Maturitas 79 (2014) 292298

Contents lists available at ScienceDirect

Maturitas

jou rn al hom ep age: www.elsev ier .com/ locate /matur i tas

Associa ealbarrier aof retir

Jose LaraHuman Nutriti esearNewcastle upo

a r t i c l

Article history:Received 30 May 2014Received in revised form 24 June 2014Accepted 1 July 2014

Keywords:Mediterranean dietHealth behaviBarriers to heaPerceived heaOnline survey

Objectives: Health behaviours including diet, smoking, alcohol consumption, and physical activity, predicthealth risks at the population level. We explored health behaviours, barriers to healthy eating and self-rated health among individuals of retirement age.

Study design 82 men and 124 women participated in an observational, cross-sectional online survey.Main outcome measures A 14-item Mediterranean diet score (MDPS), perceived barriers to healthy eat-

ing (PBHE), self-reported smoking, physical activity habits, and current and prior perceived health status

1. Introdu

There is diet, smokihigh body mTogether, thin total motial contribself-rated hciated with

In an eretirement oping effec

CorresponE-mail add

http://dx.doi.o0378-5122/ by-nc-nd/3.0/)ourslthy eatinglth

(PHS) were assessed. A health behaviours score (HBS) including smoking, physical activity, body massindex (BMI) and MDPS was created to evaluate associations with PHS. Two-step cluster analysis identiednatural groups based on PBHE. Analysis of variance was used to evaluate between group comparisons.Results: PBHE number was associated with BMI (r = 0.28, P < 0.001), age (r = 0.19; P = 0.006), and MDPS(r = 0.31; P < 0.001). PHBE cluster analysis produced three clusters. Cluster-1 members (busy lifestyle)were signicantly younger (57 years), more overweight (28 kg/m2), scored lower on MDPS (4.7) andreported more PBHE (7). Cluster-3 members (no characteristic PBHE) were leaner (25 kg/m2), reportedthe lowest number of PBHE (2), and scored higher on HBS (2.7) and MDPS (6.2). Those in PHS categories,bad/fair, good, and very good, reported mean HBS of 2.0, 2.4 and 3.0, respectively (P < 0.001). Comparedwith the previous year, no signicant associations between PHS and HBS were observed.Conclusions: PBHE clusters were associated with BMI, MDPS and PHS and could be a useful tool to tailorinterventions for those of peri-retirement age.

2014 The Authors. Published by Elsevier Ireland Ltd. This is an open access article under the CCBY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

ction

strong evidence that lifestyle behaviours including poorng, low physical activity, high alcohol consumption, orass index (BMI), are strongly associated with health [1].ese behaviours are associated with four-fold increasesrtality in men and women [2,3] and make a substan-ution to the global burden of disability [4]. In addition,ealth, an independent predictor of mortality, [5] is asso-

health behaviours [6].ra when the worlds population is ageing and theage is being extended in many countries, devel-tive lifestyle-based interventions, including dietary

ding author. Tel.: +44 0 1912481141.resses: jose.lara@ncl.a.uk, jjlaragallegos@yahoo.com (J. Lara).

interventions, offers considerable potential to promote healthyageing [7,8]. Interventions targeting people of retirement age maybe particularly useful since this life-transition is associated withother behavioural changes [9,10], and improved eating habitsmay alter the ageing-trajectory and reduce the burden of age-related disability and disease. The health benets of consuming aMediterranean diet (MD) have been demonstrated in numerousepidemiological studies [1113]; and these observational nd-ings are now supported by evidence from primary [14,15], andsecondary prevention randomised clinical trials [16]. However,interventions promoting the MD outside Mediterranean countriesare scarce [1719] and current evidence suggests that most dietaryinterventions result in only small to moderate effect-size [20].

Identifying and addressing perceived barriers to healthy eating(PBHE) is a critical step in developing effective dietary interven-tions. There are no studies of PBHE in people of retirement age buta European survey of young and middle-age adults reported that

rg/10.1016/j.maturitas.2014.07.0032014 The Authors. Published by Elsevier Ireland Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/.tion of Mediterranean diet and other hs to healthy eating and perceived healthement age, Leigh-Ann McCrum, John C. Mathers

on Research Centre, Institute for Ageing and Health, Newcastle University, Biomedical Rn Tyne NE4 5PL, UK

e i n f o a b s t r a c tth behaviours withmong British adults

ch Building, Campus for Ageing and Vitality,

J. Lara et al. / Maturitas 79 (2014) 292298 293

concerns with time and taste were the most frequently reportedPBHE [21]. In addition, a small study on 39 older men (mean age75 years) identied poor cooking skills and low motivation tochange eating habits, as important PBHE [22].

