Exercise and Diabetes

Physical Activity of Moderate Intensity andRisk of Type 2 DiabetesA systematic review

CHRISTIE Y. JEON, BA1

R. PETER LOKKEN, BA1FRANK B. HU, MD1,2,3

ROB M. VAN DAM, PHD2

OBJECTIVE To systematically evaluate the evidence for an association between physicalactivity of moderate intensity and risk of type 2 diabetes.

RESEARCH DESIGN AND METHODS We searched EMBASE and Medline throughMarch 2006 and examined reference lists of retrieved articles. We excluded studies that did notassess physical activity of moderate intensity independent of activities of vigorous intensity(more than six times the resting metabolic rate). Information on study design, participantcharacteristics, assessment of physical activity, and outcomes and estimates of associations wereextracted independently by two investigators. We calculated summary relative risks (RRs) usinga random-effects model for the highest versus the lowest reported duration of activities.

RESULTS We identified 10 prospective cohort studies of physical activity of moderateintensity and type 2 diabetes, including a total of 301,221 participants and 9,367 incident cases.Five of these studies specifically investigated the role of walking. The summary RR of type 2diabetes was 0.69 (95% CI 0.580.83) for regular participation in physical activity of moderateintensity as compared with being sedentary. Similarly, the RR was 0.70 (0.580.84) for regularwalking (typically 2.5 h/week brisk walking) as compared with almost no walking. Theassociations remained significant after adjustment for BMI. Similar associations were observed inmen and women and in the U.S. and Europe.

CONCLUSIONS These findings indicate that adherence to recommendations to partici-pate in physical activities of moderate intensity such as brisk walking can substantially reduce therisk of type 2 diabetes.

Diabetes Care 30:744752, 2007

The prevalence of type 2 diabetes ishigh and expected to increase dra-matically in the U.S. and worldwide(1). Type 2 diabetes is a chronic diseaseassociated with premature mortality andvarious debilitating complications (2). In-tensive treatment regimens can preventsome but not all complications (3). There-fore, primary prevention efforts areclearly needed.

Moderately intense physical activi-ties, such as walking and gardening, arethe most common forms of activity

among adults in the U.S. (4) and may bean easily adoptable, relatively safe meansto reduce the risk of type 2 diabetes. Ran-domized trials have shown that physicalactivity alone or in conjunction with di-etary changes can reduce the incidence oftype 2 diabetes (58). However, the in-tensity of activity required remains un-clear because the independent role ofmoderately intense activities has not beendirectly examined in these trials.

Observational studies have consis-tently reported an inverse association

between physical activity and type 2 dia-betes, but most of these studies focusedon vigorous activities or physical activityof various intensities combined (e.g.,9,10). In this article, we systematicallyreview the epidemiological evidence onthe association between physical activityof moderate intensity and risk of type 2diabetes.

RESEARCH DESIGN ANDMETHODS We searched EMBASEand Medline through March 2006 forprospective cohort and cross-sectionalstudies investigating the association be-tween moderately intense physical activ-ity and incidence and prevalence of type 2diabetes. The search terms physical ac-tivity, exercise, and walking wereused in combination with noninsulin de-pendent diabetes mellitus, NIDDM,and type 2 diabetes, and references listsof retrieved articles were examined. Phys-ical activity of moderate intensity was de-fined as requiring a metabolic equivalenttask (MET) score of 3.06.0 (11). OneMET corresponds to the energy expendi-ture during rest (quiet sitting). A typicalactivity of moderate intensity is briskwalking at 5.6 km/h (3.5 miles/h) on a flatsurface requiring 3.8 MET (12). Othercommon activities of moderate intensityinclude playing golf, leisure bicycling at16 km/h (10 miles/h), and gardening(12).

The EMBASE and Medline searcheswere performed independently by two in-vestigators (C.Y.J. and R.P.L.), yielding491 and 488 articles, respectively. Twoadditional relevant articles were found byexamination of reference lists of retrievedarticles. Studies were excluded if they didnot involve human subjects, did notpresent age-adjusted estimates, involvedstudy populations overlapping with otherstudies, or combined moderately intenseactivity with strenuous or light activity.Ten prospective cohort studies (1322)and six cross-sectional studies (2328)were eligible. Because of the large hetero-geneity in exposure and outcome mea-sures for the cross-sectional studies, a

From the 1Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts; the2Department of Nutrition, Harvard School of Public Health, Boston, Massachusetts; and the 3ChanningLaboratory, Harvard Medical School and Brigham and Womens Hospital, Boston, Massachusetts.

Address correspondence and reprint requests to Rob M. van Dam, Department of Nutrition, HarvardSchool of Public Health, 665 Huntington Ave., Bldg. 2, Boston, MA 02115. E-mail: [email protected].

