Vous êtes sur la page 1sur 11

REVIEW ARTICLE

Physical Activity and Mortality in Individuals With Diabetes Mellitus


A Prospective Study and Meta-analysis
Diewertje Sluik, MSc; Brian Buijsse, PhD; Rebecca Muckelbauer, PhD; Rudolf Kaaks, PhD; Birgit Teucher, PhD; Nina Fns Johnsen, PhD; Anne Tjnneland, PhD; Kim Overvad, PhD; Jane Nautrup stergaard, PhD; Pilar Amiano, MSc; Eva Ardanaz, PhD; Benedetta Bendinelli, PharmD; Valeria Pala, PhD; Rosario Tumino, MD; Fulvio Ricceri, MSc; Amalia Mattiello, MD; Annemieke M. W. Spijkerman, PhD; Evelyn M. Monninkhof, PhD; Anne M. May, PhD; Paul W. Franks, PhD; Peter M. Nilsson, MD, PhD; Patrik Wennberg, PhD; Olov Rolandsson, MD, PhD; Guy Fagherazzi, PhD; Marie-Christine Boutron-Ruault, MD, PhD; Franc oise Clavel-Chapelon, PhD; Jose Mar a Huerta Castan o, PhD; Valentina Gallo, PhD; Heiner Boeing, PhD; Ute No thlings, DrPH

Background: Physical activity (PA) is considered a cornerstone of diabetes mellitus management to prevent complications, but conclusive evidence is lacking. Methods: This prospective cohort study and metaanalysis of existing studies investigated the association between PA and mortality in individuals with diabetes. In the EPIC study (European Prospective Investigation Into Cancer and Nutrition), a cohort was defined of 5859 individuals with diabetes at baseline. Associations of leisure-time and total PA and walking with cardiovascular disease (CVD) and total mortality were studied using multivariable Cox proportional hazards regression models. Fixed- and random-effects meta-analyses of prospective studies published up to December 2010 were pooled with inverse variance weighting. Results: In the prospective analysis, total PA was associated with lower risk of CVD and total mortality. Compared with physically inactive persons, the lowest mortality risk was observed in moderately active persons:

hazard ratios were 0.62 (95% CI, 0.49-0.78) for total mortality and 0.51 (95% CI, 0.32-0.81) for CVD mortality. Leisure-time PA was associated with lower total mortality risk, and walking was associated with lower CVD mortality risk. In the meta-analysis, the pooled randomeffects hazard ratio from 5 studies for high vs low total PA and all-cause mortality was 0.60 (95% CI, 0.490.73).
Conclusions: Higher levels of PA were associated with lower mortality risk in individuals with diabetes. Even those undertaking moderate amounts of activity were at appreciably lower risk for early death compared with inactive persons. These findings provide empirical evidence supporting the widely shared view that persons with diabetes should engage in regular PA.

Arch Intern Med. 2012;172(17):1285-1295. Published online August 6, 2012. doi:10.1001/archinternmed.2012.3130 Lifestyle measures, including physical activity (PA), are key factors for selfmanagement in patients with diabetes to prevent macrovascular complications and premature mortality.5 Increased PA has long been considered a cornerstone of diabetes management. Persons with diabetes are recommended to engage in at least 150 minutes per week of moderateintensity aerobic PA.5,6 Walking has been of particular interest because it requires no specific facilities, can be easily implemented in the daily routine, and is relatively safe.7 In the general population, being physically active has been associated with a lower risk of overall and cardiovascular disease (CVD) mortality compared with being

jor cause of illness and premature death in most countries.1 Efforts to reduce the impact of diabetes complications have been predominantly aimed at controlling hyperglycemia, hypertension, and dyslipidemia by using

IABETES MELLITUS IS A MA-

See also pages 1283 and 1306


medication strategies, despite the lack of evidence of long-term benefits.2,3 However, diabetes management should extend to an overall intervention strategy that includes lifestyle modification to reduce the risk of complications.4

Author Affiliations are listed at the end of this article.

Author Affil the end of th

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1285

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013

inactive.8 Because persons with diabetes are at higher risk for CVD and premature death, it is important to determine whether PA can produce similar beneficial effects in this high-risk population. Indeed, a metaanalysis9 of 14 controlled trials in diabetic persons showed that exercise programs had beneficial effects on glycemic control. Several prospective cohort studies 10-21 have found that higher PA levels were associated with reduced CVD and total mortality rates, but conclusive highlevel evidence is lacking. The objective was to investigate whether PAtotal, leisure time, and walkingwas associated with CVD and total mortality in a large cohort of individuals with diabetes. A meta-analysis summarizing evidence from prospective studies was performed to put the present findings in context and to provide a higher level of evidence.
METHODS

with missing PA data (n=349, including the cohort in Umea , Sweden), the analytical sample included 5859 individuals.

PA ASSESSMENT
At baseline, participants received a lifestyle questionnaire by mail, which they completed at home. This questionnaire asked about occupational activity and duration and frequency of walking, cycling, gardening (average values of summer and winter), household work, doit-yourself activities, and sports during the past year. Total PA was investigated using the Cambridge Physical Activity Index,25 which combines self-reported occupational activity with time participating in cycling and sports. Occupational activity was categorized as sedentary, standing, manual, or heavy manual. The sum of hours per week spent on cycling and sports was categorized into 4 levels. Based on a 4 4 matrix, participants were divided into 4 categories, that is, inactive (sedentary job and no recreational activity), moderately inactive, moderately active, and active (sedentary job with 1 hour of recreational activity per day, standing or physical job with some recreational activity, or a heavy manual job). The index has been shown to have acceptable repeatability, and it was positively associated with objective measures of the ratio of daytime expenditure to resting metabolic rate and cardiorespiratory fitness.25 Leisure-time PA included walking, cycling, gardening, sports, household work, and do-it-yourself activities. Duration and frequency were directly assessed, and intensity, that is, energy expenditure, was estimated by assigning metabolic equivalents (METs), ranging from 3 for walking and household activities to 6 for sports.26 A MET is defined as the ratio of work metabolic rate to a standard metabolic rate of 1.0 kcal (4.184 kJ) kg1 h1. One MET is the energy expended by a person while sitting quietly.

style questionnaire included a question about insulin therapy. When a participant did not report the use of diabetes medication during the visit or reported insulin therapy in the questionnaire, we assumed that the participant did not take medication. Dietary intake, including alcohol consumption during the past year, was assessed using country-specific instruments. 2 2 Weight and height were measured with participants not wearing shoes. Systolic and diastolic blood pressures were measured by trained personnel at baseline except in Navarra, Spain. The HbA1c level was measured in erythrocytes of blood collected at baseline except in Denmark. Smoking behavior, educational level, and prevalence of myocardial infarction, stroke, and cancer were assessed using questionnaires.

OUTCOME ASCERTAINMENT
Sixty-one participants (1%) were lost to follow-up. For those completely followed up, causes and dates of deaths were ascertained using record linkages with local, regional, or central cancer registries, boards of health, or death indices. An exception was Germany, where deceased participants were identified with follow-up mailings and subsequent inquiries to municipality registries, regional health departments, physicians, or hospitals. Mortality data were coded according to the International Classification of Diseases, Injuries, and Causes of Death, Tenth Revision, using the codes I00-I99 for CVD mortality.

