University Track Graduate Writing Research Proposal Assignment For this assignment, you will create a 2-3 page research proposal on a topic related to your future major/area of study that is of interest to you. In this proposal, you will be expected to discuss the specific area of research you will be conducting for your future assignments in this class (annotated bibliography and final paper). You cannot write your paper on a topic that has not been approved, so this research proposal will allow you to discuss your topic in detail with your teacher, and this will also help inform your stance on the topic. Research proposals are a common aspect of graduate classes (and academic careers, if you choose to continue in this area). This research proposal follows a very basic format that is shorter and much simpler than research proposals you may have to do in the future, but this will allow you to get a good start on the process now. Please format this proposal using 12-point Times New Roman, double-spaced, with 1-inch margins. This assignment will be due in the D2L Assignments folder before class on Monday, July 9. Research Proposal Template I. II. III. IV. Proposed research topic a. Background on the topic b. Current controversies or academic debates/differing viewpoints related to the topic c. Why it interests you and is appropriate to your field of study Research questions a. Two to four research questions based on your current understanding of the topic; this will inform your selection of articles b. Questions should be focused, specific, and attainable (you should be able to find the answers, or at least come close, through completing this assignment) Research strategy a. What assumptions do I have about this topic that are informing my stance and research framework? b. Where will you find your sources? c. How will you determine which sources are appropriate? d. How will you evaluate your sources once you find them? Sources a. What sources do you already have that are informing your opinion? (Aim to have at least three for the proposal (you will need 5-7 for the final paper) Example Research Proposal 1. Proposed research topic: Architecture and water conservation a. Water conservation is a major issue in dry areas such as Tucson, and much of the water that falls on this area is not collected appropriately. Within the new focus on sustainable architecture, there have been advances in the building of homes and office buildings that now collect rainwater more efficiently, leading to better water conservation (Campana et al, 2017). In recent years, this type of architecture has moved from the fringe to the mainstream, with modern office buildings now often being designed to not only save energy but also to collect rainwater, which may be used for gardens, landscaping, or industrial uses (Campisano et al, 2017; Xu, Xiao, & Wei, 2016). b. The main controversies related to this topic surround not whether architecture should strive to be more sustainable, but how architects can work sustainable plans into their designs in a cost-effective way that will make sustainable architecture more attainable for everyone, including people who are building homes or offices on a tight budget. The considerations must include cost, efficiency, and sanitation. c. This topic is very interesting to me because I care about protecting the environment, and as an architect, I would like to incorporate this concern into my designs for my future work. Sustainable architecture is becoming more and more popular in my field around the world, and having a good understanding of ways to be sustainable in my designs will help me adapt to potential client concerns as well as follow my dreams of protecting the environment in a larger way. 2. Research questions a. What forms of passive water collection (use of berms, swales, and other landscaping features) are most effective and efficient in the environment of Southern Arizona? b. What forms of active water collection (use of cisterns, etc.) are most effective for small to medium homes in the environment of Southern Arizona, and how can these water collection methods be incorporated into initial home design? 3. Research strategy a. Assumptions: My initial assumption about this topic is regarding the fact that water conservation is: 1) a possible and beneficial aspect of architecture; and 2) that potential clients will be interested in learning more about how they can incorporate sustainable facets such as passive and active water collection methods into their small to medium homes or offices. b. Sources: I plan to find most of my sources in online academic journals related to architecture and sustainability, which are available through the University of Arizona library website. I may also use one or two popular sources, such as articles from reputable newspapers or magazines, to provide a popular insight into the topic. c. For the academic articles, I will need to ensure that the articles address the research questions I have listed above. If the articles do not address these questions directly, I will only use them if they address some aspect of these questions that is immediately relevant to my research topic. I will also need to evaluate each of these sources and provide critiques of them, including critiques of their methodology, theoretical framework, or basic assumptions. If I use any popular sources, I will need to check the reputation of the sources and understand any bias that may be present in their writing. 4. Sources Campana, P. E., Quan, S. J., Robbio, F. I., Lundblad, A., Zhang, Y., Ma, T., & ... Yan, J. (2017). Optimization of a residential district with special consideration on energy and water reliability. Applied Energy, 194751-764. doi:10.1016/j.apenergy.2016.10.005 Campisano, A., Butler, D., Ward, S., Burns, M. J., Friedler, E., DeBusk, K., & ... Han, M. (2017). Urban rainwater harvesting systems: Research, implementation and future perspectives. Water Research, 115195-209. doi:10.1016/j.watres.2017.02.056 Xu, J., Xiao, D., & Wei, C. (2016). Architecture Water-saving Design and Building Technique of Sponge Yard. Journal Of Landscape Research, 8(6), 16-19. doi:10.16785/jissn1943-989x.2016.6.005 Nutrition, Metabolism & Cardiovascular Diseases (2013) 23, 122e129 Available online at www.sciencedirect.com journal homepage: www.elsevier.com/locate/nmcd Impact of overweight and obesity on cardiac benefit of antihypertensive treatment E. Gerdts a,*, G. de Simone b, B.P. Lund a, P.M. Okin c, K. Wachtell d, K. Boman e, M.S. Nieminen f, B. Dahlöf g, R.B. Devereux c a Institute of Medicine, University of Bergen, and Department of Heart Disease, Haukeland University Hospital, N-5021, Bergen, Norway b Department of Clinical and Experimental Medicine, Federico II University Hospital, Naples, Italy c Division of Cardiology, Weill Medical College of Cornell University, NY, USA d Department of Cardiology, Gentofte University Hospital, Copenhagen, Denmark e Department of Medicine Skeleftaa Hospital, University of Umeaa Skelleftaa, Sweden f Department of Medicine, Helsinki University Central Hospital, Helsinki, Finland g Department of Medicine, Sahlgrenska University Hospital e Östra Gothenburg, Sweden Received 9 November 2010; received in revised form 6 February 2011; accepted 29 March 2011 Available online 19 July 2011 KEYWORDS Body mass index; Obesity; Left ventricular hypertrophy Abstract Background and aims: Increased body mass index (BMI) has been associated with increased cardiovascular morbidity and mortality in hypertension. Less is known about the impact of BMI on improvement in left ventricular (LV) structure and function during antihypertensive treatment. Methods and results: Annual BMI, echocardiograms and cardiovascular events were recorded in 875 hypertensive patients with LV hypertrophy during 4.8 years randomized treatment in the Losartan Intervention For Endpoint reduction in hypertension (LIFE) echocardiography substudy. Patients were grouped by baseline BMI into normal (n Z 282), overweight (n Z 405), obese (n Z 150) and severely obese groups (n Z 38) (BMI
24.9, 25.0e29.9, 30.0e34.9, and 35.0 kg/m2, respectively). At study end, residual LV hypertrophy was present in 54% of obese and 79% of severely obese patients compared to 31% of normal weight patients (both p < 0.01). In regression analyses, adjusting for initial LV mass/height2.7, higher BMI predicted less LV hypertrophy reduction and more reduction in LV ejection fraction (both p < 0.05), independent of blood pressure reduction, diabetes and in-study weight change. During follow-up, 91 patients suffered cardiovascular death, myocardial infarction or stroke. In Cox regression analysis 1 kg/m2 higher baseline BMI predicted a 5% higher rate of cardiovascular events and 10% higher cardiovascular mortality over 4.8 years (both p < 0.05). * Corresponding author. Tel.: þ47 55972170; fax: þ47 55975150. E-mail address: eva.gerdts@med.uib.no (E. Gerdts). 0939-4753/$ - see front matter ª 2011 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.numecd.2011.03.008 Cardiac benefit of BP reduction in obesity 123 Conclusions: In hypertensive patients in the LIFE study, increased BMI was associated with less reduction of LV hypertrophy and less improvement in LV systolic function which may contribute to the observed higher cardiovascular event rate of treated hypertensive patients. ª 2011 Elsevier B.V. All rights reserved. Introduction The increased prevalence of hypertension and diabetes in obese subjects has been suggested to at least partly explain the increased cardiovascular risk observed in obesity [1,2]. In hypertensive populations obesity has been associated with higher prevalence of left ventricular (LV) hypertrophy as well as lower LV systolic function [3e5]. It is well known that systematic antihypertensive treatment induces regression of hypertensive LV hypertrophy [6]. However, as documented from the Losartan Intervention For Endpoint reduction in hypertension (LIFE) study that randomized 9193 hypertensive patients with signs of LV hypertrophy on the electrocardiogram to 4.8 years blinded losartan- or atenolol-based antihypertensive treatment, concomitant obesity predicted increased cardiovascular mortality during follow-up independent of other individual cardiovascular risk factors and type of antihypertensive treatment [7]. Furthermore, clustering of metabolic abnormalities has been associated with blunted reduction in hypertensive LV hypertrophy assessed by electrocardiography [8]. Less is known about the impact of obesity on cardiac benefit, including regression of LV hypertrophy and improvement of LV systolic function measured by serial echocardiograms during systematic long-term antihypertensive treatment. Thus, the aim of the present analysis was to assess whether obesity itself attenuates the known beneficial effects of aggressive antihypertensive treatment on LV structure and systolic function. echocardiogram before a primary study endpoint and thus were eligible for the present analysis. From these, 2 patients were excluded who did not have body mass index (BMI) measured at baseline. The study population was grouped by baseline BMI into underweight (n Z 4, BMI 3.5 and 35 mg/mmol, respectively [14]. All patients gave informed consent to participate in the LIFE echocardiography substudy, which was approved by regional ethics committees in all participating countries. Methods Echocardiography Patient population The present analysis was added to the LIFE echocardiography analysis plan before unblinding of data. The LIFE echocardiography study, a prospectively planned substudy of the main LIFE study, enrolled 960 of the 9193 participants in the parent trial for annual echocardiographic follow-up [9]. Results from the main LIFE study that randomized patients aged 55e80 years with essential hypertension (baseline blood pressure 160e200/ 95e115 mm Hg) and electrocardiographic (ECG) LV hypertrophy (according to Cornell voltage-duration or SokolowLyon voltage criteria) from Norway, Sweden, Denmark, Finland, Iceland, the United Kingdom and the United States of America to a mean of 4.8 years double-blind treatment with losartan compared to atenolol demonstrated the superiority of losartan-based treatment in reduction of cardiovascular events and signs of LV hypertrophy on ECG [10,11]. Of the 960 patients enrolled in the LIFE echocardiographic substudy, 881 patients had LV dimensions measured at enrollment and on at least one follow-up Organization, patient recruitment, protocol, and echocardiographic methods used in the LIFE echocardiographic substudy have been previously published [9,15]. All echocardiograms were sent to the Cornell Echocardiography Reading Center for blinded interpretation. Measurements of LV diameter and wall thicknesses were made on two-dimensional parasternal long-axis views according to American Society of Echocardiography standards and used for calculation of endocardial fractional shortening, LV ejection fraction by the Teichholz method and LV mass using an autopsy-validated equation [16]. LV mass showed excellent inter-study reliability in a separate study of 183 patients from the Reading Center [17]. LV hypertrophy was considered present if LV mass indexed to height2.7 exceeded 49.2 g/m2.7 in men or 46.7 g/m2.7 in women [18]. Diastolic relative wall thickness was calculated as posterior wall thickness/internal radius and considered increased if 0.43 [19]. Individual LV geometry was assessed from LV mass/height2.7 and relative wall thickness in combination and patients 124 E. Gerdts et al. grouped in the 4 following LV geometric patterns: In patients without LV hypertrophy, LV geometry was considered normal if relative wall thickness also was normal and characterized as concentric remodeling if relative wall thickness was increased. In patients with increased LV mass, hypertrophy was classified as eccentric or concentric based on the presence of normal or increased relative wall thickness, respectively [20]. Midwall shortening and its relation to circumferential end-systolic stress at the level of the LV minor axis (stress-corrected midwall shortening) were calculated using a previously validated formula [21,22]. Mitral and aortic regurgitation were assessed by color Doppler using a previously described 4-category grading system [23,24]. Supine blood pressure measured by arm cuff sphygmomanometer at the end of the echocardiogram was used in calculation of hemodynamic variables. Heart rate was measured from the echocardiographic recordings. Endpoints The primary composite endpoint in the LIFE trial included cardiovascular death and fatal and nonfatal myocardial infarction and stroke. The components of the primary endpoint were also prespecified secondary endpoints. All endpoints were validated by an independent endpoint classification committee [11]. Statistical analyses Data management and analysis were performed using SPSS 17.0 (SPSS, Chicago, IL) software. Data are presented as Table 1 mean (standard deviation) for continuous variables and as percentages for categorical variables. Between-BMI class comparisons were made by ANOVA with Scheffe’s post-hoc test or chi-square statistics, as appropriate. LV dimensions, mass and systolic functional variables were compared between BMI classes using a general linear model including gender and African American ethnicity as covariates. Intreatment changes in variables were taken as the difference between the value at baseline visit and the value at the final study visit or the last study visit before occurrence of a primary composite endpoint and calculated in each individual patient. Impact of obesity on in-treatment reduction in LV mass/height2.7 was assessed by multiple linear regression analysis using an enter procedure with colinearity diagnostics controlling for baseline LV mass/ height2.7, age, gender, race, smoking, history of myocardial infarction, in-treatment systolic blood pressure and albuminuria and randomized study treatment. A similar model was used for assessing the impact of obesity on intreatment improvement in LV systolic function (LV ejection fraction, midwall shortening and stress-corrected midwall shortening) adding in-treatment reduction in LV mass/height2.7 to the covariates. In subsequent models an indicator variable for prevalent and incident diabetes mellitus and in-treatment change in body weight as a continuous variable were added as covariates. The associations of baseline BMI with the primary composite endpoint and with cardiovascular death, myocardial infarction, and stroke as separate endpoints were evaluated in Cox regression models including systolic blood pressure and LV mass as time-varying covariates, updated at each annual visit, and baseline Framingham risk Baseline characteristics in normal, overweight, obese and severely obese groups of patients. Variable Baseline Age (years) Women (%) Diabetes (%) AfricaneAmerican (%) Weight (kg) Body mass index (kg/m2) Systolic BP (mmHg) Diastolic BP (mmHg) Mean BP (mmHg) Pulse pressure (mmHg) Heart rate (bpm) Albuminuria (%) Last study visit Weight (kg) Body mass index (kg/m2) Systolic BP (mmHg) Diastolic BP (mmHg) Mean BP (mmHg) Pulse pressure (mmHg) Heart rate (bpm) Albuminuria (%) New-onset diabetes (%) Normal weight (n Z 282) Overweight (n Z 405) Obese (n Z 150) Severely obese (n Z 38) 67  7 40 6 11 67.0  8.5 23.0  1.5 174  14 97  9 123  8 77  16 71  11 26 66  7 34* 11* 10 79.0  9.0** 27.3  1.4 173  14 99  8** 124  8 73  15** 72  12 26 65  6* 55* 15* 15 89.7  11.0** 31.9  1.5 175  14 99  9 124  8 76  15 73  11 30 62  6** 71* 21* 45* 105.3  16.9** 39.4  4.9 174  15 97  9 123  8 76  17 71  8 50 68.2  10.5 23.6  2.8 146  17 80  9 102  10 65  16 67  12 23 3 80.3  10.6** 27.6  2.4** 144  16 83  10* 104  10 61  15* 67  11 23 6** 89.6  12.0** 31.7  3.0** 145  16 83  9** 104  10 62  15 70  12 25 9* 104.9  20.0** 38.8  5.0** 143  16 80  11 101  10 62  16 63  10 32 11* Abbreviations: LV Z left ventricular; BP Z blood pressure. *p < 0.05; **p < 0.01 vs normal weight group. Cardiac benefit of BP reduction in obesity 125 Table 2 Baseline LV dimensions, wall thicknesses and systolic function in normal weight, overweight, obese and severely obese groups of patients. Variable Baseline LV diastolic diameter (cm) LV systolic diameter (cm) Septal thickness (cm) Posterior wall thickness (cm) Relative wall thickness LV ejection fraction (%) Midwall shortening (%) Stress-corrected midwall shortening (%) LV mass/height2.7 (g/m2.7) LV hypertrophy (%) Aortic regurgitation (%) Mitral regurgitation (%) Final echocardiogram LV diastolic diameter (cm) LV systolic diameter (cm) Septal thickness (cm) Posterior wall thickness (cm) LV ejection fraction (%) Relative wall thickness Midwall shortening (%) Stress-corrected midwall shortening (%) LV mass/height2.7 (g/m2.7) LV hypertrophy (%) Aortic regurgitation (%) Mitral regurgitation (%) Normal weight (n Z 282) Overweight (n Z 405) Obese (n Z 150) Severely obese (n Z 38) 5.18  3.43  1.15  1.06  0.41  62  9 15.5  97.3  52.8  67 18 26 5.31  3.56  1.16  1.07  0.41  61  8 15.5  97.1  55.7  73* 12 23 5.34  3.61  1.16  1.08  0.41  60  8 15.2  95.8  59.3  85* 11 20 5.63  3.86  1.22  1.13  0.41  59  9 15.0  95.6  74.5  97* 11 37 0.54 0.61 0.16 0.13 0.06 2.2 12.9 10.9 5.28  0.54 3.61  0.64 0.98  0.16 0.87  0.12 59  9 0.33  0.05 16.7  2.5 105.5  13.7 44.1  11.8 31 23 46 0.57* 0.61* 0.14 0.12 0.07 2.0 12.6 11.7* 5.47  0.51** 3.75  0.54** 0.99  0.13 0.89  0.11 59  7 0.33  0.06 16.7  2.0 106.3  11.5 46.5  9.3* 41* 17* 38* 0.56* 0.61* 0.16 0.13 0.07 2.1 13.8 11.3** 5.51  0.53** 3.82  0.57** 1.00  0.14 0.89  0.13 58  8 0.33  0.06 16.4  2.3 104.9  13.6 49.5  10.7** 54* 14* 23** 0.67** 0.69** 0.21* 0.16** 0.07 2.1 12.5 20.2** 5.68  0.59** 3.97  0.64** 1.04  0.18* 0.95  0.18** 57  8 0.34  0.07 15.9  2.0 100.9  11.5 60.2  19.7** 79* 8* 45 Abbreviations: LV Z left ventricular. *p < 0.05; **p < 0.01 vs normal weight group. 70 63 60 54 49 50 41 30 34 30 40 26 23 20 24 18 10 11 9 5 10 3 0 Results 0 Normal Overweight Normal 80 Obese Conc remod E-LVH Severely obese C-LVH 71 68 70 58 49 60 50 44 38 28 40 30 18 8 20 10 1 3 1 3 2 5 3 0 Normal score (based on age, gender, diabetes, total and high density lipoprotein, and baseline systolic blood pressure and electrocardiographic LV hypertrophy) [25], race and treatment allocation as fixed covariates [26]. Two-tailed P < 0.05 was considered significant both in univariate and multivariate analyses. Overweight Obese Severely obese Figure. 1 The prevalence of eccentric LV hypertrophy was greater in higher BMI classes at baseline (upper panel) and the increase in its prevalence across BMI categories became steeper at final study visit (lower panel) (all p < 0.05). At baseline, obese patients groups were younger, and included more women and patients with diabetes and of African American ethnicity (all p < 0.01), while blood pressure was comparable between groups (Table 1). Baseline LV dimensions and wall thicknesses increased progressively with increasing BMI class, and prevalence of LV hypertrophy was significantly higher in overweight and obese groups than in the normal weight group (Table 2, Fig. 1). During antihypertensive treatment, BMI and body weight on average increased by 0.2  2.4 kg/m2 and 0.6  kg, respectively, in the total study population, the normal and overweight groups had slight increases in body weight and BMI, while the obese and severely obese groups had slight decreases in weight and BMI during study follow-up (Table 2). In-treatment blood pressure was comparably reduced 126 E. Gerdts et al. Table 3 Changes in LV geometric patterns in obese (upper panel) and severely obese (lower pane) patients during antihypertensive treatment. A. Obese patients (n Z 150) Baseline LV geometry Normal geometry (n Z 16) Concentric remodeling (n Z 7) Eccentric LV hypertrophy (n Z 82) Concentric LV hypertrophy (Z45) Final LV geometry Normal (n) Concentric remodeling (n) Eccentric LV hypertrophy (n) Concentric LV hypertrophy (n) 12 0 4 0 4 1 2 0 31 0 39 2 19 2 18 6 66 3 73 8 B. Severely obese patients (n Z 38) Baseline LV geometry Concentric remodeling (n Z 0) Eccentric LV hypertrophy (n Z 24) Concentric LV hypertrophy (n Z 13) Final LV geometry 0 0 0 0 4 0 20 0 Normal geometry (n Z 1) Normal (n) Concentric remodeling (n) Eccentric LV hypertrophy (n) Concentric LV hypertrophy (n) 1 0 0 0 2 1 7 3 7 1 27 3 Abbreviations: LV Z left ventricular. in all BMI classes while absolute reduction in LV mass/ height2.7 was higher in overweight and obese groups (all p < 0.01 vs. normal weight group). However, compared to the normal weight group, the prevalence of LV hypertrophy, predominantly of eccentric type, was 1.7 and 2.5 times higher in the obese and morbid obese groups on final study echocardiogram (Fig. 1). Of note, a considerable reduction in prevalence of concentric LV hypertrophy, the prognostically most unfavorable LV geometric pattern, was also demonstrated in the obese groups (Table 3). In multiple regression analyses, higher baseline BMI predicted less intreatment reduction of LV mass/height2.7 (b Z 0.08), more in-treatment reduction in LV ejection fraction (b Z 0.07) and less improvement in midwall shortening Table 4 Impact of BMI on in-treatment changes of LV mass/height2.7 and LV systolic function measured as ejection fraction, midwall shortening and stress-corrected midwall shortening in multiple linear regression analyses. Variable In-treatment reduction in LV mass/height2.7 In-treatment reduction in EF In-treatment increase in MWS In-treatment increase in scMWS Multiple R2 of model Baseline dependent variable BMI (kg/m2) Age (years) AfricaneAmerican ethnicity Previous myocardial infarction In-treatment decrease in LV mass/height2.7 (g/m2.7) In-treatment decrease in body weight (kg) In-treatment albuminuria (mg/mmol) Diabetes mellitus Female gender Atenolol-based treatment 0.30** 0.56** 0.10** 0.10** 0.32** 0.56** 0.08* 0.31** 0.52** 0.06* 0.07* 0.40** 0.58** 0.10** 0.07* 0.07* 0.09** 0.16** 0.06* 0.10** 0.10** 0.14** 0.07* 0.12** 0.07* 0.06* 0.13** 0.09** 0.08** 0.09** 0.07* Abbreviations: BMI Z body mass index; EF Z ejection fraction; LV Z left ventricular; MWS Z midwall shortening; scMWS Z stresscorrected midwall shortening. *p < 0.05; **p < 0.01. Data are presented as Multiple R2 for individual models, and b coefficients and level of significance for independent predictors and covariates. Additional variables that did not enter any of the models: In-treatment change in systolic blood pressure and smoking. Cardiac benefit of BP reduction in obesity Table 5 stroke. 