T H E I E N O F L I F E S T Y L E C H A N

Nutrition; Awareness/Education/Communications; Culture Change/Supportive Environments;Underserved Populations: Low Income

Health and Eating Behavior Differs BetweenLean/Normal and Overweight/Obese Low-IncomeWomen Living in Food-Insecure EnvironmentsHeidi Dressier, RD; Chery Smith, PhD, MPH, RD

AbstractPurpose. To explore diffeiens and shnilarities in environmental, personal, and behavioral

factors influencing eating behavior among loio-income women of varying weight status.Design. Focus gtvups (n = 16) were used to collect qualitative data. Quantitative data

collected included demographic, lifestyle, and anthropovieiiic (heights and weights).Setting. Community centers, libraes, and homeless shelters in loxiyincome neighborhoods.Participants. Of the 83 participants, 58% were African-Ameiican, 28% were white, and

the remainder wee Ainerican Indian, Hispanic, or mixed race. A total of 35% of participantswere lean/normal (body mass index 25 kg/m ) .

Method. Focxts group methodology was iwied to collect data on 83 participants. SocialCognitive Theory (SCT) provided the theoretical basis. Transcripts were analyzed using theopen-coding method and comments between lean/normal and ovenoeight/obese women weresystematically compared.

Results. Focus group themes xoerefood environment, aberrant eating behavior, health values,and beliefs. Differences in personal and behavioral factm s were apparent between ovenueight/obese and lean/nomial women, with the former givup frequently discussing emotional eating,overeating, and stashing food, and the latta- group communicating greater' nutrition Imowledgeand describing regular physical activity. Both groups similarly expressed experiences xuithfoodinsecurity and lived in loxo-income environvients.

Conclusion. For loxo-income xuomen, personal and behavioral factors may modify theinfluence of their obesogenic food environment. Further research xuithin the context of SCTshould examine differ'ences between lean/normal and ovenueight/obese xoomen living in thesame environment. (AmJ Health Promot 2O3;27[6]:358-365.)

Key Words: Low-Income, Food Insecurity, Food En\nronment, Obesity, Lean, SCT,Prevention Research. Manuscript fomiat: research; Research purpose: descriptive;Study design: qualitative; Outcome measure: qualitative data; Setting: localcommunity; Target population age: women age 18-64; Target populationcircumstances: low-income, lean/normal, ovenveight/obese

Heidi Dressier, RD, is a Craduate Student and Chery Smith, PhD, MPH, RD, is anAssociate Professor at the Department of Food Science and Nutrition, University ofMinnesota, St. Paul, Minnesota.Send repritu requests to Cheiy SniiLh, PhD, MPH, RD, UnivcrsiLy of Minnesota, Department ofFood Science and Nutrition, 225 FScN, 1334 Eckles Avenue, St. Paul, MN 55108-6099; csmithumn.edu.

Tim manmcripl was submitted February 3, 2012; rettiswns were requested March 20, May 7, and July 19, 2012; themantiscripl was accefHed for publication July 30, 2012.Cof/yiighl 2013 by American Jount a I of Health Promotion, nc.0S90-I171/13/S5.00--0DOI: iO.427S/a}hp. 2019-QUAU55

PURPOSE

Obesity continues to be a leadingpublic health problem, partictilarlyamong low-income women.'" Foodinsecurit)', characterized by inadequateaccess to food that supports a healthy,active lifestyle, has been paradoxicallylinked to a higher prevalence ofovenveight and related complicationsin low-income women.'"'"^ Moreover,participation in the Supplemental Nti-trition Assistance Program (S^AP) hasbeen associated with increased,weight,^"' with research suggesting thatthe feast and famine ofthe food stampcycle contributes to overweight byaltering metabolism to promote weightgain7 Further implicating the foodenxironment, low-income indi\idtialshave easier access to cheap, energy-dense foods than healthier, moreexpensive alternatives. In addition,several barriers related to healthy foodchoice, such as cost, have been found

1 2 9to exist among lo\v-mcome women. '

Despite the obesity-promoting foodand economic en\aronments in whichwomen of low socioeconomic status(SES) live, there are many who are ableto maintain a normal weight. Al-though factors like genetics, energyexpenditure, and personal food pref-erence^'"^'^ may partly explain weightvariation, many underlying questionsremain as to how it is possible forsome low-income women to maintaina healthy weight while others do not.To design more effective preventionand treatment programs, researchersneed to learn how personal andbeha\noral factors protect low-income

358 American Journal of Health Promotion July/August 2013, Vol. 27, No. 6

women from overweight and obesitywhile those women are living in obe-sogenic environments. Therefore, thepurpose of this study was to qualita-tively investigate differences and com-monalities in food-relatedenvironmental, personal, and behav-ioral factors influencing eating behav-ior among low-income women ofvarying weight status.