STUDY DESIGN AND POPULATION


The EPIC study (European Prospective Investigation Into Cancer and Nutrition) is an ongoing prospective study of 519 978 men and women aged 35 to 70 years from 23 study centers in 10 European countries.22 Within the EPIC study, a cohort was established of participants with a confirmed diagnosis of diabetes mellitus at baseline between 1992 and 2000. As described previously,23,24 15 study centers from 6 countries provided additional data on diabetes diagnosis and medications. No information was available to distinguish type 1 and 2 diabetes mellitus. To be considered diabetic, a self-reported diagnosis at baseline had to be confirmed by at least 1 additional information source. Depending on the available options in the study centers, these sources included contact with a physician, self-reported use of diabetes medication, confirmation of self-reported diabetes status during follow-up, linkage to diabetes registries, and a baseline glycated hemoglobin (HbA1c) level greater than 6%. The cohort comprised 6412 individuals with confirmed diabetes at study enrollment. After the exclusion of participants without follow-up information on vital status (n = 27), participants with extreme or implausible energy intakes (n=177), and participants

STATISTICAL ANALYSIS
Hazard ratios (HRs) and 95% CIs of total and CVD mortality were calculated using Cox proportional hazards regression and a commercially available software program (SAS, version 9.2; SAS Institute, Inc).27 Total PA was analyzed in 4 categories, and leisure-time PA (METhours per week) and walking (hours per week) were analyzed in quartiles. The lowest category or quartile was the reference. Center and age at recruitment in 1-year categories were entered as stratum variables.28 Age was used as the underlying time scale, with entry time defined as the participants age (in years) at recruitment and exit time defined as the participants age (in years) at death or censoring. The HRs were adjusted for sex (model 1); disease duration (years); use of diabetes-related medication (none, insulin, oral hypoglycemic agents, or both); self-reported myocardial infarc-

COVARIATE ASSESSMENT
Diabetes duration was calculated by subtracting the self-reported year of diagnosis or, when available, the exact date of diagnosis supplied by the physician from the year of baseline examination. Insulin therapy or use of oral hypoglycemic agents was either self-reported during the visit at the study center or was obtained through medical verification; this information was not collected in Spain and Denmark. Moreover, the life-

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1286

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013

tion, stroke, or cancer; alcohol consumption (grams per day); smoking status (never, former [quit 10, 11-20, or 20 years ago], or current), smoking duration (10, 11-20, 21-30, 31-40, or 41 years), and number of cigarettes currently smoked (15, 15-24, or 25 per day); education (5 categories); energy intake (kilocalories per day); and factor scores for the first 3 patterns derived from factor analysis on 16 food groups (model 2). Among the covariates, proportions of missing data were 31% for diabetes medication, 8% for disease duration, 5% for prevalence of cancer, 1% for prevalence of myocardial infarction or stroke, and less than 1% for diet and education. These missing values were imputed using the multiple imputation technique.29 All the variables included in the multivariable adjustment model were included in the imputation procedure, and 20 duplicate data sets were sampled.30 In sensitivity analyses, it was checked whether additional adjustment for HbA1c level, BMI, and systolic blood pressure affected the risk estimates. Prevalent cases of myocardial infarction, stroke, and cancer and participants with follow-up of less than 2 years (n = 5039) were excluded. Statistical interaction by sex, body mass index (BMI), and insulin therapy was tested by adding a product term to the multivariable adjustment model. Finally, results were compared with analyses without multiple imputation (n = 5376), in which participants with missing values for continuous variables were excluded, and missing values for categorical variables were modeled as a separate indicator variable.

verse variance weighting were performed using R package meta (version 2.12.2). Heterogeneity was assessed by the Q statistic and the I2 index. RESULTS

BASELINE CHARACTERISTICS Individuals who were more physically active were younger, were more likely to be male, had a lower BMI and a lower HbA1c level, and had a shorter diabetes duration than did those who were inactive (Table 1). They were also more likely to use insulin and to report fewer comorbidities. PROSPECTIVE ANALYSIS After median follow-up of 9.4 years, 755 participants had died (13%). Death due to CVD accounted for 28% of all deaths (n = 212). Total PA was inversely associated with total and CVD mortality (Table 2). The lowest HR was observed in persons categorized as moderately active: the HR in the multivariable adjustment model was 0.62 (95% CI, 0.490.78) for total mortality. When excluding heavy manual workers and nonworkers from the analyses, the lowest risk was still observed in the moderately active group. Leisuretime PA was also associated with a lower risk of CVD and total mortality: the HR in the highest category was 0.73 (95% CI, 0.57-0.93) for total mortality. The association with CVD mortality was weaker in magnitude and nonsignificant but showed the same trend. Participants who walked more than 2 hours per week had lower CVD mortality risk compared with those in the lowest activity group: the HR in the category of 2 to 4.5 hours per week was 0.54 (95% CI, 0.36-0.82). The relationship of walking with total mortality was less pronounced. Additional adjustment for vigorous PA did not alter the association. Adjustment for intermediate factors did not affect the risk estimates. For total PA, the HR in moderately active persons was 0.62 (95% CI, 0.50-0.79) after additional adjustment for HbA1c and 0.62 (95% CI, 0.49-0.78) after additional ad-

justment for BMI or systolic blood pressure. Excluding participants with comorbidities at baseline led to lower HRs: for total PA, the HR in moderately active persons was 0.58 (95% CI, 0.43-0.77) for total mortality and 0.31 (95% CI, 0.15-0.60) for CVD mortality. Sex seemed to modify the association between total PA and total mortality (P for interaction = .04, multivariable model). Women had a lower HR across quartiles than did men, but the trend showed the same direction (the HR in the highest category was 0.79 [95% CI, 0.59-1.05] in men and 0.59 [95% CI, 0.36-0.96] in women). The analyses without multiple imputation showed similar results (the HR for total mortality in the highest category of total PA was 0.70 [95% CI, 0.54-0.90]), indicating that missing observations did not influence the effect estimates. META-ANALYSIS In the 12 cohort studies10,13-15,17-21,32-34 included in the meta-analyses, verification criteria of diabetes status ranged from self-report to an oral glucose tolerance test; 4 studies used official diagnostic criteria (Table 3). Sample sizes ranged from 29221 to 5859 (the present study), and mean follow-up was 12.5 years. Physical activity was assessed by questionnaire in 8 studies10,14,17,19,20,32-34 and by interview in 4.13,15,18,21 Participants were divided into PA categories ranging from inactive to active. All but 1 study13 adjusted for a wide range of multiple risk factors. Study quality ranged from 613,14,19,34 to 918 stars of a maximum of 9. Three studies19,20,33 reported on total PA but measured only leisure-time activities and were, therefore, classified as such. The meta-analyses of prospective studies showed that the highest levels of total and leisure-time PA and walking were associated with a lower risk of total and CVD mortality compared with a low activity level ( Figures 1 , 2 , and 3 ). Magnitudes of associations were similar for total and CVD mortality. Significant heterogeneity was found for total PA and total mortality (I2 = 69%, P = .01) and for walking and total mortality (I2 = 71%;

