JRHS

Journal of Research in Health Sciences

doi:10.34172/jrhs.2024.162
JRHS 2024; 24(4):e00627
http://jrhs.umsha.ac.ir

Original Article
Investigation of the Link Between Food Assistance Programs
and Physical Activity Among Children and Adolescents
ID
Pardis Noormohammadpour (MD, MPH)1* , Nicole Robertson (MPH)1
1
Dalla Lana School of Public Health, University of Toronto, Toronto, Ontario, Canada

Article history: Abstract

Received: April 2, 2024
Background: While the direct benefits of food assistance programs are well-documented, there
Revised: July 10, 2024
Accepted: July 27, 2024 is a need to explore indirect benefits like increased physical activity. This study examined
ePublished: September 30, 2024 whether participation in the Supplemental Nutrition Assistance Program (SNAP) was associated
with improved physical activity levels in children and adolescents aged 2-17 in the United States
Keywords: during 2017-2018.
SNAP programs, Physical activity, Study Design: A cross-sectional study.
Preschool children, Adolescents
Methods: This cross-sectional study used a subset of National Health and Nutrition Examination
*Corresponding author: Survey (NHANES) data (n = 2620). In the NHANES 2017-2018 dataset, physical activity was
Pardis Noormohammadpour, measured through self-report questionnaires, which captured participants’ frequency, duration,
Email: pardis. and intensity of various activities. We used weighted logistic regression and the Hosmer -
noormohammadpour@mail. Lemeshow - Sturdivant forward model - building strategy to investigate this hypothesized
utoronto.ca association using SAS version 9.4.
Results: In the adjusted model, controlling for the other variables in the model, we found that
children and adolescents from households that had received SNAP/Food Stamps had 1.53
times higher odds (odds ratio [OR] = 1.53, 95% CI: 1.24-1.89) of achieving the recommended
guidelines of 60 minutes of daily physical activity compared to those who had not received
benefits. Each additional year of age resulted in 0.82 times lower odds (OR = 0.82; 95% CI:
0.79, 0.85) of meeting the recommended amounts of physical activity. Additionally, each unit
increase in BMI was associated with 0.96 times lower odds (OR = 0.96, 95% CI: 0.93, 0.98) of
engaging in recommended physical activity.
Conclusion: These findings suggest that participation in the SNAP/Food Stamps program may
indirectly benefit participants by increasing physical activity levels.

Please cite this article as follows: Noormohammadpour P, Robertson N. Investigation of the link between food assistance programs and
physical activity among children and adolescents. J Res Health Sci. 2024; 24(4):e00627. doi:10.34172/jrhs.2024.162

Background the benefits of physical activity, it is estimated that only

Many studies have established the link between physical one in four children in the United States (US) meet this
inactivity and poor health outcomes. Studies have recommendation, which has been further worsened by
indicated that physical inactivity is associated with many the COVID-19 pandemic.1 It is crucial to examine the
chronic conditions, such as diabetes, obesity, heart disease, factors contributing to physical inactivity in children and
and premature mortality.1 In contrast, physical activity adolescents. Understanding these factors can help develop
has shown numerous benefits. Specifically in children and targeted interventions to overcome barriers to achieving
adolescents, physical activity has been shown to decrease recommended activity levels.
the risk of the mentioned morbidities, improve cognition Previous studies conducted in the US have suggested
and performance in school, decrease stress, improve sleep, an association between lower socioeconomic status
improve overall mental health, and ensure healthy growth and physical inactivity in children5-7 and an inverse
and development, which can extend into adulthood.1-3 In relationship between food insecurity and physical activity
order to maintain or improve overall health and reduce in children.8,9 Programs like Supplemental Nutrition
the risk of chronic disease, the Physical Activity Guidelines Assistance Program (SNAP) might indirectly increase
for American Children recommend a minimum of 60 children’s physical activity by improving nutrition and
minutes of aerobic, muscle-strengthening, or physical freeing up household resources for activity-related
exercise daily.4 Despite this comprehensive knowledge of expenses. Although these programs are not typically

