Trust in federal statistical agencies and vaccine uptake among U.S. adults Authors: Nathaniel J Maxey, Jemar R Bather, Taliyah S Griffin, Plamena P Powla, Felix M Pabon-Rodriguez Published: 2026 ABSTRACT AbstractBackground:In the United States (US), COVID-19 vaccination saved 3.2 million lives from 2020 to 2023, yet vaccine uptake remains a public health challenge. Seasonal influenza vaccination also remains an important public health priority because influenza contributes substantially to preventable morbidity and mortality each year. Examining the administration of COVID-19 and influenza vaccines provides an opportunity to understand the factors influencing vaccine uptake.Methods:Repeated cross-sectional data were obtained from the Household Pulse Survey. COVID-19 and influenza vaccine uptake were based on self-reported vaccination status since August 2024. Independent variables of interest were (1) trust in federal statistics, (2) confidence in the US Census Bureau, and (3) confidence in US statistical agencies. Weighted logistic regression quantified the association between trust measures and non-vaccine uptake, adjusting for sociodemographic factors.Results:Compared to those who trusted federal statistics, individuals who did not trust federal statistics had significantly higher odds of COVID-19 non-vaccine uptake (adjusted odds ratio [aOR]: 2.61, 95% CI: 2.05–3.32) and influenza non-vaccine uptake (aOR: 1.87, 95% CI: 1.58–2.21). Similarly, individuals with very little confidence in the U.S. Census Bureau (COVID-19: aOR: 2.16, 95% CI: 1.16–4.02; influenza: aOR: 2.67, 95% CI: 1.74–4.10) or U.S. statistical agencies (COVID-19: aOR: 2.28, 95% CI: 1.52–3.43; influenza: aOR: 1.86, 95% CI: 1.35–2.57) exhibited a greater likelihood of non-vaccine uptake.Conclusion:Distrust in federal statistical agencies was associated with lower COVID-19 and influenza vaccine uptake, highlighting the importance of institutional trust in vaccination behavior. Unlike prior studies focused on trust in healthcare systems or government broadly, this study specifically identifies trust in federal statistical agencies as a factor associated with vaccine decision-making. Future longitudinal and mixed-methods research should examine how institutional trust develops over time and whether communication and transparency-focused interventions can improve vaccine uptake and public confidence in health information.Keywords: Vaccine uptake, COVID-19, Vaccination, Influenza vaccination, Weighted logistic regression, Survey Introduction In the United States (US), the COVID-19 vaccination saved an estimated 3.2 million lives between December 2020 and November 2022 [1]. Despite this success, vaccine uptake remains an ongoing public health challenge, with an estimated 240,797 COVID-19 deaths between December 2020 and June 2021 potentially preventable through vaccination [2–4]. By December 2021, one year after COVID-19 vaccines became available, only 63% of the US population had completed the primary series of an approved vaccine [5]. While COVID-19 vaccination rates showed some progress, similar challenges in vaccine uptake persist with other preventable diseases. Among adults 18 years and older, influenza vaccination coverage during the 2023–24 season was approximately 45%, compared with approximately 47% during the 2022–23 influenza season [6]. With this vaccination disparity, the success of federal, state, and local efforts to curb hesitancy largely depends on public trust [7]. Such trust is vital for encouraging the public to adopt protective measures that limit disease transmission and mitigate negative health impacts [7]. However, sociodemographic factors such as age, race, and educational attainment significantly affect the trust required for vaccine uptake [8–12]. Racial/ethnic differences are evident, with African American and Hispanic populations demonstrating higher levels of vaccine hesitancy compared to their White counterparts [11,12]. This hesitancy often reflects distrust in government, shaped by historical and ongoing discrimination [8–14]. One potential explanation for these findings is that federal statistical agencies play a central role in generating and communicating population health