Frequency of Eating Homegrown Produce Is Associated with Higher Intake among Parents and Their Preschool-Aged Children in Rural Missouri
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
RESEARCH
Current Research
Frequency of Eating Homegrown Produce Is
Associated with Higher Intake among Parents
and Their Preschool-Aged Children in Rural
Missouri
MARILYN S. NANNEY, PhD, RD; SHELDON JOHNSON; MICHAEL ELLIOTT, PhD; DEBRA HAIRE-JOSHU, PhD
favorites, sometimes⫽3.7 favorites, almost always/al-
ABSTRACT ways⫽4.8 favorites) and parental role modeling (rarely/
Objective The purpose of this study was to identify never⫽5.3 fruit and vegetable eating observations, some-
whether or not there are associations between frequency times⫽6.2 fruit and vegetable eating observations, almost
of eating homegrown produce among rural parents and always/always⫽6.3 fruit and vegetable eating observa-
their preschool children and overall intake. tions). Interactions within the larger community food envi-
Study design A cross-sectional study, including parents ronment were not significantly affected as measured by
(n⫽1,658) and their preschool children (aged 2 to 5 years) weekly food dollars spent (rarely/never⫽$136, some-
enrolled in a parent education program, in eight rural times⫽$117, almost always/always⫽$137) or eating out the
Southeast Missouri counties was conducted. previous month (rarely/never⫽7.7 times, sometimes⫽7.2
Main outcome measures Parents completed a telephone in- times, almost always/always⫽7.0 times).
terview answering questions for themselves and their Conclusions Our findings suggest that educational pro-
preschool child about their fruit and vegetable intake grams promoting awareness of local produce sources and
during the past 7 days using a 29-item food frequency
facilitating the development of gardening programs may
questionnaire (82% response rate).
be a worthwhile investment.
Statistical analysis performed Frequency of eating home-
J Am Diet Assoc. 2007;107:577-584.
grown fruits and vegetables was examined and catego-
rized as almost always/always (n⫽226), sometimes
(n⫽871), and rarely/never (n⫽546). Odds ratios with 95%
C
hildren living in rural settings have different food
confidence intervals, independent sample t tests (two- and nutrition challenges than children living in
sided) analyses, and analysis of covariance (with post hoc other settings. Children in rural locations tend to be
tests) were conducted. slightly food insecure and rely on federal food assistance
Results Significant differences were found in the overall programs such as food stamps (15%) and free or reduced
fruit and vegetable diets and nutrient quality between all lunches (40%) (1,2). Food security challenges have been
three groups for both parents and their preschool chil- positively linked to reduced intakes of fruits and vegeta-
dren. Furthermore, frequency of eating homegrown fruits bles and overweight among children (3-5). Whereas nu-
and vegetables promoted a positive home environment trition and overweight concerns are found throughout the
with increased availability of produce (rarely/never⫽4.3 United States, the problem may be especially severe in
items, sometimes⫽4.7 items, almost always/always⫽5.2 rural areas citing complex cultural and structural chal-
items), preschooler’s preference for them (rarely/never⫽3.5 lenges to maintaining healthful lifestyles like consuming
adequate fruits and vegetables (6,7). It appears that rural
residents experience as much as a 55% increased preva-
M. S. Nanney is an assistant professor and S. Johnson lence of overweight and obesity compared to their urban
is a student, Health Promotion and Education Depart- counterparts (8-12). Furthermore, McMurray and col-
ment, University of Utah, Salt Lake City. M. Elliott is a leagues (12) reported on differences in urban (n⫽962) and
data analyst and D. Haire-Joshu is professor and direc- rural (n⫽1,151) third- and fourth-grade children and
tor of the Obesity Prevention Center, Saint Louis Uni- found significant increases in sum of skinfolds among
versity School of Public Health, St Louis, MO. rural children after controlling for race, sex, and socio-
Address correspondence to: Marilyn S. Nanney, PhD, economic status (rural 27.6 mm, urban 24.0 mm
RD, University of Utah Health Promotion & Education [P⫽0.0001]). Although the empirical literature on the
Department, 1901 E So Campus Drive Annex 2115, Salt dietary habits of rural families is scarce (13), a study of
Lake City, Utah 84112. E-mail: susie.nanney@hsc. the dietary intake of a sample of rural Native American
utah.edu and non-Hispanic white children showed some dietary
Copyright © 2007 by the American Dietetic concerns. The main sources of nutritional intake among
Association. these children were whole milk, cheese, white breads,
0002-8223/07/10704-0002$32.00/0 salty snacks, soft drinks, hot dogs, candy, and sweetened
doi: 10.1016/j.jada.2007.01.009 fruit drinks. The children had poor food variety. The main
© 2007 by the American Dietetic Association Journal of the AMERICAN DIETETIC ASSOCIATION 577sources of energy were low-nutrient-dense, high-fat foods card for completing a survey. Data were collected over 3
and refined carbohydrates. Solid fruits and vegetables years during the fall (2002 through 2004). The scheduling
were seldom used to increase nutrient and fiber intake. of the survey was designed to be consistent with the
Stroehla (14) discovered that the intake of fruits and Parents as Teachers program schedule, which follows the
vegetables among a rural sample was lower than national academic calendar. Phase 1: sites one through six, Phase
averages. For example, fruits and vegetables accounted 2: sites seven though 12, and Phase 3: sites 13 though 16.