© 2024 The Author(s); Published by Hamadan University of Medical Sciences. This is an open-access article distributed under the terms of
the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
Noormohammadpour and Robertson

designed to boost physical activity, we aimed to explore excluded participants who had missing responses, refused
this potential benefit. No prior research has examined to respond, or responded with “I don’t know” to the
the link between food benefit programs and physical exposure of interest, namely the participation/receipt of
activity levels in US children and adolescents. We focus the SNAP/Food Stamp program (ever received versus
on the SNAP program, the most significant anti- hunger never received), and the outcome of interest, the number
initiative in the US, which supports over 41 million low- of days of being physically active for at least 60 minutes.
income individuals in accessing nutritious food.10
The SNAP/Food Stamp program provides nutrition Measurements
assistance to low-income individuals and families, Exposure
offering a Benefits Card to purchase eligible food based The exposure to be measured was the participation or
on income and resource limits.11 Studies have suggested receipt of SNAP/Food Stamp benefits. One adult in
that the SNAP/Food Stamps Program has reduced food the participating household responded on behalf of all
insecurity by as much as 30% while also putting money household members. This variable was assessed via the
back into the economy, lessening the extent and severity of following question: “Have you or any member in your
poverty, supporting low-paid workers, promoting healthy household ever received SNAP or Food Stamp benefits?”.
eating and improved health, and lowering health care Questions related to SNAP/Food Stamp benefits were
expenditures.10 The program has been reported to have collected at the household level.14 This will be treated
positive effects on the physical activity levels of children, as a binary variable with yes/no responses. Subjects
which may be an important pathway to improve health who answered yes were considered to be exposed.14
outcomes. Respondents who refused to answer, answered “I don’t
This study used a cross-sectional research design to know”, or did not provide a response to the question were
explore whether participation in the SNAP/Food Stamps considered to have missing data and were not included in
program was linked to improved physical activity in the analysis.15
children and adolescents aged 2-17 years in the US
during 2017-2018. Using publicly available data from Outcome
the National Health and Nutrition Examination Survey The outcome to be measured was physical activity levels
(NHANES), the study aimed to quantify the relationship and the results were presented according to whether
between SNAP/Food Stamp program participation and participants have met the physical activity guidelines of
meeting recommended physical activity guidelines. The 60 minutes of daily physical activity.16 This variable was
hypothesis was that exposure to the program would be assessed via the following question, “During the past 7
associated with increased physical activity levels in the days, on how many days were you physically active for a
study population. total of at least 60 minutes per day?”. Participants or their
proxies (for children under the age of 12 or those who
Materials and Methods were unable to respond to the question themselves) were
Study population and data collection instructed to add up the time spent in any kind of physical
This cross-sectional analysis used a subset of the activity that resulted in an increased heart rate and made
NHANES cohort, specifically all children and adolescents the respondent breathe hard some of the time.17,18 Physical
aged 2 to 17 years (2620 subjects) from the 2017-2018 activity questions were asked by trained interviewers in
cycle. The NHANES program, which has been designed the home using the computer-assisted personal interview
to assess the health of children and adults across the system for participants aged 2 to 11 and 16 to 17 years;
US, combines interviews and physical examinations to participants aged 12 to 15 were asked this question in
examine a sample of approximately 5000 people per year the Mobile Examination Centre (MEC). Responses were
across the country.12 The sampling methods aim to create then collapsed and dichotomized into a binary measure
a nationally representative sample of the US population, (yes or no) to decide whether they had met the guidelines;
with oversampling of selected demographics as needed respondents were considered to have met the guideline if
to produce reliable statistics12; the 2017-2018 NHANES they reported being physically active for at least 60 minutes
cycle includes an oversampling of Asian Americans.13 We on seven days per week. Any lesser value was considered
have chosen to study children and adolescents aged 2-17 to have not met the guideline. Respondents who refused
because we believe that interventions targeting this priority to answer, answered “I don’t know”, or did not provide a
group would greatly benefit population health now and response to the question were considered to have missing
into the future. The NHANES dataset used for this study data and were not included in the analysis.15
can be accessed at https://wwwn.cdc.gov/nchs/nhanes/
continuousnhanes/default.aspx?BeginYear = 2017. Covariates
We have used demographics, physical activity levels, Potential confounders were identified a priori through
food insecurity, and physical examination datasets a literature review and conceptualized using a directed
retrieved from the National Centre for Health Statistics acyclic graph. Potential confounders were retrieved from
and merged them based on the participant ID. We the Food Security, Demographic Variables, and Body