127 Association between baseline BMI and rates of combined and individual cardiovascular death, myocardial infarct and Variable Primary composite endpoint (n Z 91) Cardiovascular death (n Z 22) Myocardial infarction (n Z 36) Stroke (n Z 53) BMI (kg/m2) In-treatment systolic blood pressure (mmHg) In-treatment LV mass (g) Framingham risk score 1.05 (1.01e1.10)* 0.98 (0.97e0.99)** 1.10 (1.01e1.20)* 1.00 (0.97e1.03) 1.02 (0.95e1.10) 0.97 (0.95e0.99)** 1.04 (0.98e1.10) 0.99 (0.98e1.01) 1.00 (1.00e1.01)* 1.05 (1.03e1.07)* 1.01 (1.00e1.01)* 1.09 (1.04e1.14)** 1.01 (1.00e1.02)** 1.02 (0.98e1.05) 1.00 (0.99e1.01) 1.04 (1.01e1.07)** Abbreviations: LV Z left ventricular; BMI Z body mass index. *p < 0.05; **p < 0.01. Data are presented as hazard ratio (95% confidence intervals) and significance from Cox regression analyses. Additional covariates that did not enter any of the models: African American ethnicity and study treatment allocation. (b Z 0.08), all (p < 0.05) independent of other covariates including baseline LV mass/height2.7 (Table 4). Of note, these associations remained significant also with intreatment change in body weight and prevalent and incident diabetes added to the models. During 4.8 years follow-up, a total of 91 primary endpoints occurred within the study population. In Cox regression analysis, higher baseline BMI was independently associated with a higher incidence of combined cardiovascular mortality, myocardial infarction and stroke [HR 1.05 per 1 kg/m2 higher BMI (95% CI 1.01e1.10), p Z 0.032] which was primarily driven by higher incidence of cardiovascular mortality [HR 1.10 per 1 kg/m2 higher BMI (95% CI 1.01e1.20), p Z 0.028] (Table 5). Discussion The purpose of this study was to evaluate the impact of obesity on cardiac benefit including regression of LV hypertrophy and improvement of LV systolic function, during systematic long-term antihypertensive treatment. As expected from previous publications, obese patient groups had higher prevalence of LV hypertrophy before starting antihypertensive treatment [3e5]. However, our findings add to previous knowledge by demonstrating that regression of LV hypertrophy and improvement in LV systolic function measured by echocardiography were attenuated in obese patients in spite of comparable blood pressure reduction induced by systematic, aggressive antihypertensive treatment. Of note, higher baseline BMI predicted less LV hypertrophy regression independent of effects of the higher prevalence of diabetes and African American ethnicity in the obese patient groups. As demonstrated by our results, prevalence of residual LV hypertrophy on the final study echocardiogram was 1.7 and 2.5 times higher in the obese and severely obese groups, respectively, approximating the 2.7 fold higher prevalence of LV hypertrophy associated with obesity recently published from the observational Multi-Ethnic Study of Atherosclerosis cohort using magnetic resonance imaging [27]. However, a significant shift from the prognostically more severe concentric LV hypertrophy pattern to eccentric LV hypertrophy pattern was seen during treatment in both obese and non-obese patient groups, consistent with a beneficial impact of antihypertensive treatment on pressure-overload [28], while volumeoverload remained relatively unchanged as a consequence of the small change in BMI and weight observed during follow-up in the present study population. Of note, higher baseline BMI predicted less LV hypertrophy regression independent of effects of the higher prevalence of diabetes and African American ethnicity in the obese patient groups [29]. It is well known that obesity may promote LV systolic dysfunction by a number of mechanisms, including comorbidities like hypertension, diabetes, coronary artery disease or LV hypertrophy. Although LV systolic function did not differ significantly among BMI classes in the present study, obesity predicted less increase in midwall shortening and more decrease in LV ejection fraction during follow-up, independent of concomitant diabetes, history of previous myocardial infarction, African American ethnicity, in-treatment LV mass and study treatment allocation. Our findings add to previous epidemiologic reports from the Olmsted County study, finding central obesity to be associated with lower LV ejection fraction and a higher risk of development of clinical heart failure, and from a Belgian community study finding subclinical systolic dysfunction in obesity demonstrated by lower longitudinal strain with higher waist-hip ratio, and lower radial strain with greater body weight [30,31]. In the present study population, waisthip ratio and body fat distribution was not measured, thus central obesity could not be directly studied. However, others have demonstrated that severe obesity, corresponding to BMI 35 kg/m2, mostly is of the central type [32]. The increased morbidity and mortality observed in obesity has often been attributed to the clustering of other cardiovascular risk factors in obese subjects, in particular increased prevalences of insulin resistance and type-2 diabetes [33]. As proof of concept, the Swedish Obesity Study, which prospectively followed 2010 obese subjects who underwent bariatric surgery to reduce weight and 2037 matched controls for 10 years, demonstrated that intentional, persistent weight reduction in obesity was not only associated with a substantial reduction in obesityassociated cardiovascular risk factors like diabetes, hypertriglyceridemia and hyperuricemia [34], but also with reduced total mortality, including reduced cardiovascular death and death from cancer [35]. 128 However, others have found an independent association between higher BMI and increased rate of cardiovascular events, as also demonstrated by the present study results. This was first reported more than 25 years ago from the Framingham Heart Study [36], and more recently from a prospective cohort study including observations from 184,697 persons in Austria [37]. In this epidemiologic study, severe obesity was associated with a 2-fold higher risk of cardiovascular mortality, independent of other covariates [37]. The complex relation between body weight and morbidity and mortality was reflected in a recent publication from the US National Center for Health Statistics [38]. In this report, obesity was particularly associated with increased cardiovascular mortality, as well as death from diabetes and kidney disease [38]. Findings in the present analysis from the LIFE echocardiography substudy extend previous LIFE publications by showing that the 11% higher rate of cardiovascular mortality per 1 kg/m2 higher BMI was observed despite beneficial treatment effects on echocardiographic LV mass and systolic blood pressure. E. Gerdts et al. [5] [6] [7] [8] [9] Conclusions In hypertensive patients, obesity is associated with less regression of echocardiographic LV hypertrophy as well as less improvement of LV systolic function during antihypertensive treatment independent of concomitant diabetes. These findings may help explain the increased rate of cardiovascular events observed among obese hypertensive patients participating in the LIFE study despite a beneficial shift from concentric to eccentric LV hypertrophy pattern. [10] Disclosures [12] E.G., K.W., K.B. and M.S.N. have all received grant support from Merck & Co, the sponsor of the LIFE study. R.B.D. and B.D. have received grant support and served as consultants for Merck & Co. B.P.L. and G.d.S. have no disclosures relating to this manuscript. Acknowledgment The LIFE echocardiography substudy was supported in part by grant COZ-368 from Merck & Co, Inc., West Point, PA, USA (the sponsor of the parent LIFE study). [11] [13] [14] [15] [16] References [1] Wilson PW, D’Agostino RB, Sullivan L, Parise H, Kannel WB. Overweight and obesity as determinants of cardiovascular risk: the Framingham experience. Arch Intern Med 2002;162: 1867e72. [2] Calle EE, Thun MJ, Petrelli JM, Rodriguez C, Heath Jr CW. Body mass index and mortality in a prospective cohort of U.S. adults. N Engl J Med 1999;341:1097e105. [3] de Simone G, Devereux RB, Mureddu GF, Roman MJ, Ganau A, Alderman MH, et al. Influence of obesity on left ventricular midwall mechanics in arterial hypertension. Hypertension 1996;28:276e83. [4] Mureddu GF, de Simone G, Greco R, Rosato GF, Contaldo F. Left ventricular filling in arterial hypertension. Influence of [17] [18] obesity and hemodynamic and structural confounders. Hypertension 1997;29:544e50. Gottdiener JS, Reda DJ, Materson BJ, Massie BM, Notargiacomo A, Hamburger RJ, et al. Importance of obesity, race and age to the cardiac structural abd functional effects of hypertension: the department of veterans affairs cooperative study group on antihypertensive agents. J Am Coll Cardiol 1994;24:1492e8. Devereux RB, Dahlöf B, Gerdts E, Boman K, Nieminen MS, Papademetriou V, et al. Regression of hypertensive left ventricular hypertrophy by losartan compared with atenolol: the Losartan Intervention for Endpoint Reduction in Hypertension (LIFE) trial. Circulation 2004;110:1456e62. de Simone G, Wachtell K, Palmieri V, Hille DC, Beevers G, Dahlöf B, et al. Body build and risk of cardiovascular events in hypertension and left ventricular hypertrophy: the LIFE (Losartan Intervention For Endpoint reduction in hypertension) study. Circulation 2005;111:1924e31. de Simone G, Okin PM, Gerdts E, Olsen MH, Wachtell K, Hille DC, et al. Clustered metabolic abnormalities blunt regression of hypertensive left ventricular hypertrophy: the LIFE study. Nutr Metab Cardiovasc Dis 2009;19:634e40. Devereux RB, Bella J, Boman K, Gerdts E, Nieminen MS, Rokkedal J, et al. Echocardiographic left ventricular geometry in hypertensive patients with electrocardiographic left ventricular hypertrophy: the LIFE Study. Blood Press 2001;10:74e82. Dahlöf B, Devereux RB, Julius S, Kjeldsen SE, Beevers G, de Faire U, et al. Characteristics of 9194 patients with left ventricular hypertrophy: the LIFE study. Losartan Intervention For Endpoint Reduction in Hypertension. Hypertension 1998; 32:989e97. Dahlöf B, Devereux RB, Kjeldsen SE, Julius S, Beevers G, de Faire U, et al. LIFE study group. Cardiovascular morbidity and mortality in the Losartan Intervention For Endpoint reduction in hypertension study (LIFE): a randomised trial against atenolol. Lancet 2002;359:995e1003. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults e the evidence report. Obes Res 1998;6(Suppl. 2):51Se209S. National Institutes of Health. WHO Study Group. Diabetes mellitus (technical report series 727). Geneva: WHO; 1985. Jensen JS, Clausen P, Borch-Johnsen K, Jensen G, FeldtRasmussen B. Detecting microalbuminuria by urinary albumin/creatinine concentration ratio. Nephrol Dial Transplant 1997;12(Suppl. 2):6e9. Wachtell K, Smith G, Gerdts E, Dahlof B, Nieminen MS, Papademetriou V, et al. Left ventricular filling patterns in patients with systemic hypertension and left ventricular hypertrophy (the LIFE study). Losartan Intervention For Endpoint. Am J Cardiol 2000;85:466e72. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, et al. Chamber quantification writing group; American society of echocardiography’s guidelines and standards committee; European association of echocardiography. Recommendations for chamber quantification: a report from the American society of echocardiography’s guidelines and standards committee and the chamber quantification writing group, developed in conjunction with the European association of echocardiography, a branch of the European society of cardiology. J Am Soc Echocardiogr 2005;18:1440e63. Palmieri V, Dahlöf B, DeQuattro V, Roman MJ, RT Hahn, Sharpe N, et al. Reliability of echocardiographic assessment of left ventricular structure and function: the PRESERVE study. J Am Coll Cardiol 1999;34:1625e32. Wachtell K, Bella JN, Liebson PR, Gerdts E, Dahlof B, Roman MJ, et al. Impact of different partition values on the prevalences of left ventricular hypertrophy and concentric Cardiac benefit of BP reduction in obesity [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] geometry in a large hypertensive population: the LIFE study. Hypertension 2000;35:6e12. Palmieri V, de Simone G, Arnett DK, Bella JN, Kitzman DW, Oberman A, et al. Relation of various degrees of body mass index in patients with systemic hypertension to left ventricular mass, cardiac output, and peripheral resistance (The hypertension genetic epidemiology network study). Am J Cardiol 2001;88:1163e8. Roman MJ, Pickering TG, Schwartz JE, Pini R, Devereux RB. Relation of arterial structure and function to left ventricular geometric patterns in hypertensive adults. J Am Coll Cardiol 1996;28:751e6. Gaasch WH, Zile MR, Hoshino PK, Apstein CS, Blaustein AS. Stress-shortening relations and myocardial blood flow in compensated and failing canine hearts with pressure-overload hypertrophy. Circulation 1989;79:872e83. de Simone G, Devereux RB, Roman MJ, Ganau A, Saba PS, Alderman MH, et al. Assessment of left ventricular function by midwall fractional shortening/end-systolic stress relation in human hypertension. J Am Coll Cardiol 1994;23: 1444e51. Jones EC, Devereux RB, Roman MJ, Liu JE, Fishman D, Lee ET, et al. Prevalence and correlates of mitral regurgitation in a population-based sample (the strong heart study). Am J Cardiol 2001;87:298e304. Lebowitz NE, Bella JN, Roman MJ, Liu JE, Fishman DP, Paranicas M, et al. Prevalence and correlates of aortic regurgitation in American Indians: the strong heart study. J Am Coll Cardiol 2000;36:461e7. Anderson KM, Wilson PW, Odell PM, Kannel WB. An updated coronary risk profile. A statement for health professionals. Circulation 1991;83:356e62. Snapinn SM, Jiang Q, Iglewicz B. Illustrating the impact of a time-varying covariate with an extended Kaplan-Meier estimator. Am Stat 2005;59:301e7. Burke GL, Bertoni AG, Shea S, Watson KE, Blumenthal RS, Chung H, et al. The impact of obesity on cardiovascular disease risk factors and subclinical vascular disease. Arch Intern Med 2008;168:928e35. Gerdts E, Cramariuc D, de Simone G, Wachtell K, Dahlöf B, Devereux RB. Impact of left ventricular geometry on prognosis 129 [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] in hypertensive patients with left ventricular hypertrophy (the LIFE study). Eur J Echocardiogr 2008;9:809e15. Gerdts E, Okin PM, Omvik P, Wachtell K, Dahlöf B, Hildebrandt P, et al. Impact of diabetes on treatment-induced changes in left ventricular structure and function in hypertensive patients with left ventricular hypertrophy. The LIFE study. Nutr Metab Cardiovasc Dis 2009;19:306e12. Ammar KA, Redfield MM, Mahoney DW, Johnson M, Jacobsen SJ, Rodeheffer RJ. Central obesity: association with left ventricular dysfunction and mortality in the community. Am Heart J 2008;156:975e81. Kuznetsova T, Herbots L, Richart T, D’hooge J, Thijs L, Fagard RH, et al. Left ventricular strain and strain rate in a general population. Eur Heart J 2008;29:2014e23. Ferreira I, Twisk JWR, van Mechelen W, Kemper HCG, Seidell JCS, Stehouwer CDA. Current and adolescent body fatness and fat distribution: relationship with carotid intimamedia thickness and large artery stiffness at age 36. J Hypertens 2004;22:145e55. Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F, Bales VS, et al. Prevalence of obesity, diabetes and obesityrelated health risk factors. JAMA 2001;2003(289):76e9. Sjöström L, Lindroos AK, Peltonen M, Torgerson J, Bouchard C, Carlsson B, et al. Swedish obese subjects study. Lifestyle, diabetes and cardiovascular risk factors 10 years after bariatric surgery. N Engl J Med 2004;351:2683e93. Sjöström L, Narbro K, Sjöström CD, Karason K, Larsson B, Wedel H, et al. Swedish obese subjects study. Effects of bariatric surgery on mortality in Swedish obese subjects. N Engl J Med 2007;357:741e52. Hubert HB, Feinleib M, McNamara PM, Castelli WP. Obesity as an independent risk factor for cardiovascular disease: a 26year follow-up of participants in the Framingham heart study. Circulation 1983;67:968e77. Klenk J, Nagel G, Ulmer H, Strasak A, Concin H, Diem G, et al. VHM&PP study group. Body mass index and mortality: results of a cohort of 184,697 adults in Austria. Eur J Epidemiol 2009; 24:83e91. Flegal KM, Graubard BI, Williamson DF, Gail MH. Causespecific excess deaths associated with underweight, overweight, and obesity. JAMA 2007;298:2028e37. diabetes research and clinical practice 107 (2015) 37–45 Contents available at ScienceDirect Diabetes Research and Clinical Practice journ al h ome pa ge : www .elsevier.co m/lo cate/diabres Active life expectancy of Americans with diabetes: Risks of heart disease, obesity, and inactivity Sarah B. Laditka b,*, James N. Laditka a,1 a Department of Public Health Sciences, and Associate Professor of Public Policy, University of North Carolina at Charlotte, 9201 University City Boulevard, Charlotte, NC 28223, United States of America b Department of Public Health Sciences, and Associate Professor of Public Policy, University of North Carolina at Charlotte, 9201 University City Boulevard, Charlotte, NC 28223, United States of America article info abstract Article history: Aims: Few researchers have studied whether diabetes itself is responsible for high rates of Received 29 April 2014 disability or mortality, or if factors associated with diabetes contribute importantly. We Received in revised form estimated associations of diabetes, heart disease, obesity, and physical inactivity with life 23 September 2014 expectancy (LE), the proportion of life with disability (DLE), and disability in the last year of life. Accepted 17 October 2014 Methods: Data were from the Panel Study of Income Dynamics (1999-2011 and 1986, African Available online 23 October 2014 American and white women and men ages 55+, n = 1,980, 17,352 person-years). Activities of daily living defined disability. Multinomial logistic Markov models estimated disability Keywords: transition probabilities adjusted for age, sex, race/ethnicity, education, and the health African Americans factors. Microsimulation measured outcomes. Aging Results: White women and men exemplify results. LE was, for women: 3.5 years less with Diabetes diabetes than without (95% confidence interval, 3.1–4.0), 11.1 less (10.3–12.0) adding heart Disability disease, 21.9 less with all factors (15.3–28.5), all p < 0.001. Corresponding results for men: 1.7 Mortality years (0.9–2.3, not significant), 8.2 (6.8–9.5) and 18.1 (15.6–20.6), both p < 0.001. DLE was, for Physical activity women: 23.5% (21.7–25.4) with no risk factors, 27.1% (25.7–28.6) with diabetes alone, 34.6% (33.1–36.1) adding heart disease, 52.9% (38.9–66.8) with all factors, all p < 0.001; for men: 13.2% (11.7–14.6), 16.3% (14.8-17.8, p < 0.01); and 22.1% (20.5–23.7), 36.4% (25.0–47.8), both p < .001. Among people with diabetes, those with other conditions were much less likely to have no disability in the final year of life. Conclusions: Much of the disability and mortality with diabetes was due to heart disease, obesity, and inactivity, risks that can be modified by health behaviors and medical care. # 2014 Elsevier Ireland Ltd. All rights reserved. 1. Introduction About 26 million people in the United States have diabetes, a common cause of disability and death [1]. Lifetime risk of having diabetes is over 33% for men and 39% for women [2]. Diabetes often causes difficulty with activities of daily living (ADLs) such as walking and dressing [3,4]. Many people with diabetes also have heart disease, or are obese or physically inactive. Each of those health factors reduces life expectancy and increases the risk of disability. Those factors may confound the association of diabetes with disability and life * Corresponding author. Tel.: +1 704 687 5390; fax: +1 704 687 1644. E-mail addresses: sladitka@uncc.edu (S.B. Laditka), jladitka@uncc.edu (J.N. Laditka). 1 Tel.: +1 704 687 8742; fax: +1 704 687 1644. http://dx.doi.org/10.1016/j.diabres.2014.10.008 0168-8227/# 2014 Elsevier Ireland Ltd. All rights reserved. 38 diabetes research and clinical practice 107 (2015) 37–45 expectancy, yet no study has examined that association after controlling for them. This study adds to knowledge about diabetes by estimating the association of diabetes with disability and life expectancy with and without heart disease, obesity, and inactivity. Heart disease is one of the most prevalent diseases in the United States [5–7]. Diabetes causes vascular damage that substantially increases the risk of developing heart disease, and doubles the risk of dying from it [6–9]. It is nonetheless useful to consider the health risks of diabetes that are independent of heart disease because people can control diabetes and reduce heart disease risk by avoiding smoking, maintaining healthy weight, eating a heart healthy diet, controlling blood pressure and low-density lipoprotein cholesterol, and following the treatments recommended by their physicians [6–9]. One study has examined associations of both diabetes and heart disease with life expectancy [9]. The researchers found that although diabetes increased heart disease risk, the substantial reduction in life expectancy associated with diabetes was not significantly different for those with and without heart disease [9]. However, that study represented a limited population in a small area, measured diabetes and heart disease only every 12 years, and did not consider effects of disability on mortality [9]. The growing prevalence of overweight and obesity is contributing to a dramatic increase in type 2 diabetes [10– 13]. Obesity is also linked with a greater risk of disability [3,4,14]. Results of research associating obesity with mortality among people with diabetes are not consistent. There is evidence suggesting an ‘‘obesity paradox,’’ higher mortality with normal weight than with overweight or obesity for people with diabetes [15], although a recent study did not support this theory [16]. Some studies have found greater mortality risk only with severe obesity [17,18]. Other researchers have found that being physically inactive, which is common among people who are obese, increases mortality with diabetes [11,12,19]. No related research has measured activity and obesity with adequate time before measuring disability to limit the possibility that disability caused the inactivity or obesity, rather than being caused by them. We address that substantial research gap. Measuring active life expectancy is a useful way to understand effects of diabetes on health. This central measure of population health estimates the proportions of remaining life from a given age with and without disability, as well as life expectancy [20,21]. Relevant studies have found that diabetes reduces life expectancy and increases disability [22–26]. However, given substantial evidence that heart disease, obesity, and inactivity are associated with both diabetes and active life expectancy, we addressed two hypotheses. The first was that people with diabetes and one or more of those health factors would have a larger proportion of older life with disability and shorter life expectancy than people with diabetes alone. Thus, we hypothesized that the impact of diabetes on disability and life expectancy would be less than previously estimated because researchers have not controlled for those factors. Our second hypothesis was that these associations would last to the end of life, resulting in a markedly smaller percentage of those with diabetes and one or more of the other health conditions having no month of disability in the last year of life, compared with those with diabetes without the other factors. Researchers call this outcome successful aging to the end of life [27]. 