METHODS

Participants and DesignSixteen heterogeneous focus groups

(n = 83) were conducted with lean/normal and overweight/obese low-income, SNAP-eligible women (

TableParticipant Characteristics for Lean/Normal and Overweight/Obese

Low-Income Women*

Characteristics

Race, No. (%)WhiteAfrican-AmericanAmerican IndianHispanicOther

Annual income, No. {%)t$40,000

Education, No. (%)tEighth grade or lessSome high schoolHigh school graduate/GED/equivalentSome CO liege/tech/vocational schoolCompleted college/tech/vocational schoolCompleted grad/professional school

Currently employed. No. (%)tYesNo

Currently on food stamps. No. (%)iYesNo

Living situation, No. (%)tShelter-basedHome-based

Current cigarette use, No. (%)Age, y, mean SDHousehold size, mean SDtNo. of children, mean SDPerception of diet, mean SDPerception of health, mean SD

Total Frequency(n = 83)

23 (27.7)48 (57.8)

5 (6.0)2 (2.4)5 (6.0)

38 (45.8)27 (32.5)12(14.5)4 (4.8)1 (1.2)

3 (3.6)14 (16.9)28 (33.7)25(30.1)

8 (9.6)2 (2.4)

10(12.0)72 (86.7)

68(81.9)14(16.8)

32 (38.6)50 (60.2)43(51.8)36 10.227

3.76 + 1.93.1 2.1

2.52 0.982.67 0.93

Frequency byBMI < 25(n = 29)

9 (31.0)16 (55.2)

1 (3.4)

3 (10.3)

16(55.2)8 (27.6)1 (3.4)3(10.3)1 (3.4)

2 (6.9)7(24.1)8 (27.6)9 (31.0)3 (10.3)

5(17.2)23 (79.3)

24 (82.8)5(17.2)

5(17.2)24 (82.8)16(55.2)

36.93 12.03.45 1.92.93 1.92.93 0.703.00 0.80

BMI StatusBMI > 25(n = 54)

14(25.9)32 (59.3)4 (7.4)2 (3.7)2 (3.7)

22 (40.7)19 (35.2)11 (20.4)

1 (1.9)

1 (1.9)7 (13.0)

20 (37.0)16(29.6)5 (9.3)2 (3.7)

5 (9.3)49 (90.7)

44(81.5)9(16.7)

27 (50.00)26 (48.2)27 (50)

35.5 9.203.93 1.93.19 2.22.3 1.02.5 0.97

* BMI indicates body mass index; and GED, general educational development,t Data not reported forn = 1.t Data not reported for n = 3.

include: (1) aberrant eating behavior,(2) health, beliefs and values, and (3)food environment. The results arereported according to themes andsubthemes identified using the open-coding method.'^'^ Components ofthe various SCT constructs emergedfrom discussions as data were analyzedand existed within themes, includingsituation for the emaronmental con-staict; emotional coping response, self-control, reinforcements, and expec-tancies for the personal construct; andbehavioral capability and outcome ex-

pectations for the behavioral construct;thus underscoring the reciprocal basisof SCT within the focus group discus-sion.

Aberrant Eating BehaviorTopics within this theme included

emotions influencing eating beha\'ior,stashing or hoarding of food, andovereating behavior. Ovenveight/obese women in the sample describedengaging in weight-promoting, aber-rant eating behavior, whereas lean/normal women did not commonlyexpress these tendencies.

Emotions Influence Eating Behavior. Mostovenveight/obese indi\'iduals talkedabout eating more as a result ofemotions like depression, stress, andboredom, whereas lean/normal wom-en reported that they ate less or thatemotions had no effect on their eatingpatterns. Ovenveight/obese partici-pants frequently described strugglingwith emotional disorders and thesubsequent effects that this had onappetite and food intake. As onewoman said, "I get depressed with nothaxing a job or income. Like me beingyoung and being by myself taking careof my kids, I go straight to foods thatcomfort me" (BMI, 36.7). However,many commented that their tendencyto eat or to not eat was the sameregardless of the specific emotion theywere experiencing, as one womanreflected, "I'm an extremely emotionaleater and it doesn't matter whichemotion it is, I want to eat" (BMI,40.5). Conversely, comments by lean/normal participants were similar towhat one woman shared, "I just eat toeat but I be hungry for it. I would crybefore I'd pick up food for depression"(BMI, 22.5), demonstrating more self-control, a component of the personalconstruct. Eating to satisfy hungerinstead of emotions was a commentheard from several lean/normal women.

Emotional eating was also discussedwithin the context of situational influ-ences, with many overweight/obeseparticipants mentioning boredom as amajor impetus for eating. Being aloneand unemployed were both mentionedas contributors to tedium. For in-stance, one woman shared her experi-ences with emotional eating, saying,"I'm used to taking care of otherpeople and have other people livingwith me.. .now that it'sjust me it's likeI'm lost.. .so I just like feed myself(BMI, 56.7). For several others, watch-ing television was a trigger for eating,as one woman mentioned, "I will dothe mindless eating thing.. .1 will eatand then like 5 minutes later if I'mwatching TV a commercial will come,and I'll be like hmm I'm hungr;/ again"(BMI, 36.4). Additionally, some womendiscussed watching television fre-quently, because of boredom.