2 J Res Health Sci, 2024, Volume 24, Issue 4

 Food assistance programs and physical activity

Measures NHANES datasets for the 2017-2018 cycle. ensuring a robust logistic regression model. As we were
Interview responses were collected through respondent- only interested in participants aged 2-17, we created a
level and household-level interview data, and body subset of the sample population to use for our analysis,
measures were collected by trained health technicians in as NHANES recommends subsetting data instead of
the MEC. We considered potential confounders related removing observations that do not meet the inclusion
to the age of the participants and/or the household criteria to provide correct variance estimates. Given that
reference persons (numeric continuous variable), gender the outcome variable was initially coded as an integer value
(categorical binary variable), height and weight or body from 0 days to 7 days of physical activity of 60 minutes or
mass index (BMI) (numeric continuous variables), race/ more, crude and adjusted ordinal regression models were
ethnicity (categorical nominal variable), number of originally fitted. Variations of this ordinal model, including
people in the household or in the family (numeric discrete collapsing outcome categories and adding or removing
variable), education level (categorical ordinal variable), covariates, were explored to check the appropriateness
marital status (categorical nominal variable), household of this model and to determine if the proportional odds
income (categorical ordinal variable), and household food assumption was met. Then, multinomial regression and
security category (categorical binary variable). logistic regression models were used to compare the
models to determine which would be the most appropriate.
Statistical Analysis As there is a recommended guideline for physical activity
All statistical analyses were conducted using SAS version in children and adolescents, it was determined that a
9.4 (SAS Institute Inc.). Given that the NHANES study binary logistic model would be used, assuming that the
includes a complex multistage probability sampling design literature supports the dichotomizing of the outcome
to be representative of the civilian, non-institutionalized variable. Accordingly, crude and multivariable logistic
US population,19 examination sample weights provided by regression models were used to determine the odds ratio
the NHANES database were used. The NHANES database and 95% confidence intervals for the crude and adjusted
recommends the use of the 2-year sample weights for relationships between the SNAP and meeting the physical
all 2017-2018 analyses, and it recommends the use of activity guidelines. An alpha level of 0.05 was used. All
the exam sample weights if household questionnaire reported P values are one-sided.
data are merged with exam data.13,17 Further, as the
primary sampling units and true design strata are not Results
released to reduce the risks of disclosure and to protect Participant characteristics
the confidentiality of information provided by survey Table 1 presents the baseline characteristics of participants
respondents,13 a masked variance unit (MVU) pseudo- included in the subset of the study population. In total,
primary sampling unit variable and an MVU pseudo- 2620 participants were eligible for inclusion based on
stratum variable were incorporated into the model for the age restrictions. In this age-restricted subset of the
variance estimation. sample, the mean age of participants was 9.61 years.
We ran descriptive statistics on our variables selected Additionally, 1304 (49.8%) of participants were male, and
for possible inclusion based on our literature review and the mean BMI was 20.07 kg/m2. Of the 2527 participants
frequency, distributions, and missing data were assessed. or proxies that provided a response to the SNAP/Food
NHANES recommends that if 10% or less of the data for Stamp Program question, 1451 (57.4%) of the participants
the outcome variable are missing, it is usually acceptable had received SNAP/Food Stamp Benefits. Out of 2581
to conduct the analyses without further adjustment15; responses to the physical activity question, 1112 (43.1%)
therefore, we used 10% as our cut-off for missing data. We had met the physical activity guidelines. The majority of
also reviewed the NHANES codebooks to confirm that participants came from households that were considered
skip patterns were not responsible for missing responses. to have full food security (n = 1404, 55.3%). Household
Then, a Chi-squared test was used to compare participants reference persons were more likely to be married and/
with complete data on the exposure and outcome with or live with a partner (n = 1854, 71.8%) and to have a
participants who had missing data (those who responded high school diploma, a General Education Development
“I don’t know”, refused to respond, or provided no (GED) certificate, a college degree, or an Associate’s
response) for one or both of these variables. degree (n = 1426, 57.3%). The median household size was
five people (n = 662, 25.3%).
Model selection
We used the Hosmer-Lemeshow-Sturdivant (H-L-S) Model fitting
forward model building strategy20 to build our models. Upon testing the ordinal regression model and its
The H-L-S forward model-building strategy involves variations, it was evident that the proportional odds
starting with a minimal set of predictors and systematically assumption was violated (Brant test P < 0.001); therefore,
adding variables based on statistical criteria to improve this model was not an appropriate selection. Then, we
model fit . This approach optimizes predictor selection explored the use of a multinomial logistic regression
and model performance while avoiding overfitting, model to address the complex relationships in our dataset.

