Cross-Sectional Evaluation of WIC Participation Status and Massachusetts Children’s Nutrition after the Cash-Value Benefit Increase

Abstract

Background

The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) increased the Cash-Value Benefit (CVB) for fruits and vegetables in 2021. Although WIC participants reported increased purchasing, consumption, and variety of fruits and vegetables after the CVB increase, it is unknown whether this resulted in higher fruit and vegetable intake relative to nonparticipants.

Objective

The objective of this study was to estimate differences in fruit and vegetable behaviors and total diet quality by WIC participation status in a convenience sample of children aged 1 to 4 years who participated in the Understanding Pandemic WIC-Waivers: Access, Redemption, Disparities (UPWARD) study.

Design

This was a cross-sectional study.

Participants/setting

From January 2024 to February 2025, caregivers of 280 Massachusetts children proxy-reported their child’s intake after the CVB increase via a frequency-based dietary screener and 24-hour recalls.

Main outcome measures

Primary outcomes were fruit and vegetable frequency of consumption (times per week) and mean daily intake (cup-equivalents). Secondary outcomes were diet quality scores using the Healthy Eating Index-2020 (HEI-2020) (range, 0–100 points).

Statistical analyses

Inverse probability weighted regression models evaluated the association between WIC participation status and dietary outcomes.

Results

WIC-participating children aged 1 to 4 years consumed fruits a mean (SE) of 4.0 (1.2) more times per week (P < .01) and consumed 0.6 (0.4) cups more fruits per day (P < .001) than nonparticipating children. WIC children also consumed vegetables a mean (SE) of 2.5 (0.9) more times per week (P < .01), although there was no difference in vegetable intake (mean [SE], −0.02 [0.3] cup-equivalents; P > .05), There were no differences in toddler or child diet quality by WIC participation status, although power was limited in these age-stratified models.

Conclusions

After the CVB increase, WIC-participating children consumed fruits and vegetables more often and consumed higher quantities of fruits than nonparticipants.

Fruits and vegetables are important FOR health promotion and chronic disease prevention across the lifespan.¹ However, most Americans do not meet federal dietary guidelines,² and many children aged 1 to 5 years do not eat fruits and vegetables daily; 1 in 2 do not eat a vegetable daily, and 1 in 3 do not eat a fruit daily.³ Income-driven disparities in fruit and vegetable intake are well documented,⁴ although income-based federal nutrition assistance programs have the potential to close this gap.^5,6 The Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) is a federal nutrition assistance program that provides food benefits, nutrition education, breastfeeding support, and referrals to social and medical services for infants, children, and pregnant and postpartum people who live in households with low income (≤185% poverty income ratio [PIR]) or adjunctively qualify via participation in other needs-based programs and are deemed at nutritional risk.⁷ WIC food benefits are administered via the WIC food package, which, based on expert recommendations,⁸ provides participants with specific quantities for 100% juice, milk and milk substitutes, cereal, whole-grain products, eggs, legumes, and canned fish, as well as a Cash-Value Benefit (CVB) that can be used to purchase fruits and vegetables.⁹

Fruits and vegetables have been a prominent focus of WIC research since the introduction of the CVB in the 2007 interim rule.¹⁰ It is well established that this policy change improved fruit and vegetable consumption among WIC participants,^11–13 although it is less clear whether it changed intake relative to eligible nonparticipants.¹³ After the introduction of the CVB, research from 2009 to 2014 showed no differences in fruit intake between WIC participants and nonparticipants^14,15 and no difference in total vegetable intake,¹⁴ although 1 study reported higher intake of the red/orange vegetable subgroup in WIC children than nonparticipating children.¹⁵ In addition, nationally representative data confirmed no differences in fruit intake, vegetable intake, or total diet quality by WIC participation status for children before the COVID-19 pandemic (2017 to March 2020).¹⁶ During the COVID-19 pandemic, the CVB for children was increased substantially, initially from $9 per month (prepandemic) to $35 per month (June 2021), and was later decreased to $24 per month (October 2021), where it has been sustained with annual inflation adjustments,^9,17 and was permanently codified as part of the 2024 WIC food package revisions.⁹ By effectively increasing the “dose” of fruit and vegetable benefits that WIC participants receive relative to prepandemic levels, the CVB change has the potential to substantially increase WIC children’s fruit and vegetable intake, although existing research is limited.

