Healthy eaters beat unhealthy eaters in prototype evaluation among men, but abstinence may pose a risk for social standing

Full Te ms & Condi ions o access and use can be ound a h p://www. and online.com/ac ion/jou nalIn o ma ion?jou nalCode= hpb20 Download by: [Tampe e Uni e si y] Da e: 19 Sep embe 2016, A : 23:58 Heal h Psychology and Beha io al Medicine An Open Access Jou nal ISSN: (P in ) 2164-2850 (Online) Jou nal homepage: h p://www. and online.com/loi/ hpb20 Heal hy ea e s bea unheal hy ea e s in p o o ype e alua ion among men, bu abs inence may pose a isk o social s anding Ma ja Kinnunen, Nelli Hankonen, A i Haukkala, B i a Renne , Piia Jallinoja, Cla issa M. L. Bingham & Pil ikki Abse z To ci e his a icle: Ma ja Kinnunen, Nelli Hankonen, A i Haukkala, B i a Renne , Piia Jallinoja, Cla issa M. L. Bingham & Pil ikki Abse z (2015) Heal hy ea e s bea unheal hy ea e s in p o o ype e alua ion among men, bu abs inence may pose a isk o social s anding, Heal h Psychology and Beha io al Medicine, 3:1, 323-336, DOI: 10.1080/21642850.2015.1095097 To link o his a icle: h p://dx.doi.o g/10.1080/21642850.2015.1095097 © 2015 The Au ho (s). Published by Taylo & F ancis. Published online: 19 Oc 2015. Submi you a icle o his jou nal A icle iews: 280 View ela ed a icles View C ossma k da a Heal hy ea e s bea unheal hy ea e s in p o o ype e alua ion among men, bu abs inence may pose a isk o social s anding Ma ja Kinnunen a *, Nelli Hankonen b,c , A i Haukkala c , B i a Renne d , Piia Jallinoja a,e , Cla issa M. L. Bingham a, and Pil ikki Abse z a,g a Wel a e and Heal h P omo ion Uni , Na ional Ins i u e o Heal h and Wel a e, Helsinki, Finland; b Depa men o Social Psychology, School o Social Sciences and Humani ies, Uni e si y o Tampe e, Tampe e, Finland; c Depa men o Social Resea ch/Social Psychology, Uni e si y o Helsinki, Helsinki, Finland; d Depa men o Psychology, Uni e si y o Kons anz, Kons anz, Ge many; e Consume Socie y Resea ch Cen e (KTK), Depa men o Poli ical and Economic S udies, Uni e si y o Helsinki, Helsinki, Finland; Cance Socie y o Finland, Helsinki, Finland; g School o Heal h Sciences, Uni e si y o Tampe e, Helsinki, Finland (Recei ed 28 Janua y 2015; accep ed 14 Sep embe 2015) Al hough p e ious s udies ha e shown ha unheal hy ea ing is associa ed wi h a posi i e image o a ypical unheal hy ea ing pee (p o o ype), hei ocus on p o o ypes is ypically na ow/ limi ed. The p esen s udy add esses his gap in he li e a u e and in es iga es wo aspec s in pee e alua ions: (1) heal hy s. unheal hy and (2) abs aine s. choose . Mo eo e , hei mean di e ences, in e ela ionships and associa ions wi h ea ing beha io will be examined. Me hods: Men in mili a y se ice (N= 1824, Age M = 19.8) epo ed hei ea ing beha io s and e alua ed ei he (a) Vege able choose and Vege able abs aine p o o ypes (N= 920) o b) Fa choose and Fa abs aine p o o ypes (N= 904) on a scale con aining 17 an onyms. Explo a o y and confi ma o y ac o analyses sugges ed a h ee- ac o s uc u e: Sel - egula ion, Social s anding in pee g oup and Appea ance (Compa a i e Fi Indexes 0.82– 0.87; Tucke –Lewis index 0.78–0.84; Roo Mean Squa e E o o App oxima ion 0.07– 0.08). Resul s: Heal hy ea e s (i.e. Vege able choose and Fa abs aine ) we e e alua ed highe on Sel - egula ion and Appea ance han espec i e unheal hy ea e s (i.e. Vege able abs aine and Fa choose ) (ps < .001). Howe e , Fa abs aine was a ed lowe han Fa choose on Social s anding in pee g oup (p< .001). Nei he Fa choose no Fa abs aine p o o ype was associa ed wi h a y ood consump ion. F ui and ege able consump ion was associa ed wi h highe Sel - egula ion a ings o Vege able choose (β= .10, p< .01) and lowe Sel - egula ion a ings (β=−.13, p< .001) and Appea ance a ings (β=−.08, p< .05) o Vege able abs aine . Conclusions: Among young men, heal hy ea ing pee s a e e alua ed as mo e sel - egula i e and be e looking han unheal hy ea ing pee s. Ra ing heal hy ea e s mo e posi i ely is ela ed o highe ui and ege able consump ion. P o o ypes play a ole in young men’s ea ing beha io , al hough he ole di e s o ege ables and a y oods. In e en ions o inc ease ui and ege able consump ion should conside add essing he ege able ea e p o o ype. Keywo ds: p o o ypes; social image; ea ing beha io ; ood choice; pee s © 2015 The Au ho (s). Published by Taylo & F ancis. *Co esponding au ho . Email: ma ja.kinnunen@ hl.fi This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/ licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. Heal h Psychology and Beha io al Medicine, 2015 Vol. 3, No. 1, 323–336, h p://dx.doi.o g/10.1080/21642850.2015.1095097 1. In oduc ion The consump ion o ui s and ege ables among young men is in equen and a om he na ional ecommenda ion’s minimum o 500 g pe day (Bingham e al., 2010; Na ional Nu i ion Council, 2005). Ea ing beha io is d i en by biological needs such as sa is ying hunge . None- heless, ood choices a e also dependen on cul u al and social no ms (S ead, McDe mo , MacKin osh, & Adamson, 2011) and also p o ide a way o exp ess g oup membe ship like e hnici y o social class, and hus, need o be examined om wide pe spec i es. Unde s anding he an eceden s o heal hy and unheal hy ood consump ion would enable e idence-based in e - en ion design. An influen ial heo e ical amewo k depic ing social influences on heal h- ela ed beha io is he p o o ype-willingness model, emphasizing he ole o p o o ypes in he decision-making p ocess. P o o ypes a e defined as social images o a ypical pee who engages in ce ain beha io (e.g. “a ypical smoke o you age”) (Gibbons & Ge a d, 1995; Hahn & Renne , 1998). Acco d- ing o he model, posi i e o nega i e alence o he p o o ype p edic s he willingness o engage in he espec i e beha io . Two ecen me a-analyses ha e shown suppo o he p o o ype-will- ingness model (Todd, Ko he, Mullan, & Monds, 2014; an Le ow, de V ies, Bu do , & an Empelen, 2014). One o he s eng hs o he p o o ype-willingness model is he in es iga ion o he social impac o pee s upon beha io h ough social images ( a he han di ec sel - epo ed influence). Van Le ow e al.’s(2014) me a-analysis shows ha p o o ypes ha e e ec no only on in en ion and willingness, bu on beha io as well. Fo example, young smoke s end o e alua e a ypical pee smoke mo e posi i ely han a non-smoke (Piko, Bak, & Gibbons, 2007). Also, a mo e posi i e pe cep ion o a ypical pee d inke p edic s highe willingness o d ink in social si ua ions (Zimme mann & Sie e ding, 2011) as well as ac ual alcohol consump ion (Ge a d e al., 2002). S udies explo ing social images o heal hy and unheal hy ea e s show ha heal hy ea e s a e gene ally pe cei ed mo e posi i ely compa ed o unheal hy ea e s (Ba ke , Tandy, & S ookey, 1999; Ge i s, de Ridde , de Wi , & Kuije , 2009; Ge i s e al., 2010; Oakes & Slo e back, 2005). When asked o desc ibe ypical heal hy ea e s, people gene ally gi e posi i e desc ip ions, bu also end o associa e some nega i e cha ac e is ics wi h heal hy ea e s. While heal hy ea e s a e desc ibed wi h i uous sel - egula i e cha ac e is ics like “ac i e”(Ge i s e al., 2009), “du i ul”and “sel -con olled”(Ba ke e al., 1999), hey a e also pe cei ed as “sel -cen e ed” (F ies & C oyle, 1993). Unheal hy ea e s in u n ge mo e nega i e desc ip ions ela ed o hei appea ance, o example, “chubby”, bu a e also pe cei ed posi i ely wi h cha ac e is ics, o example, “sociable”,“en e aining”and “masculine”(F ies & C