for only 15% of the children’s total fiber intake, compared The Institutional Review Board of Saint Louis University
to nearly 30% among children nationally. approved this study, and informed consent was obtained
Related to nutrition and overweight, rural areas report from all participants.
a higher prevalence of chronic diseases, including diabe-
tes and heart disease (15). Diabetes is 17% higher in rural
areas than in central cities and 12% higher than urban Measurement
and suburban areas combined (16). Despite a 50% reduc- Parents completed a telephone interview answering
tion in coronary heart disease and stroke during the past questions for themselves and their preschool child about
30 years (17), disparities among rural populations have their fruit and vegetable intake, child preference for each
become more exaggerated (18-20). Compounding the fruit or vegetable item, and home and community food
problems in rural areas are limited resources to effec- environment (ie, parental role modeling, frequency of eat-
tively provide health education to prevent, diagnose, and ing fast food). Trained, professional interviewers con-
treat these illnesses, making health promotion and dis- ducted the interviews. The interview took about 30 min-
ease prevention particularly challenging, yet all the more utes to complete (average 29 minutes). Completed
critical. interviews resulted in an 82% response rate.
The aims of this report are to present cross-sectional
data that identify the association between frequency of
eating homegrown produce among rural parents and Saint Louis University for Kids Food Frequency
their preschool children and overall intake. Questionnaire (FFQ)
Parent respondents were asked to recall for themselves
and their preschool child if they had eaten specific fruits
METHODS and vegetables (n⫽29 items) during the past 7 days, and
High 5 For Kids is a collaboration between Parents as if so, how many times. For example, “Did you have can-
Teachers National Center, the Obesity Prevention Center taloupe in the past 7 days? If yes, how often? One time,
at Saint Louis University’s School of Public Health and two times, three or four times, five or six times, seven or
the Parents as Teachers affiliates in the Bootheel (south- more times?” Each question was repeated using the
east) region of Missouri. Parents as Teachers National child’s name. In addition to gathering the frequency of
Center, Inc, is an international child development pro- consumption for each fruit or vegetable eaten during the
gram committed to assisting parents in preparing their previous week, preference was established for the pre-
infant to 5-year-old children for kindergarten. The phi- school child. The parent was asked to select from the
losophy of the organization is that “parents are a child’s following response choices to indicate the child’s prefer-
first and most influential teacher” (21). Program sites ence for that food: favorite, likes it, hates it, or never
self-report that they are mostly located in rural commu- had it.
nities (55%) and further classify themselves as small Development. The Saint Louis University FFQ for Kids
towns (33%). Fewer program sites identify themselves as was modified from our previous work (23) and specifically
located in urban (20%) and suburban (15%) areas. Ac- identified foods most relevant to families living in the
cording to the program’s 2001 annual report, rural pro- rural Midwest (24). Foods selected for the FFQ, the av-
grams served 87,985 families and urban programs served erage serving size, and nutrient content were based on
52,387 families (21). the foods and amounts eaten by a nationally represented
sample from the 1994 and 1996-1997 Continuing Survey
of Food Intakes by Individuals interviews for 20- to 59-
Sample/Setting year-old women and 2- to 5-year-old children living in the
Parents (n⫽1,658) of preschool-aged children (aged 2 to 5 rural Midwest. The serving size and nutrient content was
years) enrolled in one of 16 Parents as Teachers program determined for each food for both the parent and the child
sites located in eight rural Southeast Missouri counties using the median serving size (grams). The median was
were recruited by their parent educator to participate in used instead of the mean due to serving sizes that were
the study. These eight counties represent a distinct geo- not normally distributed. Methods used to develop a pre-
graphic region for Parents as Teachers services and over- vious FFQ resulted in acceptable validity and reliability
all, residents experience the poorest chronic disease and were repeated for this study.
health outcomes in the state (22). Recruitment strategies Validation. Pilot work with the Saint Louis University for
included personal invitations and flyers distributed the Kids FFQ demonstrated that parents can serve as accu-
summer before the start of the fall school year. All fami- rate proxies to report fruit and vegetable consumption for
lies enrolled in Parents as Teachers with a child who their preschool children and themselves. Linneman and
would be 2 to 5 years old sometime during the following colleagues (25) conducted an observational study with
school year were eligible to participate. When a family independent observers and is explained elsewhere.
had more than one child in that age range, the oldest Briefly, a convenience sample of 64 parents and their 2- to
child only was recruited. Parents were given a $20 gift 5-year-old children were recruited from parenting fairs to
578 April 2007 Volume 107 Number 4Table 1. Demographic characteristics of rural Missouri parents by frequency of eating homegrown fruits and vegetables
All parents Almost always/always eats Rarely/never eats
Characteristic (nⴝ1,658) homegrown FVa (nⴝ207)b homegrown FV (nⴝ525)c P value
Age (y), meanⴞSD d
29.0⫾6.9 31.0⫾7.9 27.8⫾6.7 ⬍0.001
Race
African American (%) 17.5 13.6 25.8 ⬍0.001
White (%) 81.2 86.4 74.2
Income
⬍$25,000 (%) 48.5 42.7 50.8 0.050
ⱖ$25,000 (%) 51.5 57.3 49.2
Education
⬍High school (%) 19.9 14.2 22.2 0.012
ⱖHigh school (%) 80.1 85.8 77.8
a
FV⫽fruits and vegetables.