2. Subjects, materials and methods 2.1. Data source and study sample We used data from the Panel Study of Income Dynamics (PSID) [28]. We followed a nationally representative sample ages 55 and over who identified themselves as African American or non-Hispanic white (hereafter white) from 1999 through 2011 (n = 1,980). We excluded races/ethnicities with samples too small for analysis (n = 78). 2.2. Dependent variables We identified disability from participants’ reports of having ‘‘any difficulty. . .because of a health or physical problem,’’ doing any ADL by themselves and without special equipment: bathing, eating, dressing, getting into or out of a bed or chair, walking, getting around outside, and getting to and using the toilet. Death was another measured outcome. The dependent variable indicated one of several transitions: remaining nondisabled, becoming disabled, recovering from disability, remaining disabled, dying when non-disabled, or dying when disabled. Each transition was defined by a participant’s responses about each of the ADLs in a pair of successive survey waves, or in one wave followed by death. With responses in up to seven survey waves plus death, each individual could have up to seven measured transitions. The PSID identified death dates using the National Death Index, compiled by the National Center for Health Statistics from state vital records. The analytic data represented 9039 transitions occurring through 17,352 person-years. 2.3. Measuring diabetes and heart disease In all seven waves interviewers asked, ‘‘Has a doctor ever told you that you have or had [diabetes/heart disease]?’’ Interviewers were instructed: ‘‘Do not accept self-diagnosed or diagnosed by a person who is not a doctor or other health professional.’’ We updated the data with each new diagnosis, but assumed that diagnosed individuals did not recover from the disease. We examined the consistency of disease reports across survey waves. 2.4. Measuring obesity We calculated body mass index (BMI, kg/m2) using height and weight reported by participants in 1986. We used 1986 data for BMI both to examine effects of earlier-life weight status and to limit the likelihood that disability measured by the outcome variable contributed to the person’s BMI. We defined obesity as BMI  35.0, using a World Health Organization threshold. This definition identified individuals whose BMIs were most likely to affect health, disability, and mortality, excluding those with lower levels of obesity diabetes research and clinical practice 107 (2015) 37–45 that may not be associated with poor health outcomes in older populations [17,18]. We examined the sensitivity of the results to defining obesity as BMI  30.0. 2.5. Measuring physical inactivity In 1986 the PSID asked, ‘‘Do you get any regular exercise, such as doing hard physical work, or walking a mile or more without stopping, or playing an active sport?’’ We considered those responding negatively to be inactive. Disability reported in 1999 would rarely cause inactivity in 1986, although permanent disability might do so for a small number of participants. For the few participants (n = 39) who did not provide BMI or activity measures in 1986, we used equivalent self-reports from 1999. We examined correlations among the 1986 obesity and activity measures with reports from that year of having only fair or poor health, and with a measure of disability defined by health conditions that ‘‘keep you from doing some types of work’’ (‘‘somewhat,’’ ‘‘a lot,’’ or completely). 2.6. Controlling for education We controlled for six education levels [29]: less than 8 years, completion of grade 8, completion of grades 9 through 12 without a high school diploma, high school graduation, some education after high school, and at least a 4-year college degree. 2.7. The model associating diabetes and related health factors with disability and death Our regression model represented increasing risks of disability and death with age, and allowed for an accelerating increase, including controls for: age, age-squared, age85plus, and (age85plus  age), where age85plus indicated that age level (yes/no). The latter two terms defined a spline function that allowed a shift in disability and death risks beginning at age 85 [30]. The model included sex, race/ethnicity (African American or white), education, and the health factors. To provide specific probabilities for each population, the model included all 2-way interactions of sex and race/ethnicity with the health factors, and all 3-way interactions of those variables. Likelihood ratio tests indicated the interactions were significant ( p < 0.001). 2.8. Discrete-time Markov chains The regression model applied maximum likelihood methods to the measured interval of each disability transition to identify Markov chains in the observed data, estimating disability and death transition probabilities for each month of life conditional on all measures in the previous month. Researchers who study active life expectancy often use this approach [22,24–26,30–37]. 2.9. Variance-adjusted standard errors Repeated measures for individuals in longitudinal studies produce underestimated standard errors [30,34,38]. To account for repeated measures, we re-estimated the model with 39 a subject-specific random effect [30,34]. There is no accepted method for adding random effects to microsimulations [39]. We therefore used the estimates from the standard regression model described in sections 2.7 and 2.8 for the microsimulations, to which we turn in Section 2.10. However, for covariates with larger variance in the random effect model, we adjusted the corresponding standard errors of the regression results to reflect that greater variance. This procedure was analogous to standard error adjustments for survey design effects [30,34]. It did not alter the point estimates from the microsimulations, but did provide considerably more conservative standard errors for those results. Of the 80 covariates, 77 were statistically significant (p < 0.001). 2.10. Microsimulation Microsimulation helps researchers understand complex phenomena that cannot be studied with more common statistical methods. Microsimulation used the transition probabilities to create large populations in which each individual had a complete record of monthly disability measures from age 55 until death. The expected age at death was the average age at death in the microsimulated population. We provide summary results for that measure, the proportion of remaining life with disability, and successful aging. We also exemplify the life course successful aging results with a detailed figure for one population, for which we used the Wilcoxon-Mann-Whitney test to compare the distributions. We held education constant at the high school level in the microsimulations, the most common educational attainment of the sampled population. Details of the methods are published [20,26,30,33–37]. 2.11. Estimating variation in the microsimulation results We used bootstrapping to estimate variation in the microsimulation results, accounting for parameter uncertainty (represented by 95% confidence intervals for the estimated parameters) and the Monte Carlo variation that affects most simulation research [30,34,35]. Bootstrapping repeated the microsimulation for each population 500 times. For each repetition we made a random selection for each parameter from its variance-adjusted 95% confidence interval (CI). The final CIs we report are the 2.5th and 97.5th percentiles of the 500 results [30,34,35]. We conducted the analyses using SAS IML (Cary, North Carolina). The Institutional Review Board (IRB) at the University of North Carolina at Charlotte determined that this research, which used de-identified secondary data, did not require IRB review. 3. Results 3.1. Sample characteristics The mean age in the analytic sample was 78.3 (weighted for national representativeness, 78.5). Women were 60.5% (59.1%). The PSID over-sampled African Americans, who were 20.1% (9.0%). The percentages with diabetes, heart disease, obesity, and physical inactivity, respectively, were: 24.2% (22.5%), 40 diabetes research and clinical practice 107 (2015) 37–45 21.5% (22.4%), 7.8% (6.2%), and 8.8% (8.2%) (results not shown in tables). There was little evidence of correlation in the 1986 measures between inactivity and either obesity (r = 0.03, p = 0.07), fair/poor health (r = 0.03, p = 0.12), or work disability (r = 0.01, p = 0.39). Obesity was weakly correlated with fair/ poor health (r = 0.04, p < 0.05) and work disability (r = 0.04, p < 0.05). Of participants with diabetes, 16% reported in a later survey response that they did not have diabetes. These participants confirmed the diabetes diagnosis with an average of 3.4 survey responses, and 54% had responded that they had the disease for more than two years. Of those reporting heart disease, 21% reported in a later survey that they did not have heart disease. They confirmed the diagnosis with an average of 3.5 diagnosis reports, and 76% said that they had the disease for more than two years. 3.2. Patterns of active life expectancy Table 1 shows the average age at death from age 55, and the percentage of remaining life with disability, together with the CIs. The estimate for people without diabetes is for those without heart disease or earlier-life obesity or inactivity. Compared with people without diabetes, those with diabetes but none of the other risk factors had shorter lives and a greater proportion of remaining life with disability. For white women, compared to those without diabetes, the average age at death for those with diabetes but none of the other factors was 3.5 years less (CI 3.1–4.0, p < 0.001, difference in years not shown in table). In the comparable result for disability, white women without diabetes were disabled 23.5% of remaining life (CI 21.7–25.4), those with diabetes 27.1% (CI 25.7–28.6, p < 0.01). Among African American women the analogous life Table 1 – Diabetes, associated factors, life expectancy and percentage of remaining life disableda White Women No Diabetes Diabetes Diabetes, Heart Disease Diabetes, Obese Diabetes, Inactive Diabetes, Heart Disease, Obese Diabetes, Heart Disease, Inactive Diabetes, Obese, Inactive Diabetes, Heart Disease, Obese, Inactive African American women No Diabetes Diabetes Diabetes, Heart Disease Diabetes, Obese Diabetes, Inactive Diabetes, Heart Disease, Obese Diabetes, Heart Disease, Inactive Diabetes, Obese, Inactive Diabetes, Heart Disease, Obese, Inactive White Men No Diabetes Diabetes Diabetes, Heart Disease Diabetes, Obese Diabetes, Inactive Diabetes, Heart Disease, Obese Diabetes, Heart Disease, Inactive Diabetes, Obese, Inactive Diabetes, Heart Disease, Obese, Inactive African American Men No Diabetes Diabetes Diabetes, Heart Disease Diabetes, Obese Diabetes, Inactive Diabetes, Heart Disease, Obese Diabetes, Heart Disease, Inactive Diabetes, Obese, Inactive Diabetes, Heart Disease, Obese, Inactive a Life Expectancy (Average Age at Death) Percent of Remaining Life Disabled Mean (95% CI) Mean (95% CI) 88.6 85.1 77.5 84.1 69.4 76.6 65.8 70.9 66.7 (86.7–90.5) (83.6–86.5) (76.4–78.5) (82.7–85.4) (63.2–75.7) (75.4–77.8) (60.3–71.2) (61.3–80.6) (58.2–75.2) 23.5 27.1 34.6 42.1 28.4 50.8 34.6 45.9 52.9 (21.7–25.4) (25.7–28.6) (33.1–36.1) (39.0–45.1) (21.7–35.1) (47.9–53.6) (25.5–43.6) (33.6–58.1) (38.9–66.8) 85.2 82.7 75.7 83.6 66.8 76.4 64.4 70.3 67.1 (82.5–87.9) (80.8–84.6) (74.3–77.1) (82.2–84.9) (61.1–72.5) (75.3–77.5) (59.3–69.6) (61.0–79.7) (58.4–75.8) 22.5 27.0 38.3 43.6 27.3 53.6 34.8 46.3 54.5 (20.3–24.7) (25.2–28.8) (36.5–40.1) (40.3–46.9) (21.7–33.0) (50.3–56.9) (26.3–43.2) (35.5–57.2) (41.8–67.2) 83.2 81.5 75.0 82.4 64.7 75.1 62.7 68.1 65.1 (79.9–86.4) (79.0–84.1) (73.1–76.9) (80.5–84.2) (60.8–68.6) (73.5–76.6) (59.5–66.0) (61.8–74.3) (59.3–70.8) 13.2 16.3 22.1 29.2 15.0 37.4 19.7 29.4 36.4 (11.7–14.6) (14.8–17.8) (20.5–23.7) (26.5–32.0) (11.6–18.5) (34.2–40.6) (14.1–25.2) (20.7–38.2) (25.0–47.8) 77.1 76.7 71.4 79.9 61.7 73.4 60.6 65.5 63.5 (73.2–81.1) (73.7–79.8) (69.2–73.6) (77.8–82.1) (58.8–64.6) (72.0–74.8) (57.8–63.4) (59.7–71.3) (58.1–68.9) 13.5 17.4 25.1 31.6 18.7 41.8 24.7 33.1 41.7 (11.9–15.0) (15.5–19.3) (23.0–27.2) (28.3–34.8) (16.7–20.7) (38.2–45.4) (20.2–29.1) (26.4–39.7) (32.5–50.9) Data Source: Panel Study of Income Dynamics, 1999–2011; 1986 for obesity and physical activity. CI = Confidence Interval. 