Several women from both weightgroups talked about other emotionalcoping mechanisms that they turned to

360 American Journal of Health Promotion July/August 2013, Vol. 27, No. 6

instead of eating, such as cigarettesmoking. Others chose more healthfuloutiets, as one woman who had previ-ously been an emotional eater de-scribed, "When I used to be sad orangry or depressed I wanted to munch,munch, munch. I don't do that any-more. Now I pray, I go outside andwalk, you know to try to get theendorphins going, get some sweatgoing, that's good for me" (BMI, 22.3),demonstrating more self-control.Healthful coping strategies like walk-ing or eating fruit instead of sweets orhigh-fat foods were also shared byseveral other lean/normal women.

Stashing or Hoarding Food. Keeping astash of food was a practice reported bymany overweight/obese participantsbut only by a few lean/normal-weightwomen. Most described their stashes asconsisting of high-calorie snack foods.For instance, one woman describedher stash as: ".. .cheetos, honey buns,hostess chocolate donuts, junk food"(BMI, 36.3). Others did not reportkeeping a stash per se but wouldconsume junk food and hide it fromtheir children. As one woman said, "Ilike Twizzlers and Hershey's bars withalmonds and I buy them and I eatthem in the car, eat them all up beforeI pick up my kids from school usual-ly. .. Hide the wrappers and clean upany evidence" (BMI, 36.4).Overeating Behamor. Overweight/obeseparticipants talked about their ten-dencies to overeat, especially wheneating at buffet restaurants, thus dem-onstrating the reciprocal associationsamong environmental, personal, andbehavioral constructs of SCT. Mostreported that they typically consumedthree to four plates of food, althoughone participant stated that she nor-mally ate six plates of food while at thebuffet (BMI, 41.9). In contrast, manylean women reported that they did notlike to go to buffets because as onewoman said, "I don't eat my money'sworth" (BMI, 18.8). Self-control wasalso an issue overweight/obese womenhad when eating meals at home. Onewoman recognized this but explainedher difficulty in managing this behav-ior, stating, ".. .it's like if I cook, whichI tried before, portion control, to haveportion control, I'm ticked off that Ididn't fix more. It's like I gotta go find

me something else or a hot pocket orsomething." Further describing herbehavior, she said, "It's not that I'mstill hungry, it's the taste of food. It'sjust so good. I just want to keep eating"(BMI, 33.4). Although there were somelean/normal women who talked aboutovereating on occasion, most ex-pressed behavior that is summarized bythe following participant's comment:"I just eat until I'm full and then afterI'm full I stop, I don't just eat all thefood 'cause it's on the plate" (BMI,19.3).Health Beliefs and Values

Subthemes within health beliefs andvalues include nutrition knowledge,importance of exercise, and influ-encers of body weight. Lean/normalwomen exhibited nutrition knowledgeand discussed being more active thanoverweight/obese women. Women inboth groups felt that body weight wasmore related to genetics than toexercise and diet alone. In this sub-theme both personal and behavioralconstructs interact.Nutrition Knowledge. Many lean/normalwomen demonstrated basic nutritionknowledge, suggesting they may havehad greater behavioral capability tomake healthier food choices. Severaleven referred to specific nutrients,such as one woman who said, "I'vealways tried to eat healthy but I guessI'm more aware of it you know satu-rated fats, and sodium and really eatingthe fruits" (BMI, 21.4). Another wom-an shared, ".. .when I was pregnantwith my kids I used to read a lot ofnutrition books so that's why I alwaystry to balance out healthy stuff becausethat's always in my head" (BMI, 22.5),which illustrates the impact that herknowledge had on her behavior.Whereas more of the overweight/obese women indicated indifference,like one participant who said, "Oh, Ireally don't pay attention to the foodpyramid" (BMI, 36.4). Others demon-strated a lack of knowledge, as thecomment by one woman suggested."Coffee's got calories in it point blank,just plain coffee, it's got a lot ofcalories, it's got a lot of caffeine in it soit's a lot of fat too" (BMI, 25.9).However, there were a few overweight/obese participants who did exhibitbasic nutrition knowledge that they

had learned as a result of dealing withtheir own or a family member's healthcondition, like diabetes or high bloodpressure.Importance of Exercise. Beliefs and valuessurrounding physical activity were alsodiscussed, with many of the lean/normal women describing the positiveexpectation that exercise helps tomaintain a healthy weight. Here, per-sonal values intertwine with behavior,and many lean/normal women dis-cussed the importance of using physi-cal activity as a means to maintainweight, whereas overweight/obesewomen rarely discussed being physi-cally active. Comments from lean/normal women included: "I work outfour times a week" (BMI, 21.58); "I liketo go for my jogs when I don't have thekids with me" (BMI, 22.7); and "I'llwalk all the way downtown.. .and I dothis every day, sometimes twice a day. Ilove to walk" (BMI, 22.69). Otherlean/normal women said that al-though they did not engage in regular,planned physical activity, they "movearound a lot," "fidget," "keep busy," or"walk to the store."