b
Parents who responded that they almost always/always eat homegrown fruits and vegetables (Likert scale: almost always/always, sometimes, rarely, never) in a telephone interview.
c
Parents who responded that they rarely or never eat homegrown fruits and vegetables (Likert scale: almost always/always, sometimes, rarely, never) in a telephone interview.
d
SD⫽standard deviation.
participate in a predetermined lunchtime meal (79% re- homegrown?” Four response choices included almost al-
sponse rate). Trained observers discretely monitored ways/always, sometimes, rarely, and never.
the children and their parents before, during, and after
the meal and recorded if the fruits and vegetables taken
were consumed. The following day, the participants were Data Analysis Plan
called and were asked to complete an adapted 29-item Frequency of eating homegrown fruits and vegetables
fruit and vegetable FFQ that looked at consumption the was examined: Almost always/always (n⫽207, 13%),
previous day (95% response rate). Kappa statistics were sometimes (n⫽871, 54%), and rarely/never (n⫽525, 33%).
computed as a measure of agreement between observed
To identify if dietary intake and food environment influ-
and reported food consumption. Interobserver agreement
indicated that parents accurately reported their chil- ences varied by frequency of eating homegrown fruits and
dren’s intake on most fruit and vegetables (⫽0.59 to vegetables, odds ratios (ORs) with 95% confidence inter-
0.61) and for themselves (⫽0.61 to 1.0). vals (CIs), independent sample t tests (two-sided) analy-
ses, and analysis of covariance were conducted. The 95%
Reliability. The psychometric properties of the Saint
Louis University for Kids instrument were examined. CIs provide an interval that, with 95% probability, will
Average days between initial and retesting were 13 days. contain the true value of the OR. An OR of 1.00 would
Cicchetti’s guidelines for acceptable interclass correlation indicate no difference in rates between the group of in-
statistics were used to consider the scale poor (below terest and the reference group. A number ⬍1.00 indicates
0.40), fair (0.40 to 0.59), good (0.60 to 0.74), or excellent that the group of interest is less likely to be affected by
(0.75 to 1.00) (26). The Saint Louis University for Kids the factor being analyzed than the reference group. An
FFQ demonstrated excellent test–retest with intraclass OR ⬎1.00 shows that the group of interest is more likely
correlation coefficients ranging from 0.75 to 0.82. Parent to be affected by the factor being analyzed than the ref-
reporting total fruit and vegetable intake for themselves erence group. Lack of statistical significance of these ORs
was 0.75 and that of their preschool child was 0.82. Indi- is determined by the presence of 1.00 in the 95% CIs.
vidual child servings of fruit (0.78) and vegetables (0.80) Independent sample t tests (two-way) were performed
and preference for fruit (0.73) and vegetables (0.83) was initially for demographic data to identify potential con-
good to excellent. Individual parent reporting for servings founding variables. Once identified, analysis of covari-
of fruit (0.70) and vegetables (0.73) was good. ance was performed controlling for those confounders.
For analysis of covariance, categorical confounding vari-
Other Demographic and Behavior Measures ables (eg, race, income, and education) were dichoto-
Other variables measured in this cross-sectional study mized. Adjusted means and standard errors were re-
included self-reported height and weight, frequency of ported for each group. Post hoc tests were conducted to
fast food visits, parental role modeling (“How many times examine between the three groups and estimated mar-
in the past week did child’s name see you eating fruits ginal means reported. A predetermined significance value
and vegetables? Times per week/day”), and home food of P⬍0.05 was established. All data were analyzed using
environment (“Did you have broccoli in your home in the the Statistical Package for Social Sciences software (ver-
past week? Yes/no”). Frequency of eating homegrown sion 13.0, 2004, SPSS Inc, Chicago, IL) where frequency
fruits and vegetables, the independent variable for this of eating homegrown fruits and vegetables (almost al-
study, was assessed by the question: “How often do you ways or always, sometimes, and rarely or never) was the
and your family eat fruits and vegetables that have been independent variable.