The aimbetween PBhealth behacross-sectio

2. Method

This stuCommitteeconforms toParticipants

2.1. Study d

ObservaPBHE and pSTROBE sta

2.2. Sample

We recrfrom amonthis study and, apartwere emplthe comm(http://wwwand the Unistudy was Queens Unsampling pforward theticipants ofwide rangelevels, mari

2.3. E-surve

The survtion Systemof web formreceived thtesting of thtaken to cosurvey.

2.4. Measur

The e-su

(1) A questlifestylereporteactivityhigh. BParticip(PHS) ugood, athe year

(2) MD adherence was assessed using the 14-item PREDIMED score(MDPS) [24]. The range of possible scores for MDPS is 014 withhigher scores indicating greater MD adherence.

(3) A questionnaire asking participants to identify PBHE from aviously published list derived from a pan-EU consumer atti-inal survey [21].

atisti

ealthted puate. Paron thoke

n the fromealth-steE. Stuctinre-cling. Cetere m

E clinvesnce ariated asearmassocst fousingth IB

ults

rticip

ticipath m

hade parant dtly (Petireed toot cate a

edite

rall f samdjust(P = 0

stat). Agecorin

SE B 0.294s of the present study were to evaluate the associationHE and adherence to the MD and the associations ofviours with perceived health based on responses to anal online-survey among older UK adults.

s

dy was approved by the Human Psychology Ethics- at Newcastle University (Registration no. 000436) and

the principles embodied in the Declaration of Helsinki. provided consent via an e-form.

esign

tional, cross-sectional online-survey of dietary habits,erceived health. This paper is reported following thetement [23].

uited 82 men and 124 women, aged 50 years,g the general public. By design, participants ofhad to have regular access to internet and email

from age, no other inclusion/exclusion criteriaoyed. Recruitment was supported by Voice North,unity engagement panel at Newcastle University.ncl.ac.uk/changingage/engagement/voicenorth/),

versity of the Third Age (http://www.u3a.org.uk/). Thealso advertised among staff at Newcastle University,iversity Belfast and University of Ulster. A snowballrocedure was attempted by asking participants to

invitation to join the study to others. Although par- this study were a convenience sample, we recruited a

of individuals in terms of education, physical activitytal and socioeconomic status and sex.

y

ey was developed using Newcastle University Informa-s and Services Form Builder resource for the creations. Participants volunteering to take part in this survey

e uniform resource locator (URL) for the survey by Pilote survey materials for clarity, understanding and time

mplete the survey was undertaken prior to the current

es

rvey included the following questionnaires:

ionnaire requesting information on demographic and factors and self-rated health status. Participants self-d their smoking habits and self-rated their physical

in a ve-point Likert scale from very low to veryMI was estimated from self-reported weight and height.ants self-rated their current perceived health statussing a ve-point Likert scale from very bad to verynd reported perceived changes in their health during

preceding the survey.

pretud

2.5. St

A hself-rato evalhealthbased non-smbeing irangedmore h

Twoof PBHconstrtwo, pclusterwere ddistancrion.

PBHwhen adherefor covreport(SE). Spassess to adjupared out wi

3. Res

3.1. Pa

Parage wisampleaveragsignicnicanwere reducatwere nmoder

3.2. M

Ove(31% oAfter acantly inverse(Fig. 1pants s(LSM 26.7 cal analysis

behaviours score (HBS) including tobacco smoking,hysical activity level, BMI and the MDPS was created

associations between health behaviours and self-ratedticipants were assigned one-point for each variablee following four criteria: BMI < 25; currently being ar; having moderate/high physical activity status; and

top MDPS-tertile (i.e. MDPS 7). Values for this score 0 to 4, with higher values indicating the presence ofy behaviours.