Other Inuencers on Body Weight. Al-though lean/normal women reportedthat eating healthy and exercising hada positive impact on health, mostbelieved that weight status and theonset of disease were controlled bygenetics. Overweight/obese womenalso expressed this sentiment, butwhen speaking personally about theirweight, many cited previous pregnan-cies and an inability to lose the weightafter as a major contributor. Further-more, many believed that metabolisminfluenced weight, as shown in thefollowing statement: "It's their [leanwomen] metabolism. If their [leanwomen] metabolism breaks down foodfaster, you know that's what it consistsof and you now see some people theirmetabolism don't eat it down fastenough, so it gets builds on and holdson to it" (BMI, 47.2). Others had moreunconventional beliefs, like one wom-an who shared, ".. .one side of mystomach is bigger than the other onebecause of what side I lay on after Ieat.. .because that meat or whatever,especially meat, anything soupy youknow goes right through you but meat,it will stick" (BMI, 33.4).

American Journal of Health Promotion July/August 2013, Vol. 27, No. 6 361

Food EnvironmentWithin food en\ronment the sub-

themes included food insecurit)', gro-ceiy shopping, food preparation, andrestaurant foods. This theme con-tained components of all constructs ofSCT (personal, behavioral, and enxa-ronmental factors). Women in bothweight groups expressed similar senti-ments when talking about the foodenvironment; however, lean/normalwomen cited health as being influen-tial when purchasing groceries.

Food Insecurity. Both lean/nomial andovenveight/obese women described ex-periences invoUing inadequate food,suggesting that all ofthe women lived ina food-insecure environment. Wiientalking about their experiences withfood insecurit)', many of them sharedtheir struggles with not ha\'ing enoughto eat growing up. One participantrecalled difficult times, saying, "We werehungiy a lot as kids and my mom was awelfare mom and she would blow thewhole first month's check on cereal youknow and TV dinnei"s, and then you'dbe kinda hungiy for about two weeks"(BMI, 20.9). A few participants furtherdescribed the lasting impact of theseexperiences, as one woman said, ".. .theway my husband and I were raised ifthere ain't that much food we feed ourkids and we'll go hungr)'" (BMI, 29.5).

Much ofthe conversation regardingfood insecurit)' revolved around copingmechanisms during periods when ad-equate food was not available, withmany reporting that this was often thecase the week before they got theirfood stamps. This meant eating what-ever was available within the householden\nronment, as one woman said, "I'malways worried about food.. .a lot oftimes we get creative, we eat a lot ofstuff we may not want to eat but we eatit any way, we like it cause we have nochoice, and the kids are not picky...you hungry, you got to eat" (BMI,49.6). Several also relied on simplemeals that would last, as one partici-pant commented, ".. .when you're atyour low, low, low, you gotta keep you abag of beans.. .some cornbread andmake you a big old pot and that shouldlast you at least 2-3 days beforestamps" (BMI, 22.5). Safety nets werealso described, and for one womanthese included ".. .borrow some mon-ey or go to a nice friend or relative's

hotise and grab something, or I'll go toa food pantry" (BMI, 29.6).Restaurants. Both lean/normal andoverweight/obese women said theyregularly consumed fast food, withsome mentioning that they went morefrequently toward the end of themonth, when food stamps and fundswere low. However, there were a fewwomen within both weight groups whosaid they preferred to eat meals pre-pared at home because of the cost ofeating out and for health reasons.Among those who did eat out fre-quently, limited cooking capabilitiesand a dislike of cooking were com-monly mentioned as reasons. Onewoman mentioned that her lixTngsituation greatly influenced her con-sumption of fast food, sa)'ing, ".. .youknow like when I was down at theDorothy Day, at the shelter, all I woulddo is eat junk food, 'cause we alwayswent to McDonalds or Burger King orwhatever 'cause I wasn't eating theirfoods" (BMI, 56.7).Grocery Shopping. All participants re-ported shopping for food at a variet)' ofplaces, including supermarkets, dis-coimt stores, meat markets, and con-venience stores, with stipermarkets anddiscount stores mentioned the most.Cost was most influential when shop-ping at supermarkets or chain discountstores, especially among ovenveight/obese women, as illustrated by thestatement: "First and foremost for me,well if I'm shopping for food it's costs,and then taste. Wliat can I make that'scheap and tastes good" (BMI, 40.4).Although for some, personal prefer-ence and taste superseded cost, asreflected by one participant who com-mented, "I go by brand, as far as thetaste and what not, I just won't getanything, or whatever's available, orwhat's in stock, or what's cheap orwhatever. I like to get a certain brand. Igotta have good quality taste" (BMI,41.9). Several participants reportedthat their child's preference greatlydictated what they purchased andcooked, otherwise, ".. .that's going tobe a waste of money so I buy things thatI know she will eat" (BMI, 40.5).

