Bhattarai, N., Prevost, A, Wright, AJ, Charlton, J., Rudisill, C., and Gulliford, MC

15 Pages·2013·0.97 MB·English

Bhattarai, N., Prevost, A, Wright, A.J., Charlton, J., Rudisill, C., and
Gulliford, M.C. (2013) Effectiveness of interventions to promote healthy
diet in primary care: systematic review and meta-analysis of randomised
controlled trials. BMC Public Health, 13 (1203). ISSN 1471-2458

Copyright © 2013 The Authors.

http://eprints.gla.ac.uk/90443/

Deposited on: 31 January 2014

Enlighten – Research publications by members of the University of Glasgow
http://eprints.gla.ac.uk
Bhattaraietal.BMCPublicHealth2013,13:1203
http://www.biomedcentral.com/1471-2458/13/1203
RESEARCH ARTICLE Open Access
Effectiveness of interventions to promote healthy
diet in primary care: systematic review and
meta-analysis of randomised controlled trials
Nawaraj Bhattarai1,3*, A Toby Prevost1, Alison J Wright1, Judith Charlton1, Caroline Rudisill2 and Martin C Gulliford1
Abstract
Background: Adietrich infruit,vegetables and dietary fibre and low infat is associated with reduced risk of
chronic disease. This review aimed to estimate theeffectiveness ofinterventions to promote healthy diet for
primary prevention among participants attending primary care.
Methods: A systematic review of trials using individual or cluster randomisation ofinterventions delivered in
primary care to promote dietary change over 12 months inhealthy participants free from chronic disease or
defined high risk states. Outcomeswere change in fruit and vegetable intake, consumption of total fat and fibre
and changes inserum cholesterol concentration.
Results: Ten studies were included with12,414 participants. The design and delivery of interventionswere diverse
withrespect to grounding in behavioural theory and interventionintensity. A meta-analysis of threestudies showed
anincreasein fruit consumptionof 0.25 (0.01 to 0.49) servings per day, with an increase invegetable consumption
of 0.25 (0.06 to 0.44) serving per day. A further three studies that reported on fruit and vegetable consumption
together showed a pooled increment of 0.50 (0.13 to 0.87) servings per day. The pooled effect on consumption
of dietary fibre, from four studies, was estimated to be 1.97 (0.43 to 3.52) gm fibre per day. Data from five studies
showed a mean decrease in total fat intake of 5.2% of total energy (1.5 to 8.8%). Data from three studies showed
a mean decrease in serum cholesterol of 0.10 (−0.19 to 0.00) mmol/L.
Conclusion: Presently-reported interventions to promote healthy diet for primary prevention in primary care,
which illustrate a diverse range of intervention methods, may yield small beneficial changes in consumption of
fruit, vegetables, fibre and fat over 12 months. The present results do not exclude the possibility that more
effective intervention strategies might be developed.
Keywords: Diet, Health promotion, Primary care, Systematic review, Meta-analysis
Background intake of fruit, and 59% of the recommended intake of
An increase in intake of fruit and vegetables of one por- vegetables[2].Ahigherintakeofdietaryfibreisassociated
tionperday(80g/day)maybeassociatedwitha10%rela- with lower risk of all-cause mortality [3], as well as lower
tive reduction in risk of ischaemic heart disease and 6% incidenceofcolorectalcancer[4]andstroke[5].Theesti-
reductioninstroke,withbetween1%and6%reductionin matedmeanfibreintakeforAmericanadultsis15.9gram
risk of certain cancers [1]. However, a typical American per day, lower than the recommended intake of at least
dietincludesonlyincludes42%oftherecommendeddaily 25–38 gram per day [6]. Cardiovascular diseases and dia-
betes are associated with obesity and high dietary intakes
of fat and sugars [7] but a typical American diet includes
*Correspondence:[email protected]
1DepartmentofPrimaryCareandPublicHealthSciences,King’sCollege 280% of the recommended intake of calories from solid
London,London,UK fats and sugars [2]. Obesity imposes a significant burden
3HealthEconomicsandHealthTechnologyAssessmentUnit,Instituteof
of morbidity and mortality on populations. The health
HealthandWellbeing,UniversityofGlasgow,1LilybankGardens,Glasgow
G128RZ,UK care costs associated with obesity are substantial and the
Fulllistofauthorinformationisavailableattheendofthearticle
©2013Bhattaraietal.;licenseeBioMedCentralLtd.ThisisanopenaccessarticledistributedunderthetermsoftheCreative
CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and
reproductioninanymedium,providedtheoriginalworkisproperlycited.