J Res Health Sci, 2024, Volume 24, Issue 4 3

Noormohammadpour and Robertson

Table 1. Baseline characteristics of participants

Quantitative variables Mean SE Median, IQR

Age of child (y) 9.61 0.14 9.14 (7.94)

Weight of child (kg) 42.03 0.54 37.13 (35.33)

Standing height of child (cm) 138.59 0.65 141.19 (43.08)

Body mass index (kg/m2) 20.07 0.14 18.35 (6.45)

Qualitative variables Sample frequency Weighted frequency Weighted percent (%)

Household food security benefit: ever received 2527

Yes 1451 32 379 058 50.37

No 1076 31 901 452 49.63

Days of being physically active for at least 60 minutes 2581

0 186 4 415 213 6.74

1 94 2 484 076 3.79

2 187 4 966 843 7.59

3 238 6 533 551 9.98

4 201 5 983 895 9.14

5 396 10 800 671 16.50

6 167 4 335 677 6.62

7 1112 25 939 777 39.63

Meet the physical activity guideline (60 minutes, 7 days per week) 2581

No 1469 39 519 926 60.37

Yes 1112 25 939 777 39.63

Household food security category 2537

Full food security: 0 1404 39 543 770 61.31

Marginal food security: 1-2 426 9 463 506 14.67

Low food security: 3-7 464 10 087 493 15.64

Very low food security: 8-18 243 5 407 194 8.38

Gender of child 2620

Male 1304 33 682 101 50.82

Female 1316 32 590 322 49.18

Race of child 2620

Mexican American 446 11 109 403 16.76

Other Hispanic 205 5 302 781 8.00

Non-Hispanic White 811 32 692 978 49.33

Non-Hispanic Black 620 9 006 956 13.59

Non-Hispanic Asian 277 3 269 246 4.93

Other races (including multi-racial) 261 4 891 059 7.38

Total number of people in the household 2620

2 93 2 427 198 3.66

3 381 10 909 194 16.46

4 728 20 000 666 30.18

5 662 16 320 053 24.63

6 388 9 149 356 13.81

7 or more people in the household 368 7 465 957 11.27

Total number of people in the household 2620

0-3 474 13 336 392 20.12

≥4 2146 52 936 032 79.88

Marital status of household reference person 2582

Married/living with partner 1854 48 965 199 74.62

Widowed/divorced/separated 381 9 857 803 15.02

4 J Res Health Sci, 2024, Volume 24, Issue 4

 Food assistance programs and physical activity

Table 1. Continued.

Qualitative variables Sample frequency Weighted frequency Weighted percent (%)

Never married 347 6 797 182 10.36

Education level of household reference person 2490

Less than high school degree 470 10 816 710 17.06

High school/GED/Associate’s degree 1426 33 750 332 53.24

College graduate or above 594 18 821 687 29.69

Household income ($) 2445

0-4999 83 1 712 717 2.75

5000-9999 71 1 352 173 2.17

10 000- 14 999 100 2 271 593 3.64

15 000-19 999 149 3 203 459 5.13

20 000-24 999 144 2 893 754 4.64

25 000-34 999 290 6 062 935 9.72

35 000- 44 999 258 5 809 825 9.31

45 000-54 999 166 3 838 468 6.15

55 000-64 999 188 4 157 757 6.66

65 000-74 999 122 3 259 108 5.22

75 000-99 999 261 8 090 109 12.97

≥ 100 000 501 17 586 957 28.19

< 20 000 31 578 075 0.93

≥ 20 000 81 1 571 412 2.52

IQR: Interquartile range; SE: Standard error; GED: General Education Development.