p cluster-analysis was used to identify natural groupsep one involved creation of pre-clusters from cases byg an algorithm known as the Cluster feature tree. In Stepusters were merged using agglomerative hierarchicalluster-membership and the optimal number of clusters

mined using model t indices including log-likelihoodeasure and the Schwarzs Bayesian information crite-

uster-membership was used as a categorical variabletigating links with lifestyle behaviours including MDusing the General Linear Model. Results were adjustedes including age, sex, and BMI. Adjusted results were

the Least Squares Means (LSM) and standard errorsan and Pearsons correlation coefcients were used toiations between variables. Partial correlation was usedr additional variables. Categorical variables were com-

the chi-square test. Statistical analyses were carriedM SPSS v19 for Windows.

ant characteristics

nts in this survey were on average 61 (SD 7) years ofales being 2.3 years older than females (P = 0.03). The

a sex ratio of 2:3 of male to female respectively. Onticipants were overweight (BMI 26 4 kg/m2), with noifference in mean BMI between men and women. A sig-

= 0.01) higher proportion of males reported that theyd from work and men were more likely (P = 0.02) to be

degree level or above (Table 1). Most participants (95%)urrent tobacco smokers and over 70% reported beinglcohol consumers.

rranean diet adherence

mean MDPS was 5.6 2; participants in the top tertileple) scored 7, and only 6% of participants scored 10.ment for age and BMI, women (n = 124) scored signi-.016) higher than men (n = 82) (Table 2). There was anistically signicant association between MDPS and BMI- and sex-adjusted comparisons showed that partici-g 9 in the MDPS were signicantly (P = 0.004) leanerMI 23.8 0.926) than those scoring 8 (LSM SE BMI).

294 J. Lara et al. / Maturitas 79 (2014) 292298

Table 1Characteristics of participants.

Men(n = 82)

Women(n = 124)

Mean SD Mean SD PaAge (years) 62.2 7.5 59.9 7.1 0.029Body mass index (kg/m2) 26.9 3.9 26.0 7.1 0.122

N (%) N (%) Pb

Place of birthUK born 77 (93.9) 110 (88.7) 0.21Non-UK born 5 (6.1) 14 (11.3)

EducationBelow degree 35 (42.5) 73 (58.9) 0.02Degree and above 47 (57.3) 51 (41.1)

Marital statusMarried 63 (76.8) 82 (66.1) 0.10Not married 19 (23.2) 42 (33.9)

Retirement statusRetired 52 (63.4) 54 (43.5) 0.01Not retired 30 (36.6) 70 (56.5)

Smoking statusCurrent smoker 6 (7.3) 4 (3.2) 0.18Non-smoker 76 (92.7) 120 (96.8)

Drinking habDrinks alcohDoes not dri

Health statuNo medical Medical pro

High-choldisease, Ty

Pa P-value for

Participacomponentor rabbit mand

J. Lara et al. / Maturitas 79 (2014) 292298 295

Table 3Characteristics of PBHE clusters.

Non-adjusted least squares means SE P Adjusted least squares means SE P

Cluster 1(n = 42)

Cluster 2(n = 93)

Cluster 3(n = 65)

Cluster 1(n = 42)

Cluster 2(n = 93)

Cluster 3(n = 65)

Age (years) 57.2 1.1 61.6 0.7 61.7 0.9 0.001 57.2 1.06 61.5 0.71 61.8 0.86 0.001aBMI (kg/m2) 28.0 0.62 26.4 0.42 25.5 0.50 0.007 28.0 0.64 26.4 0.42 25.5 0.50 0.010bPREDIMED score (MDPS) 4.71 0.30 5.61 0.20 6.20 0.24 0.001 4.8 0.31 5.6 0.20 6.1 0.24 0.005cNo. of perceived barriers to healthy

eating (PBHE)6.9 0.24 3.5 0.16 1.9 0.19

296 J. Lara et al. / Maturitas 79 (2014) 292298

Given the strong evidence supporting its efcacy and effec-tiveness in enhancing health, the MD is the dietary pattern ofchoice when promoting a healthy diet [1416,25]. The PREDIMEDstudy show that closer adherence to a MD, e.g. a MDPS of 9 orhigher, waprevalence Importantlytwo-point health riskswith thesenicantly lcurrent stuas reportedline) [15]. Ein the currthe PREDIMting relativfor improveplied with food groupas well as wsuch as redrecommendbenets [1interventio

Several eating pattidentied the EU popwas the moNorthern Irness to chanoffering a middle-agevenience a[28]. In the1 characterand the becomprised lowest MDPtherefore liactivity) anfactors (i.e.this clustering it hardcomprised fewer PBHEno specic included olest MDPS that membfrom clustebut not HBin HBS thalacking coticipants ofcost of mative afuenlevel or abple were frtaught cookfrom worksadopting hechoice of he[30]. Based1 members

benet from the availability of healthier food in workplaces andfrom the development of a wider range of healthier, convenientfood-products [31].