Bhattaraietal.BMCPublicHealth2013,13:1203 Page2of14
http://www.biomedcentral.com/1471-2458/13/1203
vastmajorityofthecostsareattributabletotreatinghealth in primary care, including dietary counselling, motiv-
consequences of obesity including type 2 diabetes, cancer ational interviews,adviceforbehaviourchange,computer-
andcardiovasculardiseases[8]. delivered dietary information, reminder telephone calls
There is evidence for the effectiveness of primary and postal newsletters. Primary care in this context
care-based interventions to promote physical activity refers to interventions delivered through the first point
[9], alcohol reduction [10] and smoking cessation [11]. of contact in a health care system, where the service
The regularity of patient consultations in primary care provider acts as the principal sourceofadvicetopatients,
[12],andthe valuethatpatientsplaceonmedicaladvice rather than through specialist referral. Dietary promotion
[13],offeropportunitiesforgeneralpractitionerstoplay intervention in this context means any methods which
important roles in promoting health and preventing are used to promote healthy diet, including healthy
disease. This may include the provision of advice on eating advice and counselling, telephone calls, group
healthy eating. Several randomised trials have evaluated lecturesoruseofanyotherdietaryeducationmaterialsin-
the potential to modify patients dietary habits through cluding posters, booklets and guidelines. We excluded
primarycarebasedinterventions.However,earliersystem- multifaceted interventions including those with physical
atic reviews of the effectiveness of dietary interventions activity promotion along with diet promotion and we
for primary prevention are limited in their applicability did not setanythresholdamountofphysicalactivityfor
to primary care by the inclusion of studies set in work exclusion.Targetpopulationsincludedthegeneralpopula-
places, shopping centres and churches. Some reviews tion of adults aged 16 years or over, including both men
have included non-randomised studies [14] and trials and women. We excluded studies in pregnant women,
with short follow-up, as well as participants with estab- patients at high risk of, or diagnosed with, cardiovascu-
lished medical conditions, or patients with defined high- lar diseases, type 2 diabetes, cancer or other chronic
risk status [14-17], with the possibility of diet restrictions conditions, as well as studies in first or second degree
andwhichwouldlimittheparticipationindietpromotion relatives of affected individuals. We also excluded studies
intervention. thatincludedparticipantsathighriskofcolorectalcancer
Recent work on understanding the effectiveness of (because of adenomatous polyps) [20] or breast cancer
interventions to increase healthy diet has focused on (with mammographic abnormalities) [21,22]. In order to
the importance of using behaviour science theory to focusonapopulationapproachtoprimaryprevention,we
understand determinants of behaviour. Use of specific excluded trials which included participants who were
intervention techniques including setting goals, moni- pregnant, or with existing chronic conditions, or at high
toring behaviour, and reviewing progress towards goals riskofdiseasessuchascolorectalorbreastcancer,orwith
in the light of feedback may be key to dietary behaviour participants who were relatives of family members with
change [18,19]. The effectiveness of behavioural interven- chronic health problems linked to diet. Such high risk
tionsmayalsodependonfactorssuchasthefrequencyof participants or those with established chronic conditions
contacts, the type of professional involved, and whether linked with diet, may be more motivated to make dietary
deliveredindividuallyorinagroupsetting. behaviour changes which may apparently show higher
We report a systematic review and meta-analysis of effectiveness of interventions promoting healthy diet in
randomised controlled trials of primary care-based diet primary care. Also, there may be possibilities of diet
promotion interventions for primary prevention in adults restrictions which may limit the participation in diet
with minimum 12 months follow up. The aim was to promotion intervention and may apparently show lower
quantifywhetherdietpromotionforprimaryprevention effectiveness. Comparators included usual care or no
in primary care is effective in sustained dietary modifi- intervention. We excluded those trials comparing one
cations over at least one year. We also aimed to charac- type of diet promotion intervention with another only
terise existing interventions in terms of their theoretical because our aim was to estimate the effect size differ-
basis and intervention techniques employed, and explore ence between a diet promotion intervention and the
whetherthesewererelatedtointerventioneffectiveness. existing usual care or no intervention; we did not aim to
compare any two methods of diet promotion interven-
Methods tions. A minimum follow- up period of 12 months after
Eligibilitycriteria randomisationwasrequired.OnlyEnglishlanguagepubli-
The review included reports of randomised or cluster cationswereincluded.