oyle, 1993; Ge i s e al., 2009; Oakes & Slo e back, 2005). Gende di e ences in he pe cep ions o he p o o ype con en ha e no always been ound, (Ge i s e al., 2009), bu male adolescen s ha e been ound o hold mo e posi i e unheal hy ea e p o o ypes and mo e nega i e heal hy ea e p o o- ypes (S einhilbe , Fuchs, & Dohnke, 2013). The associa ions o ea e p o o ypes and ea ing seem o be complex, al hough se e al s udies ha e shown a posi i e ela ionship be ween posi i e p o o ypes o unheal hy ea e and ela ed ea ing beha io s (Ba ke e al., 1999; Ge i s e al., 2009,2010). Fo example, gende seems o mode a e he e ec o ea e p o o ypes on ea ing beha io s. P o o ype pe cep ion pa ially explains gende di e ences in ea ing among adolescen s, especially among olde age adolescen s (S einhilbe e al., 2013). Also, a o able images o unheal hy ea e s associa e wi h highe a y ood consump ion among gi ls, bu wi h highe so d ink consump ion among boys (Ge i s e al., 2009). Dohnke, S einhilbe , and Fuchs (2015) es ed he ull p o o ype-willingness model on ea ing beha io and ound ha among adolescen s, only he unheal hy ea e p o o ype was associa ed 324 M. Kinnunen e al. wi h ea ing. The unheal hy ea e p o o ypes p edic ed subsequen unheal hy ea ing, and no only h ough in en ion and willingness as he model p edic s, bu also di ec ly. This finding is ac ually in line wi h an Le ow e al.’s(2014)finding ha heal h- isk p o o ypes gene ally exhibi s on- ge associa ions wi h ou come a iables han heal h-p o ec i e p o o ypes ( an Le ow e al., 2014). In e ms o ea ing, simila ly a o able images o heal hy ea e s ha e no always been ound o ela e o one’s own ood choices (Ge i s e al., 2009), p obably because heal hy ea e s end o ge posi i e a ings no only om o he heal hy ea e s bu om unheal hy ea e s as well (Ba ke e al., 1999). So a , he ques ion in li e a u e has been whe he and how p o o ypes influence he heal hi- ness o ea ing beha io . Howe e , Ri is, Shee an, and A mi age (2006) no ed ha p o o ypes could ac ually e e ei he o a ypical pe son engaging in a beha io (ac o s) o o a ypical pe son abs aining om pe o ming a ce ain beha io (abs aine s). Hence, p o o ypes can be inhibi o y (i.e. no wan ing o be like a ypical pee who a oids heal hy oods) o acili a ing (i.e. wan ing o gain he a o able cha ac e is ics o he heal hy ood choosing pee ). Combining he di e en ypes o p o o ypes, ou p o o ype dimensions can be dis inguished: (1) isky beha - io ac o p o o ype ep esen ing a ypical pee who engages in isky beha io s (e.g. ea s unheal hy oods), (2) a isky beha io abs aine p o o ype ep esen ing a ypical pee who abs ains om isky beha io (e.g. does no ea unheal hy oods), (3) a heal hy beha io ac o p o o ype ep e- sen ing a ypical pee who ac i ely engages in heal hy beha io (e.g. ea s heal hy oods) and finally (4) a heal hy beha io abs aine p o o ype ep esen ing a ypical pee who abs ains om heal hy beha io s (e.g. does no ea heal hy oods). In e es ingly, he ( ew) posi i e desc ip ions ob ained o desc ibe unheal hy ea e s o en eflec social and elaxing aspec s, like “ un-lo ing”(Ba ke e al., 1999)o “sociable”(Ge i s e al., 2009). Also, he ( ew) nega i e desc ip ions ob ained o heal hy ea e s end o eflec abs inence om some hing, o example, “se ious”o “highly s ung”(Ba ke e al., 1999). I is possible ha as espec ed and eligible heal hy ea ing may be, abs aining om unheal hy oods is no a desi able s a egy o pu suing heal h in social si ua ions. One o he co e ideas o he p o o ype-willingness model is ha heal h beha io s ake place in social con ex , in which in en ion o do some hing ei he does no exis a all o is so weak ha i can be o e idden by willingness o do some hing en i ely di e en (Gibbons & Ge a d, 1995). Mos ecen findings sugges ha in ea ing con ex , he ac ual beha io can be p edic ed di ec ly by p o o ypes (Dohnke e al., 2015). The e o e, he dis inc ion ou lined by Ri is e al. (2006) aises in e es ing ques ions; does he abs aining o ac i e na u e o (heal hy o unheal hy) ea ing beha io make a di e ence in e ms o p o o ype a o abili y? Heal hy ea ing di e s om o he heal h beha io s in a ma ked way: Fo smoking, he e is only one way o choosing a heal hy op ion, ha is, by abs aining om smoking. Simila ly, o physical ac i i y, he only heal hy way is o ac i ely engage in being physically ac i e. Howe e , one can ea heal hily bo h by choosing heal hy oods and by abs aining om unheal hy ones. I is unclea how his influences p o o ypes. P e ious s udies ha e examined associa ions be ween unheal hy o heal hy ea e p o o ypes and ac ual unheal hy o heal hy ea ing beha io . To ou knowledge, none has in es iga ed all he ou abo e-men ioned di e en p o o ypes, and hei associa ions wi h espec i e ea ing be- ha io . The p esen s udy will add ess his gap in he li e a u e. The aims o he s udy a e (1) o explo e heal hy beha io ac o o abs aine p o o ypes (“Veg- e able choose ”and “Vege able abs aine ”) and isky beha io abs aine o ac o p o o ype (“Fa abs aine ”and “Fa choose ”), (2) o iden i y whe he hese ou p o o ypes ha e simila ac o s uc u es and (3) whe he hese ac o s a e in e ela ed. The final aim (4) is o explo e he associ- a ions o hese ob ained ac o s wi h ea ing beha io . Heal h Psychology and Beha io al Medicine 325 Based on p e ious esea ch among adolescen s (Ge i s e al., 2009,2010), we expec ha he p o o ype e alua ions will depic he ollowing ypes o p o o ype ac o s among young adul men oo: one ha ep esen s sel - egula o y cha ac e is ics, ano he ha con ains i ems ha ela e o sociable cha ac e is ics and a hi d ha ela es o appea ance. Conside ing he p e ious esea ch on e alua ions o heal hy and unheal hy ea e s (e.g. Ba ke e al., 1999; Ge i s e al., 2009), we p esen he ollowing hypo heses: (1) he heal hy ea e p o- o ypes (Vege able choose and Fa abs aine ) a e a ed mo e posi i ely by hei appea ance and sel - egula i e cha ac e is ics han he espec i e unheal hy ea e p o o ypes (Vege able abs aine and Fa choose ). (2) Assuming ha he sociable cha ac e is ic ac o is ound, choose s a e a ed o e abs aine s in he sociable cha ac e is ics ac o wi hin bo h heal hy and unheal hy ea ing be- ha io . (3) As mos heal hy o unheal hy ea e p o o ypes ound in ea lie li e a u e seem o eflec sel - egula i e ai s i idly (Ge i s e al., 2009), we expec sel - egula i e p o o ypes o play a c ucial ole in p edic ing esponden s’own ea ing beha io . And (4) finally, assuming ha abs aining om ood is no socially accep able, and knowing om p e ious esea ch ha unheal hy p o o ypes ha e mo e p edic i e powe on beha io (Dohnke e al., 2015; Ge i s e al., 2009), we assume ha Vege able abs aine p o o ypes will ha e he s onges associa ion wi h espec i e ea ing beha io . 2. Me hods The p esen s udy is a pa o he De enceNu i S udy (Bingham e al., 2012; Hankonen, Kinnunen, Abse z, & Jallinoja, 2014) ha ook place in wo ga isons: Kainuu B igade in No h-Eas e n and A mou ed B igade in Sou he n Finland. In Finland, e e y man is liable o mili a y se ice. Du ing hei se ice ime, consc ip s li e in ga isons. A ga isons, men a e p o- ided h ee meals plus an op ional e ening snack. The p o ided meals a e planned o mee he na ional nu i ional ecommenda ions as well as he demands he consc ip s ace du ing he se ice (e.g. i ha ing a physically ough camping, he p o ided meals con ain mo e ene gy). Fa y oods, such as pizzas, kebabs o bu ge s and so on a e ea en as snacks a he han