Compliance with more health behaviours including non-g, bptio

wer 3]. Inong

healtsultswedbles

yea(25. In aalthalthcade

variaconoen he

[35]f die

he pr, butctiverevioecrea

reti in mr life

reshersentiis anlf-ratershintioing ttifyinong tationility

is lim is liked byeneright

l as patio

ted.onclnd ethy e. Ouely trceive nundinults d bevelydieta-retirs associated with signicantly lower odds ratios forof obesity and risk of cardiovascular diseases [15,26]., current evidence shows that small changes such as

increase in MDPS are strongly associated with lower [14,15]. Results from the present study were in line

ndings; individuals scoring 9 in MDPS were sig-eaner than those scoring 8. MD adherence in thedy was lower than that of Mediterranean populations

in the PREDIMED study (mean MDPS of 8.7 2 at base-ven the MDPS for those classied in the upper tertileent study (7.9 1.2) was below the mean values forED population, showing that even among those repor-ely healthier eating patterns, there is signicant roomment. Low numbers of participants in this study com-the MD recommendations of health benecial food ors such as nuts, sh, wine, fruit (men), and olive oil,ith recommendations to limit the consumption foods

meat and butter/margarine. These are food groups andations commonly associated with signicant health5,26] and therefore should be targeted in dietaryns.barriers can potentially limit adoption of healthiererns. The pan-European consumer attitudinal surveytime and taste as the most frequent PBHE amongulation [21]; for UK-based respondents, willpowerst frequent barrier [21]. A recent telephone-survey ineland reported that lack of willpower and willing-ge, were the most common PBHE [27]. A recent studyhealthy diet for 3 days to a small group of Scottish

adults, identied time pressures, desire for con-nd lack of motivation to cook as important PBHE

present study, three PBHE clusters emerged. Cluster-ised by busy lifestyle, irregular working hours,lief that healthy eating involves lengthy preparation,the younger, more overweight individuals with theS and HBS. These time pressure related PBHE werekely to impact other health behaviours (e.g. physicald contribute to the greater prevalence of health risk

BMI) and its associated risks among participants in. Cluster-2 identied lack of willpower and nd-

to give-up liked foods as characteristic PBHE andolder, but leaner participants with higher MDPS and. In contrast, those in cluster-3 were characterised byPBHE and reported the lowest number of PBHE andder, and the leanest, individuals who scored the high-and HBS. Interestingly, our adjusted results showeders of clusters-2 and -3 were signicantly differentr-1 in terms of BMI, MDPS, and number of PBHE,S; only members of cluster-2 however scored lowern members of cluster-3. Our ndings that cost andoking skills were not prominent PBHE among par-

this study may be explained by the relatively lowny foods in the MD eating pattern [29], the rela-ce of the participants (48% were educated to degreeove; Table 1) and a cohort effect these older peo-om generations who were more likely to have beening at school and/or from a parent. Recent evidenceite studies in the USA suggest that workers views onalthier eating patterns focus on availability of a wideralthful, convenient, more affordable food in cafeterias

on these results and the characteristics of cluster- in the current study, those in employment might

smokinconsumfold loage [2,PHS amrated Our reies of Svegeta(5564adults healththat heand hetwo denomicsocioebetwehealthnants o[36].

In tor HBSically awith pwith damongweightin late

TheresearcOur idanalysand semembintervescreenin stra

Aminformdesirabstudy whichreectwith gistics mas welConrmwarran

In ca fast ato healpeopleare likself-pewith thconfouour resassesseobjectioping of perieing physically active, moderate alcohol intake, andn of fruit and vegetables, is associated with up to four-mortality-risk, equivalent to 14 years in chronological

the present study, our HBS was associated with better participants. Signicant evidence indicates that self-h is an independent predictor of mortality [5,32,33].