controlled trialstudydesigns.Outcomemeasuresincluded
fruit and vegetable intake (servings/day), fat (% of total Searchmethods,studyselectionanddataextraction
energy intake), fibre consumption (gram per day) and We searched Medline, PsycINFO, EMBASE, Centre for
change in serum cholesterol level (mg/dl or mmol/l). Reviews and Dissemination, and the Cochrane Library,
Interventions included any diet promotion intervention withnorestrictionsinthedateandyear,usingthecombined
Bhattaraietal.BMCPublicHealth2013,13:1203 Page3of14
http://www.biomedcentral.com/1471-2458/13/1203
search terms “dietary intervention AND primary care”, in mg/dl, the conversion factor of 38.6598 was used to
“diet promotion intervention AND primary care”, “diet expressthemasmmol/L.
advice AND primary Care”, “counselling AND diet AND We used random effects meta-analysis to pool the
primary care”, “diet promotion AND primary care”, “diet estimates from individual studies. We used the I2 stat-
advice AND behaviour change”, “advice in primary care istic to describe the variation in effect size attributable
AND behaviour ANDdiet”, “nutritionalcounselling AND to heterogeneity among studies; higher values suggest-
primarycare”,“lifestylecounsellingANDcardiovascularrisk ing greater heterogeneity. We also constructed funnel
AND primary care”, “fruit AND vegetable AND primary plots to assess for the publication bias for the studies
care”, “nutritional counselling AND general practice”, included in the review. We used the metan command
“dietary intervention AND general practice”, “dietary in STATA version 12 for the analysis.
intervention AND primary care practice”, “nutritional
advice AND general practice”, “primary care and diet Results
modification”, “fruit and vegetables consumption AND Identificationoftrials
general practice” “primary care AND fiber consump- We screened titles and abstracts of 2,932 papers and
tion” and “fruit and vegetable consumption AND diet identified 49 full text articles. We further identified 3
intervention”. We also reviewed reference lists of relevant full text articles from cross-checking references of 49
articles and previous systematic reviews. The search was full text articles. We then identified 10 trials [27-36] for
carriedoutinitiallyinSeptember2012andagaininMarch inclusioninthisstudyafter excluding trials notmeeting
2013.NBcarriedoutinitialscreeningoftitleandabstracts the eligibility criteria. The details are shown in the flow
against inclusion criteria and retrieved those potentially diagram(Figure1).
eligible. NB and MCG independently assessed the re-
trieved fulltextarticlesandanydifferenceswerereviewed Studyandparticipantscharacteristics
and agreed. NB extracted data concerning participants, We present the study and participant characteristics in
interventions, and outcomes in a tabular form designed Table 1. Ten studies included in the systematic review
for this review. AJW and NB extracted data on the were published between 1988 and 2006. These studies
nature of each intervention, including total number of were conducted in samples representing the primary
contacts with participants, mode(s) of administration care general population in Japan [27] (1 study), USA
and intervention techniques used [23] and coded the [28,31,32,34-36] (6 studies), Italy [30] (1 study) and UK
extent to which intervention was based on psychological [29,33] (2 studies).The randomised study sample size
theories of the determinants of behaviour change, using varied among studies and ranged from 213 to 3,179
a published coding scheme [24]. NB and MCG cross participants with12,414participantsrandomisedintotal.
checkedtheextracteddata. Participants were men and women, but three studies
[28,31,36] included women only, with the age ranging
from 18 to 79 years. Participants were generally healthy
Methodologicalqualityassessment
without established chronic diseases. Participants in one
NB appraised each study for methodological quality using
study[28]werepostmenopausalwomenconsumingatleast
a standard guidance and checklist [25]. We assessed the
36% energy from fat; participants in another study [31]
methodological quality and risk of bias in terms of ran-
had serum cholesterol values of 200 mg/dl (5.17 mmol/l)
domisation, allocation concealment, blinding, loss to fol-
ormorebutas thisisclosetothepopulationmeanvalue,
lowupandoutcomeassessmenttoolvalidity.
this study was not excluded as being directed at high risk
individuals.