Lean/normal women commonlydiscussed that purchasing healthyfoods, such as fruits and vegetables, wastheir top priority while groceiy shop-

ping. They discussed this within thecontext of achie\'ing positive healthoutcomes. For example, one womandiscussed that she purchased a lot offnnts and vegetables to maintainhealth because she had a heart condi-tion. Another participant said that sheate healthy to set a good example forher children. One woman simply said,"I ti7 to watch what I eat just for myown weight" (BMI, 22.7). Although afew ovenveight/obese women said thathealth was influential in the foodchoices they made, many of thesewomen expressed frustration that theyhad to compromise on their prefer-ences for healthful foods because ofthe costs, like one participant who said,"Yeah, fruits and vegetables within mybudget, I can't afford to buy a lot. I canbuy some but I can't purchase ever)'-thing that I would like to purchase'cause it's expensive" (BMI, 49.6).Although lean/normal women ac-knowledged cost as a barrier, theyvoiced greater confidence in tjieirabilities to purchase healthier foods, asone woman concisely stated, "[ willsacrifice other things to get the good,healthy foods" (BMI, 21.4).Food Preparation. Some women report-ed limited cooking skills or a dislike ofcooking and a preference for ready-made foods, like frozen dinners or fastfood. This had an impact on what theyfed their children, as one womanexpressed, "I do not like to ccok.. .forlunch they'll eat sandwiches and chipsor something or otherwise I'll buysome lunchables.. .for dinner I'll buychicken nuggets and fries or some-thing like that" (BMI, 23.5). Com-ments regarding a like or dislike ofcooking were heard in both weightgroups, although several overv/eight/obese participants were very descriptivewhen talking about food preparationand seemed to derive a lot of pleasurefrom cooking. This was apparent inone woman who excitedly said, "I stayin my kitchen. I stay creative, differentdishes I'm adventurous in the kitch-en, veiy adventurous.. .1 cook ever)'day, trust me" (BMI, 33.4).

DISCUSSION

SCT guided this qualitative study toexamine the relationship between en-

362 American Journal of Health Promotion July/August 2013, Vol. 27, No. 6

vironmental, personal, and behavioralconstructs and weight status among agroup of low-income women. Althoughboth lean/normal and overweight/obese participants lived in similar foodenvironments, the results suggest thatpersonal and behavioral factors dif-fered between weight groups. Over-weight/obese focus group participantsmore often reported keeping a stash offood and overeating, and overwhelm-ingly described themselves as emo-tional eaters. Lean/normal womencommonly expressed nutrition knowl-edge, reported regular exercise, andmore often cited health as a prioritywhen grocery shopping.

Findings that overweight/obesewomen tended to engage in aberranteating behavior may imply that expe-riences with food insecurity did varybetween weight groups, because emo-tional eating, binge eating, andhoarding food have all been associatedwith food insecurity. '" ^ Furthermore,the literature has suggested that psy-chological attachments to food thatdevelop from childhood food insecu-rity persist into adulthood.^''^^ Howev-er, it may be that the eating behaviorsdescribed by overweight/obese womenin this study have other origins, be-cause research has linked eating forpsychological reasons in obese indi-viduals across all socioeconomicgroups.^''"^^ Thereby, experiences withfood insecurity may only exacerbateundesirable eating behaviors in certainwomen. Concurrent with the behaviorsdescribed by overweight/obese partic-ipants in this study, research hasreported that obese individuals areunable to distinguish between emo-tional cues and hunger cues, therebyincreasing the likelihood of eatingbecause of emotional disturbances.^^Obese women living in low-incomeenvironments may be especially vul-nerable to emotional eating,^^'^^ be-cause research has found that thisdemographic suffers from a higherprevalence of depression and a greateroccurrence of stress. '^^ '^^ Consideringthis well-established link between neg-ative emotional states and pover-

3 29 30 1 . . 1

ty, ' ' the interaction betweenemotional coping responses and thelow-income environment should befurther investigated with this popula-tion.

Conversely, most lean/normal wom-en in the present study describedeating for hunger rather than foremotional reasons. Generally, they re-ported coping in more healthful ways,such as walking or eating fruit instead.Cigarette smoking was also mentionedas a coping mechanism, although thiswas a remark heard by women in bothweight groups. Additionally, quantita-tive results revealed that smoking wassimilar between both weight groups,suggesting that smoking was not amodifying factor when examiningemotional eating between the lean/normal and overweight/obese women.Consistent with the findings of thisstudy, Nuss et al.^ ^ reported thatnormal-weight, low-income, postpar-tum women were more responsive tophysiological aspects of eadng, such ashunger and crawngs, whereas over-weight/obese women were more in-fluenced by psychological factors, suchas emotions. In a study with bothwomen and men, researchers foundthat subjects with a lower BMI were lesslikely to eat during negative emotionalstates. Lack of an eating response tonegative emotions may provide a pro-tective effect on low-income womenliving in stressful, low-income, food-insecure environments.