are in line with those from recent cross-sectional stud-ish adults indicating that modest increases in fruit andintake, or leisure-time physical activity by older adultsrs) [34], as well as exercise and total physical activity in64 years) [35], are strongly associated with self-ratedddition, a longitudinal study of Swedish adults found

behaviours including smoking, exercise, social supporty eating were associated with self-rated health up tos later [6]. Self-rated health is associated with socioeco-bles such as education and employment status [35] butmic status does not necessarily explain the relationshipalth behaviours such as physical activity and self-rated. However economic factors may be important determi-tary choices in later life, particularly around retirement

esent study retirees did not differ in terms of BMI, MDPS, were more likely to report less PBHE and being phys-

than non-retired participants. These results contrastus ndings suggesting greater obesity risk associatedsed consumption of fruits and vegetables particularlyrees from manual occupations [37]. Increased bodyid-life predicts poorer physical and cognitive health

[38].ults of this study have important implications for

interested in behaviour change and for policy makers.cation of natural groups of PBHE through cluster-

d their association with body mass, health behavioursed health suggests that pre-screening for PBHE cluster-p may have utility in personalising lifestyle-basedns. These ndings may help in the development ofools in the form of standardised questionnaires usefulg individuals.he limitations of this study is the fact that self-reported

is susceptible to recall and other biases such as social [39]. In addition, generalisation of the ndings from thisited by the nature of the population sample surveyedely to have included health-motivated individuals, as

the low rates of smoking and mean BMI in comparisonal populations of retirement age [40]. These character-

have inuenced our ndings on the clustering of PBHEresence or absence of other health-related behaviours.n of these ndings using a larger population sample is

usion, this e-survey was well received and proved to befcient way to screen dietary habits, perceived barriersating, and perceptions of health status among UK olderr study showed that individuals reporting more PBHEo have lower MDPS, and higher BMI and HBS. Currented health was signicantly and positively associatedmber of health behaviours. Although we cannot excludeg from different sources of bias (e.g. optimistic bias),are consistent with evidence indicating these readily-havioural factors are associated with perceived, and

measured, health. These ndings may help in devel-ry interventions promoting the MD among individualsement age.

J. Lara et al. / Maturitas 79 (2014) 292298 297

Contributors

Jose Lara and John C. Mathers conceived and designed thestudy and developed the study objectives. Leigh-Ann McCrum wasresponsiblewrote the tation of da

Competing

The auth

Funding in

The auth

Acknowled

The authsurvey.

JL and JCa research Lifelong HeMedical Resnology andPhysical SciCouncil, Metish GovernResearch (NCare Researern IrelandHealth and no. G09006

Appendix A

Supplemin the onl2014.07.003

References

[1] US Burdeof disease

[2] Khaw KT,health betive popu

[3] van Damlifestyle f2008;337

[4] Murray CUK healthLancet 20

[5] Idler EL, Bcommun

[6] Mood C. Lysis span

[7] Mathers JNutr Soc

[8] Kiefte-deingredien

[9] Schfer Mnity for ch2012;35:

[10] Wilson Dhealth prInt Honor

[11] So F, Cesand healt

[12] So F, Abence to thmeta-ana

[13] So F, Macchi C, Abbate R, Gensini GF, Casini A. Mediterranean diet and healthstatus: an updated meta-analysis and a proposal for a literature-based adher-ence score. Public Health Nutr 2013:114.

[14] Salas-Salvado J, Bullo M, Babio N, Martinez-Gonzalez MA, Ibarrola-Jurado N,Basora J, et al. Reduction in the incidence of type 2 diabetes with the Mediter-

ean dil. Diabruch Rventio3;368orgert, tradocard9;99:adakiluatiot. Br J let J,

promealth7;61:an KJ,ptionrt disiffere

J Br Da J, Hctivenatic r4;12:rney a pan37.ghes lthy 76.

von repont: gu8.rdervad t adh05.z-Canociatidomizrtneziales

obes2;7:eleton

vegeiking 0;13:cdiarmerienetite wnowre. Puine Cd: ass

2007bet E

za as alth Nay EJlth as2;41:yaminrent k1;26:ergreween

self-rdy. BMergre

and slth 20klin er adu7.yens

cheleneightlth N for acquisition of data and quality assurance. Jose Lararst manuscript. All authors contributed to the interpre-ta, writing and approved the nal manuscript.

interests

ors declare no conict of interest.

formation

ors have received no funding for this article.

gements

ors are grateful to the volunteers who took part in this

M acknowledge support from the LiveWell Programmeproject funded through a collaborative grant from thealth and Wellbeing (LLHW) initiative, managed by theearch Council (MRC) on behalf of the funders: Biotech-

Biological Sciences Research Council, Engineering andences Research Council, Economic and Social Researchdical Research Council, Chief Scientist Ofce of the Scot-ment Health Directorates, National Institute for HealthIHR)/The Department of Health, The Health and Socialch & Development of the Public Health Agency (North-), and Wales Ofce of Research and Development forSocial Care and the Welsh Assembly Government (Grant86).