April 2007 ● Journal of the AMERICAN DIETETIC ASSOCIATION 579Table 2. Parent and child daily nutrient intake and food environment influences by frequency of eating homegrown fruits and vegetables (FV),
as self-reported by parents via telephone interview
Almost always/always Sometimes eats Rarely/never eats
eats homegrown FVa homegrown FV homegrown FV
Variable (nⴝ207) (nⴝ871) (nⴝ525) P b value
4™™™™™™™™™™™™™™™™™™™™™™™™ mean c⫾SE d ™™™™™™™™™™™™™™™™™™™™3
e
Parent daily intake
Total FV servings (excluding french fries) 4.8⫾0.1 4.2⫾0.1 3.5⫾0.1 ⬍0.001
Fruit servings 2.3⫾0.1 1.9⫾0.05 1.5⫾0.1 ⬍0.001
Vegetable servings (excluding french fries) 2.6⫾0.1 2.4⫾0.1 2.1⫾0.04 ⬍0.001
Vitamin A (g retinoic acid equivalent) 295.8⫾13.8 244.9⫾6.8 204.8⫾8.7 ⬍0.001
Vitamin C (mg) 83.4⫾3.5 73.7⫾1.7 59.9⫾2.2 ⬍0.001
Fiber (g) 7.5⫾0.2 6.5⫾0.1 5.5⫾0.1 ⬍0.001
Preschooler daily intakee
Total FV servings (excluding french fries) 5.6⫾0.2 5.0⫾0.1 4.5⫾0.1 ⬍0.001
Fruit servings 3.9⫾0.2 3.5⫾0.06 3.1⫾0.1 ⬍0.001
Vegetable servings (excluding french fries) 1.7⫾0.1 1.5⫾0.03 1.4⫾0.04 ⬍0.001
Vitamin A (g retinoic acid equivalents) 294.4⫾13.6 251.29⫾6.6 204.2⫾8.6 ⬍0.001
Vitamin C (mg) 83.8⫾2.6 75.8⫾1.3 65.6⫾1.6 ⬍0.001
Fiber (g) 6.7⫾0.2 6.0⫾0.1 5.4⫾0.1 ⬍0.001
Preferred FV (n)f 4.8⫾0.2 3.7⫾0.1 3.5⫾0.1 ⬍0.001
Preferred fruits (n) 3.0⫾0.02 2.9⫾0.1 2.9⫾0.01 0.016
Preferred vegetables (n) 2.9⫾0.02 2.8⫾0.1 2.7⫾0.01 0.002
Food environment
Child observed parent eating FV (times
past week) 6.3⫾0.2 6.2⫾0.1 5.3⫾0.2 0.001
Parent monthly fast food (times) 7.0⫾0.5 7.2⫾0.2 7.7⫾0.3 0.190
Home FV availability (count last week) 5.2⫾0.1 4.7⫾0.1 4.3⫾0.1 ⬍0.001
Weekly grocery expense ($) 137.0⫾15.6 116.9⫾7.2 136.0⫾9.0 0.186
a
FV⫽fruits and vegetables.
b
Post hoc test between three groups.
c
Analysis of covariance compared difference between the three groups. Means adjusted for age, race, income, and education of parent.
d
SE⫽standard error of the mean.
e
Calculated from a 29-item food frequency questionnaire.
f
Parent identified fruit or vegetable as his or her child’s favorite (response options: favorite, likes it, doesn’t like it, hates it, has never had it).
RESULTS Table 2 identifies socioeconomic differences at baseline
Table 1 identifies that parents enrolled in the study were between the almost always and rarely/never homegrown
mostly women (98% women, n⫽1,631), younger than age fruit and vegetable eaters for age, race, income, and ed-
30 years (mean age 29 years), mostly married (69%, ucation. Sometimes homegrown fruit and vegetable eat-
n⫽1,141), and white (81%, n⫽1,347). Educational attain- ers were not significantly different from the almost al-
ment included 46% completing more than a high school ways homegrown fruit and vegetable eaters for these
diploma (n⫽763), 31% high school diploma or equivalent demographic variables. Compared to rarely homegrown
(n⫽514), and 13% not completing high school (n⫽216). fruit and vegetable eaters, almost always homegrown
The median income was between $25,000 and $34,999. fruit and vegetable parents were more likely to be older
Most parents believed children should eat fruit and veg- (31 vs 28 years), white, have completed high school, and
etables three times each day for good health. Parents have annual incomes greater than $25,000. Based on
reported eating about 3.8 (median) servings a day them- these findings, results presented hereafter have been con-
selves, excluding french fries. Using self-reported height trolled for age, race, income, and educational attainment
and weight, parents were moderately overweight for their of the parent.
height (mean body mass index 27.1)
The adult respondent served as a proxy for the pre-
school child in the home. The majority (97%) completing Individual Dietary and Nutrient Intakes
the baseline survey reported being the parent of the child There were significant differences in the overall fruit and
enrolled in the study. There was an equal distribution of vegetable diets and nutrient quality between groups by
male and female children (52% vs 48%). The average age frequency of consumption among parents and their pre-
of the child participants was 3.2 years (range 1 to 6.6 school children, independent of parent age, race, income,
years). Parent report indicates that children ate about 4.9 and educational attainment.
servings of fruits and vegetables a day, excluding french Parent Intake. Intakes of both fruits and vegetables were
fries. significantly different between all three groups. Parents
580 April 2007 Volume 107 Number 4almost always eating homegrown fruit and vegetables seven times during the previous month (7.0 vs 7.7 times,
were 3.2 times more likely to eat five daily servings of P⫽0.190). Similarly, both groups spent about the same
fruit and vegetable (P⬍0.001, 95% CI 2.20 to 4.59) com- amount of money on their weekly grocery bill the previ-
pared to the rarely/never eaters. The almost always ous week ($137 vs $136, P⫽0.186).
homegrown fruit and vegetable families ate an average of
1.3 additional fruit and vegetable servings (4.8 vs 3.5, DISCUSSION
P⬍0.001). Almost always homegrown fruit and vegetable The food environment has been described in a variety of
parents ate nearly a serving more fruits (2.3 vs 1.5, ways, including the influences within the home and com-
P⬍0.001) and nearly a half a serving more vegetables (2.6 munity among families of young children (27). Most no-
vs 2.1, P⬍0.001) than rarely homegrown fruit and vege- tably in the home, the availability and accessibility of
table eaters. Furthermore, the quality of that intake im- healthful foods and parental role modeling of eating those
proved as frequency of homegrown fruit and vegetable foods positively influences the child’s dietary intake (28).