For most overweight/obese womenin this study, price was the primaryconcern when deciding what foods tobuy. Previous research has reportedprice tends to exceed other factors forfood purchasing decisions among low-income women '^ '^ '^ ;^ however, theseresearchers grouped all women to-gether, not specifying differences inpractices between normal-weight wom-en and obese women. By examiningweight groups separately, results indi-cate that lean/normal-weight womenplace a greater emphasis on healthwhen purchasing food for theirhousehold. This suggests that theselean/normal women may value pro-moting and maintaining health morestrongly than overweight/obese wom-en. Although this may be influential intheir purchasing decisions, behavioralcapability may also be a factor, becauselean/normal women exhibited morenutrition knowledge. Additionally,these findings may explain why theywere able to avoid the food insecurity-obesity link, because nutrition educa-

tion improves food security,^* andresearch has shown that increasingnutrition knowledge leads to improve-ments in diet and weight loss.'"*^Although not quantitatively measuredin this study, self-efficacy may beinvolved in the food purchasing deci-sions of these women, because normal-weight women have been shown tohave a higher baseline self-efficacythan obese women. Furthermore,others have found that African-American women with greater self-efficacy report increased fruit andvegetable intake and decreased fatintake.^' Additional research examin-ing food purchasing practices amonglean/normal low-income women isnecessary to elucidate which SCTconstructs, such as self-efficacy, aremost influential in their decisions.

Congruent to their described foodpurchasing behavior, perceived diethealthfulness was significantly greateramong lean/normal women in quan-titative results of this study. Supportingthis, research has reported normal-weight indi\nduals are more likely toconsider themselves to be a health-conscious consumer and consider theirweight when making food choices.^^'^^However, it is uncertain whether thisintention has an impact on actualbehavior; Chang et al.^ ^ found similarfat intake between normal-weight andobese WIC mothers. Because of thepaucity of research, studies examiningdifferences in food intake motivatorsand food intake behavior betweennormal-weight and overweight/obeselow-income women are warranted.

Many lean/normal women in thisstudy reported a tendency to engage inregular exercise, whereas overweight/obese women did not. Here too, lean/normal women have behavioral attri-butes found to be inversely related toweight status,'' '*' unlike their heavnercounterparts living in the same envi-ronment. MacFarlane et al.''^ foundthat healthy-weight status in low-income women was associated withwomen reporting an ability to maketime for physical activity. Differences innonexercise activity thermogenesis be-tween lean and obese individuals'*' mayalso offer a partial explanation for whysome of the lean/normal women inthe present study were able to maintaina healthy weight, because many of

American Journal of Health Promotion July/August 2013, Vol. 27, No. 6 363

them said they were generally active by"mo\ing around," even if they did notengage in regular, planned physicalacd\it)'. Exercise programs that includenonexercise activity thermogenesis-increasing activities, such as walking,have been shown to be effective withlow-income populations.'*'* Althoughlean/normal women voiced the im-portance of physical actixnty in main-taining weight, women in both weightgroups expressed similar sentimentsthat weight and diseases, such as highblood pressure and diabetes, weremostly a result of genetics, rather thanlifestyle factors. In contrast, previousresearch with low-income womenfound that those who believed weightto be genetically predetermined hadhigher BMIs.'^ Present results suggestthat both normal-weight and over-weight/obese women may not becompletely receptive to preventativeand treatment efforts if the roles ofdiet and other lifestyle factors are notemphasized by the health professional.

Comments regarding fast food in-take were comparable between weightgroups, with both lean/normal andovenveight/obese women describingfreqtient consumption. Research hasfound an association between weightstatus and fast food consumption inwomen of higher socioeconomic sta-tus, but no relationship exists betweenfast food consumption and weightstatus among low-income women.'However, neither type of fast food noramount eaten was measured in theresearch by Lin et al.' or in thecurrent sttidy. It may be that lean low-income women eat fast food as fre-quently as their overweight counter-parts but choose lower-calorie optionsand eat smaller portion sizes.

To our knowledge, this research isthe first to qualitatively examine varia-tion in environmental, personal, andbeha\'ioral factors between lean/nor-mal and ovenveight/obese womenliving in obesity-promoting, low-income environments. Although simi-larities existed between women in thisstudy, there were notable differences inparticipants' comments regardingemotional eating, nutrition knowledge,and physical activity. These resultssuggest that personal and beha\'ioralfactors, in part, may modify the influ-ence of the food environment, pro-

13

tecting some women from overweightand obesity. However, genetic differ-ences may have contributed to theweight variation in this sample, butevaluation of such factors was beyondthe scope of this study. Although theresults of this study contribute to theliterature examining the food insecu-rity-obesity paradox, results are limitedby the generalizability of the sample,because regional variation can occur.Furthermore, although focus groupscan elicit rich conversation that wouldnot be possible in a structtired settingor through surveys, participants' re-sponses may be subject to bias within agrotip setting. Recognizing these limi-tations, focus group methodology re-mains suitable for collecting formativedata and filling in knowledge gaps inthe literature, thus providing a basis foradditional research.