information used to guide public health recommendations [7,11]. Distrust in these institutions may therefore reduce confidence in the validity of vaccine-related data, perceived disease risk, or the credibility of public health messaging more broadly [7,11]. This may be associated with how individuals interpret vaccination recommendations and engage with preventive health interventions [13,14]. Given widespread distrust in federal statistics, further research is needed on its link to vaccine uptake [13–15]. Using data from the Household Pulse Survey, we investigated the relationships of COVID-19 and influenza vaccine uptake with trust in federal statistics, confidence in the US Census Bureau, and confidence in US statistical agencies, controlling for sociodemographic factors. We hypothesized lower trust in federal statistical agencies would be significantly associated with increased odds of non-vaccine uptake. Methods Study design and setting We used cross-sectional data from three waves of the Household Pulse Survey, a national survey conducted by the US Census Bureau collected on October 15–29, 2024; December 17, 2024–January 3, 2025; and February 21–March 7, 2025 [16]. This 20-min survey included various topics, including mental health, social isolation, and socio-demographic factors. Households were randomly selected from the Master Address File, and one individual from each household was invited to participate. As with many large national online surveys, participation in the Household Pulse Survey was voluntary, which may introduce the potential for nonresponse bias despite the application of survey weighting procedures. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology guidelines [17]. Analytic sample Between October 15 and October 29, 2024, 9355 individuals participated in the Household Pulse Survey. Of these, 1.5% (n = 144) were excluded due to missing information on COVID-19 and influenza vaccination status, 1.2% (n = 111) due to unavailable trust or confidence measures, and 0.8% (n = 78) due to missing covariate data (marital status: n = 37, number of children: n = 34, and employment status: n = 7). These exclusions resulted in a final unweighted sample of 9022 participants. Between December 17, 2024, and January 3, 2025, 9404 individuals participated in the Household Pulse Survey. Of these, 1.9% (n = 178) were excluded due to missing information on COVID-19 and influenza vaccination status, 1.2% (n = 115) due to unavailable trust or confidence measures, and 0.9% (n = 77) due to missing covariate data (marital status: n = 38, number of children: n = 24, and employment status: n = 15). These exclusions resulted in a final unweighted sample of 9034 participants. Between February 21, 2025, and March 7, 2025, 6740 individuals participated in the Household Pulse Survey. Of these, 1.8% (n = 119) were excluded due to missing information on COVID-19 and influenza vaccination status, 1.3% (n = 90) due to unavailable trust or confidence measures, and 1.0% (n = 68) due to missing covariate data (marital status: n = 26, number of children: n = 36, and employment status: n = 6). These exclusions resulted in a final unweighted sample of 6463 participants. COVID-19 and influenza vaccine uptake We assessed COVID-19 and influenza vaccination uptake using one item: Have you received the following vaccines this fall (i.e., since August 2024)? This item was asked separately for COVID-19 and influenza vaccination. We derived each outcome as a binary variable, denoting ‘1’ for not receiving the vaccine and ‘0’ for receiving the vaccine. Trust in federal statistical agencies The Household Pulse Survey included questions about trust in federal statistics, confidence in the US Census Bureau, and confidence in US statistical agencies. The Household Pulse Survey did not provide respondents with a formal definition of ‘U.S. statistical agencies.’ For interpretive purposes, this study contextualized the term using publicly available Household Pulse Survey documentation describing federal statistical agency partners involved in survey development and data collection, including agencies such as the Centers for Disease Control and Prevention and the National Center for Health Statistics [14]. For trust in federal statistics, participants were asked, “The population count, the crime rate, and the unemployment rate are examples of statistics produced by the federal government. Personally, how much trust do you have in federal statistics in the United States? Would you say that you tend to trust federal statistics, or you tend not to trust them?” Response options included “Tend to trust federal statistics” or “Tend not to trust federal statistics.” We treated this as a binary variable in the analysis. For confidence in the US Census Bureau and US statistical agencies, participants were asked, “Please indicate how much confidence you, yourself, have in the US Census Bureau” and “Please indicate how much confidence you, yourself, have in the US statistical agencies.” Response options included “A great deal,” “Quite a lot,” “Some,” and “Very little.” We dichotomized each confidence variable as very little versus other categories. Dichotomization was used to distinguish respondents expressing particularly low institutional confidence from all other respondents and to improve interpretability across survey waves. Covariates We controlled for the following sociodemographic factors: age (measured continuously in years), Sex (male, female), race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, non-Hispanic Asian, non-Hispanic Other/Multiracial), Hispanic (yes, no), educational attainment (bachelor’s degree or higher, some college education/associate’s degree, high school or less), employed (yes, no), marital status (married, not married but in a domestic partnership, divorced/separated/widowed, never married), number of children (measured continuously), and health insurance coverage (private, Medicare, Medicaid, other, unknown). Statistical analysis We used the survey and gtsummary R packages to produce design-weighted descriptives [18,19]. The Household Pulse Survey provides design-based person-level survey weights developed by the U.S. Census Bureau to account for unequal probabilities of selection, nonresponse, and population calibration, and analyses conducted using the survey package incorporated these weights into both point and variance estimation to generate design-adjusted standard errors and confidence intervals [16,18]. Bivariate analyses were based on Wilcoxon rank-sum and chi-squared tests. Multivariable analyses included survey-weighted logistic regression modeling each outcome (COVID-19 and influenza vaccine uptake) as a function of each primary condition variable. We fit unadjusted models and models adjusted for age, gender identity, race/ethnicity, educational attainment, employment status, marital status, number of children, and health insurance coverage. In wave-specific analyses, we stacked the October 2024, December 2024, and February 2025 waves (October as reference), built a pooled survey design, estimated weighted vaccine uptake prevalence by wave, and modeled uptake on wave; analyses were conducted in R Version 4.5.1 (R Core Team, R Foundation for Statistical Computing). Results Weighted descriptive statistics In Table 1, for the wave starting in October 2024, the weighted population estimate represented approximately 243,558,789 U.S. adults (mean age 48 years [SD = 17]), with 51% female, 76% non-Hispanic white, and 18% Hispanic. Compared to individuals who received the COVID-19 vaccine, those who did not were younger (mean age = 46 vs. 55 years, p < 0.001), more likely to have a high school diploma or less (38% vs. 30%, p < 0.001), more likely to be employed (65% vs. 48%, p < 0.001), more likely to be never married (22% vs. 16%, p = 0.006), and more likely to receive Medicaid (10% vs. 5%, p < 0.001). Regarding trust in federal agencies, unvaccinated individuals (vs. vaccinated) tended not to trust federal statistics (35% vs. 17%, p < 0.001) and had higher percentages reporting very little confidence in the US Census Bureau (4% vs. 2%, p = 0.001) and US statistical agencies (9% vs. 4%, p < 0.001). Corresponding unweighted descriptive statistics are presented in Supplementary Table S1 and show similar patterns. Table 1Weighted descriptive statistics, overall and by COVID-19 Vaccination Uptake, Household Pulse Survey, October 2024.CharacteristicOverallN = 243,558,789COVID-19 Vaccination Uptakep-value1YesN = 48,890,176NoN = 194,668,614 Influenza vaccination uptake, No. (%) <0.001 Yes73,303,693(30%)41,219,570(84%)32,084,123(16%) No170,255,096(70%)7,670,605(16%)162,584,491(84%) Age, Mean (SD) 48.0 (17.0)54.7 (17.4)46.4 (16.5) <0.001 Sex, No. (%) 0.25 Male117,931,362(48%)24,596,229(50%)93,335,132(48%) Female125,627,427(52%)24,293,946(50%)101,333,481(52%) Race/Ethnicity, No. (%) 0.24 White186,101,056(76%)36,529,160(75%)149,571,896(77%) Black30,921,731(13%)6,844,918(14%)24,076,813(12%) Asian13,384,773(5%)3,214,143(7%)10,170,630(5%) Other/Multiracial13,151,229(5%)2,301,954(5%)10,849,274(6%) Hispanic, No. (%) 0.63 No200,682,315(82%)40,648,525(83%)160,033,790(82%) Yes42,876,474(18%)8,241,651(17%)34,634,823(18%) Educational attainment, No. (%) <0.001 Bachelor’s degree or higher83,049,225(34%)20,624,638(42%)62,424,587(32%) Some college education/associate’s degree71,646,818(29%)13,482,216(28%)58,164,602(30%) High school or less88,862,746(36%)14,783,322(30%)74,079,425(38%) Employed, No. (%) <0.001 No92,978,984(38%)25,481,176(52%)67,497,808(35%) Yes150,579,805(62%)23,409,000(48%)127,170,805(65%) Marital status, No. (%) 0.006 Married134,744,768(55%)28,928,490(59%)105,816,279(54%) Not married but in a domestic partnership19,691,342(8%)3,537,778(7%)16,153,564(8%) Divorced/Separated/Widowed38,646,527(16%)8,646,413(18%)30,000,114(15%) Never married50,476,152(21%)7,777,495(16%)42,698,656(22%) Number of kids, Mean (SD) 0.7 (1.1) 0.5 (0.9) 0.7 (1.1) <0.001 Health insurance, No. (%) <0.001 Private171,884,476(71%)35,513,729(73%)136,370,748(70%) Medicare28,332,694(12%)8,235,931(17%)20,096,763(10%) Medicaid22,186,167(9%)2,664,036(5%)19,522,131(10%) Other3,003,239(1%)493,761(1%)2,509,478(1%) Unknown18,152,213(7%)1,982,718(4%)16,169,495(8%) Trust in federal statistics, No. (%) <0.001 Tend to trust federal statistics (ref)166,188,574(68%)40,369,811(83%)125,818,763(65%) Tend not to trust federal statistics77,370,215(32%)8,520,364(17%)68,849,851(35%) Confidence in the US Census Bureau, No. (%) 0.001 Other (ref)234,469,593(96%)48,050,986(98%)186,418,607(96%) Very little9,089,196(4%)839,190(2%)8,250,007(4%) Confidence in US statistical agencies, No. (%) <0.001 Other (ref)224,807,611(92%)47,090,283(96%)177,717,328(91%) Very little18,751,178(8%)1,799,892(4%)16,951,285(9%)1Pearson’s χ2: Rao & Scott adjustment; Design-based Kruskal Wallis test. Weighted descriptive statistics, overall and by COVID-19 Vaccination Uptake, Household Pulse Survey, October 2024. Pearson’s χ2: Rao & Scott adjustment; Design-based Kruskal Wallis test. In Table S2, for the wave starting in December 2024, the weighted population estimate represented approximately 242,170,689 U.S. adults (mean age 48.1 years [SD = 17.0]), with 51% female, 76% non-Hispanic white, and 18% Hispanic. Compared to individuals who did not receive the COVID-19 vaccine, those who did were older (mean age = 54.4 vs. 45.7 years, p < 0.001), more likely to have a bachelor’s degree or higher (44% vs. 30%, p < 0.001), less likely to be employed (50% vs. 66%, p < 0.001), more likely to be married (59% vs. 54%, p < 0.001), and had fewer children on average (0.4 vs. 0.8, p < 0.001). Regarding trust in federal agencies, vaccinated individuals (vs. unvaccinated) more often tended to trust federal statistics (83% vs. 64%, p < 0.001) and had lower percentages reporting very little confidence in the US Census Bureau (2% vs. 5%, p < 0.001) and US statistical agencies (3% vs. 9%, p < 0.001). Race/ethnicity distributions differed modestly (p = 0.030), while sex did not (p = 0.59). Similar sociodemographic differences were observed across COVID-19 and influenza uptake groups (Tables S3–S6). Associations between wave and vaccine uptake Across survey waves, vaccination uptake declined relative to October 2024. Specifically, for COVID-19 (Table 2), odds of uptake were lower in December 2024 (unadjusted OR = 0.66, 95% CI: 0.61–0.71; aOR = 0.64, 95% CI: 0.59–0.69) and February 2025 (unadjusted OR = 0.60, 95% CI: 0.55–0.66; aOR = 0.57, 95% CI: 0.52–0.62), all p < 0.001. Consistently, influenza vaccination showed a similar pattern (Table 2) compared with October 2024, uptake was lower in December 2024 (unadjusted OR = 