Statisticalanalysis
For each trial, we extracted the intervention effects at Interventionandcontrolcharacteristics
12 months. We estimated the intervention effect as the The diet promotion interventions in the trials varied in
difference in the change in mean outcome values (follow number of contacts with participants, mode of delivery
up value minus baseline value) between the intervention and behaviour change techniques employed (Table 2).
group and control group. If the baseline data were not The number of scheduled contacts for intervention with
reportedinthetrials,weusedthedifferenceinthemean theparticipantsintheinterventiongroupsrangedbetween
outcomes between groups at follow up. If not supplied, one and twenty. Most involved at least one face to face
we followed standard procedure [26] to derive standard contact, but two [32,34] involved only a combination
error (SE) for each measure. Where fruit and vegetable of telephone calls and mailed intervention materials. Of
consumption was expressed in grams, the conversion theinterventionsusingface-to-facesessions,onlyone[28]
factor of one serving=80 gm was used to express them was solely delivered in a group format, while the others
asservingsandwheretheserumcholesterolwasexpressed used a combination of group and individual contacts.
Bhattaraietal.BMCPublicHealth2013,13:1203 Page4of14
http://www.biomedcentral.com/1471-2458/13/1203
Figure1Studyselectionflowdiagram.
Many of the interventions also involve printed mate- In each study, the control group was not enrolled in
rials. The interventions involved between two and eight any intervention, but four had minimal interventions:
intervention techniques. Interventions that involved a dietary guidelines [28], standard health education from
greater number of contacts with participants did not leaflets [29],a non-personalised conversation withoutdiet
necessarily employ a greater number of techniques. counselling[30]andbreastself-examinationcounselling
Four [31,32,34,35] of the ten interventions were expli- [31](Table2).
citly described as being based on at least one psycho- Ineachtrial,previouslyvalidatedself-administeredfood
logical theory of behaviour change. While none of the frequency questionnaires, or modified simpler versions,
fourreportsexplicitlylinkedallcomponentsoftheinter- were used to measure the study outcomes. Diet intakes
vention to all the relevant constructs of the theoretical were estimated using the average of the diet consump-
model(s) upon which they claimed to be based, they tion in the previous 24 hours to 1 month, collected
all explicitly linked at least one intervention technique usingfoodfrequencyquestionnaires.
to at least one determinant of behaviour specified by
relevant psychological theory. In three [31,32,34] of the Methodologicalqualityofincludedstudies
interventions, the intervention was tailored for partici- Thefunnelplots and Egger’s testfor potentialpublication
pants according to how they varied on a psychological bias were not informative as insufficient studies were
constructspecifiedbyatheory. identified for each outcome. Table 3 presents a summary
hB
ttpha
Table1Studyandparticipantcharacteristics ://wwttarai
w e
Study(Year) Country Study Selectionofparticipants Numberof Participantsrandomised Eligibleagerange Ethnicityandsocioeconomicstatus Dietassessmenttool .b ta
design practices (%Female) (mean)years io l.
m B
Baron(1990) UK RCT Randomlyselectedparticipants Onegroup 437randomised368 25-60(41.7) Socialclass1or2:controls,30%men, Self-administeredfood ed MC
[33] registeredwithafamilypractice gperancetircael participated(49) 2443%%wwoommeenn;.intervention39%men, frequencyquestionnaire centra Public
Beresford(1997) USA Cluster Participantsattendingroutine 28physician 2121(68) 26%>65years White:91%;Somecollegeeducation: Telephoneinterview l.co He
[35] RCT visitswithoutmajorillness pwclriianthcictinisce6s 7p3e%ry.eFaarm:2il8y%in.comebelow$25000 afrdemquineinscteyreqduefostoiodnnaire m/1471-22013alth
Coates(1999) USA RCT Postmenopausalwomen University 2208(100) 50-79(60) White(55%),Black(28%),Hispanic(16%); Self-administeredfood 45,1
83
[28] vleoalsutn3te6e%rse,ncoerngsyumfroinmgfaatt ccleinnitcraelsin <PoHsitghhigSchhsocohlo(o1l1w%i)t,hHnigohcSoclhleogoeld(2e0g%re),e frequencyquestionnaire /13/1:1203
threestates (35%),graduate/postgraduate(33%) 2
0
3
Fries(2005) USA RCT Randomlyselectedparticipants Threerural 754(64) 18-72(46.34) White:61%,AfricanAmerican:37%; Telephoneinterview
[34] fromphysicians’lists Virginia 8thgrade:Collegedegree:24%; administeredfatand
physician Income<$10,000:14.69%,≥$41,000:19%. fibrebehaviour
practices questionnaire
Gann(2003) USA RCT Womenvolunteersaged Oneclinic 213(100) 20-40(33.4) 76%White,13.5%Black,4%Hispanic, Telephoneinterview
[36] 20–40yearsrecruitedthrough 5.5%Asian,1.5%other;85%completed administeredfood
advertisinganddirectmail college frequencyquestionnaire,
inChicago. basedon24hrdietrecall
oneachofthreedays
Kristal(2000) USA RCT Randomlyselectedpatients Health 1459(50) 18-69(45.8) White(85.9%),Black(4.5%),Asian(5.8%), Telephoneinterview
[32] enrolledwithanHMO. maintenance Hispanic(3.0%),Other(0.8%); administeredFood
organisation Householdincome<$25,00012.2%, FrequencyQuestionnaire
≥$70,00021.7%. (FFQ)andDietHabits
Questionnaire
Roderick(1997) UK Cluster Unselectedpatientsattending 8family 956(50) 35-59(47.3) Non-manualoccupation,intervention Self-administeredfood
[29] RCT GPsurgerypractices practices 60%,control49%;rented frequencyquestionnaire.