Existing research on the pandemic-era CVB change indicated that WIC participants were highly satisfied with the CVB increase.^18–25 WIC participants qualitatively perceived that the CVB increase contributed to positive fruit and vegetable behaviors such as increased purchasing, consumption, and variety,^18,19,21 and the change has also been linked to greater benefit redemption^24,26,27 and longitudinal improvements in food security.^24,25 However, there are limited and conflicting findings regarding the extent to which the pandemic-era CVB increase led to changes in fruit and vegetable intake, with 1 study reporting a slight decrease by 0.03 cups/day in total fruit and vegetable intake 1 year after the CVB increase,²⁴ and another study reported gains in a fruit and vegetable diversity score, which was ultimately not associated with fruit and vegetable intake.²⁶ Both studies lacked a comparison group, making it difficult to attribute findings to the WIC CVB increase as opposed to secular changes in intake that were occurring among the population at large during the pandemic due to supply chain disruptions.²⁸ Therefore, the objectives of this study were to evaluate differences in fruit and vegetable behaviors (ie, frequency and quantity) and to explore any related changes to total diet quality (eg, if fruits and vegetables displaced other unhealthy foods in the diet) between WIC children and a comparison group of WIC-eligible but nonparticipating children after the CVB change was made permanent in Massachusetts. It was hypothesized that fruit and vegetable intake behaviors would be higher among WIC participants than eligible nonparticipants, but little overall change in total diet quality was expected, given other recent reports of no differences in Massachusetts children’s total diet quality by WIC status.²⁹

METHODS

Data Collection

This cross-sectional study, after the CVB increase, collected dietary data from January 2024 to February 2025 among a convenience sample of WIC participants aged 1 to 4 years and eligible nonparticipants in Massachusetts, who were participating in the Understanding Pandemic WIC Waivers: Access, Redemption, Disparities (UPWARD) study. The UPWARD study was a mixed-methods evaluation of the COVID-19 pandemic WIC waivers on participation rates, redemption rates, and diet-related disparities.³⁰ Participants’ parents and guardians were recruited through flyers distributed to community sites (eg, WIC clinics, Head Start programs, grocery stores, and libraries), in-person recruitment events (eg, at WIC clinics), and targeted social media advertisements. Prospective study participants completed an online eligibility screener, which was reviewed using a comprehensive fraud detection system.³¹ Eligibility criteria included English or Spanish language capabilities, parent or guardian (at least 18 years old) of a child aged 1 to 4 years, eligible for or participating in WIC, lives in Massachusetts, and had access to a computer or smartphone. Eligible individuals were then enrolled in the study, provided electronic written informed consent, and completed an online demographic survey and a frequency-based dietary screener, as well as online 24-hour dietary recalls on behalf of their child. All study participants were invited to complete the dietary screener and 24-hour recall a second time. Participants received a $25 gift card per administration (ie, $25 for the first screener and recall, an additional $25 for the second screener and recall). All study procedures were approved by the Harvard T.H. Chan Institutional Review Board (IRB23–1484).

Analytic Sample

Using prior estimates of the SD of fruit and vegetable intake in WIC participants and eligible nonparticipants,³² 80% statistical power was estimated to detect an effect size of 0.23 cups of fruits and vegetables for 150 children in each group at a double-sided α of .05. In total, 5458 individuals completed online eligibility screening, of which 711 were deemed eligible and contacted with study details (Figure 1). Most were subsequently lost to follow-up or declined to participate (n = 313), although some (n = 107) were later deemed ineligible (eg, due to online survey fraud, misreported eligibility, or duplicate survey attempts). Ultimately, 291 respondents completed all study procedures, including a demographic survey and at least 1 set of frequency- and intake-based dietary assessments. After exclusion of participants with incomplete dietary recalls (n = 1) and implausibly low (<500 kcal, n = 9) or high (>5000 kcal, n = 1) energy intakes,³³ the final analytic sample was 280 children aged 1 to 4 years (Figure 1).

Figure 1.

Flowchart of participant selection for a cross-sectional study of Special Supplemental Nutrition Program for Women, Infants, and Children (WIC)-enrolled children aged 1—4 years, showing recruitment, eligibility assessment, survey and dietary recall completion, exclusions, and derivation of the final analytic sample used to evaluate dietary intake and diet quality in relation to changes in the WIC fruit and vegetable cash value benefit.

Outcome Assessment

Dietary data were collected using both a frequency-based dietary screener (the rapid Prime Diet Quality Screener Score for Kids [rPDQS-Kids]) adapted from the rPDQS for adults,³⁴ and 24-hour recalls collected via the Automated Self-Administered 24-Hour Dietary Assessment Tool (ASA24),³⁵ described in more detail below. Children’s dietary data were proxy-reported by a parent or guardian.³⁶ Although proxy-reporting situations may be limited in accuracy due to shared feeding responsibility of the child (eg, daycare provider, nanny, and parent),³⁷ it is expected that such measurement error nondifferentially affects all children in the sample, regardless of WIC participation status, thereby not biasing overall conclusions. In addition, there are known challenges to using ASA24 in low-income populations like WIC, such as usability issues³⁸ and overestimation of portion size.³⁹ However, the provision of assistance (ie, interview administration rather than self-administration) does not substantially impact the accuracy of dietary recalls,⁴⁰ so the ASA24 remained self-administered in this study.