. Supplementary data

entary material related to this article can be found,ine version, at http://dx.doi.org/10.1016/j.maturitas..

n of Disease Collaborators. The state of us health, 1990-2010: burdens, injuries, and risk factors. J Am Med Assoc 2013;310:591606.

Wareham N, Bingham S, Welch A, Luben R, Day N. Combined impact ofhaviours and mortality in men and women: the EPIC-Norfolk prospec-lation study. PLoS Med 2008;5:e12.

RM, Li T, Spiegelman D, Franco OH, Hu FB. Combined impact ofactors on mortality: prospective cohort study in US women. Br Med J:a1440.JL, Richards MA, Newton JN, Fenton KA, Anderson HR, Atkinson C, et al.

performance: ndings of the Global Burden of Disease Study 2010.13;381:9971020.enyamini Y. Self-rated health and mortality: a review of twenty-sevenity studies. J Health Soc Behav 1997;38:2137.ife-style and self-rated global health in Sweden: a prospective anal-ning three decades. Prev Med 2013;57:8026.C. Nutrition and ageing: knowledge, gaps and research priorities. Proc2013;72:24650.

Jong JC, Mathers JC, Franco OH. Nutrition and healthy ageing: the keyts. Proc Nutr Soc 2014;73:24959., Jaeger-Erben M, Bamberg S. Life events as windows of opportu-anging towards sustainable consumption patterns? J Consum Policy

6584.M, Palha P. A systematic review of published research articles onomotion at retirement. J Nurs Scholarsh: Off Publ Sigma Theta Tau

Soc Nurs/Sigma Theta Tau 2007;39:3307.ari F, Abbate R, Gensini GF, Casini A. Adherence to Mediterranean dieth status: meta-analysis. Br Med J 2008;337:a1344.bate R, Gensini GF, Casini A. Accruing evidence on benets of adher-e Mediterranean diet on health: an updated systematic review andlysis. Am J Clin Nutr 2010;92:118996.

rantria

[15] Estpre201

[16] de Ldiemy199

[17] Papevadie

[18] Goutionin h200

[19] LogAdoheaof dOff

[20] LarEffetem201

[21] Keaof S13

[22] Huhea269

[23] Eriktheme806

[24] SchSaldie114

[25] RuiAssran

[26] MaCosand201

[27] Appandof l201

[28] MaExpApp

[29] Dremo

[30] Devsalasity

[31] CompizHea

[32] Gilthea201

[33] Bencur201

[34] Sodbetandstu

[35] SodciseHea

[36] Conold721

[37] NooMein wHeaet: results of the PREDIMED-Reus nutrition intervention randomizedetes Care 2011;34:149., Ros E, Salas-Salvado J, Covas MI, Corella D, Aros F, et al. Primaryn of cardiovascular disease with a Mediterranean diet. N Engl J Med:127990.il M, Salen P, Martin JL, Monjaud I, Delaye J, Mamelle N. Mediterraneanitional risk factors, and the rate of cardiovascular complications afterial infarction: nal report of the Lyon Diet Heart Study. Circulation77985.

A, Scott JA. The Mediterranean eating in Scotland experience project:n of an Internet-based intervention promoting the MediterraneanNutr 2005;94:2908.Lapointe A, Lamarche B, Lemieux S. Effect of a nutritional interven-oting the Mediterranean food pattern on anthropometric prole

y women from the Quebec city metropolitan area. Eur J Clin Nutr1293300.