accommodationintervention11%,
control25%.
Sacerdote(2006) Italy RCT Unselectedpatients,not 33general 3179(50) 18-65(44.5) Notreported Familyphysician
[30] obese,nochronicdisease practitioners administeredfood
frequencyquestionnaire
Stevens(2003) USA RCT Womenwithrecentnegative Health 616(100) 40-70(53.8) Minoritygroups:7%;Collegegraduates: Self-administeredfatand
[31] mammogramandtotal maintenance 40% fibrebehaviour
cholesterol≥200mg/dl organization questionnaire(FFBQ)
(HMO)
Takahashi Japan RCT Healthyvolunteersintwo Notreported 550(68) 40-69(56) Notreported Self-administereddiet
(2006)[27] ruralvillages,advicegiven historyquestionnaire P
afterannualhealthchecks (DHQ) ag
e
5
o
f
1
4
Table2Interventioncharacteristics hB
ttpha
Study(Year) aMdomdien(iss)troaftion Interventionintensity aSptaptreodacthheoretical Useoftheory Iunsteedrv*entiontechniques Tuosetadltechniques CusTetdechniques Ccoonndtritoilon ://wwttarai
w e
Baron(1990) Facetoface, Dietaryadviceandabooklet none N/A 1.provideinformationon 3 0 Nodietary .b ta
[33] individuallyorin withadviceondiet,promotional consequencesofthe advice io l.
m B
smallgroups, materialsdisplayedatthepractice. behaviour e M
sbbuyoponpkuolerrstte,esdd.eblivyered 3iwn0egrmeroisnucphpsee,drbusreileesfdsifooanltlo,oiwnndeuivapindsdueastslhliyorenoesr 2h1o.wPrtoovipdeerfionrsmtrutchteionon dcentra CPublic
monthsafterentryintothestudy behaviour l.co He
m a
Beresford(1997) 1)Facetoface- Self-helpbookletandphysician Sociallearning No 12p7.ro.pmurospevtisdofefoinllfoowrm-uaptionon 8 2 Nointervention/ /1471-22013lth
[35] physicianintro- endorsementtopromotedietary theory consequencesof usualcare 45,1
83
dbuocoeksleste;lf-help cfohlalonwgiengsutchheacshiamnpgrinogvinsogchiaelanltohr,m b3.ephraovvioiduerinformation /13/1:1203
toeatlowerfat,higherfibrefoods, 2
2)mailedreminder anddoingsomethingpositivefor regardingothers’approval 03
letter
oneself.Introductionofbooklettaking 5.goalsetting(behaviour)
lessthan3minutes,2weekslater,a
8.Barrieridentification
reminderlettersignedbyphysician
andproblemsolving?