Fruits and Vegetable Intake Behaviors.

The primary study outcome was fruit and vegetable behaviors, including weekly consumption frequency (times per week from rPDQS-Kids) and mean daily quantities (cup-equivalents from ASA24). Both fruit and vegetable assessments excluded juice because the WIC CVB cannot be used to purchase juice. Vegetables excluded potatoes because, unlike other vegetables, the Massachusetts WIC program prohibits the use of the CVB to purchase canned or frozen varieties of potatoes.⁴¹ In addition, the validated rPDQS and the rPDQS-Kids frequency screeners define vegetables as excluding potatoes,^34,42 so potatoes were excluded from the total vegetable mean intake to uphold a consistent definition of vegetables.

Frequency data were collected using the rPDQS-Kids, a 16-item frequency-based dietary screener comprised of 7 healthy components (ie, fish; whole grains; legumes; vegetables, excluding potatoes; fruits; nuts; and water), 8 unhealthy components (ie, processed meats, beef, full-fat dairy, fast food, refined grains, sweets, 100% juice, and sugar-sweetened beverages), and a neutral unscored component (ie, fluid milk of any fat percentage). The rPDQS-Kids can be used to assess estimates of weekly consumption frequencies, or scoring standards can be applied to calculate a frequency-based diet quality score. Scoring standards are available in Figure 2.

Figure 2.

Traffic light scoring standards of the Rapid Prime Diet Quality Score Screener-Kids diet quality score. Total scores range from 0 to 30 points; higher scores reflect higher diet quality. Scores are calculated by assigning 0 points for red, 1 for yellow, and 2 for green frequencies of food group consumption. ^aIndicates inclusion in the 2024 Special Supplemental Nutrition Program for Women, Infants, and Children food packages for children aged 12 to 23 months or 2 to 4 years. Exceptions: fish: includes canned varieties only; dairy: 12- to 23-month-old package includes full-fat varieties and 2- to 4-year-old package includes low-fat or nonfat varieties; legumes: includes 1 lb dry or 64 oz canned legumes or 18 oz peanut butter; vegetables: white potatoes included. ^bNot scored.

Mean intake data were proxy-reported by a parent or caregiver using ASA24’s 24-hour recalls, which uses a modified version of the US Department of Agriculture Automated Multiple Pass Method.⁴³ An optional second dietary recall was administered at least 2 weeks later and was completed by 73.9% of study participants (WIC: n = 177 [73.8%]; nonparticipants: n = 30 [75%]). The mean of 2 recalls was used when available. Each food item reported in the ASA24 is linked to the US Department of Agriculture’s Food Patterns Equivalents Database,⁴⁴ which converts reported food and beverage items into disaggregated US Department of Agriculture food pattern components (eg, cup-equivalents of fruits, ounce-equivalents of whole grains, and teaspoon-equivalents of added sugars), thus enabling calculation of mean fruit and vegetable quantity (cupequivalents) and diet quality scores.

Diet Quality Scores.

In an exploratory analysis, diet quality scores were calculated using the Healthy Eating Index (HEI). The HEI measures alignment with the 2020–2025 Dietary Guidelines for Americans (DGA). Like the DGA, the HEI is updated every 5 years to incorporate the latest scientific evidence.^45,46 The most recent version of the HEI is the 2020 version, including the HEI-2020 for 2 years of age and older⁴⁶ and the HEI-Toddlers-2020 for ages 12 to 23 months old.^47–49

Both the HEI-2020 and HEI-Toddlers-2020 include the same dietary components and are scored from 0 to 100 points; a score of 100 indicates a diet that is optimally aligned with the DGA. The following 9 dietary components focus on adequacy, that is, foods to eat more of for good health: total fruits, whole fruits, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafood and plant proteins, and fatty acids. High HEI adequacy component scores reflect high dietary intakes. The following 4 dietary components focus on moderation, that is, foods to limit intake of for good health: refined grains, sodium, saturated fats, and added sugars. High moderation component scores reflect low dietary intakes. The tandem indices for toddlers and for ages 2 years and older include the same dietary components on the same scales to allow for bridging of components and to support examination of healthy eating across the lifespan.⁴⁷ However, the scoring standards differ, reflecting distinct dietary guidance for toddlers relative to children and adults. In this study, the HEI-2020 and HEI-Toddlers-2020 scores were constructed using the simple scoring algorithm, which is the available method for use in regression analyses.^50,51 Sensitivity analyses were performed for children aged 2 to 4 years to also assess diet quality using the rPDQS-Kids total diet quality score. Evaluation of diet quality scores was an exploratory secondary outcome.