Received for publication 1 September 2006 and accepted in revised form 2 December 2006.C.Y.J. and R.P.L. contributed equally to this work.Abbreviations: MET, metabolic equivalent task.A table elsewhere in this issue shows conventional and Systeme International (SI) units and conversion

factors for many substances.DOI: 10.2337/dc06-1842 2007 by the American Diabetes Association.

R e v i e w s / C o m m e n t a r i e s / A D A S t a t e m e n t sR E V I E W

744 DIABETES CARE, VOLUME 30, NUMBER 3, MARCH 2007

Table

1C

ohortstudies

ofm

oderatelyintense

physicalactivityand

riskof

type2

diabetes

Leadauthor,

year(ref.)

Country

SexA

ge(years)

N/n

casesM

oderate-intensityactivity

Diabetes

assessment

Follow-up

(years)A

djustments

Helm

rich,1991(13)

U.S.

M3968

5,990/202N

onvigorous(

5M

ET

h)sports

onlyvs.no

sportsparticipation;w

alking:

15vs.

5blocks/day

Self-reportofphysician

diagnosis14

Age

Lynch,1996(14)

FinlandM

4260897/46

Activities

requiring

5.5M

ET

hfor

atleast

2h/w

eek(yes

vs.no)

OG

TT

orself-report

ofantidiabetesm

edication

4.2A

ge,BMI,baseline

glucoselevel,

durationofvigorous

activities

Hu

FB,1999(15)

U.S.

F4065

70,912/1,419W

alkingin

ME

Th/w

eekam

ongthose

who

didnot

performvigorous

activities(highest

10

ME

Th/w

eekvs.low

est

0.5M

ET

h/week

quintile)


diagnosis(validated)

8A

ge,BMI,sm

oking,menopausal

status,hormone

therapy,parentalhistory

ofdiabetes,alcoholuse,hypertension,hypercholesterolem

ia

Folsom,2000

(16)U

.S.F

556934,257/1,997

Frequencyofactivities

6.0

ME

Th

(4

times/

week

vs.rareor

never)


diagnosis12

Age,BM

I,waist-to-hip

ratio,education,sm

oking,alcoholuse,horm

onetherapy

use,fam

ilyhistory

ofdiabetes,dietary

factorsO

kada,2000(17)

JapanM

35606,013/444

Weekend

leisure-time

activity(gardening,

home

repairs,shopping)

vs.beingsedentary

duringw

eekends

OG

TT

10A

ge,BMI,alcoholuse,sm

okingstatus,hypertension,parentalhistory

ofdiabetes,weekday

physicalactivity

Wannam

ethee,2000

(18)U

.K.

M4059

5,159/196R

egularw

alkingor

recreationalactivities(gardening

andhom

erepairs)

vs.sedentary

Primary

carerecords

16.9A

ge,BMI,sm

oking,alcoholuse,socialclass,preexistingcoronary

heartdisease

Hu

FB,2001(19)

U.S.

M4075

37,918/1,058W

alkingin

ME

Th/w

eek(highest

vs.lowest

quintile)


diagnosis(validated)

10A

ge,smoking,parentalhistory

ofdiabetes,hypertension,BM

I,dietary

factors,vigorousexercise

Hu

G,2003

(20)Finland

M/F

356414,290/373

Leisure-time

physicalactivity

(walking,

cycling,lightgardening)

4

h/week

vs.less

Nationalregistries

12A

ge,BMI,study

year,systolicblood

pressure,smoking

status,education,other

physicalactivity

Weinstein,2004

(21)U

.S.F

45

37,878/1,361W

alking(

4h/w

eekvs.

none)Self-report

ofphysiciandiagnosis

(validated)6.9

Age,BM

I,family

historyof

diabetes,alcoholuse,smoking,

hormone

therapy,hypertension,hypercholesterolem

ia,dietaryfactors,treatm

entgroup,other

physicalactivityH

sia,2005(22)

U.S.

FPost-

menopausal

87,907/2,271W

alking(highest

10

ME

Th/w

eekvs.low

est0

ME

Th/w

eekquintile)

Self-reportof

antidiabetesm

edicaton

5.1A

ge,BMI,alcoholuse,education,

smoking

status,hypertension,hypercholesterolem

ia,dietaryfactors

ME

T

work

metabolic

rate/restingm

etabolicrate.O

GT

T,oralglucose

tolerancetest.

Jeon and Associates

DIABETES CARE, VOLUME 30, NUMBER 3, MARCH 2007 745

meta-analysis was only conducted for thecohort studies.

Information on study design, partici-pant characteristics, definition of moder-ately intense physical activity, adjustmentsfor potential confounders, and estimatesof associations were abstracted indepen-dently by two investigators (C.Y.J. andR.P.L.). In all studies, estimates of relativerisks (RRs) with information about theirvariance were reported. For studies thatreported more than one level of moder-ately intense activity, the RR for the high-est as compared with the lowest level ofactivity was abstracted.