intake increased as determined by vitamins A and C and Of all the determinants of fruits and vegetables studied
fiber intake. Specifically, fruit and vegetable sources of among children, the availability and accessibility of fruits
vitamin A (296 g vs 205 g retinoic acid equivalents, and vegetables and taste preferences were most consis-
P⬍0.001), vitamin C (84 vs 60 mg, P⬍0.001), and fiber tently and positively related to consumption (29,30). Our
(7.5 vs 5.5 g, P⬍0.001) were significantly higher for al- study results find positive associations with eating home-
most always homegrown fruit and vegetable diets. Paren- grown produce and an increase in home availability and
tal obesity did not differ between the groups (mean body number of preschooler favorite fruit and vegetables. More
mass index 27.6 vs 27.2, P⫽0.441). recently, research highlights the influence of the larger
Preschool Children’s Intake. The effects of homegrown pro- community environment on dietary intake of families
duce eating frequency upon intake were not limited to (23). For example, consumption of fast food among chil-
parents. Preschool children of parents who indicated that dren in the United States seems to have an adverse effect
they almost always eat homegrown fruits and vegetables on the diet quality in ways that could increase risk for
were 2.3 times as likely (P⬍0.001, 95% CI 1.64 to 3.23) to obesity (31). However, we did not find that consuming
eat five daily servings of fruits and vegetables. Similar to homegrown fruits and vegetables served as a buffer
that of their parents, there was more than a serving against common community barriers such as expense and
difference between the groups (5.6 vs 4.5). Separately, fast-food eating. Weekly grocery expenses and number of
intake included nearly a serving increase in fruits (3.9 vs fast-food restaurant visits were similar across groups.
3.1, P⬍0.001) and nearly a half a serving increase in Our study found no association by frequency of eating
vegetables (1.7 vs 1.4, P⬍0.001). Significantly higher homegrown produce and parent body mass indexes. Stud-
daily dietary intakes of vitamins A and C and fiber from ies to date report inconsistent findings of the relationship
fruits and vegetables are shown in Table 1. In addition, between fruit and vegetable consumption and weight. In
children from families who consumed homegrown fruits a prospective cohort study among children and adoles-
and vegetables more frequently preferred more fruits and cents (N⫽14,918), Field and colleagues (32) found no
vegetables. Parents of almost always homegrown fruit association between fruit and vegetable intake and
and vegetable-eating children reported more fruits and change in body mass index. A review of the epidemiologic
vegetables as their child’s favorite (4.8 vs 3.5). These evidence shows clinical evidence that combining advice to
findings are independent of parent age, race, income, and increase fruit and vegetable consumption with energy
educational attainment. restriction is an effective strategy for weight manage-
ment (33). An American Dietetic Association Evidence
Analysis Library workgroup examined the best available
Food Environment evidence on the relationship between child overweight
In our study, almost always homegrown fruit and vege- and fruits and vegetable consumption. Their summary
table category parents were more likely to have a colorful determined the current evidence to be limited in finding
variety of fruits and vegetables in the home the previous an association and offered that one explanation for the
week; more than three times as likely to have tomatoes inconsistency in associations may be the high intakes of
(OR⫽3.6, 95% CI 1.06 to 12.10); more than two times as fried vegetables such as french fries and chips (34).
likely to have cantaloupe (OR⫽2.8, 95% CI 1.85 to 4.18) Limitations to consider when drawing conclusions from
and broccoli (OR⫽2.2, 95% CI 1.55 to 3.07); and nearly this work include dietary measurement and generaliz-
twice as likely to have carrots (OR⫽1.7, 95% CI 1.06 to ability of study results. Although objective methods con-
2.62) and beans (OR⫽1.5, 95% CI 1.02 to 2.31). Table 2 sidering personal and geographic variables were used to
presents the home availability of fruits and vegetables as select fruits and vegetables for inclusion (and serving
a summary score for each group. Almost always home- sizes) in the questionnaire, (24) an exhaustive list would
grown fruit and vegetable category families had more not have been practical. Furthermore, FFQs tend to over-
fruit and vegetable options available in the home the inflate dietary intakes. Therefore, the actual number of
previous week (5.2 vs 4.3 choices, P⬍0.001, range 0 to 7). servings and nutrient values may be larger than ex-
In our study, almost always homegrown fruit and vege- pected, but equally so for all groups. Milligrams of vita-
table parents reported that their preschooler saw them min C, micrograms of vitamin A, and grams of fiber are
eat fruit and vegetables more times during the past week from fruits and vegetables only. Other nonfruit and non-
(6.3 vs 5.3 times, P⫽0.001). vegetable sources of theses nutrients were not captured
These study results identified that parents of both for this study. One question was used to characterize the
groups reported eating at fast-food restaurants about frequency of consuming homegrown produce in general.