Exposure Assessment

The exposure—binary WIC participation status—was self-reported and defined as responding affirmatively to a question about whether the child is currently receiving WIC benefits, akin to WIC exposure assessment in large national surveys like the National Health and Nutrition Examination Survey and the Pregnancy Risk Assessment Monitoring System.^52,53 Self-reported indicators for WIC participation status in such surveys generally have high reliability (eg, Pregnancy Risk Assessment Monitoring System κ = .81) and validity (eg, Pregnancy Risk Assessment Monitoring System sensitivity of 90.8%, specificity of 90.6%) for retrospective self-report in comparison with WIC status on birth certificates.⁵³ Eligible nonparticipants were defined as children not currently receiving WIC benefits who lived in households that self-reported eligibility in the past 30 days for WIC. Eligibility was determined using income-eligibility (ie, ≤185% PIR) and adjunctive eligibility for WIC via participation in other needs-based programs such as Supplemental Nutrition Assistance Program, Head Start, Massachusetts Transitional Aid to Families with Dependent Children, MassHealth/Medicaid, or Section 8 Housing Choice Voucher Program.^54–58

Statistical Analysis

Descriptive statistics were used to characterize the sample. To empirically test for differences in dietary outcomes between WIC participating and nonparticipating children, an inverse probability weighted (IPW) linear regression model was used, which is useful in addressing nonexchangeability between WIC participating and nonparticipating groups. The IPWs were constructed using information on child-level (ie, age, gender, race, and ethnicity), parent-level (ie, education and employment), and household-level (ie, Supplemental Nutrition Assistance Program participation in the past 30 days and food security assessed via the 2-item Hunger Vital Sign⁵⁹) variables, consistent with prior WIC literature that used IPW models.⁶⁰ Household-clustered robust variance estimators (95% CI and SEs) were used to account for within-household correlation among children, as up to 2 children per household could participate in the study. All analyses were performed in R software, version 4.2.1⁶¹ with a 2-sided α of .05.

RESULTS

Sample Characteristics

There were 280 Massachusetts children aged 1 to 4 years with at least 1 complete set of dietary assessments, unevenly distributed between WIC participants (n = 240 children from 192 unique households) and nonparticipants (n = 40 children from 27 households) (Table 1). Compared with nonparticipating children, WIC children were less likely to identify as Black race (20.8% vs 40.0%) and non-Hispanic ethnicity (49.6% vs 67.5%), and more likely to identify as White (61.3% vs 55.0%). Both child and parent or guardian age and gender were comparable between groups. Parents of WIC children had a lower prevalence of employment (55.8% vs 70.3% employed either part- or full-time) and higher rates of unemployment while looking for work (17.2% vs 7.4%) relative to parents of nonparticipating children. Both WIC-participating and nonparticipating children lived in households with low incomes and were income-eligible or adjunctively eligible for WIC, and more than one-half of the households in both groups experienced a high prevalence of food insecurity (WIC: 59.4%, nonparticipants: 51.9%). However, WIC participants were more likely than nonparticipants to be simultaneously enrolled in other needs-based programs, especially the Supplemental Nutrition Assistance Program (72.9% vs 33.3%) and Medicaid (74.0% vs 29.6%) (Table 1).

Table 1.

Characteristics of Massachusetts children aged 1 to 4 years in the UPWARD^a study (n = 280)^b

