuals.17 We also observed a decline in the prevalence of smoking among both individuals with and those without diabetes, which has been documented in prior studies.17
In the present study, individuals with diabetes had greater declines in total and LDL cholesterol than those without diabetes and similar declines in blood pressure levels. The improvement in risk factor levels was paralleled by an increase in drug treatment, which likely contributed to the observed trends. Given that diabetes mellitus is still associated with approximately an overall 2- to 3-fold increased risk of CVD mortality,18–20 these findings highlight the need for further improvements in blood pressure and cholesterol control among individuals with diabetes. Previous work in the Framingham Heart Study has demonstrated the increasing burden of CVD due to type 2 diabetes mellitus.21 Lifestyle modifications that result in weight loss have also been shown to result in an improved cardiometabolic profile among individuals with diabetes; however, long-term weight maintenance is often not successful.22,23
Recent clinical trials have focused on intensive blood glucose control as a way of reducing the morbidity and mortality associated with type 2 diabetes mellitus24–27; however, the benefits of intensive control of glucose levels on major CVD end points were not observed in 3 recently published major trials.24,25,27 In addition, in the ACCORD trial (Action to Control Cardiovascular Risk in Diabetes), it has been shown that intensive blood glucose control is associated with increased mortality compared with more standard blood glucose control regimens.24,25
Although improved blood glucose control may have benefits with respect to microvascular complications, intensified efforts to control other major CVD risk factors are currently the most effective way to reduce the risk of CVD among patients with diabetes.28,29 In this context, it is notable that we observed a steep increase in the use of glucose-lowering medications that outweighed the pharmacological management for elevated LDL cholesterol. Trials have shown repeatedly that reduction of cholesterol and blood pressure in individuals with diabetes results in reduced rates of CVD.1,28,30–32 Several trials of statin use among individuals with diabetes have shown that lipid lowering, particularly via intensive regimens, is associated with a reduction in CVD events.1,30,31 Trials of intensive blood pressure lowering among individuals with diabetes have demonstrated reductions in both microvascular and macrovascular events.2,28
Strengths and Limitations
One of the major strengths of the present study is that we were able to assess temporal trends over a 35-year period from 1970 to 2005. Additionally, data on diabetes status and CVD risk factor levels were collected rigorously at periodic examinations. One of the limitations of the present study is that the study sample was predominantly white; thus, our results may not be completely generalizable to other ethnic or racial groups. For some of the categorical characteristics, we had a relatively small sample size. We had relatively few data points in the 2000 to 2005 time period relative to the other periods; this may have affected the precision of the results. We did not have a sufficient sample size to examine the data separately by sex. Additionally, because the participants in the present study underwent routine screening for diabetes mellitus and other CVD risk factors, the improvements in CVD risk factors observed in the present study may represent an optimistic picture relative to the general US population.
We evaluated treatment and control relative to current guidelines that were not in place when much of the clinical data were collected. It is possible that the results may be explained by the emergence of new diabetes treatments and the development of new treatment guidelines over time. Lastly, because we wanted to apply a uniform definition of diabetes mellitus across all of the time periods, we used a fasting blood glucose cutoff of 126 mg/dL, which was not adopted until 1998. Thus, it is likely that the low proportion of individuals in the earlier decades who were untreated for diabetes mellitus may be explained by the fact that they were simply not diagnosed at the time because of the different criteria that were in place. It is also possible that they received their diabetes diagnosis at their examination visit and thus may not have had the opportunity to be treated.
Individuals 50 and 60 years old with diabetes have had declines in major CVD risk factors over the past several decades; however, these improvements have not been sufficient to overcome their CVD risk factor burden compared with those without diabetes. Further efforts are needed to aggressively control CVD risk factors among individuals with diabetes.
Sources of Funding
This work was supported by the National Heart, Lung, and Blood Institute’s Framingham Heart Study (contract No. N01-HC-25195).
Shepherd J, Barter P, Carmena R, Deedwania P, Fruchart JC, Haffner S, Hsia J, Breazna A, LaRosa J, Grundy S, Waters D. Effect of lowering LDL cholesterol substantially below currently recommended levels in patients with coronary heart disease and diabetes: the Treating to New Targets (TNT) study. Diabetes Care. 2006; 29: 1220–1226.
Hansson L, Zanchetti A, Carruthers SG, Dahlof B, Elmfeldt D, Julius S, Menard J, Rahn KH, Wedel H, Westerling S. Effects of intensive blood-pressure lowering and low-dose aspirin in patients with hypertension: principal results of the Hypertension Optimal Treatment (HOT) randomised trial: HOT Study Group. Lancet. 1998; 351: 1755–1762.
Imperatore G, Cadwell BL, Geiss L, Saadinne JB, Williams DE, Ford ES, Thompson TJ, Narayan KM, Gregg EW. Thirty-year trends in cardiovascular risk factor levels among US adults with diabetes: National Health and Nutrition Examination Surveys, 1971–2000. Am J Epidemiol. 2004; 160: 531–539.
Kannel WB, Feinleib M, McNamara PM, Garrison RJ, Castelli WP. An investigation of coronary heart disease in families: the Framingham offspring study. Am J Epidemiol. 1979; 110: 281–290.
Splansky GL, Corey D, Yang Q, Atwood LD, Cupples LA, Benjamin EJ, D'Agostino RBSr, Fox CS, Larson MG, Murabito JM, O'Donnell CJ, Vasan RS, Wolf PA, Levy D. The Third Generation Cohort of the National Heart, Lung, and Blood Institute’s Framingham Heart Study: design, recruitment, and initial examination. Am J Epidemiol. 2007; 165: 1328–1335.