April 2007 ● Journal of the AMERICAN DIETETIC ASSOCIATION 581Also, it is unclear from our study if the homegrown pro- the interaction within the larger food environment (ie,
duce was grown by the participants, relative or neighbor, effects on food costs and frequency of fast-food restaurant
or purchased from a farmer’s market. Finally, this study visits). School-based gardening programs are an ideal
gathered data from parents of young children in rural situation in which children are exposed to fresh produce
southeast Missouri; therefore, the results may not be and contribute to the overall health of the residents,
generalizable to older populations living in non-Midwest- especially in rural areas. Research has shown that when
ern states and suburban areas. children garden, their attitude toward fruits and vegeta-
These study results have practical application when bles improves, in addition to their preference for health-
considering the overall poor dietary and health outcomes ful snacks. In addition to improved nutrition attitudes,
among rural residents. Specifically, they highlight the the children learn more about environment and science
need to address the low levels of fruits and vegetables and improve their self-esteem and attitudes toward
being consumed and the disparaging rates of chronic school. These programs can be especially beneficial to
diseases among rural populations. These data suggest at-risk children like those living in rural areas (42).
that education programs emphasizing the benefits of con- Schools around the country have various garden educa-
suming homegrown produce, promoting awareness of lo- tion programs, but they are few. Feasibility studies have
cal sources (ie, farmers’ stands), and facilitating the de- demonstrated that small school-based gardening pro-
velopment of gardening programs may be a worthwhile grams are a practical way to increase fruit and vegetable
investment. consumption among youth (43). The National Gardening
Association has recently adopted a new program to in-
CONCLUSIONS crease the number of school garden curriculums. The
Epidemiologic evidence has been translated to numerous Adopt a School Garden program directs donations from
professional association guidelines that underscore the businesses and individuals to schools that want to create
importance of a variety of colorful fruits and vegetables as learning gardens. The association offers publications and
important to overall health and disease prevention (35). programs focused on five core areas: education, health
The latest Dietary Guidelines for Americans calls for five and wellness, environmental stewardship, community
to 13 servings of fruits and vegetables per day, yet the development, and home gardening (44).
average American consumes only three, omitting pota- Limited resources to effectively provide health educa-
toes (36). The need for improvement can be seen globally, tion to prevent, diagnose, and treat chronic conditions
and location can contribute to the difficulty in creating make health promotion and disease prevention particu-
opportunity and education to promote fruit and vegetable larly challenging in rural areas, yet all the more critical.
consumption, especially for the nearly 20% of the popu- Our findings contribute to the current body of literature
lation residing in nonmetropolitan areas of this country that identifies the promotion of home, school, and com-
(37). Our study results suggest that eating homegrown munity gardens as a viable strategy to improve the struc-
fruits and vegetables improves intake and the quality of tural food environment. Designing more nutritious envi-
that intake among rural parents and their preschool chil- ronments encourages more healthful diets that can lead
dren even after adjusting for sociodemographic variables. to good health and may help avoid obesity.
Total fruit and vegetable servings were higher by more
than one serving for the almost always homegrown fruit This work was funded by the National Cancer Institute
and vegetable eaters for both the parent and his or her (grant No. 5 R01 CA68398).
preschool-aged child. Furthermore, frequency of eating The authors thank the Parents as Teachers National
homegrown fruits and vegetables promoted a positive Center, Inc, Missouri Parents as Teachers-Region E
home environment in terms of availability of quality pro- (Nancy Hale, coordinator), and the Obesity Prevention
duce, parental role modeling of eating fruits and vegeta- Center Project Team (Kimberly Hessler, project manager,
bles, and the preschooler’s preference for them. and Amanda Harrod and An Huynh, research assistants).
Future directions may include exploring ways to
strengthen the broader structural environment to pro-
mote consumption of quality fruits and vegetables among References
rural families. Current literature suggests that facilitat- 1. Stuff J, Horton J, Bogle M, Connell C, Ryan D,
ing the development of school and community gardens Zaghloul S, Thornton A, Simpson P, Gossett J, Szeto
may be effective. Overall, gardeners have greater con- K, Lower Mississippi Delta Nutrition Intervention
sumption of fresh vegetables compared with nongarden- Research Consortium. High prevalence of food inse-
ers, and lower consumption of sweet foods and drinks curity and hunger in households in the rural Lower
(38). Programs that encourage gardening may not only Mississippi Delta. J Rural Health. 2004;20:173-180.
increase access to fresh fruits and vegetables, but in 2. US Department of Agriculture Economic Research
addition gardening can provide psychological well-being, Service. Rural children at a glance. Available at:
social well-being, food security, and financial savings http://www.ers.usda.gov/publications/EIB1/EIB1.pdf.
(39,40). Although our study found no association with Accessed January 18, 2007.
frequency of eating homegrown produce and lowered food 3. Olson C. Nutrition and health outcomes associated
costs, one project estimated savings of nearly $500 per with food insecurity and hunger. J Nutr. 1999;129:
garden per season in food costs for community gardeners 521S-524S.
(41). More work is needed to better understand the vari- 4. Casey P, Szeto K, Lensing M, Bogle M, Weber J.
ous ways to increase consumption of homegrown produce Children in food-insufficient low-income families.
(ie, home gardening or farmers’ market purchases) and Arch Pediatr Adolesc Med. 2001;155:508-514.
582 April 2007 Volume 107 Number 45. Alaimo K, Olson C, Frongillo E. Low family income heart disease mortality among blacks and whites in
and food insufficiency in relation to overweight in US the United States. Am J Public Health. 2001;91:
children. Arch Pediatr Adolesc Med. 2001;155:1161- 1499-1506.
1167. 19. Pearson TA, Lewis C. Rural epidemiology: Insights
6. Tai-Seale T, Chandler C. Nutrition and overweight from a rural population laboratory. Am J of Epide-
concerns in rural areas. In: Rural Healthy People miol. 1998;148:949-957.