Characteristic	WICc participants	Nonparticipants	Overall
Child-level	240	40	280
Gender
Female	124 (51.7)	19 (47.5)	143 (51.1)
Male	113 (47.1)	21 (52.5)	134 (47.9)
Other	3 (1.3)	0 (0)	3 (1.1)
Age, y	2.35 (1.05)	2.50 (1.22)	2.38 (1.07)
Age group
Toddlers 1 to <2 y	62 (25.8)	12 (30)	74 (26.4)
Children 2–4 y	178 (74.2)	28 (70)	206 (73.6)
Race
Asian	10 (4.1)	1 (2.5)	11 (3.9)
Black	50 (20.8)	16 (40.0)	66 (23.6)
Other	33 (13.8)	1 (2.5)	34 (12.1)
White	147 (61.3)	22 (55.0)	169 (60.4)
Ethnicity
Mexican	5 (2.1)	0 (0)	5 (1.8)
Non-Hispanic	119 (49.6)	27 (67.5)	146 (52.1)
Other Hispanic	37 (15.4)	3 (7.5)	40 (14.4)
Puerto Rican	79 (32.9)	10 (25.0)	89 (31.8)
Parent or guardian-level	192	27	219
Gender
Female	188 (97.9)	25 (92.6)	213 (97.3)
Male	2 (1.0)	2 (7.4)	4 (1.8)
Other	2 (1.0)	0 (0)	2 (0.9)
Age group
18–30 y	74 (38.5)	12 (44.4)	86 (39.3)
31–45 y	109 (56.8)	13 (48.1)	122 (55.7)
46–60 y	7 (3.6)	2 (7.4)	9 (4.1)
60+ y	2 (1.0)	0 (0)	2 (0.9)
Highest education
Less than a high school diploma	7 (3.6)	1 (3.7)	8 (3.7)
High school degree or equivalent (eg, GEDd)	50 (26.0)	5 (18.5)	55 (25.1)
Some college, no degree	64 (33.3)	10 (37.0)	74 (33.8)
Associate degree (AA, AS)	25 (13.0)	1 (3.7)	26 (11.9)
Bachelor's degree (BA, BS)	31 (16.1)	8 (29.6)	39 (17.8)
Graduate degree (MS, MPH)	15 (7.8)	2 (7.4)	17 (7.8)
Employment status
Employed and working full-time (35+ h/wk)	51 (26.6)	10 (37.0)	61 (27.9)
Employed and working part-time (1–34 h/wk)	56 (29.2)	9 (33.3)	65 (29.7)
Not currently employed, and not looking for work	32 (16.7)	1 (3.7)	33 (15.1)
Not currently employed, but looking for work	33 (17.2)	2 (7.4)	35 (16.0)
Other	10 (5.2)	2 (7.4)	12 (5.5)
Prefer not to say	10 (5.2)	3 (11.1)	13 (5.9)
Household-level	192	27	219
Food-security statuse
Food insecure	114 (59.4)	14 (51.9)	128 (58.4)
Food secure	78 (40.6)	13 (48.1)	91 (41.6)
SNAPf use in past 30 d
No	52 (27.1)	18 (66.7)	70 (32.0)
Yes	140 (72.9)	9 (33.3)	149 (68.0)
TAFDCg use in past 30 d
No	150 (78.1)	25 (92.6)	175 (79.9)
Yes	42 (21.9)	2 (7.4)	44 (20.1)
Medicaid use in past 30 d
No	50 (26.0)	19 (70.4)	69 (31.5)
Yes	142 (74.0)	8 (29.6)	150 (68.5)
Section 8 housing voucher in past 30 d
No	159 (82.8)	24 (88.9)	183 (83.6)
Yes	33 (17.2)	3 (11.1)	36 (16.4)

^aUPWARD = Understanding Pandemic WIC Waivers: Access, Redemption, Disparities Study.

^bUPWARD study eligibility criteria permitted up to 2 children per household to participate in the study, resulting in 280 unique children from 219 households/caregivers.

^cWIC = Special Supplemental Nutrition Program for Women, Infants, and Children.

^dGED = General Educational Development credential.

^eRepresents food-security status over the past 30 days.

^fSNAP = Supplemental Nutrition Assistance Program.

^gTAFDC = Transitional Assistance for Families and Dependent Children.⁵⁶

Fruit and Vegetable Outcomes

Crude estimates showed that WIC participants consumed fruit 9.0 (95% CI, 8.3 to 9.6) times per week, whereas non-participants consumed fruit 6.7 (95% CI, 5.2 to 8.2) times per week (Table 2). In IPW-adjusted regression models, WIC children consumed fruit 8.9 (95% CI, 8.2 to 9.6) times per week, whereas nonparticipants consumed fruit 4.9 (95% CI, 2.6 to 7.2) times per week, reflecting an adjusted mean (SE) difference of 4.0 (1.2) times per week (P < .01). Moreover, WIC children consumed higher quantities of fruits (mean [SE], 0.6 [0.4] cup-equivalents more per day; P < .001). Despite no significant differences in vegetable mean intake, WIC children consumed vegetables more frequently (mean [SE], 2.5 [0.9] more times per week; P < .01) (Table 2, Figure 3).

Table 2.

Mean intakes and mean differences in fruit and vegetable frequency and intake of Massachusetts children aged 1 to 4 years (n = 280) by WIC^a participation status^b

	Unadjusted Estimates			Adjusted Estimates
	Crude Intakes		Unadjusted difference, mean (SEc)	Adjusted Intakes		IPWd-adjusted difference,e mean (SEc)
Variable	WIC participants (n = 240)	Nonparticipants (n = 40)		WIC participants (n = 240)	Nonparticipants (n = 40)
CVBf outcomes
Fruit frequency (times per week)gh	9.0 (8.3 to 9.6)	6.7 (5.2 to 8.2)	2.3 (0.8)**	8.9 (8.2 to 9.6)	4.9 (2.6 to 7.2)	4.0 (1.2)**
Fruit quantity (c-eqi)g	1.3 (1.1 to 1.4)	1.0 (0.6 to 1.4)	0.2 (0.2)	1.3 (1.1 to 1.4)	0.7 (−0.02 to 14)	0.6 (0.4)***
Vegetable frequency (times per week)hj	5.8 (5.1 to 6.4)	5.1 (3.7 to 6.6)	0.6 (0.8)	5.7 (5.1 to 6.4)	3.3 (1.5 to 5.0)	2.5 (0.9)**
Vegetable quantity (c-eq)hj	0.6 (0.6 to 0.7)	0.9 (0.7 to 1.1)	−0.2 (0.1)	0.7 (0.6 to 0.8)	0.7 (0.09 to 1.3)	−0.02 (0.3)
Non-CVB outcomes
Juice frequency (times per week)h	5.2 (4.5 to 5.9)	2.3 (1.6 to 3.1)	2.8 (0.5)***	4.9 (4.3 to 5.6)	3.6 (1.8 to 5.3)	1.4 (1.0)
Juice quantity (c-eq)	0.5 (0.4 to 0.6)	0.4 (0.2 to 0.5)	0.2 (0.1)*	0.5 (0.4 to 0.6)	0.4 (0.01 to 0.7)	0.2 (0.2)