Statistical analysisWe used STATA version 9.1 (STATA,College Station, TX) for all statistical anal-yses. Summary measures were calculatedusing random-effects models, which al-low each of the studies to estimate a dif-ferent effect size (29). P values forheterogeneity of study results were calcu-lated using Cochrans Q test (30). Whenavailable, we separately analyzed BMI-unadjusted and -adjusted results from theoriginal studies to assess the role of phys-ical activity independent of its associationwith weight. To examine sources of het-erogeneity, we conducted metaregressionanalysis with the log RR of studies as depen-dent variable and country (U.S./other)and sex (male/female) as independentvariables. We assessed potential publica-tion bias using funnel plots, plots of RRsversus precision, and the Begg (31) andEgger (32) tests.

RESULTS

Cohort studiesWe identified 10 cohort studies on phys-ical activity of moderate intensity and riskof type 2 diabetes, including a total of301,221 participants and 9,367 incidentcases. Table 1 shows the characteristics ofthe included studies. Fig. 1 shows the re-sults for moderately intense physical ac-tivity and risk of type 2 diabetes with andwithout adjustment for BMI. For onestudy only BMI-unadjusted (13) and forone study only BMI-adjusted estimateswere presented (22). The summary RR oftype 2 diabetes without BMI adjustmentwas 0.69 (95% CI 0.580.83) for thehighest as compared with the lowest cat-egory of moderate-intensity physical ac-tivity. The BMI-adjusted RR of diabeteswas 0.83 (0.760.90). The P value forheterogeneity in study results was

0.001 for the BMI-unadjusted and 0.24for the BMI-adjusted estimates.

In five of the cohort studies the spe-cific role of walking was examined(13,15,19,21,22). Figure 2 shows the re-sults for walking and risk of type 2 diabe-tes. The BMI-unadjusted summary RRcomparing the highest with the lowestwalking level was 0.70 (95% CI 0.580.84), and the BMI-adjusted RR was 0.83(0.750.91). P values for heterogeneitywere 0.08 for the BMI-unadjusted and0.68 for the BMI-adjusted estimates. Thereported amount of walking for the refer-ence category was minimal, and theamount in the highest category was atleast 10 MET h/week (15,22), which isequivalent to 2.5 h/week of brisk walk-ing. Amounts of walking of 23 h/week(21), 2.13.8 MET h/week (15), and 5.110.0 MET h/week (22) have also been as-sociated with a significantly lower risk oftype 2 diabetes as compared with beingsedentary.

All but two studies (16,22) consid-ered confounding by vigorous physicalactivity by adjusting for vigorous activityor excluding participants that engaged invigorous activities. The studies that didnot consider potential confounding byvigorous activity did not consistently re-port stronger or weaker associations thanthe other studies.

The Egger and Begg tests provided noevidence for publication bias for the BMI-unadjusted (P 0.63 and P 0.84, re-spectively) and BMI-adjusted (P 0.83and P 0.40) association between mod-erately intense physical activity and riskof type 2 diabetes.

Characteristics of the studypopulationThe inverse association between moder-ately intense physical activity and type 2diabetes was observed in populationsfrom the U.S., Finland, and the U.K. (Ta-ble 1). In the Japanese study by Okada etal. (17), no inverse association was ob-served, but this may have been due to themoderate-intensity activity definition thatwas restricted to weekends, included rel-atively light-intensity activities, and didnot consider duration. We conducted ametaregression analysis for the associa-tion between moderate-intensity physicalactivity and did not find significant differ-ences in results between U.S. and non-U.S. populations (P 0.20). Thesummary RR for the U.S. studies was 0.68(95% CI 0.560.81), while that of thenon-U.S. studies was 0.75 (0.561.00).

After excluding the study by Okada et al.(17), the RR for non-U.S. studies was 0.66(0.540.80) and the P value for differ-ence by region 0.96. For the BMI-adjusted RRs, the P values for differenceby region was 0.07 before and 0.75 afterexclusion of the study by Okada et al. Wealso evaluated potential differences bysex. The BMI-unadjusted association be-tween moderate-intensity physical activ-ity and diabetes risk was significantlystronger for female (RR 0.58 [95% CI0.510.65]) than for male (0.82 [0.700.96]) cohorts (P 0.04). However, afterexcluding the study by Okada et al., theRR for male cohorts became 0.77 (95% CI0.670.88) and the difference with thefemale cohorts no longer statistically sig-nificant (P 0.36). For the BMI-adjustedRRs, the P values for differences betweenmale and female cohorts was 0.17 beforeand 0.81 after exclusion of the study byOkada et al.