However, the complex survey design of the NHANES H-L-S method identified age, gender, BMI, and race/
dataset posed significant challenges. Specifically, in this ethnicity of the participant as important covariates to be
design, the degrees of freedom for statistical tests are included in the model in addition to the exposure and the
influenced by the number of primary sampling units outcome. The H-L-S method also provided evidence that
rather than the total number of individuals sampled. household food security level was not appropriate to be
Multinomial logistic regression is used when there is a included as an effect measure modifier or as a confounding
categorical dependent variable with more than two levels. variable and therefore was not included. The Chi-square
It requires a considerable amount of degrees of freedom test demonstrated that the missing values for the exposure
to account for the multiple categories of the dependent variable and the outcome variable were not related to one
variable and to adequately estimate the effects of the another (P = 0.41).
predictors on each category. In our case, the degrees of
freedom needed for the multinomial logistic regression Associations between SNAP and physical activity
model exceeded the 15 degrees of freedom specified by In the crude logistic regression model, participants who
NHANES21 for our dataset, which limited the robustness received SNAP/Food Stamp benefits were 1.38 times more
of our statistical analysis. The complexity of the NHANES likely (odds ratio [OR] = 1.38, 95% confidence interval
survey design, including its clustering and stratification, [CI]: 1.10, 1.72) to achieve the recommended physical
meant that the effective sample size and degrees of freedom activity guidelines than participants who did not receive
were reduced. The constraints imposed by these design SNAP/Food Stamp benefits. In the adjusted logistic
features made it challenging to fit a multinomial logistic regression model, controlling for the other variables in
regression model appropriately, as the model required the model, participants who received SNAP/Food Stamp
more degrees of freedom than were available. To address benefits were 1.53 times more likely (OR = 1.53, 95% CI:
these constraints and ensure the robustness and reliability 1.24, 1.89) to achieve the recommended physical activity
of our results, we opted for a binary logistic regression guidelines than participants who did not receive SNAP/
model. This model is more suited to situations where Food Stamp benefits. Both findings were significant at the
the dependent variable is binary and is less demanding 0.05 level.
in terms of degrees of freedom. By choosing the binary Based on the adjusted model, controlling for the other
logistic regression model, we were able to adhere to the variables in the model, male children were 1.35 times
statistical limitations of the NHANES dataset while still more likely (OR = 1.35, 95% CI: 1.11, 1.64) than female
providing meaningful insights into the relationships children to meet the recommended physical activity
between the predictors and the outcome variable. The guidelines. Some demographic characteristics, such as age

J Res Health Sci, 2024, Volume 24, Issue 4 5

Noormohammadpour and Robertson

and BMI, were related to lower amounts of recommended is also demonstrated in older age categories. Further, as
physical activity. Mexican Americans had 0.58 times participation in SNAP could increase the probability of
lower odds (OR = 0.58, 95% CI: 0.40, 0.83) of reaching the meeting the recommended amount of physical activity
recommended amount of physical activity in comparison in children and adolescents, the risk of chronic non-
to non-Hispanic white subjects. Each year of increase in communicable diseases would decrease, which could
age resulted in 0.82 times lower odds (OR = 0.82, 95% CI: lead to a decrease in the likelihood of all causes and
0.79, 0.85) of meeting recommended amounts of physical specific causes of premature morbidity and mortality as
activity. Moreover, the odds of engaging in recommended adolescents transition into adulthood.
physical activity were 0.96 times lower (OR = 0.96, 95% CI: Even though physical activity was assessed in different
0.93, 0.98) with each unit of increase in BMI. Results from formats over various years in the NHANES databases,
the crude and adjusted models can be found in Table 2. which provide a rich source of data regarding the amount
of physical activity in the US population, the number of
Discussion published studies in this field on children and adolescents
These results provide evidence of the benefits of past remains relatively small. In 2014, To et al discussed that
or current participation in SNAP, or the Food Stamps children who were food insecure did less moderate-to-
Program, on child and adolescent physical activity levels. vigorous physical activity than their counterparts who were
Those who received SNAP/Food Stamps benefits were not.8 In 2022, To et al demonstrated that physical activity,
found to have increased odds of meeting the physical measured by a wearable device, was higher on weekdays than
activity guidelines of 60 minutes of daily physical activity, on weekends in children, and physical activity was affected
thus supporting our initial hypothesis. In addition, being by age, weight, and ethnicity.23 Further, previous studies
a Mexican American or other Hispanic group, increasing have indicated that factors that influence physical activity
age, or having a higher BMI were associated with lower may operate differentially across certain demographics,
odds of achieving recommended physical activity. On such as age, gender, and race/ethnicity.24,25 To the best of
the other hand, the odds of reaching the recommended our knowledge, none of the previous studies investigated
amount of physical activity were higher in males, the effect of SNAP/Food Stamps benefits on meeting the
controlling for the other variables in the model. recommended level of physical activity in children and
These positive findings agree with the body of literature adolescents. The findings of this study could be helpful to
demonstrating the health benefits of the SNAP/Food decision-makers regarding future policies on SNAP.
Stamps program. In a recent study, Heflin et al showed To further substantiate our findings, we applied Hill’s
that SNAP participation decreased all-cause mortality considerations for causality.26 The significant odds
by about one to two percentage points throughout the ratios (OR = 1.53 for SNAP participation and physical
population and decreased certain causes of death among activity) indicate a robust association, demonstrating the
persons aged 40 to 64.22 While the authors did not discuss strength of this relationship. Our results are consistent
the possible pathways behind their findings, our finding with other studies showing the health benefits of SNAP
could explain a reason for their result if this association participation,22 supporting the consistency criterion. The
Table 2. Associations between selected variables and physical activity

Variables Unadjusted OR (95% CI) P value Adjusted OR (95% CI) P value

Age of child (y) 0.80 (0.78, 0.83) 0.001* 0.82 (0.79, 0.85) 0.001*

Body mass index (kg/m ) 2

0.87 (0.85, 0.88) 0.001* 0.96 (0.93, 0.98) 0.002*

Household food security benefit: ever received

Yes 1.38 (1.10, 1.72) 0.008 1.53 (1.24, 1.89) 0.001*

No Ref. Ref.