^aWIC = Special Supplemental Nutrition Program for Women, Infants, and Children.

^bMean intake values were derived from Automated Self-Administered 24-Hour Dietary Assessment Tool³⁵ 24-hour recall. Two recalls were available for 73.9% of the sample; the mean of 2 recalls was used when available. Frequency values were derived from 1 administration of the Rapid Prime Diet Quality Screener Score for Kids (rPDQS-Kids)³⁴ from an overlapping time period.

^cSEs and 95% CIs use household-clustered robust variance estimators to account for clustering at the household level. A maximum of 2 children per household were permitted to participate in the study.

^dIPW = inverse probability weighted.

^eAdjusted estimates using IPWs that included child-level (ie, age, gender, race, and ethnicity), parent-level (ie, education and employment), and household-level (ie, Supplemental Nutrition Assistance Program and food security) variables. Results reflect estimates for WIC children in comparison with nonparticipants (reference category).

^fCVB = Cash-Value Benefit.

^gFruit was defined as all fruits excluding juice.

^hFrequency data were collected on the rPDQS-Kids and converted to weekly frequencies using the following scoring standards: 0 = fewer than once per week or never; 1 = once per week; 3 = 2 to 4 times per week; 7 = nearly daily or daily; 14 = twice per day or more.

ⁱc-eq = cup-equivalents.

^jVegetable frequency on the rPDQS-Kids is defined as total vegetables excluding potatoes. For consistency, vegetable quantity was operationalized as the sum of all vegetables excluding potatoes. Vegetables do not include vegetable juice.

^*P < .05.

^**P < .01.

^***P < .001.

Figure 3.

Crude mean fruit and vegetable intakes of Massachusetts children aged 1 to 4 years (n = 280) with error bars by Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) participation status. Mean quantity values derived from the Automated Self-Administered 24-Hour Dietary Assessment Tool³⁵ 24-hour recall; frequency values derived from the rapid Prime Diet Quality Screener Score for Kids.

Diet Quality Exploratory Analysis

The mean unadjusted total diet quality score for children aged 2 to 4 years who participated in WIC was 61.1 (95% CI, 59.4 to 62.9) points out of 100 on the HEI-2020 compared with 60.4 (95% CI, 55.6 to 65.3) points for nonparticipating children. In IPW-adjusted regression models, there were no significant differences in diet quality by WIC participation status on either the HEI-2020 or rPDQS-Kids (Table 3). Statistical comparisons of toddler diet quality were not possible due to the small toddler sample size (n = 62 WIC participants, n = 12 nonparticipants), yielding unstable estimates.

Table 3.

Diet quality scores by WIC^a participation status in a convenience sample of Massachusetts children aged 2–4 years (n = 206)

Variable	Unadjusted Estimates			Adjusted Estimates
	Crude Intakes		Unadjusted differences mean (SEc)	Adjusted Intakes		IPWc -adjusted differences mean (SEb)d
	WIC participants (n = 178)	Nonparticipants (n = 28)		WIC Participants (n = 178)	Nonparticipants (n = 28)
HEI-2020 e
Total score (0–100 points)	61.1 (59.4–62.9)	60.4 (55.6–65.3)	0.7 (2.6)	61.2 (59.5–63.0)	55.5 (46.3–64.8)	5.7 (4.8)
Adequacy (0–60 points)	33.9 (32.8–35.1)	33.4 (30.0–36.8)	0.5 (1.8)	34.1 (33.0–35.3)	29.0 (22.8–35.2)	5.1 (3.2)
Moderation (0–40 points)	27.2 (26.3–28.1)	27.1 (25.1–29.0)	0.1 (1.1)	27.1 (26.2–28.0)	26.6 (22.9–30.2)	0.5 (1.9)
rPDQS-Kids f
Total score (0–30 points)	16.2 (15.7–16.8)	17.2 (15.6–18.8)	−1.0 (0.8)	16.3 (15.6–16.9)	15.0 (13.1–16.8)	1.3 (1.0)
Healthy (0–14 points)	6.9 (6.5–7.3)	7.0 (5.9–8.1)	−0.1 (0.6)	6.9 (6.4–7.3)	6.3 (5.1–7.5)	0.6 (0.6)
Unhealthy (0–16 points)	9.4 (9.0–9.8)	10.2 (9.2–11.2)	−0.8 (0.5)	9.4 (9.0–9.9)	8.7 (7.6–9.9)	0.7 (0.6)

^aWIC = Special Supplemental Nutrition Program for Women, Infants, and Children.