0.54, 95% CI: 0.51–0.58; aOR = 0.51, 95% CI: 0.48–0.55) and February 2025 (unadjusted OR = 0.48, 95% CI: 0.45–0.52; aOR = 0.45, 95% CI: 0.41–0.48), all p < 0.001 (Table 2). Table 2Association of survey wave with COVID-19 vaccine and influenza vaccine uptake.TermCOVID-19 vaccination UptakeInfluenza vaccination UptakeUnadjustedAdjustedUnadjustedAdjustedOR95% CIp-valueOR95% CIp-valueOR95% CIp-valueOR95% CIp-value Wave October 2024 (ref)–––––––– December 20240.660.61, 0.71 <0.001 0.640.59, 0.69 <0.001 0.540.51, 0.58 <0.001 0.510.48, 0.55 <0.001 February 20250.600.55, 0.66 <0.001 0.570.52, 0.62 <0.001 0.480.45, 0.52 <0.001 0.450.41, 0.48 <0.001 Abbreviations: CI = Confidence Interval, OR = Odds Ratio, and — indicates reference category. Association of survey wave with COVID-19 vaccine and influenza vaccine uptake. Abbreviations: CI = Confidence Interval, OR = Odds Ratio, and — indicates reference category. Associations between trust in federal statistical agencies and vaccine uptake In October 2024, weighted logistic regression models demonstrated persistent relationships between trust in federal statistical agencies and vaccine uptake, even after adjusting for sociodemographic factors. Table 2 shows both unadjusted and adjusted associations. Individuals who tend not to trust federal statistics had significantly higher odds of not receiving the COVID-19 vaccine compared to those who trust federal statistics (unadjusted OR [OR]: 2.59, 95% CI: 2.05–3.28, p < 0.001; adjusted OR [aOR]: 2.61, 95% CI: 2.05–3.32, p < 0.001). Similarly, those with very little confidence in the US Census Bureau had higher odds of non-vaccine uptake (OR: 2.53, 95% CI: 1.41–4.56, p = 0.002; aOR: 2.16, 95% CI: 1.16–4.02, p = 0.015), as did individuals with very little confidence in US statistical agencies (OR: 2.50, 95% CI: 1.69–3.68, p < 0.001; aOR: 2.28, 95% CI: 1.52–3.43, p < 0.001) relative to those with at least some confidence in these institutions (Table 3). Table 3Associations between trust in federal statistical agencies and COVID-19 vaccine uptake among US adults, Household Pulse Survey.CharacteristicCOVID-19 vaccination UptakeOctober 2024December 2024February 2025UnadjustedAdjustedUnadjustedAdjustedUnadjustedAdjustedOR95% CIp-valueOR95% CIp-valueOR95% CIp-valueOR95% CIp-valueOR95% CIp-valueOR95% CIp-value Trust in federal statistics Tend to trust federal statistics (ref)––––––––––––Tend not to trust federal statistics2.592.05, 3.28 <0.001 2.612.05, 3.32 <0.001 2.782.29, 3.38 <0.001 2.862.32, 3.51 <0.001 1.651.35, 2.00 <0.001 1.631.31, 2.02 <0.001 Confidence in the US Census Bureau Other categories (ref)––––––––––––Very little2.531.41, 4.56 0.002 2.161.16, 4.02 0.015 2.411.45, 4.02 <0.001 1.921.15, 3.21 0.013 1.861.22, 2.83 0.004 1.561.02, 2.38 0.041 Confidence in US statistical agencies Other categories (ref)––––––––––––Very little2.501.69, 3.68 <0.001 2.281.52, 3.43 <0.001 2.771.94, 3.97 <0.001 2.381.67, 3.39 <0.001 2.131.51, 3.00 <0.001 1.901.33, 2.73 <0.001 OR = Odds Ratio, CI = Confidence Interval, and — indicates reference category.Adjusted models control for age, gender identity, race, Hispanic ethnicity, educational attainment, employed status, marital status, number of kids, and health insurance coverage.Weighted COVID-19 vaccine uptake was consistently higher among respondents who tended to trust federal statistics compared with those who tended not to trust federal statistics across all survey waves (October 2024: 24.3% vs. 11.0%; December 2024: 33.2% vs. 15.2%; February 2025: 32.4% vs. 22.6%). Associations between trust in federal statistical agencies and COVID-19 vaccine uptake among US adults, Household Pulse Survey. OR = Odds Ratio, CI = Confidence Interval, and — indicates reference category. Adjusted models control for age, gender identity, race, Hispanic ethnicity, educational attainment, employed status, marital status, number of kids, and health insurance coverage. Weighted COVID-19 vaccine uptake was consistently higher among respondents who tended to trust federal statistics compared with those who tended not to trust federal statistics across all survey waves (October 2024: 24.3% vs. 11.0%; December 2024: 33.2% vs. 15.2%; February 2025: 32.4% vs. 22.6%). Moreover, the weighted