META-ANALYSIS
A systematic literature search in MEDLINE and ISI Web of Knowledge for prospective studies on PA published up to December 2010 yielded 4344 publications, of which 12 were included in the meta-analysis (eFigure 1; http://www .archinternmed.com). Study quality scores were assigned using the NewcastleOttawa Scale (eTable); quality criteria included representativeness, exposure and outcome ascertainment, adjustment, follow-up, and attrition.31 The most complete adjusted HRs and 95% CIs of the highest vs the lowest activity category were extracted. Study selection, quality assessment, and data extraction was performed by 2 of us (D.S. and B.B.) independently; any discrepancies between the 2 were resolved by discussion. Fixed- and random-effects meta-analyses with in-

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1287

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013

Table 1. Baseline Characteristics of 5859 Individuals With Diabetes Mellitus Across Categories of Total Physical Activity
Total Physical Activity a Variable Men, % Age, mean (SD), y HbA1c, mean (SD), % Disease duration, median (IQR), y Medication use, % None Insulin and OHA Insulin only OHA only Types of physical activity, MET-h/wk, median (IQR) Cycling Sports Gardening Do-it-yourself activities Housework activity Walking Stair climbing Vigorous physical activity Physical activity at work, % Sedentary occupation Standing occupation Manual work Heavy manual work Nonworker Unknown BMI, mean (SD) Men Women Waist-height ratio, mean (SD) Men Women Energy intake, mean (SD), kcal/d Alcohol consumption, median (IQR), g/d Factor scores, mean (SD) b Healthy pattern Traditional pattern Modern pattern Smoking status, % Never Former Current Educational level, % None Primary school Technical/professional school Secondary school Higher, including university degree Comorbidities, % Hypertension Myocardial infarction Stroke Cancer Inactive (n = 1793) 47 58.5 (6.6) 8.3 (2.0) 5.2 (2.3-11.0) 24 8 24 43 0 (0-0) 0 (0-0) 0 (0-8) 0 (0-4.5) 27 (6-63) 13.5 (4.5-27) 0.5 (0-1.6) 2 (1-5) 28 0 0 0 69 3 28.8 (4.4) 30.0 (5.7) 0.59 (0.07) 0.59 (0.09) 2007 (606) 3.2 (0-17.0) 0.01 (0.98) 0.13 (0.99) 0.04 (0.98) 41 33 26 6 43 22 12 16 59 8 5 4 Moderately Inactive (n = 1897) 53 58.0 (6.4) 8.2 (1.9) 5.1 (2.1-10.6) 27 11 25 38 3 (0-9) 0 (0-6) 3 (0-12) 0 (0-9) 18 (6-48) 13.5 (6-25.5) 1 (0-2.3) 2 (1-3) 26 17 0 0 54 3 28.4 (4.1) 29.1 (5.5) 0.58 (0.07) 0.57 (0.08) 2073 (613) 6.9 (0.9-21.6) 0.07 (0.98) 0.04 (0.95) 0.01 (1.01) 38 38 24 3 40 27 11 18 57 7 4 5 Moderately Active (n = 1171) 61 56.6 (6.6) 8.1 (1.9) 4.6 (1.9-9.9) 26 10 29 36 9 (0-24) 0 (0-12) 4 (0-14) 0 (0-9) 15 (3-42) 13.5 (6-25.5) 1.0 (0-2.6) 2 (1-3) 16 31 16 0 35 2 28.4 (4.4) 28.7 (5.1) 0.58 (0.07) 0.56 (0.08) 2138 (648) 8.2 (1.5-25.5) 0.004 (1.04) 0.05 (1.00) 0.01 (0.99) 37 38 25 4 40 28 11 17 52 6 3 4 Active (n = 998) 61 56.6 (6.3) 8.1 (2.0) 4.2 (2.0-8.2) 22 13 36 29 23.3 (6-45) 6 (0-21) 4 (0-12) 0 (0-9) 15 (3-36) 15 (7.5-27) 1.2 (0-2.6) 3 (1-4) 9 21 32 12 25 1 28.0 (4.1) 28.7 (5.6) 0.57 (0.06) 0.56 (0.09) 2229 (676) 9.3 (1.7-25.6) 0.13 (1.02) 0.27 (1.05) 0.03 (1.03) 37 37 26 3 43 31 9 14 52 6 2 2

Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); HbA1c, glycated hemoglobin; IQR, interquartile range; MET, metabolic equivalent; OHA, oral hypoglycemic agent. a Classification of total physical activity level according to the Cambridge Physical Activity Index based on occupational activity and duration of cycling and sports. b Derived from factor analysis of 16 food groups, the healthy pattern (eigenvalue, 2.16) was characterized by high intake of fruit, vegetables, legumes, and fish; the traditional pattern (eigenvalue, 2.00) by high intake of potatoes, dairy, eggs, meat, and sugar; and the modern pattern (eigenvalue, 1.76) by high intake of cereals, fats, cakes, and biscuits.

P = .01). For walking, an influential meta-analysis revealed that the present study contributed most

heterogeneity, although excluding the present study from the metaanalyses did not substantially influ-

ence the pooled HR. Visual inspection of funnel plots did not indicate publication bias (eFigure 2).

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1288

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013

Table 2. HRs (95% CIs) for Associations Between Total Physical Activity, Leisure-Time Physical Activity, and Walking and Total and CVD Mortality in 5859 Individuals With Diabetes Mellitus
Variable Total mortality Cases/PY Incidence rate per 1000 PY Sex-adjusted HR (95% CI) b Multivariable HR (95% CI) c Multivariable HR (95% CI) d CVD mortality Cases/PY Incidence rate per 1000 PY Sex-adjusted HR (95% CI) b Multivariable HR (95% CI) c Multivariable HR (95% CI) d Inactive Moderately Inactive Total Physical Activity a 304/15 941 19.1 1 [Reference] 1 [Reference] 1 [Reference] 99/15 941 6.2 1 [Reference] 1 [Reference] 1 [Reference] 45 Total mortality Cases/PY Incidence rate per 1000 PY Sex-adjusted HR (95% CI) b Multivariable HR (95% CI) c Multivariable HR (95% CI) d CVD mortality Cases/PY Incidence rate per 1000 PY Sex-adjusted HR (95% CI) b Multivariable HR (95% CI) c Multivariable HR (95% CI) d 271/17 230 15.7 0.64 (0.53-0.76) 0.69 (0.57-0.83) 0.68 (0.54-0.66) 61/17 230 3.5 0.59 (0.43-0.82) 0.65 (0.46-0.91) 0.46 (0.28-0.74) 115/10 768 10.7 0.53 (0.42-0.66) 0.62 (0.49-0.78) 0.58 (0.43-0.77) 27/10 768 2.5 0.40 (0.26-0.63) 0.51 (0.32-0.81) 0.31 (0.15-0.60) 119/9253 12.9 0.62 (0.50-0.78) 0.74 (0.59-0.94) 0.81 (0.61-1.08) 25/9253 2.7 0.53 (0.34-0.85) 0.62 (0.38-1.01) 0.48 (0.25-0.94) 113 121/13 544 8.9 0.64 (0.50-0.81) 0.73 (0.57-0.93) 0.62 (0.46-0.85) 29/13 544 2.1 0.57 (0.35-0.93) 0.63 (0.38-1.04) 0.30 (0.14-0.64) 9.0 161/12 183 13.2 0.95 (0.75-1.19) 0.95 (0.75-1.20) 0.90 (0.67-1.21) 48/12 183 3.9 0.69 (0.46-1.06) 0.64 (0.41-0.98) 0.65 (0.35-1.19) Moderately Active Active P Trend