2010: A Companion Document to Healthy People 20. Barnett E, Halverson J, Elmes GA. Metropolitan and
2010. Vol 1. College Station, TX: The Texas A&M non-metropolitan trends in coronary heart disease
University System Health Science Center, School of mortality within Appalachia. Ann Epidemiol. 2000;
Rural Public Health, Southwest Rural Health Re- 10:370-379.
search Center; 2003:187-198. 21. Parents as Teachers National Center. Available at:
7. Gamm L, Hutchinson L, Dabney B, Dodrsey A. Rural http://www.parentsasteachers.org/site/pp.asp?c⫽ekIRL
Healthy People 2010: A Companion Volume to cMZJxE&b⫽272093. Accessed January 26, 2006.
Healthy People 2010. Vol 2. College Station, TX: The 22. Brownson RC, Smith CA, Pratt M, Mack NE, Jack-
Texas A&M University System Health Science Cen- son-Thompson J, Dean CG, Dabney S. Preventing
ter, School of Rural Public Health, Southwest Rural cardiovascular disease through community-based
Health Center; 2003. risk reduction: The Bootheel Heart Health Project.
8. Noel M, Hickner J, Ettenhofer T. The high prevalence Am J Public Health. 1996;86:206-213.
of obesity in Michigan primary care practices. An 23. Haire-Joshu D, Brownson R, Nanney M, Houston C,
APRNet Study. Upper Peninsula Research Network. Steger-May K, Schechtman K, Auslander W. Improv-
J Fam Prac. 1998;47:39-43. ing dietary behavior in African Americans: The Par-
9. Levin S, Mayer-Davis E, Ainsworth BE. Racial/ethnic ents as Teachers High 5, Low Fat program. Prev Med.
health disparities in South Carolina and the role of 2003;36:684-691.
rural locality and educational attainment. South Med 24. Nanney M, Haire-Joshu D, Hessler K, Elliott M,
J. 2001;94:711-718. Brownson R. Evaluating changeability to improve
10. Greenlund KJ, Kiefe CI, Gidding SS. Differences in fruits and vegetables intake among school aged chil-
cardiovascular disease risk factors in black and white dren. Nutr J. 2005;4:34.
young adults: Comparisons among five communities 25. Linneman C, Hessler K, Nanney S, Steger-May K,
of the CARDIA and the Bogalusa Heart Studies. Cor- Haire-Joshu D. Parents are accurate reporters of
onary artery risk development in young adults. Ann their preschoolers’ fruit and vegetable consumption
Epidemiol. 1998;8:22-30. under limited conditions. J Nutr Educ Behav. 2004;
11. Ogunyemi D, Hullett S, Leeper J. Pre-pregnancy 36:305-308.
body mass index, weight gain during pregnancy, and 26. Cicchetti D. Guidelines, criteria, and rules of thumb
perinatal outcome in rural black population. J Ma- for evaluating normed and standardized assessment
tern Fetal Med. 1998;7:190-193. instruments in psychology. Psychol Assess. 1994;6:
12. McMurray R, Harrell J, Bangdiwala S, Deng S. Car- 284-290.
diovascular disease risk factors and obesity of rural 27. Haire-Joshu D, Nanney M. Prevention of overweight
and urban elementary school children. J Rural and obesity in children: Influences on the food envi-
Health. 1999;15:365-374. ronment. Diabetes Educ. 2002;28:415-422.
13. Stanton CA, Fries EA, Danish SJ. Racial and gender 28. Tibbs T, Haire-Joshu D, Schechtman K, Brownson R,
differences in the diets of rural youth and their moth- Nanney M, Houston C, Auslander W. The relation-
ers. Am J Health Behav. 2003;27:336-347. ship between parental modeling, eating patterns, and
14. Stroehla B, Malcoe L, Velie E. Dietary sources of dietary intake among African American parents.
nutrients among rural Native American and white J Am Diet Assoc. 2001;101:535-541.
children. J Am Diet Assoc. 2005;105:1908-1916. 29. Blanchette L, Brug J. Determinants of fruit and veg-
15. Eberhardt M, Ingram D, Makue D. Health United etable consumption among 6-12 year old children and
States, 2000, Urban and Rural Health Chartbook. effective interventions to increase consumption. J
Hyattsville, MD: National Center for Health Statis- Hum Nutr Diet. 2005;18:431-443.
tics; 2001:1-160. DHHS Publication No. PHS 2001- 30. Haire-Joshu D, Nanny MS. Behavioral interventions
1232. for cancer prevention: Dietary intake and physical ac-
16. Department of Health and Human Services. Current tivity. In: Given CW, Given B, Champion VL, Kozachik
estimates from the National Health Interview Sur- S, DeVoss DN, eds. Evidence-Based Cancer Care and
vey, United States, 1998. Hyattsville, MD: National Prevention. St Louis, MO: Springer Publishers; 2002:
Center for Health Statistics Vital and Health Statis- 17-62.
tics: Series 10, No. 207. DHHS Publication No. (PHS) 31. Bowman S, Gortmaker S, Ebbeling C, Pereira M,
20021535. Ludwig D. Effects of fast food consumption on energy
17. Sundquist J, Winkleby MA, Pudaric S. Cardiovascu- intake and diet quality among children in a national
lar disease risk factors among older black, Mexican- household survey. Pediatrics. 2004;113:112-118.