as planned meals. Howe e , a e he fi s weeks, i no on encampmen , hey o en spend weekends and some o he ee days a home. Nea ly 80% o he age coho comple es he se ice. Abou 95% o hose who se e en e se ice a he yea when hey u n 19 o 20 yea s (De ence Command Public In o ma ion Di ision, 2012). Da a we e collec ed in 2008 and 2009 om ou coho s en e ing hei se ice in Janua y and July. Two se ice uni s om bo h ga isons we e included o pa icipa e, each uni con aining om 80 o 150 men. Pa icipa ion was olun- a y, and in o med signed consen was ob ained om hose willing o pa icipa e. Pa icipa ion ook place du ing he se ice ime. The s udy p o ocol was app o ed by he e hics commi ee o he Hospi al Dis ic o Helsinki and Uusimaa. 2.1. Pa icipan s The ou coho s con ained 2.125 male consc ip s o whom 1.824 (85%, Age: M = 19.8 yea s, SD = 0.81, ange 19–28) filled in ques ionnai e in Finnish language du ing he fi s week o hei mili- a y se ice. O all pa icipan s, 98.6% we e 22 o younge . Excluding he 25 olde pa icipan s om he analyses did no change he esul s o he co ela ions o eg ession analyses. Thus, hey we e included in all analyses. D opou was due o in e up ion o se ice, being ill du ing he measu emen s, mili a y ans e s o e usal o pa icipa e in he s udy. The pa icipan s epo ed hei highes educa ion achie ed as ollows: seconda y school 12%, oca ional aining 42%, uppe seconda y school 44% and 3% had highe o some o he educa ion. Majo i y o he pa icipan s (76%) s ill li ed wi h hei pa en s, and only 22% li ed alone o wi h a spouse. 326 M. Kinnunen e al. 2.2. Measu es 2.2.1. P o o ypes and p o o ype ac o s P e ious ea e p o o ype s udies ha e measu ed “ ypical heal hy ea e /unheal hy ea e o you age”o “a ypical pe son o you age who ea s he ollowing oods …” We decided o choose oods ha (a) a e e y essen ial pa o any heal hy die (i.e. ui and ege ables) and (b) oods ha a e unheal hy and mo e likely o be consumed wi h pee s han alone and (c) a ailable in ga - ison a ea ( a y ood). Bo h ood g oups ha e been a ge ed by heal h educa ion o decades, and hence, hei associa ions wi h heal h a e ela i ely well known among popula ion oo (Jallinoja, Ni a, Helako pi, & Kahma, 2014). Mo eo e , he consump ion o ui s and ege ables among consc ip s is oo low as ega ds ecommenda ion (Bingham e al., 2009). Pa icipan s we e andomly assigned in o wo g oups; ei he on Fa g oup whe e hey e alu- a ed a ypical “Fa choose ”and a ypical “Fa abs aine ”(N= 904) o on Vege able g oup whe e hey we e asked o e alua e a ypical “Vege able choose ”and a ypical “Vege able abs aine ” (N= 920).The ques ion: “We will ask you o e alua e a ypical consc ip , who beha es in a way desc ibed below”was ollowed by he defini ions o an abs aine and an ac o : Abs aine : “A ypical consc ip who ies o a oid a y oods and snacks”(Fa condi ion)/ “ ege ables and ui s”(Vege able condi ion) and Ac o : “A ypical consc ip who s i es o choosing a y oods and snacks”(Fa condi ion)/“ ui s and ege ables”(Vege able condi ion) “a e e y meal”. Pa icipan s e alua ed each ou ypes on scales o 17 an onyms. The adjec i es we e selec ed based on ea lie s udies (Ba ke e al., 1999; F ies & C oyle, 1993) and on a quali a i e pilo s udy among male consc ip s. The adjec i e pai s we e asked o be answe ed on a 5-poin scale wi h he an onym adjec i es a he ex emes. Pa icipan s we e asked o “choose an op ion om he scale ha mos accu a ely ep esen s you opinion”. The an onyms we e he ollowing: easy- going–up igh , popula –unpopula , un eliable– eliable, con incing–uncon incing, physically fi –physically unfi , childish–g own-up, a –skinny, sel -indulgen –aus e e, masculine– eminine, ca eless–me iculous, esponsible–i esponsible, insecu e–confiden , muscula –lanky, ashionable– un ashionable, dumb–in elligen , a ac i e–una ac i eandgoodcompany–dull. App op ia e e e sions we e done o highe sco es o eflec mo e posi i e e alua ions. Explo a o y ac o analyses wi h Maximum likelihood ex ac ion and Va imax o a ion we e conduc ed o in es iga e he ac o s uc u e o he ou p o o ypes (Vege able choose , Vege able abs aine , Fa choose and Fa abs aine ). The ob ained ac o solu ion was in es iga ed wi h con- fi ma o y ac o analysis. The fi s analyses yielded a h ee- ac o solu ion o Fa choose and Vege able a oide and ou - ac o solu ions o Fa a oide and Vege able choose . Fo he h ee- ac o solu ion he c i e ia we e eigen alues ≥1 and sc ee plo s confi med he esul s. Fo he ou - ac o solu ion, eigen alues we e ≥1, bu he sc ee plo s we e less clea . Two i ems we e d opped (con incing and aus e e) due o c oss-loadings and field esea che s’ eedback om he commen s om consc ip s conce ning he meaning o he i ems. In subsequen analyses, he h ee- ac o solu ion was suppo ed by eigen alues and sc ee plo s. The h ee- ac o explana- o y ac o solu ions explained om 46.3% o 52% o a iance. The ac o s uc u e was examined wi h confi ma o y ac o analysis, wi h Full In o ma ion Maximum Likelihood o es ima ing missing da a. As an indica ion o accep able model fi , Com- pa a i e Fi Indexes anged om 0.82 o 0.87, Tucke –Lewis index om 0.78 o 0.84 and he Roo Mean Squa e E o o App oxima ion om 0.07 o 0.08. Fo all ou p o o ypes, h ee simila ac o s eme ged. The ac o s we e Sel - egula ion (i ems: eliable, g own-up, me iculous, esponsible and in elligen ), Social s anding in pee g oup (i ems: easy-going, popula , masculine and good company) and Appea ance (i ems: muscula , ashion- able, a ac i e, skinny and physically fi ). P o o ype i ems we e subsequen ly summed in o Heal h Psychology and Beha io al Medicine 327 composi e sco es o he analyses. The C onbach’s alphas we e sa is ac o y a 0.60 le el. Alphas o o he Fa abs aine sum sco es: Sel - egula ion α= 0.68 and Social s anding in pee g oup α= 0.64. Fo Fa choose sum sco es: Sel - egula ion α= 0.73, Social s anding in pee g oup α= 0.66 and Appea ance α= 0.68. Fo Vege able abs aine sum sco es: Sel - egula ion α= 0.72, Social s anding in pee g oup α= 0.60 and Appea ance α= 0.65. Fo Vege able choose sum sco es: Sel - egula ion α= 0.74, Social s anding in pee g oup α= 0.73 and Appea - ance α= 0.66.). As a and ege able p o o ypes we e measu ed in di e en g oups, he s a is ical compa isons we e made wi hin he g oup (i.e. a p o o ypes in one and ege able p o o ypes in ano he ). 2.2.2. Ea ing beha io Ea ing was measu ed by a 36-i em ood equency ques ionnai e asking on how many days du ing he pas week in ci ilian li e he esponden had consumed ce ain oods (i em ange 0–7). The ques ionnai e was based on se e al co esponding ques ionnai es among Finns (e.g. Paalanen e al., 2006) and adjus ed o consc ip s on he bases o a p e ious ood dia y s udy among Finnish consc ip s (Bingham e al., 2009). Fo he pu poses o he p esen s udy, wo ea ing indexes we e c ea ed: F ui and Vege able Index o desc ibe he consump ion o ui s o be ies and esh ege able ( he mean o he wo i ems), and Fa y Food Index o measu e a y ood consump ion ( he mean o fi e i ems: F ench ies, po a o chips, pizza and kebab, hambu - ge s and ho dogs, mea pies and sa o y pas ies). The Fa y Food Index was based on ea lie s udies o ea ing beha io among men in mili a y se ice (Bingham e al., 2011; Jallinoja e al., 2011). The F ui and Vege able Index a ied be ween 0.00 and 7.00 (M = 2.93, SD = 1.88) and he Fa y Food Index be ween 0.00 and 6.20, (M = 1.01, SD = 0.78). 