SO WHAT? Implications for HealthPromotion Practioners andResearchersWhat is already known on this topic?

The food insecurity-obesity para-dox among low-income women hasbeen well established; however,questions remain as to how somewomen are able to maintain ahealthy weight while living in a food-insecure, low-income environment.What does this article add?

This study provides insight intohow personal and behavioral factorsdiffer in low-income women of varj'-ing weight status, possibly alteringthe influence of the food environ-ment on eating behavior and health.What are the implications for healthpromotion practice or research?

Qualitative results from this studyprovide a foundation for furtherinvestigation into environmental,personal, and behavioral factoi"s thatinfluence health and eating behaviorin lean/normal vs. overweight/obesewomen.

Acknowledgm ents

We xoould like to thank oil ofthe ivometi laho participated inthis study. This project was funded through a U.S.Departinfiit of Agriculture Si\'P-Ed grant.

References1. Ogden C, Carroll M, Curiin L, el al.

Preralence of overweight and obesiiy in

the United States, 1999-2004. y A? MedAssoc. 2006:295:1549-1555.

2. Wiig K, Smith C. The art of gro:etyshopping on a food stamp budget: factoreinfluencing the food choices oflovv-income women as they tiy to make endsmeet. Pubtic Health Nutr. 2OO8;1S-:1726-1734.

3. Richards R, Smith C. Shelter en^'ironme^tand placement in communit)' affectslifestyle factors among homeless familiesin Minnesota. AmJ Health Promcl.2006:21:36-44.

4. Parker ED, Widome R, NettletonJA,Pereira MA. Food security and metabolicsyndrome in US adults and adolescents:findings from the national and nutritionexamination sur\'ey. 1999-2006. AnnEpidemiol 2010:20:364-370.

5. Townsend MS, Peerson J, Love B, et al.Food insecurit)' is positively related toovenveight in women. J Nutr.2001:131:1738-1745.

6. Franklin B, Jones A, Love D, et al.Exploring mediators of food insecurit)'and obesity: a review of recent literature./Coima Healtk 2012;37:253-264.

7. Dinour L, Bergen D, Yeh M. Ths foodinsecurity-obesity paradox: a review of theliterattire and ihe role food stamps mayplay. J Am Diet Assoc. 2007:107:1952-1961

8. DrewnowskI A. Obesity and the Foodenxironmentdieiaiy energ)' density anddiet costs. Am J Prev Med.2004:27(suppl):154-162.

9. Eikenberiy N, Smith C. Healthful eating:perceptions, motivations, barriers, andpromoters in low-income Minnesotacommunities. J Am Diet Assoc.2004:104:1158-1161.

10. Pollard J, Greenwood D, Kirk S, CadeJ.Lifestyle factors affecting fruit andvegetable consumption in the UKWomen's Cohort Study. Aftpetite.2001:37:71-79.

11. Steptoe A, Pollard TM, WardleJ.Developinenl of a measure of the motivesunderl)'ing the selection of food: the foodchoice questionnaire. Appetite.1995;25i267-284.

12. Bauer F, Elbei-s CC, Adan RA, ei al.Obesity genes identified in genome-wideassociation sttidies are associated withadiposit)' measures and potentially withnutrient-specific food preference. AmJCtin Nutr. 2009:90:951-959.

13. Lin B, Huang C, French S. Factoreassociated unth women's and children'sbody mass indices by income stiitus. IntjObes Metab Disord. 2004:28:536-542.

14. Smith C, Morton L. Rural food deserts:low-income perspectives on food access inMinnesota and Iowa. J Nutr Edite Behav.2009:41:176-187.

15. Wiig-Dammann K, Smith C. FactorsalTecting low-income women's foodchoices and the perceived impact ofdietaiy intake and socioeconomic statuson their health and w^hi. J Nutr EducBehav. 2009:41:242-253.

16. Bandura A. SelJ-EJficacy: The Exerdse ofControl. New York, N\': W.H. Freeman;1997.

364 American Journal of Health Promotion July/August 2013, Vol. 27, No. 6

17. Lee RD, Nieman DC. NutritionalAssessment. 3rd ed. St. Louis, Mo: Mosby;1996.

18. Corbin D, Strauss A. Basics of QualitativeResearch: Techniques and Procedures forDevelofnng Grounded Theory. ThousandOaks, Calif; Sage Publications; 1998.

19. Morgan D, Knieger R. The Focus Croup Kit:Volumes 1-6. Thousand Oaks, Calif: SagePublications; 1998.

SO. Morse J. The significance of saturation.Qual Health Res. ]995;5:147-149.

21. Olson C'Bove C, Miller E. Growing uppoor: long-term implications for eatingpatterns and body weight. Appetite.2007;49:198-207.