.001 .001 .03

.001 .004 .001

Leisure-Time Physical Activity, MET-h/wk e 45-74 75-113 191/11 976 15.9 0.80 (0.66-0.98) 0.85 (0.70-1.04) 0.77 (0.60-0.99) 66/11 976 5.5 1.09 (0.76-1.57) 1.18 (0.81-1.73) 0.91 (0.54-1.52) Walking, h/wk 2.0-4.5 159/11 778 13.5 0.83 (0.67-1.02) 0.88 (0.71-1.09) 0.87 (0.67-1.12) 37/11 778 3.1 0.52 (0.35-0.78) 0.54 (0.36-0.82) 0.40 (0.22-0.74) 174/12 864 13.5 0.74 (0.60-0.91) 0.80 (0.64-0.99) 0.79 (0.60-1.03) 50/12 864 3.9 0.83 (0.56-1.24) 0.90 (0.60-1.37) 0.69 (0.39-1.24) 4.6-9.0 166/13 302 12.5 0.83 (0.67-1.02) 0.86 (0.70-1.07) 0.76 (0.58-1.00) 37/13 302 2.8 0.49 (0.32-0.75) 0.50 (0.32-0.77) 0.44 (0.24-0.79)

269/14 809 18.2 1 [Reference] 1 [Reference] 1 [Reference] 67/14 809 4.5 1 [Reference] 1 [Reference] 1 [Reference] 2.0

.001 .007 .003

.02 .06 .002

Total mortality Cases/PY Incidence rate per 1000 PY Sex-adjusted HR (95% CI) b Multivariable HR (95% CI) c Multivariable HR (95% CI) d CVD mortality Cases/PY Incidence rate per 1000 PY Sex-adjusted HR (95% CI) b Multivariable HR (95% CI) c Multivariable HR (95% CI) d

269/15 930 16.9 1 [Reference] 1 [Reference] 1 [Reference] 90/15 930 5.6 1 [Reference] 1 [Reference] 1 [Reference]

.80 .70 .24

.21 .10 .06

Abbreviations: CVD, cardiovascular disease; HR, hazard ratio; MET, metabolic equivalent; PY, person-years. a Classification of total physical activity level according to the Cambridge Physical Activity Index based on occupational activity and duration of cycling and sports. b Model 1: age and center stratified and adjusted for sex. c Model 2: model 1 additionally adjusted for diabetes medication (no medication, insulin, oral hypoglycemic agents, or both); disease duration; self-reported myocardial infarction, stroke, or cancer; alcohol consumption; smoking behavior; educational attainment; energy; and scores for the first 3 dietary patterns derived from a factor analysis of 16 food groups. d Model 3: excluding participants with self-reported myocardial infarction, stroke, or cancer or follow-up of less than 2 years (n = 5039). e Leisure-time physical activity included walking, cycling, gardening, sports, and household and do-it-yourself activities.

COMMENT

In this prospective analysis and meta-analysis of individuals with diabetes, higher levels of total PA, leisure-time PA, and walking were associated with a lower risk of total and CVD mortality. In the prospective analysis, people who reported being moderately physically active had lower mortality risk compared with

those who reported being physically inactive. In persons with diabetes, an increase in PA has been shown to reduce HbA1c levels9,35 and improve insulin sensitivity.36 Moreover, PA has been shown to have beneficial effects on inflammation, hypertension, dyslipidemia, endothelial function, and abdominal adiposity in persons with and without diabetes.37-40

PROSPECTIVE ANALYSIS The association between total PA and mortality was slightly Jshaped. This could have been due to misclassification of activity levels, which may be higher in the most physically active group owing to labeling bias. This result is in contrast to the other studies15,19,21,32,34 included in the meta-analysis, which

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1289

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013

Table 3. Identified Published Prospective Cohort Studies on PA and Mortality in Individuals With Diabetes Mellitus
Patients, No./Sex/ Age, y
10

Source Batty et al, 2002, Whitehall II Study, United Kingdom

Diabetes Verification Type 2 diabetes or IGT: oral glucose tolerance test

Follow-up, Mean, y 25.0

Exposure

Outcome, No. of Cases

Outcome Ascertainment

Adjustments

Stars a 7

352/M/ 40-64

Ford and DeStefano,13 1991, National Health and Nutrition Examination Survey I Epidemiologic Follow-up Study (1982-1984), United States Gaziano et al,14 2002, Physicians Health Study enrollment cohort, United States Gregg et al,15 2003, National Health Interview Survey, United States

602/M-F/ 40-77

Self-reported diabetes

10.0

Walking pace: 215 Total National Health Age, employment mortality, Service Register grade, systolic slower/same/faster Leisure-time activity: 79 CHD blood pressure, inactive/moderately mortality, cholesterol level, active/active 39 other CVD smoking, BMI, (questionnaire) mortality forced expiratory volume in 1 second, disease at study entry Leisure-time PA: 233 Total National Death None most mortality, Index and other active/moderately 92 CHD tracing methods mortality active/inactive (interview)

2838/M/ 40-84

Self-reported diabetes

5.2

2896/MF/18

Self-reported diabetes

Vigorous exercise: 1-3 times per mo, 1 time/wk, 2-4 times/wk, 5 times/wk (questionnaire) Walking: 0, 0-1.9, 2 h/wk Total PA: 0, 0-1.9, 2 h/wk (interview)

356 Total mortality

Death certificate, Age, smoking, medical records, alcohol, history of next of kin angina or transient ischemic attack

671 Total National Death mortality, 316 Index CVD mortality

Hu et al,17 2004, six independent population surveys (1972-1997), Finland

3316/M-F/ Type 2 diabetes 25-74 confirmed by WHO criteria

18.4

18.7 Hu et al,32 2005, six 3708/M-F/ Type 2 diabetes independent 25-74 confirmed by population WHO criteria surveys (1972-1997), Finland Jonker et al,21 2006, 292/M-F/ Random blood 12.0 (3 Framingham 28-62 glucose 200 Pooled Heart Study, mg/dL or follow-up United States treatment with periods) hypoglycemic agent

Occupational: 1410 Total Statistics Finland light/moderate/ mortality, 903 active; commuting: CVD mortality 0, 1-29, 30 min/d; leisure-time activity: low/moderate/high (questionnaire) Total PA: 1423 Total Statistics Finland low/moderate/high mortality, 906 (questionnaire) CVD mortality

Age, sex, race, BMI, self-rated health, smoking status, weight loss approaches, hospitalizations, hypertension, physician visits, limitations due to CVD and cancer, level of functional limitation Age, sex, study year, BMI, systolic blood pressure, cholesterol level, smoking status

Total PA: 292 Total low/moderate/high mortality (interview)