^bSEs and 95% CI use household-clustered robust variance estimators to account for clustering at the household level. A maximum of 2 children per household were permitted to participate in the study.

^cIPW = inverse probability weight.

^dAdjusted estimates using IPWs that included child-level (ie, age, gender, race, and ethnicity), parent-level (ie, education and employment), and household-level (ie, Supplemental Nutrition Assistance Program and food security) variables. Results reflect estimates for WIC children in comparison with nonparticipants (reference category).

^eHEI-2020 = Healthy Eating Index 2020.⁴⁶

^frPDQS-Kids = Rapid Prime Diet Quality Screener Score for Kids.³⁴

DISCUSSION

In this cross-sectional convenience sample of Massachusetts children aged 1 to 4 years after the CVB increase was made permanent, WIC participants consumed fruits more frequently and in greater quantities than nonparticipants, and consumed vegetables more often, although vegetable quantity was comparable between the 2 groups. These findings differ from prepandemic studies, which found no difference in fruit and vegetable intake between WIC participants and nonparticipants before the CVB change was implemented.^14–16 In exploratory analyses, total diet quality among preschool-aged children (2–4 years) was comparable between WIC participants and nonparticipants, suggesting that despite variations in fruit and vegetable frequency and consumption patterns, total diet quality was largely similar, although the study was underpowered to definitively make this claim.

Although evaluations of the pandemic-era WIC CVB change are largely limited to samples of WIC participants that lacked a comparison group,^{18–21,23,26,27} nationally representative research with nonparticipant comparison groups that was conducted before the COVID-19 pandemic (ie, before the large increase in CVB), found no differences in fruit intake between WIC participants and non-participants^14–16 and no difference in total vegetable intake^14,16; however, 1 study reported higher intake of the red and orange vegetable subgroup in WIC children than nonparticipating children.¹⁵ In contrast to prior largely null nationally representative prepandemic findings, our study found that WIC children in Massachusetts consumed fruits 4 more times per week and averaged 0.6 cup more fruits per day, in addition to consuming vegetables 2.5 more times per week, compared with nonparticipating children. These significant findings indicate a marked departure from national null prepandemic trends,¹⁶ suggesting potential shifts in fruit and vegetable behaviors among WIC families after the CVB increase. Because higher fruit and vegetable intake in early childhood is associated with improved overall diet quality and reduced long-term cardiometabolic risk,^62,63 the observed differences underscore the potential for WIC to serve as a meaningful policy lever to narrow socioeconomic disparities in fruit and vegetable consumption among young children.^25,64

Beyond whole fruits and vegetables, WIC participants and nonparticipants had no significant differences in juice intake; WIC participants in this sample consumed 0.2 cups per day more than nonparticipants, closely aligning with national prepandemic estimates of a 0.22-cup difference.¹⁴ The role of juice in children’s diets remains contested, as reflected in the differential treatment of juice by dietary assessment tools. On diet quality indices, juice is an “unhealthy” dietary component on the rPDQS-Kids tool, yet counts in the Total Fruits component (an adequacy or “healthy” component) on the HEI. The inclusion of juice in the WIC food package has a historied debate. Proponents argue that juice consumption helps children meet daily fruit intake goals and improves total diet quality.⁶⁵ Critics, on the other hand, call for reallocating funds from juice benefits to the CVB—which is a state option in the 2024 food package update—citing concerns that WIC participants are already more likely to exceed recommended juice levels.¹⁴ Over time, the WIC program has progressively reduced juice allocations, initially in the 2007 interim rule (implemented in the 2009 food package revisions),¹⁰ and more recently in the 2024 WIC food package update, which was pending implementation in Massachusetts at the time of data collection; all state agencies must fully implement this change by April 2026.⁹ Further research is needed to evaluate the health effects and cost-effectiveness of additional CVB substitutions for juice in the WIC food package, and any related implications for children’s diet quality.