41 diabetes research and clinical practice 107 (2015) 37–45 expectancy result was not statistically significant; the analogous disability results were 22.5% (CI 20.3–24.7) and 27.0% (CI 25.2–28.8, p < 0.01). Comparing life expectancy without diabetes to life expectancy with diabetes but no additional risks, confidence intervals overlapped for both African American and white men, indicating that the comparison did not identify statistically significant differences. In the result for disability for white men, those without diabetes were disabled 13.2% of remaining life (CI 11.7–14.6), those with diabetes 16.3% (CI 14.8–17.8, p < 0.01). Among African American men, those with diabetes but no other measured risk were disabled for a greater percentage of remaining life (17.4%, 15.5–19.3) than those without diabetes (13.5%, 11.9–15.0, p < 0.01). Individuals with both diabetes and either heart disease or earlier-life inactivity had a lower average age at death than those with only diabetes. Among white women, compared to those with diabetes but none of the other risks, those with diabetes and heart disease lived 7.6 fewer years (CI 7.2–8.0, p < 0.001). Among white women with diabetes, those with heart disease had a significantly higher percentage of remaining life with disability (34.6%, 33.1–36.1) than those with none of the other risks (27.1% 25.7–28.6, p < 0.001). Analogous results for African American women for disability also indicated additional risk with heart disease, with disability in 38.3% of remaining life with diabetes and heart disease (36.5–40.1) compared with 27.0% (25.2–28.8, p < 0.001) with diabetes alone. The corresponding reductions in life expectancy for white and African American men were also significant, respectively 6.5 (5.9–7.2, p < 0.001) and 5.3 (4.5–6.2, p < 0.05) years. Among white men with diabetes, those with heart disease had a significantly higher percentage of remaining life with disability (22.1%, 20.5–23.7) than those with none of the other risks (16.3%, 14.8–17.8, p < 0.001); corresponding results for African American men were: 25.1% (23.0–27.2) and 17.4% (15.5–19.3, p < 0.001). Those with diabetes and all three risk factors had the lowest average age at death in all populations, 21.9 fewer years (15.3–28.5) comparing white women with this combination to white women with none of the other factors ( p < 0.001). Corresponding results were 18.1 fewer years for African American women, 18.1 years for white men, 13.6 for African American men (all p < .001). There was little evidence that earlier-life obesity by itself reduced life expectancy for those with diabetes. The confidence intervals for having diabetes and obesity overlapped with those for having only diabetes for all populations, indicating a lack of statistically significant differences. However, in all populations the percentage of remaining life with disability was significantly greater for those with diabetes and earlier-life obesity than for those with diabetes alone. White women who were obese in 1986 and had diabetes during the study period were disabled 42.1% of remaining life (CI 39.0–45.1), compared to the 27.1% reported above for those with diabetes only (CI 25.7–28.6, p < 0.001). Corresponding results were: for African American women, 43.6% (40.3–46.9) versus 27.0% (25.2–28.8, p < 0.001); for white men, 29.2% (26.5– 32.0) versus 16.3% (14.8–17.8, p < 0.001); for African American men, 31.6% (28.3–34.8) versus 17.4% (15.5–19.3, p < 0.001). In all populations, compared to those with diabetes alone, those with diabetes and all of the other health risks had more disability. White women with diabetes but none of the other risks lived 27.1% (25.7–28.6) of remaining life with disability, as Table 2 – Successful Aging to the End of Life: Percentage of Individuals Dying at ages 84 through 86 with No Months of Disability in the Year Preceding Death, Associations with Diabetes and Associated Factorsa White Women No Diabetes Diabetes Diabetes, Heart Disease Diabetes, Obese Diabetes, Inactive Diabetes, Heart Disease, Obese Diabetes, Heart Disease, Inactive Diabetes, Obese, Inactive Diabetes, Heart Disease, Obese, Inactive Mean (95% CI) Mean (95% CI) 25.1 13.9 6.3 4.3 15.6 1.6 7.3 3.9 1.3 (19.0–31.5) (9.5–18.0) (4.9–7.9) (3.2–6.1) (10.0–23.4) (0.9–2.5) (4.3–10.7) (1.6–7.1) (0.1–2.5) 32.7 18.6 8.1 5.7 17.5 2.1 1.2 0.5 1.5 (26.5–38.8) (14.1–23.1) (5.8–10.4) (3.7–7.7) (8.6–26.5) (1.1–3.1) (0.1–2.9) (0.1–1.5) (0.1–3.7) White Men No Diabetes Diabetes Diabetes, Heart Disease Diabetes, Obese Diabetes, Inactive Diabetes, Heart Disease, Obese Diabetes, Heart Disease, Inactive Diabetes, Obese, Inactive Diabetes, Heart Disease, Obese, Inactive a African American Women African American Men Mean (95% CI) Mean (95% CI) 47.9 30.9 16.4 11.3 32.7 4.6 17.8 11.1 4.5 (39.9–54.8) (23.7–37.3) (12.8–20.1) (8.0–16.3) (25.3–42.6) (2.9–6.7) (11.7–25.4) (5.2–18.4) (2.1–7.7) 56.4 28.0 20.0 14.8 33.5 5.8 19.1 12.1 4.1 (48.9–63.8) (17.4–38.7) (15.1–24.9) (9.4–20.2) (22.8–44.1) (3.6–8.0) (0.1–66.7) (3.2–21.0) (0.1–9.9) Data Source: Panel Study of Income Dynamics, 1999–2011; 1986 for obesity and physical activity. CI = Confidence Interval. 42 diabetes research and clinical practice 107 (2015) 37–45 Figure 1 – Diabetes, associated factors, and successful aging to the end of life. Percentage of deaths with no disability in the last year of life, non-Hispanic whites. reported above, compared with 52.9% (CI 38.9–66.8, p < 0.001) for those with diabetes and all three risks. Analogous results were: for African American women, 27.0% (25.2–28.8) versus 54.5% (41.8–67.2, p < 0.001); for white men, 16.3% (14.8–17.8) versus 36.4% (25.0–47.8, p < 0.001); for African American men, 17.4% (15.5–19.3) versus 41.7% (32.5–50.9, p < 0.001). 3.3. Successful aging to the end of life Table 2 shows results for successful aging, the percentage of individuals who died at ages 84-86 without disability in the last year of life, and the CIs. To clarify comparisons among groups, we describe the results from Table 2 using relative risks, the ratio of the rate for people with diabetes to the rate for people without diabetes. Comparing white women with diabetes to those without diabetes, the relative risk of having no disability in the last year of life was 0.55 (13.9/ 25.1, p < 0.001, relative risks not shown in table). Thus, the ‘‘risk’’ or likelihood that women with diabetes would have no disability in the last year of life was 45% lower than that likelihood for women without diabetes. Corresponding risk ratios were: 0.57 for African American women; 0.65 for white men, and 0.50 for African American men (all p < 0.001). The risk of disability in the last year of life was also greater for those with diabetes and one or more of the other risks. For all groups, among those with diabetes and all of the other risks only a small percentage were not disabled in the last year of life: 1.3% of white women (CI 0.1– 2.5), 1.5% of African American women (CI 0.1–3.7), 4.5% of white men (CI 2.1–7.7), 4.1% of African American men (CI 0.1–9.9). Fig. 1 shows analogous percentages for white women who died at ages 75 to 95. Having diabetes with most combinations of the other risks substantially limited the proportion of people without disability in the final year of life. The results at age 75 illustrate that finding, where 43% of women without diabetes or any of the other risks had no disability in their last year of life, and 27% of those with only diabetes did so, both compared to only 3% of those with diabetes and all other risks (all p < 0.001). Women with diabetes alone were considerably more likely to avoid disability in their last year of life than women with diabetes and most combinations of the other risks, an advantage that persisted throughout older life. Results for white men and African Americans were similar (not shown). 3.4. Sensitivity of results to the definition of obesity There was little evidence that obesity was associated with life expectancy for people with diabetes (Table 1). Using BMI  30.0 to define obesity instead of BMI  35.0 did not result in statistically significantly different estimates (results not shown in tables). Using BMI  30.0 increased the proportion of life with disability by 34.5% for women and 14% for men; however, these comparisons were not statistically significant different from those using BMI  35.0. 4. Discussion We studied the association of diabetes with active life expectancy and measured the individual and combined additional risks for people with diabetes of heart disease and earlier-life obesity and inactivity. Consistent with previous studies, participants with diabetes had shorter lives than others, with more disability [22–26]. Consistent with our first hypothesis, being obese or inactive earlier in life, or having heart disease, greatly reduced both active life and life expectancy for people with diabetes. One of the largest negative associations for active life and life expectancy was diabetes research and clinical practice 107 (2015) 37–45 the combination of diabetes with having been physically inactive 13 years before the study baseline, a result that is consistent with previous research [11,12,19]. Also consistent with previous studies, we found little evidence that earlier-life obesity reduced life expectancy for those with diabetes if they had none of the other health risks [17,18]. Considerable evidence suggests that obesity may have limited effects on mortality among people who maintain cardiorespiratory fitness through regular physical activity [40]. However, among people with diabetes those who were obese earlier in life were disabled for a significantly greater proportion of life than those who were not obese. Consistent with our second hypothesis, people with diabetes were much more likely than others to be disabled at the end of life. With any of the additional risks few people with diabetes avoided ADL disability in their last year. With diabetes and none of the additional risks, on the other hand, the likelihood of being disabled in the last year of life was considerably lower. It is likely that diabetes was responsible for some of the additional risks associated with heart disease because diabetes is a cause of heart disease. We examined the independent risks of diabetes for disability and death because controlling diabetes can limit its impact on heart disease [6–8]. It is therefore useful to better understand disability and mortality for people who can control their diabetes. Our finding that having heart disease greatly reduced life expectancy for people with diabetes was not consistent with the only previous related study [9]. Participants in that study were primarily non-Hispanic whites in Framingham, Massachusetts [9]. Diagnoses of diabetes and heart disease were updated about every 12 years in that study, which may have resulted in an underestimation of the associations between these diseases and mortality [9]. Also the research method in that study did not account for effects of disability on life expectancy. Thus, differences between the two studies may be due to substantial differences in data and methods. 