senttotheparticipantswhohad
receivedtheintervention. 9.Setgradedtasks
19.Providefeedbackon
performance
21.provideinformation
onhowtodothe
behaviour
27.Useoffollowup
prompts
Coates(1999) Facetoface,in Dietarycounsellingsessionsingroups None N/A 5.goalsetting–behaviour 8 2 Notcounselled,
[28] groups,delivered thatmetweeklyfor6weeks,bi-weekly butgivenDietary
8.problemsolving
bynutritionists for6weeks,monthlyfor9monthsand Guidelinesfor
thenquarterlyuntil18months.Group 12.promptrewards Americans
memberssharedexperiences. contingentoneffort/
successtowards
behaviourandon
successfulbehaviour
16.promptselfmonitoring
21.provideinformation
onhowtoperformthe
behaviour
22.model/demonstrate
thebehaviour Pa
g
e
26.promptpractice
6
o
29.plansocialsupport f
1
4
Table2Interventioncharacteristics(Continued) hB
ttpha
Fries(2005)[34] Mpipnrhfaoooiflrnempseslauicotsainololann–leognnrootuhpe Imdsneitealefti-lrah.vrIeeynlnpcfetluibeododniobnbkagylcepktt,see.lrloesPwophnho-laointlneiezreeacdcaaynlld Cssthoooeccmoiiaarmllymc,uoTnagTriMtnkyei-ttbiivnaegse,d YcthheeasnT–gTseMtafrgoemof 581..2gppororaoblmsleepmtttinrsgeowl–vaibnrdeghsaviour 8 2 Nointervention ://www.biottaraietal.
(ifany)ofstaff 2weeksafterthepersonalised contingentoneffort/ me BM
makingthephone dietaryfeedbackwithbrief successtowards d C
call cmoauilnesdelilninsgt.aIgngfoerrmedatfioornmbaot,oklet: bsuechcaevsisofuurlbanehdaovniour centra Public
oneeachweekimmediately l.c H
16.promptselfmonitoring o e
aftertheinterventionphonecall. m a
21.provideinformation /1 lth
obnehhaovwioutroperformthe 471-2452013,1
2th2e.mbeohdaevl/ioduermonstrate 8/13/13:1203
2
26.promptpractice 0
3
29plansocialsupport
Gann(2003) Facetoface- Classroomnutritioneducation None N/A 21.provideinformation 2 0 Nointervention
[36] groupsessionsplus plusindividualcounsellingwith onhowtoperformthe untilafterendof
twoindividualses- 18groupclassesand2individual behaviour study
sions–noinforma- meetingsin12months.To
22.model/demonstrate
tiononthe maximizetheimpactofintervention,
thebehaviour
professionalgroup appropriatefoodsandmealswere
(ifany)ofstaffde- preparedandservedatintervention
liveringthesessions sessionstoreinforceneweating
behavioursanddemonstratetheease
ofpreparations.Sessionsincluded
discussionandpracticeofshopping,
labelreading,andmealpreparation
techniques,eatingoutandconvenience
foods
Kristal(2000) Mailplusone Tailoreddietaryinterventionincluding Sociallearning Yes– 1.provideinformation 7 2 Usualcare
[32] phonecall i)apackageofself-helpmaterials, theory,TTM,diet intervention abouttheconsequences (Nointervention)
deliveredbya ii)dietaryanalysiswithbehavioural individuation tailoredto ofthebehaviour
“trainedhealth feedback,iii)amotivationalphonecall, model stageof
educator” andiv)‘semi-monthly’newsletters. change, 5.goalsetting–behaviour
motivesfor 9.setgradedtasks
changingdiet
19.providefeedbackon
andstated
behaviour
interestin
dietarychange 21.instructiononhowto
performthebehaviour
22.model/demonstrate
thebehaviour(?) P
a
g
27.useoffollow-up e
prompts 7
o
f
None N/A 5.goalsetting-behaviour 3or4*** 2or3** 1
4
Table2Interventioncharacteristics(Continued) hB
ttpha
R[2o9d]erick(1997) iFnadcievitdouafalcsees–sions, Dsuiebtsatrityuatidovnicaeftaeirmtheedrfeovriefowodofthe 1b0ehparvoimouprtalregvoiealws of Setdauncdaatirodnhealth ://wwttarai
dpaselusleisvsetswrmeoden“bftu”yrtnhuerrses tfdoyipeoteda,srqycuosahnnesteuittmyseawdne.drSefprgeeiqcviueaenllynocduyeto.siRfgeknveeiyedw c1ml6eo.anorirtifo1rt7ihn.(iFgsow–rasnsoomotfeqt)uhsieteelf- lheeaaflletht,yGeuaitdinegto w.biome etal.BM