Cross-sectional studiesSix cross-sectional studies that examinedthe association between moderately in-tense activities and type 2 diabetes or im-paired glucose tolerance (Table 2) wereidentified. Results from two Dutch stud-ies suggested inverse associations for gar-dening and bicycling (which tends to beof easy pace in this population) but not forwalking (24,26). In a French study, par-ticipation in moderately intense house-hold activities was inversely associatedwith type 2 diabetes (25). Two other stud-ies were consistent with an inverse asso-ciation between moderately intenseactivities and type 2 diabetes, but CIswere wide (23,27). In an Australianstudy, 2.5 h/week of moderately intenseactivities as compared with less timespent on these activities tended to be as-sociated with a lower prevalence of abnor-mal glucose metabolism in women butnot in men (28).

CONCLUSIONS In our meta-analysis of 10 prospective cohort studies,a substantial inverse association was ob-served between physical activity of mod-erate intensity and risk of type 2 diabetes.Those who regularly engaged in physicalactivity of moderate intensity had 30%lower risk of type 2 diabetes as comparedwith sedentary individuals. A similar de-crease in diabetes risk was observed whenwe specifically examined regular walking.After adjustment for BMI, the reduction indiabetes risk remained substantial (17%)

Physical activity and type 2 diabetes


for both regular moderately intense activ-ity and walking.

Significant inverse associations wereobserved in both men and women and in

both U.S. and Northern European co-horts. Most studies were conducted inpredominantly white populations. Nosignificant association between moderate-

intensity physical activity and type 2 dia-betes was observed in the two studies thatreported results for other ethnic groups,but this may have been due to the light

Figure 1RRs for total physical activity of moderate intensity and incidence of type 2 diabetes for individual cohort studies and all studies combinedwithout adjustment for BMI (A) and with adjustment for BMI (B). RR comparing the highest with the lowest reported level of moderate-intensityactivity are shown in Table 1. Filled bars and open diamonds indicate 95% CIs. The size of the squares corresponds to the weight of the study in themeta-analysis.

Jeon and Associates


definition of activity (17) and limited sta-tistical power as a result of lower numbersfor nonwhites (22). In the Diabetes Pre-vention Program, the lifestyle interven-tion that focused on both diet and

physical activity resulted in a similar re-duction in incidence of type 2 diabetes inparticipants of African-American, His-panic, American-Indian, and Asian eth-nicity as compared with whites (6,33).

Results from cross-sectional studieswere generally consistent with an inverseassociation between moderately intensephysical activity and type 2 diabetes. Re-sults were more heterogeneous than for

Figure 2RRs for walking and incidence of type 2 diabetes for individual cohort studies and all studies combined without adjustment for BMI (A)and with adjustment for BMI (B). RR comparing the highest with the lowest reported level of walking are shown in Table 1. Filled bars and opendiamonds indicate 95% CIs. The size of the squares corresponds to the weight of the study in the meta-analysis.



Table

2C

ross-sectionalstudiesof

moderately

intensephysicalactivity

andrisk

oftype

2diabetes

Leadauthor,

year(ref.)

Country

SexA

ge(years)

N/n

casesO

utcome

assessment

Classification

moderate-

intensityactivity

OR

(95%C

I)A

djustments

James,1998

(23)U

.S.M

/F3055

916/78T

ype2

diabetes:FBGor

useof

antidiabetesm

edicationW

alkingfor

15

min

ata

time

orhom

em

aintenance/gardeningvs.inactive

0.51(0.201.28)

Age,sex,education,BM

I,w

aist-to-hipratio

Baan,1999(24)

Netherlands

M5575

503/69T

ype2

diabetes:2-hpostload

glucoseor

useofantidiabetes

medication

Walking

(285

vs.60

min/w

eek)1.45

(0.633.32)A

ge,BMI,w

aist-to-hipratio,

family

historyofdiabetes,

smoking

Bicycling(yes

vs.no)0.26

(0.140.49)G

ardening(yes

vs.no)0.56

(0.291.06)F

5575513/49

Same

asm

aleparticipants

Walking

(270

vs.45

min/w

eek)1.05

(0.472.36)Sam

eas

male

participants

Bicycling(yes

vs.no)0.37

(0.180.78)G

ardening(yes

vs.no)0.46

(0.211.01)D

efay,2001(25)

FranceM

60

1,113/213T

ype2

diabetes:FBGor

useof

antidiabetesm

edicationU

sualintensityof

householdactivities

(moderate

vs.light)

0.67(0.480.94)

None

F

601,419/132

Same

asm

aleparticipants

Usualintensity

ofhousehold

activities(m

oderatevs.light)

0.61(0.400.95)

None

vanD

am,

2002(26)