Gender of child

Male 1.29 (1.09, 1.51) 0.005* 1.35 (1.11, 1.64) 0.006

Female Ref. Ref.

Race of child

Mexican American 0.65 (0.45, 0.92) 0.020* 0.58 (0.40, 0.83) 0.006

Non-Hispanic White Ref. Ref.

Non-Hispanic Black 1.00 (0.71, 1.40) 0.990 1.00 (0.71, 1.40) 0.980

Non-Hispanic Asian 0.82 (0.49, 1.37) 0.420 0.83 (0.43, 1.59) 0.540

Other Hispanic 0.71 (0.45, 1.11) 0.120 0.56 (0.34, 0.92) 0.024*

Other races (including multi-racial) 0.89 (0.56, 1.42) 0.600 0.83 (0.54, 1.26) 0.350
CI: Confidence Interval; Ref: Reference Category
*Statistically significant (P < 0.05)
Odds ratios are adjusted for age, gender, BMI, race/ethnicity.

6 J Res Health Sci, 2024, Volume 24, Issue 4

 Food assistance programs and physical activity

relationship between SNAP benefits and physical activity Highlights

could be specific, with few other explanations, which
may meet the specificity criterion. Additionally, SNAP • Children from SNAP households were 1.53 times
participation may precede the increase in physical activity, more likely to meet exercise guidelines.
which could suggest a temporal relationship requirement. • Male kids were 1.35 times more likely than females
There is some indications of a dose-response relationship, to meet physical activity guidelines.
• Mexican Americans had 0.58 times lower odds of
with higher SNAP benefits potentially correlating
meeting physical activity guidelines.
with increased physical activity, which may fulfill the
biological gradient criterion. Improved nutrition and
energy security from SNAP benefits provide a plausible Acknowledgments
mechanism for increased physical activity, addressing the We would like to acknowledge Professor David Fisman for his
valuable guidance.
plausibility criterion. The findings are coherent with the
known benefits of SNAP in improving health outcomes,27 Authors’ Contribution
fulfilling the coherence criterion. While experimental Conceptualization: Pardis Noormohammadpour, Nicole Robertson.
data is limited, related interventions support our findings, Data curation: Nicole Robertson, Pardis Noormohammadpour.
Formal analysis: Pardis Noormohammadpour, Nicole Robertson.
addressing the experiment criterion. Finally, similar social
Investigation: Nicole Robertson, Pardis Noormohammadpour.
support programs have shown causal relationships with Methodology: Pardis Noormohammadpour, Nicole Robertson.
health improvements,28,29 meeting the analogy criterion. Project administration: Nicole Robertson, Pardis Noormohammadpour.
By incorporating these considerations, we strengthen Resources: Pardis Noormohammadpour, Nicole Robertson.
the argument that SNAP/Food Stamps benefits have a Software: Pardis Noormohammadpour, Nicole Robertson.
Supervision: Pardis Noormohammadpour, Nicole Robertson.
positive impact on physical activity levels in children
Validation: Nicole Robertson, Pardis Noormohammadpour.
and adolescents, independent of other covariates, as Writing–original draft: Nicole Robertson, Pardis Noormohammadpour.
demonstrated by our adjusted analysis. Writing–review & editing: Pardis Noormohammadpour, Nicole
One notable limitation of our study is the potential for Robertson.
underreporting related to the socioeconomic status of
Competing Interests
participants . Since SNAP is primarily used by individuals
The authors have no conflicts of interests to declare.
from lower socioeconomic backgrounds , there is a risk
that some participants may be less inclined to provide Ethical Approval
accurate responses due to social stigma or discomfort with As a publicly available dataset, NHANES does not require ethical
the topic. This could affect the reliability of the data and approval for secondary analysis.