sessionsdelivered atsecondvisit.3and6month d C
bfayctGoPrsifelCeVvDateridsk rceavrdieiwovsaascnudlaGrPrisrkeffearcratolrisf elevated. b21e.hianvsitoruucrtoiornofonwehiogwht.to centra Public
performthebehaviour l.c H
o e
m a
Sacerdote Facetoface– Personalisednutritionalintervention, None N/A 1.provideinformation 2 0 Asimplerand /1 lth
(2006)[30] idnedliivveidreudalbsyesGsiPo,n, bheasaeltdhothnaatsburmocmhuarreizeadbothuetdItiaeltiaannd aobfobuethathveiocuornsequences ncoonnvpeersrastoionnalized 471-22013
sbuopopkolerttedby G1iGn9uPt9e.id8rveaelinnndteiosonfno,r2aaf1oC5lloomrwrien-cutepdNvuuicstiartitstiiootonnatlhe 2hb1oe.hwparvtooiovupidreerfionrsmtrutchteionon wofitahoburotcthhuereu.se 458/13/12,13:1203
0
Stevens(2003) Facetoface– Individual45minutecounselling Socialcognitive Yes–personal 5.goalsetting–behaviour 7 3 Nodietary 3
[31] individualsessions sessionsandtelephonesupport.Print theory,TTM barriers,self advice,however
8.barrieridentification
plusphonecalls outofthecounsellingsessionalong efficacyand advisedon
andproblemsolving
deliveredby withnutritioneducationmaterials stageof BreastSelf
master’sdegree includingdescriptionsofthedesired change 9.setgradedtasks Examination(BSE)
levelhealth dietarypatternandadvice.Second
counsellors, 45minutevisit,2–3weeksafterthe 10.promptreviewof
behaviouralgoals
supportedbyprint first.
materials 19.providefeedbackon
performance
21.provideinstructionon
howtoperformthe
behaviour
37.motivational
interviewing
Takahashi Facetoface, Two15mindietarycounselling None N/A 5.goalsetting–behaviour 3 2 Nointervention
(2006)[27] individualsessions sessions,agrouplectureandtwo
19.providefeedbackon
plusonegroup newsletters
performance
session,postal
newsletters. 21.provideinstructionon
Professionalgroup howtodothebehaviour
ofthosedelivering
theintervention
unclear
Key:CT=controltheory,TTM=transtheoreticalmodel.
*codedusingCALO-REtaxonomy[23];**Numberofinterventiontechniquesusedconsistentwithcontroltheory(outofthefollowingfourinterventiontechniques:promptspecificgoalsetting,promptreviewofbehavioural
goals,promptselfmonitoringofbehaviourandprovidefeedbackonperformance);***4techniquesusedifparticipantwasoverweight,otherwisethreetechniques.
P
a
g
e
8
o
f
1
4
hB
ttpha
://wwttarai
w e
.b ta
io l.
m B
e M
d C
centra Public
l.c H
o e
m a
TStaubdlye(Y3eRaris)kofRbainadsoamssiseastisomnemnetthfoordincluAdlelodcasttioundcieosncealment Blinding Participationat Outcomeassessment Intentiontotreat /1471-22013lth
12months validityreported (ITT)analysis 45,1
83
T[2a7k]ahashi(2006) RinanEdxcoeml numbersgenerated Notstated Partial.Nurseassessmentwasblinded 448/550(81%) Yes No /13/1:1203
2
0
Coates(1999) Blockrandomisation Notstated Notstated 1,141/2,208(52%) Yes No 3
[28]
Roderick(1997) Pairsmatchedbyregion Notstated Notstated Intervention86%;control74% Yes Yes
[29]
Sacerdote(2006) Randomnumbersgenerated Yes Outcomeassessorsandparticipants 2,977/3,179(93%) Yes Yes
[30] bycomputer statedtobeblinded
Stevens(2003) Notstated Notstated Partial.Clinicstaffconductingdata Intervention89%;control85% Yes No
[31] collectionwereblinded
Kristal(2000) Stratifiedbysexandage Notstated Notstated 1,205/1,459(83%) Partial No
[32]
Baron(1990) Notstated Notstated Notstated 329/368(89%) Notstated No
[33]
Fries(2005) Notstated Notstated Notstated 516/754(68%) Yes No
[34]
Beresford(1997) Tableofrandomnumbers Recruitersandpotentialparticipants No 1,818/2,121(86%) Yes No
[35] blindtogroupallocation
Gann(2003) Tableofrandomnumbers Notstated Notstated 177/213(83%) Yes Yes
[36]
P
a
g
e
9
o
f
1
4