Netherlands

M6989

424/90IG

Tor

newly

detectedtype

2diabetes:O

GT

TT

akingw

alks(

20m

in/day

vs.none)0.92

(0.461.88)A

ge,otherphysicalactivities

Bicycling(

20m

in/dayvs.none)

0.37(0.200.70)

Gardening

(20

min/day

vs.none)0.33

(0.160.66)

Com

be,2004

(27)France

M2074

3,759/299N

ewly

detectedtype

2diabetes:

FBGW

alking

1h/day

(yesvs.no)

0.81(0.641.04)

None

Dunstan,2004

(28)A

ustraliaM

25

8,299/1,555(m

enand

wom

en)IFG

,IGT

,ornew

lydetected

type2

diabetes:OG

TT

Walking

andother

moderate-intensity

activities(

2.5vs.

2.5

h/week)

0.96(0.661.40)

Age,sex,education,fam

ilyhistory

ofdiabetes,sm

oking,dietaryfactors,

TV

watching,w

aistcircum

ferenceF

25

Same

asm

aleparticipants

Walking

andother

moderate-intensity

activities(

2.5vs.

2.5

h/week)

0.73(0.531.01)

Same

asm

aleparticipants

FBP,fastingblood

glucose;IFG,im

pairedfasting

glucose;IGT

,impaired

glucosetolerance;O

GT

T,oralglucose

tolerancetest.

Jeon and Associates


the prospective studies, possibly as a re-sult of the smaller study sizes and differ-ences in the activity classification. In twocross-sectional studies of Dutch elderly,no association was observed for walking,whereas other moderately intense activi-ties were inversely association with type 2diabetes (24,26). This lack of associationfor walking may have reflected a loweraverage walking pace in these elderlypopulations. A brisk usual walking pacewas associated with a substantially lowerrisk of type 2 diabetes compared with aneasy pace in two U.S. studies (15,19).

Strengths and limitations of the dataBecause our meta-analysis included onlyobservational studies, it is possible thatthe summary estimates were influencedby confounding and other biases. Al-though all studies adjusted for age, not alladjusted for known or suspected diabetesrisk factors such as dietary factors, alcoholconsumption, cigarette smoking, andwaist-to-hip ratio. The inverse associationwas similar in studies with the most com-plete adjustment for confounding, but wecannot fully exclude residual confound-ing by unmeasured or imprecisely mea-sured diabetes risk factors. Lack ofadjustment for light-intensity physical ac-tivity or sedentary activities may also haveresulted in residual confounding (15,19).Furthermore, the included studies mostlyfocused on leisure time physical activity,but commuting and occupational activi-ties can also contribute importantly to theaccumulation of moderately intensephysical activity for the reduction of dia-betes risk (20,34).

The prospective design of the cohortstudies and low loss to follow-up in moststudies reduced the likelihood of selec-tion bias. Some misclassification of mod-erate-intensity physical activity probablyoccurred, but it seems unlikely that thismisclassification differed by future diabe-tes outcome, and it can thus be expectedto have biased estimates of associationstoward the null. In the studies that reliedupon self-report of a physicians diagnosisof diabetes, differences in detection of di-abetes associated with physical activitycould have led to diagnostic bias. This isunlikely given that several cross-sectionalstudies measured glucose concentrationsin all participants, and restriction tosymptomatic cases did not substantiallyaffect the results in a cohort study (15).We did not find any evidence of publica-tion bias based on the Egger and Begg testand the funnel plot. However, the power

of these tests is known to be limited (35),and we cannot fully exclude the possibil-ity that publication bias has affected ourresults.

MechanismsWe found a significant inverse associationbetween moderately intense physical ac-tivity and type 2 diabetes that persistedafter adjustment for BMI. In line with thisfinding, biological mechanisms have beenidentified for beneficial effects of physicalactivity on glucose metabolism indepen-dent of body fatness. Exercise has beenshown to increase insulin-stimulated gly-cogen synthesis through an increased rateof insulin-stimulated glucose transport byGLUT4 glucose transporters and in-creased glycogen synthase activity (36).In addition, elevated capillary prolifera-tion in muscles, increased muscle mass,and a higher proportion of more insulin-sensitive types of muscle fibers may con-tribute to beneficial effects of physicalactivity on insulin sensitivity (37).

Findings from intervention studiesNo large randomized trials have specifi-cally investigated the effect of increasingphysical activity of moderate intensity onincidence of type 2 diabetes. A random-ized trial of moderate-intensity physicalactivity in individuals with a family his-tory of diabetes did not find a significantreduction in incidence of type 2 diabetesafter 2 years, but compliance with theprogram was poor and the number of par-ticipants small (n 37 in the exerciseprogram) (38). In a 2-year randomizedcontrolled trial in 179 individuals withtype 2 diabetes, counseling to achieve 10MET h/ week of moderately intense phys-ical activity resulted in significantly re-duced body weight and fasting glucoseand A1C concentrations (39). In a posthoc analysis, a significant reduction inA1C was observed for participants thatincreased 1120 MET h/week, but greaterbeneficial effects were observed in thosewho increased their activity with 2130MET h/week (39). Several randomizedcontrolled trials tested the effects of briskwalking on glucose metabolism in indi-viduals without diabetes (40 42). In-creased walking resulted in reducedplasma insulin (40,41) or glucose (42)concentrations after 1224 weeks.