American Diabetes Association. Standards of medical care in diabetes–2008. Diabetes Care. 2008; 31 (suppl 1): S12–S54.
Ford ES, Li C, Pearson WS, Zhao G, Mokdad AH. Trends in hypercholesterolemia, treatment and control among United States adults. Int J Cardiol. Published online before print December 9, 2008. doi:10.1016/j.ijcard.2008.11.033.
Jimenez-Garcia R, Hernandez-Barrera V, Jimenez-Trujillo I, Garrido PC, Lopez de AA, Gil de MA. Trends in cardiovascular risk factors and lifestyle behaviors among Spanish adults with diabetes (1993–2003). J Diabetes Complications. 2008. Published online before print August 4, 2008. doi:10.1016/j.jdiacomp.2008.06.002.
Fox CS, Coady S, Sorlie PD, D'Agostino RBSr, Pencina MJ, Vasan RS, Meigs JB, Levy D, Savage PJ. Increasing cardiovascular disease burden due to diabetes mellitus: the Framingham Heart Study. Circulation. 2007; 115: 1544–1550.
Pi-Sunyer X, Blackburn G, Brancati FL, Bray GA, Bright R, Clark JM, Curtis JM, Espeland MA, Foreyt JP, Graves K, Haffner SM, Harrison B, Hill JO, Horton ES, Jakicic J, Jeffery RW, Johnson KC, Kahn S, Kelley DE, Kitabchi AE, Knowler WC, Lewis CE, Maschak-Carey BJ, Montgomery B, Nathan DM, Patricio J, Peters A, Redmon JB, Reeves RS, Ryan DH, Safford M, Van DB, Wadden TA, Wagenknecht L, Wesche-Thobaben J, Wing RR, Yanovski SZ. Reduction in weight and cardiovascular disease risk factors in individuals with type 2 diabetes: one-year results of the look AHEAD trial. Diabetes Care. 2007; 30: 1374–1383.
Patel A, MacMahon S, Chalmers J, Neal B, Billot L, Woodward M, Marre M, Cooper M, Glasziou P, Grobbee D, Hamet P, Harrap S, Heller S, Liu L, Mancia G, Mogensen CE, Pan C, Poulter N, Rodgers A, Williams B, Bompoint S, de Galan BE, Joshi R, Travert F. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med. 2008; 358: 2560–2572.
Duckworth W, Abraira C, Moritz T, Reda D, Emanuele N, Reaven PD, Zieve FJ, Marks J, Davis SN, Hayward R, Warren SR, Goldman S, McCarren M, Vitek ME, Henderson WG, Huang GD. Glucose control and vascular complications in veterans with type 2 diabetes. N Engl J Med. 2009; 360: 129–139.
UK Prospective Diabetes Study Group. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. BMJ. 1998; 317: 703–713.
Skyler JS, Bergenstal R, Bonow RO, Buse J, Deedwania P, Gale EAM, Howard BV, Kirkman S, Kosiborod M, Reaven P, Sherwin RS. Intensive glycemic control and the prevention of cardiovascular events: implications of the ACCORD, ADVANCE, and VA diabetes trials: a position statement of the American Diabetes Association and a Scientific Statement of the American College of Cardiology Foundation and the American Heart Association. Circulation. 2009; 119: 351–357.
Colhoun HM, Betteridge DJ, Durrington PN, Hitman GA, Neil HA, Livingstone SJ, Thomason MJ, Mackness MI, Charlton-Menys V, Fuller JH. Primary prevention of cardiovascular disease with atorvastatin in type 2 diabetes in the Collaborative Atorvastatin Diabetes Study (CARDS): multicentre randomised placebo-controlled trial. Lancet. 2004; 364: 685–696.
Adler AI, Stratton IM, Neil HA, Yudkin JS, Matthews DR, Cull CA, Wright AD, Turner RC, Holman RR. Association of systolic blood pressure with macrovascular and microvascular complications of type 2 diabetes (UKPDS 36): prospective observational study. BMJ. 2000; 321: 412–419.
Cardiovascular disease (CVD) risk factor control is critical to reducing the CVD risk of those with diabetes mellitus. Recent clinical trials have demonstrated that the use of more aggressive targets for blood pressure and cholesterol control among individuals with diabetes mellitus results in a reduced incidence of CVD events. Despite this, data from multiple sources suggest that those with diabetes do not experience optimal risk factor control compared with those without diabetes. Although individuals with diabetes have experienced substantial declines in CVD risk factors, the experience relative to individuals without diabetes has not been well described. The primary goal of this analysis was to characterize the trends in body mass index, blood pressure, cholesterol, and smoking among individuals with and without type 2 diabetes mellitus from 1970 through 2005 in the Framingham Heart Study. Our results showed that compared with those without diabetes mellitus, individuals with diabetes experienced a greater increase in body mass index, a greater decrease in low-density lipoprotein cholesterol, and a similar magnitude of decline in systolic blood pressure. Smoking prevalence has declined among both those with and those without diabetes mellitus. Prevalence of control of both hypertension and low-density lipoprotein cholesterol remains suboptimal for both individuals with and those without diabetes. Overall, individuals with diabetes have not experienced the necessary declines in CVD risk factors to overcome their increased risk of CVD. Further efforts are needed to aggressively control CVD risk factors among individuals with diabetes mellitus.
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/cgi/content/full/CIRCULATIONAHA.108.846519/DC1.
Guest Editor for this article was Gregory Y.H. Lip, MD.