Age, sex, study year, educational level, BMI, systolic blood pressure, total cholesterol level, smoking status Office and Age, sex, educational level, marital hospitalization status, smoking, records, laboratory test baseline diseases results, death (CVD, cancer, left certificates, and ventricular autopsy reports hypertrophy, arthritis, ankle edema, pulmonary disease), examination at start of follow-up

(continued)

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1290

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013

Table 3. Identified Published Prospective Cohort Studies on PA and Mortality in Individuals With Diabetes Mellitus (continued)
Patients, No./Sex/ Age, y
33

Source Nelson et al, 2010, Third National Health and Nutrition Examination Survey, United States Smith et al,18 2007, Rancho Bernardo Study, United States

Diabetes Verification Self-report

Follow-up, Mean, y 7.6

Exposure Leisure-time PA: inactive, insufficient, recommended levels (questionnaire)

Outcome, No. of Cases 642 Total mortality

Outcome Ascertainment National Death Index

Adjustments

Stars a 8

1507/MF/17

347/M-F/ 50-90

Type 2 diabetes confirmed by WHO criteria

10

Tanasescu et al,19 2003, Health Professionals Follow-up Study (1986-1998), United States

2803/M/ 40-75

Trichopoulou et al,34 2006, European Prospective Investigation Into Cancer and Nutrition, Greece

1013/MF/35

Self-reported diagnosis at 30 y confirmed with 1 classic symptoms plus raised plasma glucose or OHA use Self-reported diagnosis and use of diabetes medication

14

4.5

Wei et al,20 2000, Aerobics Center Longitudinal Study, United States

1188/M/50 Type 2 diabetes defined according to ADA criteria

11.7

Age, sex, race/ethnicity, educational level, HbA1c level, HDL cholesterol level, BMI, and smoking status Walking: nonwalker, 538 Total Annual mailings Sex, age, smoking and telephone status, BMI, 1-mile walker, mortality, 1-mile walker 143 CHD calls, death average drinks per (interview) mortality, certificates day, exercise, 138 other obtained hypertension, CVD mortality triglyceride levels, HDL cholesterol level, history of CHD Walking: 0-1.4, 355 Total Reported in or by Alcohol use, smoking 1.5-4.1, 4.2-7.9, mortality, 96 linkage with the status, family MI 8.0-16.0, 16.1 CVD mortality National Death history, vitamin E MET-h/wk Index, CVD use, disease Leisure-time PA: confirmed by duration, OHAs, review of dietary intake of 0-5.1, 5.2-12.0, trans and saturated 12.1-21.7, medical records 21.8-37.1, 37.2 or autopsy fat, fiber and folic MET-h/wk report acid, history of CVD, hypertension, (questionnaire) cholesterol level 80 Total Active follow-up Sex, age, educational Total (occupational and leisure-time) mortality by qualified level, smoking PA: 30, 30-32, physicians status, waist-height 32-34, 34-37, ratio, hip 37 MET-h/d circumference, (questionnaire) insulin use, hypertension, or hyperlipidemia treatment Leisure-time PA: 180 Total National Health Age, examination active/inactive mortality Index. Death year, CVD history, (questionnaire) certificates cholesterol level, smoking status, obtained diabetes status, glucose level, alcohol use, hypertension, overweight

Abbreviations: ADA, American Diabetes Association; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CHD, coronary heart disease; CVD, cardiovascular disease; HbA1c, glycated hemoglobin; HDL, high-density lipoprotein; IGT, impaired glucose tolerance; MET, metabolic equivalent; MI, myocardial infarction; OHA, oral hypoglycemic agents; PA, physical activity; WHO, World Health Organization. a Stars (maximum = 9) indicate the quality of the studies assessed using the Newcastle-Ottawa Scale.31

showed linear inverse associations, with the lowest observed HR in the highest quartile. In the present study, participants with heavy manual work occupations were automatically assigned to the active category. Because such people are more frequently exposed to occupational risk factors and more often have a lower socioeconomic status, they may have a more unfavorable risk profile.41 However, excluding heavy manual workers and nonworkers from the analyses did not change the find-

ings. Thus, diabetic individuals who are physically inactive seem to have a higher risk of early death, and already being moderately active may improve survival. Walking may reduce the risk of CVD in people with diabetes by improving glycemic control and other risk factors.7 In the present study and meta-analysis, persons in the highest quartiles of walking duration had a lower risk of CVD mortality compared with those in the lowest quartile. In contrast, walking was not re-

lated to significantly lower total mortality risk in this study, whereas other studies10,15,18,19 in the metaanalysis reported strong inverse relationships. In the present study, persons in the highest category reported walking more than 9 hours per week. Walking levels were lower in the other studies included: in comparison, persons in the active category in the study by Tanasescu et al19 walked more than 16 MET-hours per week. It is known that Europeans are more active than North American

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1291

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013

A Source
Gregg et al,15 2003 Hu et al,32 2005 Jonker et al,21 2006 Trichopoulou et al,34 2006 Present study

Population
National Health Interview Survey, United States Population surveys, Finland Framingham Heart Study, United States EPIC, Greece EPIC, Europe

HR (95% CI)
0.71 (0.58-0.86) 0.52 (0.45-0.60) 0.53 (0.39-0.72) 0.33 (0.16-0.71) 0.74 (0.59-0.93) 0.59 (0.54-0.66) 0.60 (0.49-0.73)

Combined fixed-effects model Combined random-effects model Heterogeneity: I 2 = 69% (95% CI, 19%-88%), Q = 12.71; (P = .01)

0.25

0.5

1.0

2.0

Hazard Ratio B Source


Gregg et al,15 2003 Hu et al,32 2005 Present study

Population
National Health Interview Survey, United States Population surveys, Finland EPIC, Europe

HR (95% CI)
0.76 (0.57-1.02) 0.52 (0.44-0.62) 0.62 (0.38-1.01) 0.58 (0.50-0.66) 0.61 (0.47-0.80)

Combined fixed-effects model Combined random-effects model Heterogeneity: I 2 = 59% (95% CI, 0%-88%), Q = 4.83; (P = .09)

0.25

0.5

1.0

2.0

Hazard Ratio

Figure 1. Hazard ratios (HRs) and 95% CIs for associations between total physical activity (highest vs lowest category) and total (A) and cardiovascular (B) mortality for individual cohort studies, including the present study, and all the cohort studies combined. EPIC indicates European Prospective Investigation Into Cancer and Nutrition; black squares, estimates for the individual studies; solid horizontal lines, 95% CIs; and white diamonds and dashed vertical lines, combined estimates for the analysis (the width of the diamond represents the 95% CI).