A Chinese trial in individuals withimpaired glucose tolerance found thatthose randomized at the clinic level to acombination of moderate and vigorousactivities had a 47% reduction in inci-

dence of type 2 diabetes as compared withthe control group (7). A post hoc analysisof the Finnish Diabetes Prevention Studyexamined the association between walk-ing during follow-up and incidence oftype 2 diabetes. After adjustment forother activities, dietary factors, and BMI,at least 2.5 h/week of walking for exercisewas associated with a 63% lower risk oftype 2 diabetes as compared with 1h/week (43). In addition, an increase inwalking intensity was associated with alower risk of type 2 diabetes.

RecommendationsOur systematic review indicates that reg-ular participation in moderately intensephysical activity is associated with a sub-stantially lower risk of type 2 diabetes.The association was partly independent ofBMI, suggesting that moderate-intensityphysical activity can reduce the risk oftype 2 diabetes even in those who do notachieve weight loss. Findings from severalprospective studies (15,21,22,43) indi-cate that 30 min or more of daily moder-ate-intensity activity, as recommended inmultiple U.S. guidelines (11,44), can sub-stantially reduce the risk of type 2 diabe-tes as compared with being sedentary.Moderately intense activity as defined inguidelines (3.0 6.0 MET h) (11) in-cludes walking at brisk pace but not walk-ing at an easy or casual pace (12), andwalking at brisk pace also seems prefera-ble for the prevention of type 2 diabetes(15,19,43). Further studies are needed todefine more specifically what combina-tions and duration and pace are optimalfor reducing the risk of type 2 diabetes.However, given that only 31% of adults inthe U.S. currently meet the general phys-ical activity recommendations (45), ef-forts to prevent type 2 diabetes shouldstrongly emphasize the benefit of moder-ately intense physical activities and en-courage wider participation in theseactivities.

References1. Wild S, Roglic G, Green A, Sicree R, King

H: Global prevalence of diabetes: esti-mates for the year 2000 and projectionsfor 2030. Diabetes Care 27:10471053,2004

2. Nathan DM: Long-term complicationsof diabetes mellitus. N Engl J Med 328:16761685, 1993

3. Intensive blood-glucose control with sul-phonylureas or insulin compared withconventional treatment and risk of com-plications in patients with type 2 diabetes



(UKPDS 33): UK Prospective DiabetesStudy (UKPDS) Group. Lancet 352:837853, 1998

4. Crespo CJ, Keteyian SJ, Heath GW, Sem-pos CT: Leisure-time physical activityamong US adults: results from the ThirdNational Health and Nutrition Examina-tion Survey. Arch Intern Med 156:9398,1996

5. Tuomilehto J, Lindstrom J, Eriksson JG,Valle TT, Hamalainen H, Ilanne-ParikkaP, Keinanen-Kiukaanniemi S, Laakso M,Louheranta A, Rastas M, Salminen V,Uusitupa M: Prevention of type 2 diabetesmellitus by changes in lifestyle amongsubjects with impaired glucose tolerance.N Engl J Med 344:13431350, 2001

6. Knowler WC, Barrett-Connor E, FowlerSE, Hamman RF, Lachin JM, Walker EA,Nathan DM: Reduction in the incidence oftype 2 diabetes with lifestyle interventionor metformin. N Engl J Med 346:393403,2002

7. Pan XR, Li GW, Hu YH, Wang JX, YangWY, An ZX, Hu ZX, Lin J, Xiao JZ, CaoHB, Liu PA, Jiang XG, Jiang YY, Wang JP,Zheng H, Zhang H, Bennett PH, HowardBV: Effects of diet and exercise in prevent-ing NIDDM in people with impaired glu-cose tolerance: the Da Qing IGT andDiabetes Study. Diabetes Care 20:537544, 1997

8. Ramachandran A, Snehelatha C, Mary S,Mukesh B, Bhaskar AD, Vijay V, IndianDiabetes Prevention Program: The IndianDiabetes Programme shows that lifestylemodifications and metformin preventtype 2 diabetes in Asian Indian subjectswith impaired glucose tolerance (IDPP-1). Diabetologia 49:289297, 2006