American, and white women and men: An analysis of 32. Field A, Gillman M, Rosner B, Rockett H, Colditz G.
NHANES III, 1988-1994. Third National Health and Association between fruit and vegetable intake and
Nutrition Examination Survey. J Am Geriatr Soc. change in body mass index among a large sample of
2001;49:109-116. children and adolescents in the United States. Int J
18. Barnett E, Halverson J. Local increases in coronary Obes. 2003;27:821-826.
April 2007 ● Journal of the AMERICAN DIETETIC ASSOCIATION 58333. Tohill B, Seymour J, Serdula M, Kettel-Khan L, Rolls 39. McBey M. The therapeutic aspects of gardens and
B. What epidemiologic studies tell us about the rela- gardening; An aspect of total patient care. J Adv
tionship between fruit and vegetable consumption Nurs. 1985;10:591-595.
and body weight. Nutr Rev. 2004;62:365-374. 40. Holben D, McClincy M, Holcomb J, Dean K, Walker
34. American Dietetic Association Evidence Analysis Li- C. Food security status of households in Appalachian
brary. Fruit and vegetable intake and child over- Ohio with children in Head Start. J Am Diet Assoc.
weight. Available at: http://www.adaevidencelibrary. 2004;104:238-241.
com/topic.cfm?cat⫽1054. Accessed January 26, 2006. 41. Patel IC. Gardening’s socioeconomic impacts. J Ex-
35. Nanney MS, Haire-Joshu D, Brownson R. Rationale tension. 1991;29:7-8.
for a ‘powerhouse’ approach to vegetable and fruit
42. National Gardening Association. Research support
interventions. J Am Diet Assoc. 2004;104:352-356.
for kids’ gardening. Available at: http://www.
36. US Department of Health and Human Services, US
Department of Agriculture. Nutrition and Your Health: kidsgardening.com/2005.kids.garden.news/research.
Dietary Guidelines for Americans, 2005. 6th ed. Wash- pdf. Accessed February 12, 2007.
ington, DC: US Government Printing Office; 2005. 43. Stables G, Young E, Howerton M, Yaroch A, Kuester S,
Home and Garden Bulletin No. 232. Solera M, Cobb K, Nebeling L. Small school-based ef-
37. Waldman H, Perlman S. Rural health issues. ASDC J fectiveness trials increase vegetable and fruit consump-
Dent Child. 2002;69:96-99. tion among youth. J Am Diet Assoc. 2005;105:252-256.
38. Blair D, Giesecke C, Sherman S. A dietary, social, 44. National Gardening Association. New adopt a school gar-
and economic evaluation of the Philadelphia Urban den program. Available at: http://assoc.garden.org/about.
Gardening Project. J Nutr Educ. 1991;10:161-167. Accessed February 13, 2007.
APPLICATION
Linking Homegrown and Locally Produced Fruits and Vegetables to Improving Access
and Intake in Communities through Policy and Environmental Change
A
pproximately three fourths of children and adults cated they ate home-grown fruits and vegetables “almost
are currently not meeting the minimum recom- always” were 2.3 times as likely to eat five daily servings
mended servings of fruit and vegetable intake, de- of fruits and vegetables.
spite extensive research demonstrating the positive im-
pact on growth, development, and disease prevention
(1,2). Limited access to fruits and vegetables is one of the UTILITY OF THE RESEARCH
hypothesized factors contributing to this inadequacy. This study reveals a possible mechanism for increasing
Studies report that low-income and predominately minor- access to fruits and vegetables through locally grown
ity communities often have limited access to these foods food. Although it was not established if “home grown” was
for a variety of reasons, including lack of access to grocery from a farmer’s market, grown by a neighbor, or from
stores and other locations that sell fresh fruits and veg- personal gardens, families consumed more fruits and veg-
etables. Strategies to increase access to fruits and vege- etables when the source was local. By establishing this
tables are potentially one powerful step toward improv- link between the consumption of locally grown food and
ing intake (3). The study by Nanney and colleagues (4) increased fruit and vegetable intake, this study identifies
examined whether the availability of home-grown pro- a potential avenue for improving access and intake. The
duce is associated with higher intake of fruits and vege- expansion of local, sustainable agriculture programs such
tables. The authors found that parents in rural commu- as farmer’s markets, community and home gardens, and
nities who report almost always eating home-grown farm to school programs is a useful and practical ap-
fruits and vegetables were 3.2 times more likely to eat proach. This approach may not only be relevant to a rural
five daily servings of fruits and vegetables compared to setting, but an increase in the presence of these programs
those who report rarely/never eating home-grown pro- in urban areas could likewise produce similar benefits.
duce. In addition, preschool children of parents who indi-
How to Apply to the Registered Dietitian’s Practice
Food and nutrition professionals can play a role in in-
This article was written by Alison Gustafson, RD, creasing access to locally grown fruits and vegetables
MPH, David Cavallo, and Amy Paxton, UNC Center through several activities. Advocacy efforts targeted to-
for Health Promotion and Disease Prevention, Chapel ward increased funding and support for farm-to-school
Hill, NC. programs is one approach, while reducing regulatory bar-
doi: 10.1016/j.jada.2007.02.023 riers and facilitating partnerships between groups with
584 April 2007 Volume 107 Number 4You can also read