A Source
Batty et al,10 2002 Ford and DeStefano,13 1991 Gaziano et al,14 2002 Hu et al,17 2004 Nelson et al,33 2010 Tanasescu et al,19 2003 Wei et al,20 2000 Present study

Population
Whitehall Study, United Kingdom NHANES, United States Physicians Health Study, United States Population surveys, Finland NHANES III, United States Health Professionals Follow-up Study, United States Aerobics Center Longitudinal Study, United States EPIC, Europe

HR (95% CI)
0.61 (0.40-0.92) 0.84 (0.48-1.47) 0.45 (0.31-0.66) 0.73 (0.57-0.94) 0.63 (0.49-0.80) 0.58 (0.41-0.83) 0.56 (0.40-0.78) 0.73 (0.57-0.93) 0.64 (0.57-0.72) 0.64 (0.57-0.72)

Combined fixed-effects model Combined random-effects model Heterogeneity: I 2 = 6% (95% CI, 0%-69%), Q = 7.43; (P = .39)

0.25

0.5

1.0

2.0

Hazard Ratio B Source


Batty et al,10 2002 Ford and DeStefano,13 1991 Hu et al,17 2004 Tanasescu et al,19 2003 Present study

Population
Whitehall Study, United Kingdom NHANES, United States Population surveys, Finland Health Professionals Follow-up Study, United States EPIC, Europe

HR (95% CI)
0.39 (0.21-0.72) 1.35 (0.49-3.71) 0.70 (0.51-0.96) 0.55 (0.28-1.08) 0.63 (0.38-1.04) 0.64 (0.51-0.80) 0.63 (0.48-0.83)

Combined fixed-effects model Combined random-effects model Heterogeneity: I 2 = 21% (95% CI, 0%-67%), Q = 5.09; (P = .28)

0.25

0.5

1.0

2.0

Hazard Ratio

Figure 2. Hazard ratios (HRs) and 95% CIs for associations between leisure-time physical activity (highest vs lowest category) and total (A) and cardiovascular (B) mortality for individual cohort studies, including the present study, and all the cohort studies combined. EPIC indicates European Prospective Investigation Into Cancer and Nutrition; NHANES, National Health and Nutrition Examination Survey; black squares, estimates for the individual studies; solid horizontal lines, 95% CIs; and white diamonds and dashed vertical lines, combined estimates for the analysis (the width of the diamond represents the 95% CI).

populations,42 and it has been observed in a Dutch population that activities of at least moderate intensity, but not lower intensity, such as walking, were related to reduced CVD incidence.43 This seems to be

in contrast to our findings on walking and CVD mortality. However, because no information on walking pace was available, we cannot draw conclusions about walking intensity. In conclusion, although these

results did not reach statistical significance, from the meta-analysis the potential benefits of walking on mortality are well established. Reverse causality could have overestimated the mortality risks if dia-

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1292

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013

A Source
Batty et al,10 2002 Gregg et al,15 2003 Smith et al,18 2007 Tanasescu et al,19 2003 Present study

Population
Whitehall Study, United Kingdom National Health Interview Survey, United States Rancho Bernardo Study, United States Health Professionals Follow-up Study, United States EPIC, Europe

HR (95% CI)
0.42 (0.25-0.69) 0.61 (0.48-0.78) 0.54 (0.33-0.88) 0.57 (0.39-0.83) 0.95 (0.75-1.20) 0.68 (0.59-0.78) 0.62 (0.47-0.83)

Combined fixed-effects model Combined random-effects model Heterogeneity: I 2 = 71% (95% CI, 26%-86%), Q = 13.77; (P = .01)

0.25

0.5

1.0

2.0

Hazard Ratio B Source


Batty et al,10 2002 Gregg et al,15 2003 Smith et al,18 2007 Present study

Population
Whitehall Study, United Kingdom National Health Interview Survey, United States Rancho Bernardo Study, United States EPIC, Europe

HR (95% CI)
0.29 (0.14-0.60) 0.66 (0.45-0.96) 0.66 (0.33-1.32) 0.64 (0.41-0.99) 0.59 (0.46-0.76) 0.58 (0.42-0.79)

Combined fixed-effects model Combined random-effects model Heterogeneity: I 2 = 31% (95% CI, 0%-75%), Q = 4.32; (P = .23)

0.25

0.5

1.0

2.0

Hazard Ratio

Figure 3. Hazard ratios (HRs) and 95% CIs for associations between walking (highest vs lowest category) and total (A) and cardiovascular (B) mortality for individual cohort studies, including the present study, and all the cohort studies combined. EPIC indicates European Prospective Investigation Into Cancer and Nutrition; black squares, estimates for the individual studies; solid horizontal lines, 95% CIs; and white diamonds and dashed vertical lines, combined estimates for the analysis (the width of the diamond represents the 95% CI).

betes or comorbidities at baseline led to inactivity. Excluding participants with comorbidities at baseline or the first 2 years of follow-up strengthened the risk estimates, indicating that reverse causality is unlikely to explain the results. However, residual confounding or misclassification cannot be excluded because measures of disease severity and comorbidities were self-reported. Adjustment for factors on the causal pathway may underestimate the magnitude of the true association between PA and mortality.11,12 However, risk estimates were not affected by additional adjustment for the intermediate factors HbA1c level, BMI, and systolic blood pressure. META-ANALYSIS Physical activity was associated with a lower total mortality risk in diabetic individuals. These associations are in line with those found in the general population, where PA relates to a 33% lower risk of overall mortality and a 35% lower risk of CVD mortality compared with inactivity.8 The present meta-analysis was a high vs low comparison. This is a common practice for metaanalyses of observational studies, but results can be difficult to interpret

because absolute levels of PA will vary between studies and are unknown.44 However, this was the best option based on the available data. Statistically significant heterogeneity was found for the associations between total PA and walking and total mortality. Because statistical heterogeneity is based only on the effect estimates and their precision, it is important to consider clinical heterogeneity. All the studies included in the meta-analysis were comparable in terms of study design, diabetes population, and outcome. However, an important issue when performing meta-analyses of PA is comparability of the exposure assessment, which was heterogeneous across the included studies. Physical activity was assessed by questionnaire or interview, with varying questions, categories, and classifications. Questionnaires, including interviews, are the most common tools for PA assessment in large epidemiologic studies because they are inexpensive and feasible. In general, PA questionnaires have a low reliability and low validity but can be adequately used to rank individuals.45 It was considered appropriate to combine the studies by meta-analyses because all measured common perceptions of PA levels.

In conclusion, evidence from the present study and from previous studies summarized by metaanalyses supports the widely held view that PA is beneficially associated with lower mortality in people with diabetes. Although these findings highlight that persons with diabetes should engage in regular PA,5 they need to be confirmed in a longterm randomized controlled trial. Also, because not many patients with diabetes adhere to this advice,46 future research should elucidate the determinants of physical inactivity and design successful strategies to promote active lifestyles. Accepted for Publication: May 12, 2012. Published Online: August 6, 2012. doi:10.1001/archinternmed.2012 .3130 Author Affiliations: Department of Epidemiology, German Institute of Human Nutrition PotsdamRehbru cke, Nuthetal, Germany (Ms Sluik and Drs Buijsse, Boeing, and No thlings); Berlin School of Public Health, Charite University Medical Center, Berlin, Germany (Dr Muckelbauer); Division of Cancer Epidemiology, German Cancer Research Center, Heidelberg, Germany (Drs Kaaks and Teucher); Danish