9. Manson JE, Nathan DM, Krolewski AS,Stampfer MJ, Willett WC, HennekensCH: A prospective study of exercise andincidence of diabetes among US malephysicians. JAMA 268:6367, 1992

10. Meisinger C, Lowel H, Thorand B, DoringA: Leisure time physical activity andthe risk of type 2 diabetes in men andwomen from the general population: theMONICA/KORA Augsburg Cohort Study.Diabetologia 48:2734, 2005

11. Pate RR, Pratt M, Blair SN, Haskell WL,Macera CA, Bouchard C, Buchner D, Et-tinger W, Heath GW, King AC, et al.:Physical activity and public health: a rec-ommendation from the Centers for Dis-ease Control and Prevention and theAmerican College of Sports Medicine.JAMA 273:402407, 1995

12. Ainsworth BE, Haskell WL, Whitt MC, Ir-win ML, Swartz AM, Strath SJ, OBrienWL, Bassett DR Jr, Schmitz KH, Emplain-court PO, Jacobs DR Jr, Leon AS: Com-pendium of physical activities: an updateof activity codes and MET intensities. MedSci Sports Exerc 32:S498S504, 2000

13. Helmrich SP, Ragland DR, PaffenbargerRS Jr: Prevention of non-insulin-depen-

dent diabetes mellitus with physical activ-ity. Med Sci Sports Exerc 26:824830, 1994

14. Lynch J, Helmrich SP, Lakka TA, KaplanGA, Cohen RD, Salonen R, Salonen JT:Moderately intense physical activities andhigh levels of cardiorespiratory fitness re-duce the risk of non-insulin-dependentdiabetes mellitus in middle-aged men.Arch Intern Med 156:13071314, 1996

15. Hu FB, Sigal RJ, Rich-Edwards JW, Cold-itz GA, Solomon CG, Willett WC, SpeizerFE, Manson JE: Walking compared withvigorous physical activity and risk of type2 diabetes in women: a prospective study.JAMA 282:14331439, 1999

16. Folsom AR, Kushi LH, Hong CP: Physicalactivity and incident diabetes mellitus inpostmenopausal women. Am J Public Health90:134138, 2000

17. Okada K, Hayashi T, Tsumura K, Sue-matsu C, Endo G, Fujii S: Leisure-timephysical activity at weekends and the riskof type 2 diabetes mellitus in Japanesemen: the Osaka Health Survey. DiabetMed 17:5358, 2000

18. Wannamethee SG, Shaper AG, AlbertiKG: Physical activity, metabolic factors,and the incidence of coronary heart dis-ease and type 2 diabetes. Arch Intern Med160:21082116, 2000

19. Hu FB, Leitzmann MF, Stampfer MJ,Colditz GA, Willett WC, Rimm EB: Phys-ical activity and television watching in re-lation to risk for type 2 diabetes mellitusin men. Arch Intern Med 161:15421548,2001

20. Hu G, Qiao Q, Silventoinen K, ErikssonJG, Jousilahti P, Lindstrom J, Valle TT,Nissinen A, Tuomilehto J: Occupational,commuting, and leisure-time physical ac-tivity in relation to risk for type 2 diabetesin middle-aged Finnish men and women.Diabetologia 46:322329, 2003

21. Weinstein AR, Sesso HD, Lee IM, CookNR, Manson JE, Buring JE, Gaziano JM:Relationship of physical activity vs bodymass index with type 2 diabetes inwomen. JAMA 292:11881194, 2004

22. Hsia J, Wu L, Allen C, Oberman A, Law-son WE, Torrens J, Safford M, LimacherMC, Howard BV: Physical activity and di-abetes risk in postmenopausal women.Am J Prev Med 28:1925, 2005

23. James SA, Jamjoum L, Raghunathan TE,Strogatz DS, Furth ED, Khazanie PG:Physical activity and NIDDM in African-Americans: the Pitt County Study. Diabe-tes Care 21:555562, 1998

24. Baan CA, Stolk RP, Grobbee DE, Witte-man JC, Feskens EJ: Physical activity inelderly subjects with impaired glucosetolerance and newly diagnosed diabetesmellitus. Am J Epidemiol 149:219227,1999

25. Defay R, Delcourt C, Ranvier M, LacrouxA, Papoz L: Relationships between physi-cal activity, obesity and diabetes mellitusin a French elderly population: the POLA

study. Int J Obes Relat Metab Disord 25:512518, 2001

26. Van Dam RM, Schuit AJ, Feskens EJ, Sei-dell JC, Kromhout D: Physical activity andglucose tolerance in elderly men: the Zut-phen Elderly study. Med Sci Sports Exerc34:11321136, 2002

27. Combe H, Vol S, Thevenot A, LasfarguesG, Caces E, Tichet J, Lecomte P: Compar-ison of men with impaired fasting glycae-mia to controls and to diabetic subjectswith fasting glycaemia from 7.0 to 7.7mmol/l: clinical, nutritional and biologi-cal status. Diabetes Metab 30:167174,2004