introduce bias into the study results. We acknowledge Funding

that this limitation may influence the validity of our This study did not receive any funding or financial support.
findings and recommend that future research incorporate
strategies to mitigate this issue, such as employing References
1. Michel J, Bernier A, Thompson LA. Physical activity in
anonymized data collection methods. Additionally, 3.5%
children. JAMA Pediatr. 2022;176(6):622. doi: 10.1001/
of the total participants had missing responses, refused to jamapediatrics.2022.0477.
answer, or responded with “I don’t know” regarding SNAP 2. World Health Organization (WHO). Physical Activity.
participation. This limitation should be considered when Geneva: WHO; 2022. Available from: https://www.who.
interpreting the results. Another limitation to consider int/news-room/fact-sheets/detail/physical-activity. Accessed
December 10, 2022 .
is that the average height and body mass index (BMI) of
3. Landry BW, Driscoll SW. Physical activity in children and
two-year-old children differ significantly from those of adolescents. PM R. 2012;4(11):826-32. doi: 10.1016/j.
17-year-old teenagers. Therefore, the results presented in pmrj.2012.09.585.
Table 1 should be interpreted considering this limitation. 4. United States Department of Health and Human Services
(HHS). Physical Activity Guidelines for Americans. 2nd
ed. Washington, DC: HHS; 2022. Available from: https://
Conclusion
health.gov/sites/default/files/2019-09/Physical_Activity_
In conclusion, the present study illustrated the higher odds Guidelines_2nd_edition.pdf. Accessed December 11, 2022 .
of reaching the recommended level of physical activity 5. Cottrell L, Zatezalo J, Bonasso A, Lattin J, Shawley S, Murphy E,
in children and adolescents who have received SNAP. et al. The relationship between children’s physical activity and
Additionally, meeting the recommended level of physical family income in rural settings: a cross-sectional study. Prev
Med Rep. 2015;2:99-104. doi: 10.1016/j.pmedr.2015.01.008.
activity was associated with the age, gender, BMI, and race 6. Tandon PS, Zhou C, Sallis JF, Cain KL, Frank LD, Saelens BE.
of participants. These results highlight the importance Home environment relationships with children’s physical
of the availability of supplemental nutrition programs activity, sedentary time, and screen time by socioeconomic
for families that may be of lower income or experience status. Int J Behav Nutr Phys Act. 2012;9:88. doi: 10.1186/1479-
food insecurity in the achievement of physical activity 5868-9-88.
7. Romero AJ, Robinson TN, Kraemer HC, Erickson SJ, Haydel KF,
guidelines in children and adolescents and for improved Mendoza F, et al. Are perceived neighborhood hazards a barrier
health outcomes that extend into adulthood. to physical activity in children? Arch Pediatr Adolesc Med.