2.3. S a is ical analysis IBM SPSS 22 was used o ob ain desc ip i e s a is ics, o explana o y ac o analysis wi h maximum likelihood ex ac ion and a imax o a ion. Fo confi ma o y ac o analyses, Mplus e sion 6.1 was used. The fi s wo hypo heses we e es ed wi h - es s. A Pea son co ela ion coe ficien was used o explo e i he ac o s a e ela ed, as well as o he hi d hypo hesis. Fo he ou h and fi h hypo heses, o p edic a o ege able ea ing by he espec i e abs aine o choose p o o ype ac o s, linea eg ession analysis was also used. 3. Resul s Ou fi s hypo hesis an icipa ed ypical heal hy ea e p o o ypes (Vege able choose and Fa abs aine ) o be a ed mo e posi i ely han espec i e unheal hy ea e p o o ypes (Vege able abs aine and Fa choose ). The findings suppo ed he hypo hesis o wo o he h ee ac o s: Sel - egula ion and Appea ance. The means and confidence in e als o he di e en p o o ype ac o s a e epo ed in Figu e 1. Acco ding o wo- ailed - es s, a ypical Vege able choose was a ed highe han Vege able abs aine on Sel - egula ion ( (909) = 22.94, p<.001) and on Appea ance ( (912) = 21.02, p<.001). Fa abs aine was a ed highe han Fa choose on Sel - egula ion ( (892) = 25.35, p<.001) and on Appea ance ( (892) = 25.88, p<.001). The a ings o Social s anding we e mo e s ongly influenced by he abs aining s. choosing na u e a he han he heal hiness o beha io . This suppo s ou second hypo hesis, which an icipa ed mo e posi i e a ings o choose s a he han abs aine s; bo h Vege able choose and Fa choose we e a ed highe on Social s anding han he espec i e abs aine s (Vege able ( (910) = 2.91, p= .004; Fa ( (892) = −8.78, p<.001). Confi ming he fi s hypo hesis, single-i em compa isons 328 M. Kinnunen e al. e ealed ha heal hy ea e p o o ypes we e pe cei ed mo e posi i ely along 16 i ems han espec - i e unheal hy ea e p o o ypes (all ps≤.001). In espec o masculini y, he associa ion was e e se: bo h Vege able choose and Fa abs aine we e seen as less masculine han Vege able abs aine (p= .001) and Fa choose (p< .001). The p ofile o single i ems can be seen om Figu e 2. The co ela ion ma ix o he p o o ype ac o s o all ou p o o ypes is shown in Table 1. Vege able choose p o o ype ac o s had significan nega i e co ela ions wi h all Vege able abs aine p o o ype ac o s wi h wo excep ions. Fa choose p o o ype ac o s had significan nega i e co ela ions wi h all Fa abs aine p o o ype ac o s wi h wo excep ions. Be ween choose s. abs aine p o o ypes, he co ela ions we e s onges o Sel - egula ion. Mo eo e , wi hin each p o o ype, he ac o co ela ions we e posi i e wi h he s onges co ela ions be ween Sel - egula ion and Appea ance (see Table 1). Figu e 1. Means and confidence in e als o he Sel - egula ion, Social s anding in pee g oup and Appea ance p o o ype ac o s. Figu e 2. P ofile diag am o he single i ems. Heal h Psychology and Beha io al Medicine 329 Table 1. Means, s anda d de ia ions and co ela ions be ween ea e p o o ype ac o s. Vege able choose Vege able abs aine Mª SD Sel - egula ion Social s anding Appea ance Sel - egula ion Social s anding Vege able choose Sel - egula ion 3.44 0.53 1 Social s anding 3.17 0.50 .35** 1 Appea ance 3.34 0.49 .53** .46** 1 Vege able abs aine Sel - egula ion 2.75 0.57 −.39** −.11* −.25** 1 Social s anding 3.10 0.47 −.06 −.20** −.05 .23** 1 Appea ance 2.74 0.55 −.38** −.20** −.39** .52** .32** Fa abs aine Fa choose Sel - egula ion Social s anding Appea ance Sel - egula ion Social s anding Fa abs aine Sel - egula ion 3.47 0.54 1 Social s anding 3.07 0.48 .26** 1 Appea ance 3.35 0.53 .42** .34** 1 Fa choose Sel - egula ion 2.75 0.49 −.35** −.04 −.22** 1 Social s anding 3.26 0.50 .15* −.11** .07 .17** 1 Appea ance 2.59 0.52 −.37** −.12* −.40** .55** .12** No es: a Min. = 1, Max. = 5. *p< .05 **p< .001. 330 M. Kinnunen e al.