4.1. Limitations The PSID’s diabetes question did not distinguish between type 1 and type 2. About 5% of adults with diabetes have type 1 [41]. Survival with type 1 has increased in recent decades but did not greatly exceed age 50 in the cohort we studied [41]. Most people with type 1 are diagnosed before age 18 [41]; 1.2% in our sample with diabetes reported diagnoses before age 18. Thus, few participants were likely to have had type 1. Given that the PSID is nationally representative its distribution of diabetes types may represent the United States population. Diabetes and heart disease were self-reported but based on physician diagnoses. Responses to questions about conditions diagnosed by physicians tend to be highly correlated with disease and physician records [3,42,43]. However, reports of diagnoses are related to health care access and symptom severity [42]. Health care access is related to socioeconomic status and race/ethnicity. It is likely that controls for education and race/ethnicity limited bias associated with socioeconomic status, access to care, and symptom severity [29]. Our approach assumed that people diagnosed with diabetes or heart disease did not recover. There is evidence that these diseases may be reversible; however, few people achieve the 43 substantial and sustained behavioral changes that are required [44,45]. The limited number of participants with inconsistent reporting of these diseases repeatedly said that they had the disease. This suggests that most ‘‘recoveries’’ represented reporting error. Height and weight were self-reported. Although many people who are obese do not recognize that fact, self-reported height and weight bias is limited, particularly during in-person interviews as in the 1986 PSID [46]. The PSID is recognized for establishing trust among participants, who report sensitive data with item-response rates over 99% [47]. Using the 1986 BMI measure greatly increased the likelihood that it would not be affected by disability reported in 1999. The small correlation of this measure with work disability in 1986 suggested that earlier functional limitation was not a notable cause of obesity. Individuals who said they were inactive in 1986 could have become active before 1999, although it is more common for people to reduce activity as they age. The small correlation of this measure with both general health and work disability in 1986 suggested that earlier health status or functional limitation were not important causes of inactivity. The activity question provided several examples of regular exercise including ‘‘doing hard physical work,’’ and ‘‘walking a mile or more without stopping.’’ Some participants who were active at more moderate levels may have responded negatively to this question. The PSID did not ask about the frequency, intensity, or duration of activity, which are needed for precise measurements. The existence and direction of any bias is uncertain because of the lack of measurement specificity. Another consideration is that African Americans may have died at a higher rate before age 55 than whites, as African Americans may have higher mortality until about age 70 [20,48]. This limitation affects most related studies. Our sample included participants beginning at age 55. Related studies typically begin at 65 or older. Our younger starting age may have provided more valid outcome comparisons for African Americans and whites than related studies with older starting ages. If African Americans died at a higher rate before age 55 than whites, and if that was due to diabetes, the results may underestimate the effect of diabetes on mortality for African Americans. In another area, although we used ‘‘successful aging’’ to describe having no disability in the final year of life, we emphasize that people with disabilities can also age successfully in many ways. 4.2. Conclusions We found that diabetes was associated with more disability and shorter lives. However, much of the greater disability and mortality among people with diabetes was attributable to risk factors that are often associated with diabetes, particularly heart disease and earlier-life inactivity, and, specifically with regard to disability, obesity earlier in life. People with diabetes who did not also have the other risks were much more likely to age successfully to the end of life, underscoring the positive message that healthy behaviors and medical care may greatly improve both the length and the quality of life for people with diabetes. 44 diabetes research and clinical practice 107 (2015) 37–45 Conflicts of Interest The authors declare that they have no conflict of interest with the article. Acknowledgments We are grateful to two anonymous reviewers for valuable comments about this research. references [1] Riffat SB Centers for Disease Control and Prevention.In: National Diabetes Fact Sheet: National Estimates and General Information on Diabetes and Prediabetes in the United States, 2011; 2011. [2] Narayan KM, Boyle JP, Thompson TJ, Sorensen SW, Williamson DF. Lifetime risk for diabetes mellitus in the United States. JAMA 2003;290:1884–90. [3] Martin LG, Freedman VA, Schoeni RF, Andreski PM. Trends in disability and related chronic conditions among people ages fifty to sixty-four. Health Aff (Millwood) 2010;29(4): 725–31. [4] Martin LG, Schoeni RF. Trends and disparities in disability and related chronic conditions among the 40 and older population. Disabil Health J 2014;7:S15–23. [5] Heidenreich PA, Trogdon JC, Khavjou OA, Butler J, Dracup K, Ezekowitz MD, et al. Forecasting the future of cardiovascular disease in the United States: a policy statement from the American Heart Association. Circulation 2011;123:933–44. [6] Fox CS, Coady S, Sorlie PD, Levy D, Meigs JB, D’Agostino Sr RB, et al. Trends in cardiovascular complications of diabetes. JAMA 2004;292:2495–9. [7] Gregg EW, Yanfeng L, Wang J, Borrows NR, Ali MK, Rolka D, et al. Changes in diabetes-related complications in the United States, 1990-2010. NEJM 2014;370:1514–23. [8] Preis SR, Hwang SJ, Coady S, Pencina MJ, D’Agostino Sr RB, Savage PJ, et al. Trends in all-cause and cardiovascular disease mortality among women and men with and without diabetes mellitus in the Framingham Heart Study, 1950 to 2005. Circulation 2009;119:1728–35. [9] Franco OH, Steyerberg EW, Hu FB, Mackenbach J, Nusselder W. Associations of diabetes mellitus with total life expectancy and life expectancy with and without cardiovascular disease. Arch Intern Med 2007;167:1145–51. [10] Cunningham SA, Riosmena F, Wang J, Boyle JP, Rolka DB, Geiss LS, et al. Decreases in diabetes-free life expectancy in the U.S. and the role of obesity. Diabetes Care 2011;34: 2225–30. [11] Gregg EW, Cadwell BL, Cheng YJ, Cowie CC, Williams DE, Geiss L, et al. Trends in the prevalence and ratio of diagnosed to undiagnosed diabetes according to obesity levels in the US. Diabetes Care 2004;27:2806–12. [12] Lee DC, Sui X, Church TS, Lee I, Blair SN. Associations of cardiorespiratory fitness and obesity with risks of impaired fasting glucose and type 2 diabetes in men. Diabetes Care 2009;32:257–62. [13] Sui X, Hooker SP, Lee I, Church TS, Colabianchi N, Lee DC, Blair SN. A prospective study of cardiorespiratory fitness and risk of type 2 diabetes in women. Diabetes Care 2008;31:550–5. [14] Sturm R, Ringel JS, Andreyeva T. Increasing obesity rates and disability trends. Health Aff (Millwood) 2004;23(2):199– 205. [15] Carnethon MR, DeChavez PJ, Briggs ML, Lewis CE, Pankow JS, Bertoni AG, et al. Association of weight status with mortality in adults with incident diabetes. JAMA 2012;308(6):581–90. [16] Tobias DK, Pan A, Jackson CL, Jackson CL, O’Reilly EJ, Ding EL, et al. Body-mass index and mortality among adults with incident type 2 diabetes. NEJM 2014;370:233–44. [17] Flegal KM, Graubard BI, Williamson DF, Gail MH. Excess deaths associated with underweight, overweight, and obesity. JAMA 2005;293:1861–7. [18] Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. JAMA 2003;289:187–93. [19] Jonker JT, De Laet C, Franco OH, Peeters A, Mackenbach J, Nusselder WJ. Physical activity and life expectancy with and without diabetes: life table analysis of the Framingham Heart Study. Diabetes Care 2006;29:38–43. [20] Laditka SB, Laditka JN. Active life expectancy: a central measure of population health. In: Uhlenberg P, editor. International Handbook of the Demography of Aging. Netherlands: Springer-Verlag; 2009. p. 543–65. [21] Robine JM, Jagger C, Mathers C, Crimmins EM, Suzman R, editors. Determining Health Expectancies. Baffins Lane, Chichester, Sussex, UK: John Wiley & Sons; 2003. [22] Andrade FCD. Measuring the impact of diabetes on life expectancy and disability-free life expectancy among older adults in Mexico. J Gerontol B Soc Sci 2010;65B(3):381–9. [23] Bélanger A, Martel L, Berthelot JM, Wilkins R. Gender differences in disability-free life expectancy for selected risk factors and chronic conditions in Canada. J Women Aging 2002;14:61–83. [24] Jagger C, Goyder E, Clarke M, Brouard N, Arthur A. Active life expectancy in people with and without diabetes. J Public Health Med 2003;25(1):42–6. [25] Jagger C, Matthews R, Matthews F, Robinson T, Robine JM, Brayne C, et al. The burden of diseases on disability-free life expectancy in later life. J Gerontol A Biol Sci Med Sci 2007;62A(4):408–14. [26] Laditka JN, Laditka SB. Effects of diabetes on healthy life expectancy: shorter lives with more disability for both women and men. In: Yi Z, Crimmins E, Carrière Y, Robine JM, editors. Longer Life and Healthy Aging. Netherlands: Springer; 2006. p. 71–90. [27] Leveille SG, Guralnik JM, Ferrucci L, Langlois JA. Aging successfully until death in old age: opportunities for increasing active life expectancy. Am J Epidemiol 1999;149:654–64. [28] Hill M. The Panel Study of Income Dynamics: A User’s Guide. Newbury CA: Sage Publications, Inc; 1991. [29] Rogers RG, Everett BG, Zajacova A, Hummer RA. Educational degrees and adult mortality risk in the United States. Biodemography Soc Biol 2010;56:80–99. [30] Laditka SB, Laditka JN. More education may limit disability and extend life for people with cognitive impairment. Am J Alzheimer’s Dis Other Demen 2014;29(5):436–47. [31] Crimmins EM, Hayward MD, Hagedorn A, Saito Y, Brouard B. Change in disability-free life expectancy for Americans 70 years old and older. Demography 2009;46(3):627–46. [32] Lièvre A, Brouard N, Heathcote C. The estimation of health expectancies from cross-longitudinal surveys. Math Popul Stud 2003;10:211–48. [33] Laditka SB, Wolf DA. New methods for analyzing active life expectancy. J Aging Health 1998;10:214–41. [34] Laditka JN, Laditka SB. Stroke and Active Life Expectancy in the United States, 1999-2009. Disabil Health J 2014;7: 472–7. diabetes research and clinical practice 107 (2015) 37–45 [35] Laditka JN, Laditka SB, Olatosi B, Elder KT. The health tradeoff of rural residence for impaired older adults: longer life, more impairment. J Rural Health 2007;23(2):124–32. [36] Laditka SB. Modeling lifetime nursing home use under assumptions of better health. J Gerontol B Soc Sci 1998;53B(4):S177–87. [37] Laditka SB, Laditka JN. Effects of improved morbidity rates on active life expectancy and eligibility for long-term care services. J Appl Gerontol 2001;20(1):39–56. [38] Cai L, Hayward MD, Saito Y, Lubitz J, Hagedorn A, Crimmins E. Estimation of multi-state life table functions and their variability from complex survey data using the SPACE program. Demogr Res 2010;22(6):129–58. [39] Matteo R, Ambra P. Imputing individual effects in dynamic microsimulation models. An application of the rank method. Department of Economics and Statistics Cognetti de Martiis. Working paper no. 13, 2012. University of Turin. Available from: http://www.unito.it/unitoWAR/ ShowBinary/FSRepo/D031/Allegati/WP2012Dip/ 13_WP_Richiardi_Poggi.pdf. Accessed 20 January 2014. [40] Sui X, LaMonte MJ, Laditka JN, Hardin JW, Chase N, Hooker SP, Blair SN. Cardiorespiratory fitness and adiposity as mortality predictors in older adults. JAMA 2007;298(21):2507–16. [41] Miller RG, Secrest AM, Sharma RK, Songer TJ, Orchard TJ. Improvements in the life expectancy of type 1 diabetes: the Pittsburgh Epidemiology of Diabetes Complications Study Cohort. Diabetes 2012;61(11):2987–92. 