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1293

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013

Cancer Society, Copenhagen, Denmark (Drs Johnsen and Tjnneland); Department of Epidemiology, School of Public Health (Drs Overvad and stergaard), and Department of Cardiology, Aalborg Hospital (Dr stergaard), Aarhus University Hospital, Aalborg, Denmark; Public Health Division of Gipuzkoa, IIS Institute BioDonostia, Health Department Basque Region, San Sebastian, Spain (Ms Amiano); Consortium for Biomedical Research in Epidemiology and Public Health, Spain (Ms Amiano and Drs Ardanaz and Huerta Castan o); Navarre Public Health Institute, Pamplona, Spain (Dr Ardanaz); Molecular and Nutritional Epidemiology Unit, Cancer Research and Prevention Institute, Florence, Italy (Dr Bendinelli); Department of Preventive and Predictive Medicine, Nutritional Epidemiology Unit, National Cancer Institute, Milan, Italy (Dr Pala); Cancer Registry and Histopathology Unit, CivileM.P. Arezzo Hospital, Ragusa, Italy (Dr Tumino); Human Genetics Foundation, Turin, Italy (Mr Ricceri); Department of Clinical and Experimental Medicine, Federico II University, Naples, Italy (Dr Mattiello); National Institute for Public Health and the Environment, Centre for Prevention and Health Services Research, Bilthoven, the Netherlands (Dr Spijkerman); Julius Centre for Health Sciences and Primary Care, University Medical Centre Utrecht, Utrecht, the Netherlands (Drs Monninkhof and May); Department of Clinical Sciences, Genetic and Molecular Epidemiology Unit, Ska ne University Hospital, Lund University, Malmo , Sweden (Dr Franks); Department of Nutrition, Harvard School of Public Health, Boston, Massachusetts (Dr Franks); Department of Clinical Sciences, Internal Medicine, Lund University, Ska ne University Hospital, Malmo (Dr Nilsson); Department of Public Health and Clinical Medicine, Family Medicine, Umea University, Umea , Sweden (Drs Wennberg and Rolandsson); Institut National de la Sante et de la Recherche Me dicale, Center for Research in Epidemiology and Population Health, and Paris-South University, Villejuif, France (Drs

Fagherazzi, Boutron-Ruault, and Clavel-Chapelon); Department of Epidemiology, Murcia Regional Health Authority, Murcia, Spain (Dr Huerta Castan o); School of Public Health, Imperial College London, London and London School of Hygiene and Tropical Medicine, London (Dr Gallo); and Epidemiology Section, Institute for Experimental Medicine, Christian-AlbrechtsUniversity of Kiel, Kiel, Germany (Dr No thlings). Correspondence: Diewertje Sluik, MSc, German Institute of Human Nutrition Potsdam-Rehbru cke, Epidemiology Arthur-ScheunertAllee, 114-116 Nuthetal 14558, Germany (Diewertje.Sluik@dife.de). Author Contributions: Ms Sluik had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analyses. Study concept and design: Sluik, Amiano, Ardanaz, Tumino, Boeing, and No thlings. Acquisition of data : Sluik, Kaaks, Teucher, Tjnneland, Overvad, Amiano, Ardanaz, Bendinelli, Pala, Tumino, Ricceri, Mattiello, Nilsson, Wennberg, Rolandsson, Boutron-Ruault, Clavel-Chapelon, Boeing, and No thlings. Analysis and interpretation of data: Sluik, Buijsse, Muckelbauer, Overvad, stergaard, Pala, Spijkerman, Monninkhof, May, Franks, Nilsson, Rolandsson, Fagherazzi, Huerta Castan o, Gallo, and No thlings. Drafting of the manuscript: Sluik. Critical revision of the manuscript for important intellectual content: Buijsse, Muckelbauer, Kaaks, Teucher, Johnsen, Tjnneland, Overvad, stergaard, Amiano, Ardanaz, Bendinelli, Pala, Tumino, Ricceri, Mattiello, Spijkerman, Monninkhof, May, Franks, Nilsson, Wennberg, Rolandsson, Fagherazzi, Boutron-Ruault, ClavelChapelon, Huerta Castan o, Gallo, Boeing, and No thlings. Statistical analysis : Sluik and Buijsse. Obtained funding: Overvad, Tumino, and No thlings. Administrative, technical, and material support: Teucher, Tjnneland, stergaard, Ardanaz, Nilsson, Rolandsson, Gallo, and Boeing. Study supervision: Buijsse, Muckelbauer, Kaaks, Overvad, Amiano, Ardanaz, Boeing, and No thlings. Financial Disclosure: None reported.

Funding/Support: This study was supported by a European Foundation for the Study of Diabetes/sanofiaventis grant (Dr No thlings). Role of the Sponsor: The European Foundation for the Study of Diabetes and sanofi-aventis had no role in the design or conduct of the study, collection or analysis of the data, or preparation or approval of the manuscript and did not have any influence on the contents. Online-Only Material: The eFigures and eTable are available at http: //www.archinternmed.com.
REFERENCES
1. International Diabetes Federation. International Diabetes Federation Diabetes Atlas, Fifth Edition. http: //www.diabetesatlas.org. Accessed June 7, 2012. 2. Montori VM, Ferna ndez-Balsells M. Glycemic control in type 2 diabetes: time for an evidence-based about-face? Ann Intern Med. 2009;150(11):803808. 3. Yudkin JS, Richter B, Gale EA. Intensified glucose lowering in type 2 diabetes: time for a reappraisal. Diabetologia. 2010;53(10):2079-2085. 4. Berry J, Keebler ME, McGuire DK. Diabetes mellitus and cardiovascular disease: Pandoras box has been opened. Herz. 2004;29(5):456-462. 5. Buse JB, Ginsberg HN, Bakris GL, et al; American Heart Association; American Diabetes Association. Primary prevention of cardiovascular diseases in people with diabetes mellitus: a scientific statement from the American Heart Association and the American Diabetes Association. Diabetes Care. 2007;30(1):162-172. 6. Sigal RJ, Kenny GP, Wasserman DH, CastanedaSceppa C, White RD. Physical activity/exercise and type 2 diabetes: a consensus statement from the American Diabetes Association. Diabetes Care. 2006;29(6):1433-1438. 7. Hu FB, Manson JE. Walking: the best medicine for diabetes? Arch Intern Med. 2003;163(12):13971398. 8. Nocon M, Hiemann T, Mu ller-Riemenschneider F, Thalau F, Roll S, Willich SN. Association of physical activity with all-cause and cardiovascular mortality: a systematic review and meta-analysis. Eur J Cardiovasc Prev Rehabil . 2008;15(3):239246. 9. Boule NG, Haddad E, Kenny GP, Wells GA, Sigal RJ. Effects of exercise on glycemic control and body mass in type 2 diabetes mellitus: a metaanalysis of controlled clinical trials. JAMA. 2001; 286(10):1218-1227. 10. Batty GD, Shipley MJ, Marmot M, Smith GD. Physical activity and cause-specific mortality in men with type 2 diabetes/impaired glucose tolerance: evidence from the Whitehall study. Diabet Med. 2002;19(7):580-588. 11. Bernsen RM, Nagelkerke NJ. Walking decreased risk of cardiovascular disease mortality in older adults with diabetes. J Clin Epidemiol. 2007; 60(10):1090-1091. 12. Caspersen CJ, Fulton JE. Epidemiology of walking and type 2 diabetes. Med Sci Sports Exerc. 2008;40(7)(suppl):S519-S528. 13. Ford ES, DeStefano F. Risk factors for mortality from all causes and from coronary heart disease among