Children’s diet quality is highest at the youngest life stages, and diet quality scores decline as children age.⁴⁵ Unadjusted diet quality for toddlers (1 to <2 years old) was lower in this sample (WIC: 62.6 points, nonparticipants: 61.3 points via simple scoring) than national averages (67.7 points via population ratio method⁶⁶ and 63.0 points via Multivariate Chain Monte Carlo method⁴⁸). Among children aged 2 to 4 years, unadjusted total diet quality ranged from 61.1 points (nonparticipants) to 60.4 points (WIC participants), which were comparable with the national average of 61 points.⁴⁵ Discrepancies between this small Massachusetts-based convenience sample and nationally representative samples may be attributed to differences in sampling schemes, age composition, or the HEI version or calculation method. In this study, HEI diet quality scores were calculated using the simple scoring method, the available method for use in regression models,⁵⁰ which was consistent with our aim to use IPW regression models to assess between-group differences by WIC participation status. Descriptive national surveillance estimates, like those mentioned above, are often generated with the population ratio method,⁵⁰ which tends to generate higher diet quality scores.⁶⁶

When comparing diet quality by WIC participation status, IPW-adjusted models were used, and found no significant differences between WIC participating and nonparticipating children aged 2 to 4 years old, which may reflect either lack of power or a true null. Nonetheless, it is consistent with null findings from prepandemic national estimates,¹⁶ and post-pandemic local research in Greater Boston, which found no differences among 69 mother—child dyads.²⁹ In addition, null findings have been observed in other WIC populations, like nationally representative samples of pregnant and postpartum individuals whose diet quality did not differ by WIC participation status.⁶⁷

Findings from the present study should be interpreted in light of several considerations. Although prior studies were conducted among WIC participants exclusively,^{18–21,23,26,27} this study included a comparison group of eligible non-participants. In research studies, WIC eligibility is often defined by income eligibility alone (ie, ≤185% PIR),^15,68,69 yet nearly three-quarters of all WIC participants initially enroll through adjunctive eligibility due to participation in other need-based programs.⁷⁰ In states with expanded Medicaid coverage like Massachusetts, income-centric definitions may misclassify eligible respondents. In this study, the inclusion of adjunctively eligible respondents in the eligible nonparticipation group improved classification accuracy and is a study strength. Nonetheless, direct comparisons of WIC participants and eligible nonparticipants remain challenging due to the lack of ethical randomization options, necessitating observational study designs like this one with inherent limitations in exchangeability. Lack of exchangeability has limited prior research on WIC’s impact, largely due to higher food insecurity among those who opt into WIC relative to eligible nonparticipants,⁷¹ but the use of IPWs in the present study enhanced comparability between these otherwise nonexchangeable groups. However, IPWs cannot fully account for nonexchangeability due to unobservable traits, like child care attendance, which can meaningfully affect dietary intake.⁷² The cross-sectional study design precludes causal inferences, and the small (n = 280) Massachusetts-based convenience sample may not be generalizable to other geographic locations or nationally-representative studies. In addition, although the rPDQS-Kids has undergone evaluation for validity and reliability, the results have not yet been published, but it was adapted from the validated rPDQS for adults.³⁴

Recruitment of eligible nonparticipants was challenging, resulting in a lower sample of nonparticipants than planned. Despite being underpowered in the nonparticipant group, this study still had a sample size comparable with some national studies¹⁶ and detected several important between-group differences in primary outcomes. However, null findings in exploratory analyses of diet quality are difficult to interpret and may also be an artifact of limited statistical power. In addition, the National Cancer Institute recommends the Multivariate Chain Monte Carlo method for regression analyses, but simple scoring was used due to implementation challenges (eg, computational intensity and the lack of publicly available SAS code for implementing it in this context).^50,73–75 In addition, the ASA24 and its associated food composition databases are not exhaustive, so reporting of ethnic foods may be limited.

Future research should further examine underlying behavioral decisions that may contribute to differences in fruit and vegetable outcomes and confirm whether the present findings observed in Massachusetts hold in other geographic locations. In addition, future research could consider how benefit redemption patterns may moderate the relationships noted herein for WIC participation and dietary outcomes. As WIC continues to evolve, understanding its role in shaping child nutrition remains critical for informing program improvements and ensuring that benefits reach those most in need.

CONCLUSIONS

In a cross-sectional convenience sample of 280 Massachusetts children after a policy change that more than doubled the cash value of WIC’s CVB for fruits and vegetables, WIC participants consumed fruits more often and in higher quantities than nonparticipants, and consumed vegetables more often, although mean vegetable intake was comparable. Taken together with other research linking the CVB increase to increased purchasing, consumption, and variety of fruits and vegetables,^18,19,21 higher benefit redemption,^24,26,27 and longitudinal improvements in food security,^24,25 there is supportive evidence for nutritional benefits at the new $24 per month level. Future research should continue to explore how to fully maximize WIC’s potential for positively impacting dietary intake among children with low incomes.

RESEARCH SNAPSHOT.

Research Question:

After the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) increased the cash value benefit for fruits and vegetables, what was the difference in fruit and vegetable intake and diet quality between WIC participants and eligible nonparticipants?

Key Findings:

In this cross-sectional convenience sample of 280 Massachusetts children aged 1 to 4 years, WIC-participating children consumed fruits and vegetables more often and consumed higher quantities of fruits than nonparticipants.

DATA AVAILABILITY

Data described in the manuscript will be made available upon reasonable request.