45 [42] Crimmins EM, Hayward MD, Seeman T. Race/ethnicity, socioeconomic status and health. In: Anderson NB, Bulatao RA, Cohen B, editors. Critical Perspectives on Racial and Ethnic Differences in Health in Later Life. Washington, DC: National Academy Press; 2004. p. 310–52. [43] Saydah SH, Geiss Tierney E, Benjamin SM, Engelgau M, Brancati F. Review of the performance of methods to identify diabetes cases among vital statistics, administrative, and survey data. Ann Epidemiol 2005;14:507–16. [44] Ornish D, Scherwitz LW, Billings JH, Gould KL, Merritt TA, Sparler SMA, et al. Intensive lifestyle changes for reversal of coronary heart disease. JAMA 1998;280(23):2001–7. [45] Taylor R. Type 2 diabetes: etiology and reversibility. Diabetes Care 2013;36(4):1047–55. [46] Ezzati M, Martin H, Skjold S. Vander Hoorn S, Murray CJL. Trends in national and state-level obesity in the USA after correction for self-report bias: analysis of health surveys. J R Soc Med 2006;99:250–7. [47] Andreski P, McGonagle K, Schoeni R. An Analysis of the Quality of the Health Data in the Panel Study of Income Dynamics. Survey Research Center, Institute for Social Research University of Michigan, 2009. Technical Series Paper #09-02. [48] Elo IT, Preston SH. Racial and ethnic differences in mortality at older ages. In: Martin LG, Soldo BJ, editors. Racial and Ethnic Differences in the Health of Older Americans. Washington, DC: National Academy Press; 1997. p. 10–42. JACC VOL. 67, NO. 19, 2016 Letters MAY 17, 2016:2310–9 endothelium of the aorta, an observation that could support a developmental effect (5). In conclusion, we here report on multiple families harboring HCN4 mutations, who show significant dilation of the aorta ascendens, besides the recently Obesity Cardiovascular Disease and the Failure of Public Health Education established combined phenotype of bradycardia, NCCM, and mitral valve disease. These results have The authors of the VIRGO (Variation in Recovery: important implications for patient care and for the Role of Gender on Outcomes of Young AMI Patients) fundamental understanding of the role of the HCN4 study (1) deserve congratulations for their impressive channel in health and disease. work. The paper covers a lot of ground, but perhaps the most important finding is the notably high prev- Alexa M.C. Vermeer, MD *Elisabeth M. Lodder, PhD Dierk Thomas, MD Floor A.M. Duijkers, MD, PhD Carlo Marcelis, MD, PhD Edwin O.F. van Gorselen, MD Philipp Fortner, MD Sebastian J. Buss, MD Derliz Mereles, MD Hugo A. Katus, MD Arthur A.M. Wilde, MD, PhD Connie R. Bezzina, PhD S. Matthijs Boekholdt, MD, PhD Patrick A. Schweizer, MD Imke Christiaans, MD, PhD *Department of Experimental Cardiology Academic Medical Center, University of Amsterdam Meibergdreef 15, Room K2-110 PO Box 22660 1100DD Amsterdam the Netherlands E-mail: E.M.Lodder@amc.uva.nl http://dx.doi.org/10.1016/j.jacc.2016.01.086 Please note: The authors acknowledge support from the Netherlands CardioVascular Research Initiative (CVON-PREDICT project) and the German Centre for Cardiovascular Research. Dr. Schweizer has received support from the MaxPlanck-Society (TANDEM project) and the Heidelberg Research Center for Molecular Medicine (Senior Career Fellowship). Dr. Thomas has received support from the German Cardiac Society and the Hengstberger Foundation (KlausGeorg and Sigrid Hengstberger Scholarship) and the Joachim Siebeneicher Foundation. The authors have reported that they have no relationships relevant to the contents of this paper to disclose. The authors thank Dr. van SpaendonckZwarts, Dr. van Haelst, and Dr. Caliskan for additional medical information and constructive discussions concerning this work. REFERENCES 1. Milanesi R, Baruscotti M, Gnecchi-Ruscone T, DiFrancesco D. Familial sinus bradycardia associated with a mutation in the cardiac pacemaker channel. N Engl J Med 2006;354:151–7. alence of obesity, particularly in young women. “Obesity” is an imprecise term that reflects body mass index, but not necessarily the deposition and distribution of fat in metabolically sensitive body compartments, such as the abdomen. This may explain the confusion in the scientific community and lay press regarding the cardiovascular risks of obesity in women. In 1997, Barrett-Connor (2) used the term “female pattern obesity” to describe non-visceral adiposity. Superficial reading of this terminology might lead some to believe that obesity is not a risk factor for women. In fact Barrett-Connor (2) specifically hypothesized that visceral obesity is the causative factor for cardiovascular disease (CVD) and that women with visceral adiposity are at higher risk than men. This hypothesis has been conclusively proven. Specific measures of visceral adiposity have been shown to be powerful predictors of CVD in women. What about the young women in the VIRGO study? Did they have visceral or peripheral adiposity? Data on waist circumference would be helpful if available. However, we know that in young women, but not older women, body mass index is a good predictor of visceral adiposity (3). In the absence of any other data, we can safely surmise that the young women experienced visceral obesity as an important risk factor for early CVD. The VIRGO study also evaluated perceived risk and health care provider discussion of risk, and it found both wanting. The VIRGO study reports that only 53% of the young patients believed that they were at risk for CVD and that obese women were significantly less likely than obese men to recognize this risk. This represents 2. Milano A, Vermeer AMC, Lodder EM, et al. HCN4 mutations in multiple families with bradycardia and left ventricular noncompaction cardiomyopathy. a compelling failure of public health education. In our J Am Coll Cardiol 2014;64:745–56. unaware of the CVD risks of tobacco. It is shameful 3. Schweizer PA, Schröter J, Greiner S, et al. The symptom complex of familial sinus node dysfunction and myocardial noncompaction is associated with mutations in the HCN4 channel. J Am Coll Cardiol 2014;64:757–67. current era, it is unlikely that there are any smokers that the same efforts have not been put into obesity education. 4. Campens L, Demulier L, De Groote K, et al. Reference values for echocar- The sociological stigma of obesity may also play diographic assessment of the diameter of the aortic root and ascending aorta spanning all age categories. Am J Cardiol 2014;114:914–20. a role. An obese woman is more likely to delay health 5. Liang X, Wang G, Lin L, et al. HCN4 dynamically marks the first heart field and conduction system precursors. Circ Res 2013;113:399–407. care if she feels that her provider holds a bias against her weight. Similarly, physicians are less likely to 2315 2316 JACC VOL. 67, NO. 19, 2016 Letters MAY 17, 2016:2310–9 perform diagnostic tests in obese patients based on the assumption that symptoms are due primarily to obesity (4). These factors are only exacerbated when mixed messages are received regarding the CVD risks of obesity. This muddle of confusion and stigma leads to denial of risk, denial of care, and ultimately, delayed diagnosis. At Southlake Regional Health Center, we Comparing the ATRIA, CHADS2, and CHA2DS2-VASc Scores for Stroke Prediction in Atrial Fibrillation studied the angiographic results in 616 young women from 2012 through 2014. In nonobese patients, the Van den Ham et al. (1) recently compared the ATRIA prevalence of normal coronary arteries was 22%, (Anticoagulation and Risk Factors in Atrial Fibrilla- whereas in obese young women, the prevalence of tion), CHADS 2 (congestive heart failure, hyperten- normal coronary arteries was only 14% (p ¼ 0.01). sion, age $75 years, diabetes mellitus, prior stroke or Similarly, 3-vessel disease was seen in 8.7% of non- transient obese young women compared with 16% of obese (congestive heart failure, hypertension, age $75 young women (p ¼ 0.005). Although these data have years, diabetes mellitus, prior stroke or transient limitations, relatively ischemic attack, vascular disease, age 65 to 74 years, delayed testing due to either health care avoidance or female) stroke risk scores in a primary care commu- underestimation of patient risk. nity cohort of patients with first-diagnosed atrial they are consistent with On the basis of the totality of evidence, the chief ischemic attack), and CHA 2DS2-VASc fibrillation (AF) not using oral anticoagulation (OAC) medical officer of the United Kingdom has declared for undefined reasons. They concluded that improved obesity to be a “national risk” (5). As a community, we risk prediction using the ATRIA score can reduce OAC have a duty to our patients to honestly recognize and overuse in low-risk patients. inform them of the risks of obesity. The VIRGO study However, the study raises major concerns. To put shows us that we are failing badly. Now is the time to their findings of a 2.99% annualized rate of stroke have an honest discussion. (excluding transient ischemic attack and systemic embolism) in appropriate context requires a better *Steven E.S. Miner, MD Lynne E. Nield, MD characterization of the cohort (e.g., a description of on- *Southlake Regional Health Center dian follow-up was only 0.74 years over a 15-year study University of Toronto period and patients were censored if given OAC, the 596 Davis Drive OAC prescription rate should have been shown. Newmarket, Ontario L3Y 2P9 OAC patients from the dataset). Also, because the me- The contemporary threshold for OAC use with non- Canada OACs can reasonably be decreased to an annual stroke E-mail: sminer@southlakeregional.org risk of approximately 1% (2). Using ATRIA in the http://dx.doi.org/10.1016/j.jacc.2016.01.084 aforementioned study, 40% of patients were catego- Please note: Both authors have reported that they have no relationships relevant to the contents of this paper to disclose. rized as low-risk (i.e., an ATRIA score #5, with REFERENCES 1. Leifheit-Limson EC, D’Onofrio G, Daneshvar M, et al. Sex differences in cardiac risk factors, perceived risk, and health care provider discussion of risk and risk modification among young patients with acute myocardial infarction: the VIRGO study. J Am Coll Cardiol 2015;66: 1949–57. 2. Barrett-Connor E. Sex differences in coronary heart disease. Why are women so superior? The 1995 Ancel Keys Lecture. Circulation 1997;95: 252–64. 3. Rankinen T, Kim SY, Perusse L, Despres JP, Bouchard C. The prediction of abdominal visceral fat level from body composition and anthropometry: ROC analysis. Int J Obes Relat Metab Disord 1999;23:801–9. 4. Drury CA, Louis M. Exploring the association between body weight, stigma of obesity, and health care avoidance. J Am Acad Nurse Pract 2002;14:554–60. annualized stroke rates of 0.40% to 1.99%), meaning that, for example, even a 74-year-old female with diabetes mellitus (i.e., 5 points) would not need OAC, nor would a 64-year-old male with hypertension and heart failure (2 points). However, a large body of evidence shows a positive net clinical benefit from OAC in such AF patients with $1 stroke risk factors (3). Their sensitivity analysis excluding renal dysfunction shows an unchanged performance of ATRIA, thus suggesting that the improvement in discriminatory ability over CHA2DS2 -VASc might be driven solely by the inclusion of more age categories. However, if the ATRIA “low-risk” category is restricted to stroke rates of
Purchase answer to see full attachment