Woodside JV, Young IS, McKinley MC, Perkins-Porras L, McKeown PP. and maintenance of a Mediterranean diet in patients with coronaryease from a Northern European population: a pilot randomised trialnt methods of delivering Mediterranean diet advice. J Hum Nutr Diet:iet Assoc 2010;23:307.obbs N, Moynihan PJ, Meyer TD, Adamson AJ, Errington L, et al.ess of dietary interventions among adults of retirement age: a sys-eview and meta-analysis of randomized controlled trials. BMC Med60.JM, McElhone S. Perceived barriers in trying to eat healthierresults-EU consumer attitudinal survey. Br J Nutr 1999;81(Suppl 2):

G, Bennett KM, Hetherington MM. Old and alone: barriers toeating in older men living on their own. Appetite 2004;43:

E, Douglas GA, Matthias E, Stuart JP, Peter CG, Jan PV. Strengtheningrting of observational studies in epidemiology (STROBE) state-idelines for reporting observational studies. Br Med J 2007;335:

H, Fit M, Estruch R, Martnez-Gonzlez MA, Corella D, Salas-J, et al. A short screener is valid for assessing Mediterraneanerence among older Spanish men and women. J Nutr 2011;141:

ela M, Estruch R, Corella D, Salas-Salvad J, Martnez-Gonzlez MA.on of mediterranean diet with peripheral artery disease: the predimeded trial. J Am Med Assoc 2014;311:4157.-Gonzlez MA, Garca-Arellano A, Toledo E, Salas-Salvad J, Buil-P, Corella D, et al. A 14-item Mediterranean diet assessment toolity indexes among high-risk subjects: the PREDIMED trial. PloS ONE43134.

KM, McGill R, Neville C, Woodside JV. Barriers to increasing fruittable intakes in the older population of Northern Ireland: low levelsand low awareness of current recommendations. Public Health Nutr51421.id JI, Loe J, Kyle J, McNeill G. It was an education in portion size.

ce of eating a healthy diet and barriers to long term dietary change.2013;71:4119.ski A, Eichelsdoerfer P. The Mediterranean diet: does it have to cost

blic Health Nutr 2009;12:16218.M, Nelson JA, Chin N, Dozier A, Fernandez ID. Pizza is cheaper thanessing workers views for an environmental food intervention. Obe-;15:57S68S., Jarlot A, Aidoo KE, Lean MEJ. Development of a nutritionally balanced

functional meal designed to meet published dietary guidelines. Publicutr 2013:110. FirstView., Vollaard AM, Kromhout D. Self-rated health and physician-rated

independent predictors of mortality in elderly men. Age Ageing16571.i Y. Why does self-rated health predict mortality? An update onnowledge and a research agenda for psychologists. Psychol Health140713.n M, McNaughton SA, Salmon J, Ball K, Crawford DA. Associationsfruit and vegetable intake, leisure-time physical activity, sitting timeated health among older adults: cross-sectional data from the WELLC Public Health 2012;12:551.

n M, Sundquist J, Johansson SE, Sundquist K. Physical activity, exer-elf-rated health: a population-based study from Sweden. BMC Public08;8:352.AI, Maguire ER, Monsivais P. Economic determinants of diet inlts: systematic review. J Epidemiol Community Health 2013;67:

AC, Visscher TL, Schuit AJ, van Rossum CT, Verschuren WM, van W, et al. Effects of retirement on lifestyle in relation to changes

and waist circumference in Dutch men: a prospective study. Publicutr 2005;8:126674.

298 J. Lara et al. / Maturitas 79 (2014) 292298

[38] Sun Q, Townsend MK, Okereke OI, Franco OH, Hu FB, Grodstein F. Adiposity andweight change in mid-life in relation to healthy survival after age 70 in women:prospective cohort study. Br Med J 2009;339:b3796.

[39] Lara JJ, Scott JA, Lean MEJ. Intentional mis-reporting of food consumption andits relationship with body mass index and psychological scores in women. JHum Nutr Diet 2004;17:20918.

[40] Sdergren M, Wang WC, Salmon J, Ball K, Crawford D, McNaughton SA. Pre-dicting healthy lifestyle patterns among retirement age older adults in theWELL study: A latent class analysis of sex differences. Maturitas 2014;77:416.

Association of Mediterranean diet and other health behaviours with barriers to healthy eating and perceived health among B...1 Introduction2 Methods2.1 Study design2.2 Sample2.3 E-survey2.4 Measures2.5 Statistical analysis

3 Results3.1 Participant characteristics3.2 Mediterranean diet adherence3.3 Health behaviours score (HBS)3.4 Perceived barriers to healthy eating (PBHE)3.5 PBHE classification and cluster characteristics3.6 Health behaviours score (HBS) and perceived health status (PHS)3.7 Retirement status

4 DiscussionContributorsCompeting interestsFunding informationAcknowledgementsAppendix A Supplementary dataReferences