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1294

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

persons with diabetes: findings from the National Health and Nutrition Examination Survey I Epidemiologic Follow-up Study. Am J Epidemiol. 1991; 133(12):1220-1230. Gaziano T, Bubes V, Gaziano J. Exercise and mortality among men with diabetes in the Physicians Health Study enrollment cohort. Eur Heart J. 2002;23(suppl):S610. Gregg EW, Gerzoff RB, Caspersen CJ, Williamson DF, Narayan KM. Relationship of walking to mortality among US adults with diabetes. Arch Intern Med. 2003;163(12):1440-1447. Hu FB, Stampfer MJ, Solomon C, et al. Physical activity and risk for cardiovascular events in diabetic women. Ann Intern Med. 2001;134(2):96-105. Hu G, Eriksson J, Barengo NC, et al. Occupational, commuting, and leisure-time physical activity in relation to total and cardiovascular mortality among Finnish subjects with type 2 diabetes. Circulation. 2004;110(6):666-673. Smith TC, Wingard DL, Smith B, Kritz-Silverstein D, Barrett-Connor E. Walking decreased risk of cardiovascular disease mortality in older adults with diabetes. J Clin Epidemiol. 2007;60(3):309-317. Tanasescu M, Leitzmann MF, Rimm EB, Hu FB. Physical activity in relation to cardiovascular disease and total mortality among men with type 2 diabetes. Circulation. 2003;107(19):2435-2439. Wei M, Gibbons LW, Kampert JB, Nichaman MZ, Blair SN. Low cardiorespiratory fitness and physical inactivity as predictors of mortality in men with type 2 diabetes. Ann Intern Med. 2000;132(8): 605-611. 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(1):38-43. Riboli E, Hunt KJ, Slimani N, et al. European Prospective Investigation into Cancer and Nutrition (EPIC): study populations and data collection. Public Health Nutr. 2002;5(6B):1113-1124. No thlings U, Boeing H, Maskarinec G, et al. Food intake of individuals with and without diabetes across different countries and ethnic groups. Eur J Clin Nutr. 2011;65(5):635-641. Sluik D, Boeing H, Montonen J, et al. Associations between general and abdominal adiposity and mortality in individuals with diabetes mellitus. Am J Epidemiol. 2011;174(1):22-34.

25. Wareham NJ, Jakes RW, Rennie KL, et al. Validity and repeatability of a simple index derived from the short physical activity questionnaire used in the European Prospective Investigation into Cancer and Nutrition (EPIC) study. Public Health Nutr. 2003;6(4):407-413. 26. Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc. 2000;32(9)(suppl):S498-S504. 27. Cox DR, Oakes D. Analysis of Survival Data. London, England: Chapman & Hall; 1984. 28. SAS Institute Inc. SAS/STAT 9.2 Users Guide, Second Edition. http://support.sas.com/documentation /cdl/en/statug/63033/HTML/default/viewer .htm#statug_phreg_sect019.htm. Accessed February 23, 2012. 29. Sterne JA, White IR, Carlin JB, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ. 2009; 338:b2393. 30. Newgard CD, Haukoos JS. Advanced statistics: missing data in clinical research, part 2: multiple imputation. Acad Emerg Med. 2007;14(7):669678. 31. Wells GA, Shea B, OConnell D, Peterson J, Welch V, Tugwell P. The Newcastle-Ottawa Scale (NOS) for assessing the quality of non-randomised studies in meta-analyses. http://www.ohri.ca/programs /clinical_epidemiology/oxford.htm. Accessed November 18, 2011. 32. Hu G, Jousilahti P, Barengo NC, Qiao Q, Lakka TA, Tuomilehto J. Physical activity, cardiovascular risk factors, and mortality among Finnish adults with diabetes. Diabetes Care. 2005;28(4):799805. 33. Nelson KM, Boyko EJ, Koepsell T. All-cause mortality risk among a national sample of individuals with diabetes. Diabetes Care. 2010;33(11):23602364. 34. Trichopoulou A, Psaltopoulou T, Orfanos P, Trichopoulos D. Diet and physical activity in relation to overall mortality amongst adult diabetics in a general population cohort. J Intern Med. 2006;259 (6):583-591. 35. Umpierre D, Ribeiro PA, Kramer CK, et al. Physical activity advice only or structured exercise training and association with HbA1c levels in type 2 diabetes: a systematic review and meta-analysis. JAMA. 2011;305(17):1790-1799.

36. Colberg SR, Sigal RJ, Fernhall B, et al; American College of Sports Medicine; American Diabetes Association. Exercise and type 2 diabetes: the American College of Sports Medicine and the American Diabetes Association: joint position statement. Diabetes Care. 2010;33(12):e147-e167. 37. Mora S, Cook N, Buring JE, Ridker PM, Lee IM. Physical activity and reduced risk of cardiovascular events: potential mediating mechanisms. Circulation. 2007;116(19):2110-2118. 38. Stewart KJ. Exercise training and the cardiovascular consequences of type 2 diabetes and hypertension: plausible mechanisms for improving cardiovascular health. JAMA. 2002;288(13): 1622-1631. 39. Thompson PD. Exercise and physical activity in the prevention and treatment of atherosclerotic cardiovascular disease. Arterioscler Thromb Vasc Biol. 2003;23(8):1319-1321. 40. Wareham NJ, van Sluijs EM, Ekelund U. Physical activity and obesity prevention: a review of the current evidence. Proc Nutr Soc. 2005;64(2):229247. 41. Hart CL, Gruer L, Watt GC. Cause-specific mortality, social position, and obesity among women who had never smoked: 28-year cohort study. BMJ. 2011;342:d3785. 42. Bassett DR Jr, Pucher J, Buehler R, Thompson DL, Crouter SE. Walking, cycling, and obesity rates in Europe, North America, and Australia. J Phys Act Health. 2008;5(6):795-814. 43. Hoevenaar-Blom MP, Wendel-Vos GC, Spijkerman AM, Kromhout D, Verschuren WM. Cycling and sports, but not walking, are associated with 10year cardiovascular disease incidence: the MORGEN Study. Eur J Cardiovasc Prev Rehabil. 2011;18 (1):41-47. 44. Berlin JA, Longnecker MP, Greenland S. Metaanalysis of epidemiologic dose-response data. Epidemiology. 1993;4(3):218-228. 45. Westerterp KR. Assessment of physical activity: a critical appraisal. Eur J Appl Physiol. 2009; 105(6):823-828. 46. Nelson KM, Reiber G, Boyko EJ; NHANES III. Diet and exercise among adults with type 2 diabetes: findings from the Third National Health and Nutrition Examination Survey (NHANES III). Diabetes Care. 2002;25(10):1722-1728.

ARCH INTERN MED/ VOL 172 (NO. 17), SEP 24, 2012 1295

WWW.ARCHINTERNMED.COM

2012 American Medical Association. All rights reserved.

Downloaded From: http://archinte.jamanetwork.com/ by a Universidad de los Andes Colombia User on 07/27/2013