22. Allen MW, Wilson M. Materialism andfood security. Appetite. 2005;45:314-323.

23. Smith C, KJosterbuer A, Levine AS.Military experience strongly influencespost-service eating behavior and BMIstatus in American veterans. Appetite.2009;52:280-289.

24. Nolan LJ, Halperin LB, Celiebter A.Emotional Appetite Questionnaire:constnict validity and relationship withBMl. Appetite. 2010;54:314-319.

25. Delahanty L, Meigs J, Hayden D, et al.Psychological and behavioral correlates ofbaseline BMI in the Diabetes PreventionProgram (DPP). Diabetes Care.2002;25:1992-1998.

26. van Strien T, Ouwens MA.Counterregulation in female obeseemotional eaters: Schachter, Goldman,and Gordon's (1968) test ofpsychosomatic theory revisited. Eat Behav.2003;3:329-340.

27. Cahill JM, Freeland-Graves J, Shah BS, etal. Development and validation of theeating stimulus index in low-income,minority women in early postpartum./AmDietAssoc. 2009;109:1593.-]598.

28. Wiig-Dammann KW, Smith C. Food-related environmental, behavioral, and

personal factors associated vWth body massindex among urban, low-income African-American, American Indian andCaucasian women. Amf Health Promot.2011;25:el-el0.

29. Miranda J, Chung JY, Green BL, et al.Treating depression in predominantly low-income young minority women: arandomized controlled trial./A? MedAssoc. 2003;290:57-65.

30. Broussard CA. Research regarding low-income single mothers' mental andphysical health: a decade in review./Poverty. 2010;14:443-451.

31. Nuss H, Clarke K, Klohe-Lehman D,Freeland-Graves J. Influence of nutritionattitudes and motivators for eating onpostpartum weight status in low-incomenew mothers. //l?n Diet Assoc.2006; 106:1774-1782.

32. Morton JF, Guthrie JF. Diet-relatedknowledge, attitudes, and practices of low-income individuals with children in thehousehold. Fam Econ Nutr Rev. 1997;10:2-15.

33. Eicher-Miller H, Mason A, Abbott A, et al.The effect of Food Stamp NutritionEducation on the food insecurity of low-income women participants./Afwr/dzicBehav. 2009;41:161-168.

34. Klohe-Lehman DM, Freeland-Graves J,Anderson E, et al. Nutrition knowledge isassociated with greater weight loss inobese and overweight low-incomemothers. J Am Diet Assoc. 2006;106:65-75.

35. Nuss H, Freeland-Graves J, Clarke K, et al.Greater nutrition knowledge is associatedwith lower 1-year postpartum weightretention in low-income women. J Am DietAssoc. 2007;107:1801-1806.

36. Richman R, Loughnan G, Droulers A, etal. Self-efficacy in relation to eatingbehaviour among obese and non-obesewomen. Intf Obes Relat Metab Disord.2011;25:907-913.

37. Henry H, Reimer K, Smith C, Reicks M.Associations of decisional balance,processes of change, and self-efficacy withstages of change for increased fruit andvegetable consumption among low-income, African-American mothers./ / lmDiet Assoc. 2006;106:841-849.

38. Chang MW, Baumann LC, Nitzke S, BrownRL. Predictors of fat intake behavior differbetween normal-weight and obese W Cmothers. Am f Health Promot. 2005;19:269-277.

39. Robinson R, Smith C. Associationsbetween self-reported health consciousconsumerism, body-mass index, andattitudes about sustainably producedfoods. AgHum Values. 2003;20:177-187.

40. Lahd-Koski M, Pietinen P, Helivaara M,Vartiainen E. Associations of body massindex and obesity with physical activity,food choices, alcohol intake, and smokingin the 1982-1997 FINRISK Studies. AmJClin Nutr. 2002;75:809-817.

41. Chan CB, Spangler E, Valcour J, Tudor-Locke C. Cross-sectional relationship ofpedometer-determined ambulatoryactivity to indicators of health. Obes Res.2003;ll:1563-1570.

42. MacFarlane A, Abbott G, Crawford D, BallK. Personal, social, and environmentalcorrelates of health weight status amongstmothers from socioeconomicallydisadvantages neighborhoods: findingsfrom the READI study. Int j Behav NutrPhys Act. 2010;7:23.

43. Levine J. Nonexercise activitythermogenesis - liberating the life-force. /Intern Med. 2007;262:273-287.

44. Clarke K, Freeland-Craves, J, Klohe-Lehman D, et al. Promotion of physicalactivity in low-income mothers usingpedometers.//lm Diet Assoc. 2007;107:962-967.

American Journal of Health Promotion July/August 2013, Vol. 27, No. 6 365

Copyright of American Journal of Health Promotion is the property of American Journal ofHealth Promotion and its content may not be copied or emailed to multiple sites or posted to alistserv without the copyright holder's express written permission. However, users may print,download, or email articles for individual use.