J Res Health Sci, 2024, Volume 24, Issue 4 7

Noormohammadpour and Robertson

2001;155(10):1143-8. doi: 10.1001/archpedi.155.10.1143. 18. Centers for Disease Control and Prevention (CDC) . The
8. To QG, Frongillo EA, Gallegos D, Moore JB. Household NHANES National Youth Fitness Survey. Atlanta: CDC; 2022.
food insecurity is associated with less physical activity Available from: https://www.cdc.gov/nchs/nnyfs/index.htm.
among children and adults in the U.S. population. J Nutr. Accessed December 11, 2022.
2014;144(11):1797-802. doi: 10.3945/jn.114.198184. 19. Centers for Disease Control and Prevention (CDC). National
9. Fram MS, Ritchie LD, Rosen N, Frongillo EA. Child experience Health and Nutrition Examination Survey, Module 2: Sample
of food insecurity is associated with child diet and physical Design. Atlanta: CDC; 2022. Available from: https://wwwn.
activity. J Nutr. 2015;145(3):499-504. doi: 10.3945/ cdc.gov/nchs/nhanes/tutorials/module2.aspx. Accessed
jn.114.194365. December 9, 2022.
10. Center on Budget and Policy Priorities (CBPP). Policy Basics: 20. Hosmer DW Jr, Lemeshow S, Sturdivant RX. Model-building
The Supplemental Nutrition Assistance Program (SNAP). strategies and methods for logistic regression. In: Applied
Washington, DC: CBPP; 2022. Available from: https://www. Logistic Regression. John Wiley & Sons; 2013. p. 89-151.
cbpp.org/research/food-assistance/the-supplemental-nutrition- 21. Centers for Disease Control and Prevention (CDC).
assistance-program-snap. Accessed December 12, 2022. NHANES Reliability of Estimates. Atlanta: CDC; 2022.
11. Food and Nutrition Service. Supplemental Nutrition Assistance Available from: https://wwwn.cdc.gov/nchs/nhanes/tutorials/
Program (SNAP). Washington, DC: USDA; 2022. Available reliabilityofestimates.aspx. Accessed December 12, 2022.
from: https://www.fns.usda.gov/snap/supplemental-nutrition- 22. Heflin CM, Ingram SJ, Ziliak JP. The effect of the
assistance-program. Accessed December 10, 2022. supplemental nutrition assistance program on mortality.
12. Centers for Disease Control and Prevention (CDC). About the Health Aff (Millwood). 2019;38(11):1807-15. doi: 10.1377/
National Health and Nutrition Examination Survey. Atlanta: hlthaff.2019.00405.
CDC; 2022. Available from: https://www.cdc.gov/nchs/ 23. To QG, Stanton R, Schoeppe S, Doering T, Vandelanotte
nhanes/about_nhanes.htm. Accessed December 1, 2022. C. Differences in physical activity between weekdays and
13. Centers for Disease Control and Prevention (CDC). National weekend days among U.S. children and adults: cross-
Health and Nutrition Examination Survey, 2017-2018 Data sectional analysis of NHANES 2011-2014 data. Prev Med Rep.
Documentation, Codebook, and Frequencies: Demographic 2022;28:101892. doi: 10.1016/j.pmedr.2022.101892.
Variables and Sample Weights. Atlanta: CDC; 2022. Available 24. Navarro SM, Tsai MM, Ritchie LD, Frongillo EA, Laraia BA,
from: https://wwwn.cdc.gov/Nchs/Nhanes/2017-2018/ Pate RR, et al. Household food insecurity and children’s
DEMO_J.htm#INDHHIN2. Accessed December 1, 2022. physical activity and sedentary behaviour in the United
14. Centers for Disease Control and Prevention (CDC). National States: the Healthy Communities Study. Public Health Nutr.
Health and Nutrition Examination Survey, 2017-2018 Data 2022;25(2):381-8. doi: 10.1017/s1368980021002536.
Documentation, Codebook, and Frequencies: Food Security. 25. Katzmarzyk PT, Denstel KD, Beals K, Carlson J, Crouter SE,
Atlanta: CDC; 2022. Available from: https://wwwn.cdc.gov/ McKenzie TL, et al. Results from the United States 2018 report
Nchs/Nhanes/2017-2018/FSQ_J.htm. Accessed December 1, card on physical activity for children and youth. J Phys Act
2022. Health. 2018;15(S2):S422-4. doi: 10.1123/jpah.2018-0476.
15. Centers for Disease Control and Prevention (CDC). National 26. Höfler M. The Bradford Hill considerations on causality:
Health and Nutrition Examination Survey, Module 1: Datasets a counterfactual perspective. Emerg Themes Epidemiol.
and Documentation. Atlanta: CDC; 2022. Available from: 2005;2:11. doi: 10.1186/1742-7622-2-11.
https://wwwn.cdc.gov/nchs/nhanes/tutorials/module1.aspx. 27. Mande J, Flaherty G. Supplemental Nutrition Assistance
Accessed December 11, 2022. Program as a health intervention. Curr Opin Pediatr.
16. Centers for Disease Control and Prevention (CDC). How Much 2023;35(1):33-8. doi: 10.1097/mop.0000000000001192.
Physical Activity Do Children Need? Atlanta: CDC; 2022. 28. Foster EM, Jiang M, Gibson-Davis CM. The effect of
Available from: https://www.cdc.gov/physicalactivity/basics/ the WIC program on the health of newborns. Health
children/index.htm. Accessed December 10, 2022. Serv Res. 2010;45(4):1083-104. doi: 10.1111/j.1475-
17. Centers for Disease Control and Prevention (CDC). National 6773.2010.01115.x.
Health and Nutrition Examination Survey, 2017-2018 Data 29. Braga B, Blavin F, Gangopadhyaya A. The long-term effects of
Documentation, Codebook, and Frequencies: Physical Activity childhood exposure to the earned income tax credit on health
- Youth. Atlanta: CDC; 2022. Available from: https://wwwn. outcomes. J Public Econ. 2020;190:104249. doi: 10.1016/j.
cdc.gov/Nchs/Nhanes/2017-2018/PAQY_J.htm#PAQ706. jpubeco.2020.104249.
Accessed December 11, 2022.

8 J Res Health Sci, 2024, Volume 24, Issue 4