28. Dunstan DW, Salmon J, Owen N, Arm-strong T, Zimmet PZ, Welborn TA, Cam-eron AJ, Dwyer T, Jolley D, Shaw JE:Physical activity and television viewing inrelation to risk of undiagnosed abnormalglucose metabolism in adults. DiabetesCare 27:26032609, 2004

29. DerSimonian R, Laird N: Meta-analysis inclinical trials. Control Clin Trials 7:177188, 1986

30. Higgins JP, Thompson SG: Quantifyingheterogeneity in a meta-analysis. Stat Med21:15391558, 2002

31. Begg CB, Mazumdar M: Operating char-acteristics of a rank correlation test forpublication bias. Biometrics 50:10881101, 1994

32. Egger M, Davey Smith G, Schneider M,Minder C: Bias in meta-analysis detectedby a simple, graphical test. BMJ 315:629634, 1997

33. Orchard TJ, Temprosa M, Goldberg R,Haffner S, Ratner R, Marcovina S, FowlerS, Diabetes Prevention Program ResearchGroup: The effect of metformin and inten-sive lifestyle intervention on the meta-bolic syndrome: the Diabetes PreventionProgram randomized trial. Ann Intern Med142:611619, 2005

34. Nakashini N, Takatorige T, Suzuki K:Daily life activity and risk of developingimpaired fasting glucose or type 2 diabe-tes in middle-aged Japanese men. Diabe-tologia 47:17681775, 2004

35. Peters JL, Sutton AJ, Jones DR, AbramsKR, Rushton L: Comparison of two meth-ods to detect publication bias in meta-analysis. JAMA 295:676680, 2006

36. Perseghin G, Price TB, Petersen KF, Ro-den M, Cline GW, Gerow K, Rothman DL,Shulman GI: Increased glucose transport-phosphorylation and muscle glycogensynthesis after exercise training in insulin-resistant subjects. N Engl J Med 335:13571362, 1996

37. Goodyear LJ, Kahn BB: Exercise, glucosetransport, and insulin sensitivity. AnnuRev Med 49:235261, 1998

38. Wing RR, Venditti E, Jakicic JM, PolleyBA, Lang W: Lifestyle intervention inoverweight individuals with a family his-tory of diabetes. Diabetes Care 21:350359, 1998

Jeon and Associates


39. Di Loreto C, Fanelli C, Lucidi P, MurdoloG, De Cicco A, Parlanti N, Ranchelli A,Fatone C, Taglioni C, Santeusanio F, DeFeo P: Make your diabetic patients walk:long-term impact of different amounts ofphysical activity on type 2 diabetes. Dia-betes Care 28:12951302, 2005

40. Aldred HE, Hardman AE, Taylor S: Influ-ence of 12 weeks of training by briskwalking on postprandial lipemia and in-sulinemia in sedentary middle-agedwomen. Metabolism 44:390397, 1995

41. Furukawa F, Kazuma K, Kawa M, Mi-yashita M, Niiro K, Kusukawa R, KojimaM: Effects of an off-site walking programon energy expenditure, serum lipids, and

glucose metabolism in middle-agedwomen. Biol Res Nurs 4:181192, 2003

42. Asikainen TM, Miilunpalo S, Kukkonen-Harjula K, Nenonen A, Pasanen M, RinneM, Uusi-Rasi K, Oja P, Vuori I: Walkingtrials in postmenopausal women: effect oflow doses of exercise and exercise frac-tionization on coronary risk factors. ScandJ Med Sci Sports 13:284292, 2003

43. Laaksonen DE, Lindstrom J, Lakka TA,Eriksson JG, Niskanen L, Wikstrom K,Aunola S, Keinanen-Kiukaanniemi S,Laakso M, Valle TT, Ilanne-Parikka P,Louheranta A, Hamalainen H, Rastas M,Salminen V, Cepaitis Z, Hakumaki M,Kaikkonen H, Harkonen P, Sundvall J,

Tuomilehto J, Uusitupa M: Physical activ-ity in the prevention of type 2 diabetes:the Finnish diabetes prevention study. Di-abetes 54:158165, 2005

44. Eyre H, Kahn R, Robertson RM: Prevent-ing cancer, cardiovascular disease, anddiabetes: a common agenda for the Amer-ican Cancer Society, the American Diabe-tes Association, and the American HeartAssociation. Diabetes Care 27:18121824,2004

45. Schoenborn CA, Adams PF, Barnes PM,Vickerie JL, Schiller JS: Health behaviorsof adults: United States, 19992001. VitalHealth Stat 10:179, 2004



Documents

Exercise and Diabetes