logistic regression models showed consistent associations between lower institutional trust and COVID-19 non-vaccine uptake. In December 2024 (Table 3), tending not to trust federal statistics was associated with increased odds of not receiving the COVID-19 vaccine (unadjusted OR [OR]: 2.78, 95% CI: 2.29–3.38, p < 0.001; adjusted OR [aOR]: 2.86, 95% CI: 2.32–3.51, p < 0.001); similarly, very little confidence in the U.S. Census Bureau (OR: 2.41, 95% CI: 1.45–4.02, p < 0.001; aOR: 1.92, 95% CI: 1.15–3.21, p = 0.013) and in U.S. statistical agencies (OR: 2.77, 95% CI: 1.94–3.97, p < 0.001; aOR: 2.38, 95% CI: 1.67–3.39, p < 0.001) showed elevated odds. By February 2025 (Table 2), these patterns persisted; specifically, tending not to trust federal statistics (OR: 1.68, 95% CI: 1.35–2.09, p < 0.001; aOR: 1.63, 95% CI: 1.31–2.02, p < 0.001), very little confidence in the US Census Bureau (OR: 1.61, 95% CI: 1.06–2.44, p = 0.025; aOR: 1.56, 95% CI: 1.02–2.38, p = 0.041), and very little confidence in US statistical agencies (OR: 1.97, 95% CI: 1.38–2.82, p < 0.001; aOR: 1.90, 95% CI: 1.33–2.73, p < 0.001) each remained associated with greater odds of not receiving the COVID-19 vaccine. Table 4 shows comparable patterns regarding influenza vaccine uptake. Those who tend not to trust federal statistics had higher odds of not receiving the influenza vaccine (unadjusted OR [OR]: 1.92, 95% CI: 1.63–2.25, p < 0.001; adjusted OR [aOR]: 1.87, 95% CI: 1.58–2.21, p < 0.001) compared to those who tend to trust federal statistics. Very little confidence in the US Census Bureau (OR: 3.16, 95% CI: 2.07–4.81, p < 0.001; aOR: 2.67, 95% CI: 1.74–4.10, p < 0.001) or US statistical agencies (OR: 2.09, 95% CI: 1.52–2.87, p < 0.001; aOR: 1.86, 95% CI: 1.35–2.57, p < 0.001) was also associated with increased odds of not receiving the influenza vaccination. Table 4Associations between trust in federal statistical agencies and influenza vaccine uptake among US adults, Household Pulse Survey.CharacteristicInfluenza vaccination UptakeOctober 2024December 2024February 2025UnadjustedAdjustedUnadjustedAdjustedUnadjustedAdjustedOR95% CIp-valueOR95% CIp-valueOR95% CIp-valueOR95% CIp-valueOR95% CIp-valueOR95% CIp-value Trust in federal statistics Tend to trust federal statistics (ref)––––––––––––Tend not to trust federal statistics1.921.63, 2.25 <0.001 1.871.58, 2.21 <0.001 2.161.84, 2.52 <0.001 2.111.79, 2.49 <0.001 1.641.38, 1.96 <0.001 1.631.35, 1.97 <0.001 Confidence in the US Census Bureau Other categories (ref)––––––––––––Very little3.162.07, 4.81 <0.001 2.671.74, 4.10 <0.001 2.231.53, 3.26 <0.001 1.691.14, 2.51 0.009 1.991.34, 2.95 <0.001 1.651.13, 2.39 0.009 Confidence in US statistical agencies Other categories (ref)––––––––––––Very little2.091.52, 2.87 <0.001 1.861.35, 2.57 <0.001 2.281.74, 2.99 <0.001 1.871.40, 2.48 <0.001 1.881.39, 2.53 <0.001 1.631.22, 2.19 0.001 OR = Odds Ratio, CI = Confidence Interval, and — indicates reference category.Adjusted models control for age, gender identity, race, Hispanic ethnicity, educational attainment, employed status, marital status, number of kids, and health insurance coverage.Weighted influenza vaccine uptake was consistently higher among respondents who tended to trust federal statistics compared with those who tended not to trust federal statistics across survey waves (October 2024: 34.2% vs. 21.3%; December 2024: 49.9% vs. 31.6%; February 2025: 50.8% vs. 38.6%). Associations between trust in federal statistical agencies and influenza vaccine uptake among US adults, Household Pulse Survey. OR = Odds Ratio, CI = Confidence Interval, and — indicates reference category. Adjusted models control for age, gender identity, race, Hispanic ethnicity, educational attainment, employed status, marital status, number of kids, and health insurance coverage. Weighted influenza vaccine uptake was consistently higher among respondents who tended to trust federal statistics compared with those who tended not to trust federal statistics across survey waves (October 2024: 34.2% vs. 21.3%; December 2024: 49.9% vs. 31.6%; February 2025: 50.8% vs. 38.6%). Extending to later waves, December 2024 (Table 4) showed similarly elevated odds of not receiving the influenza vaccine among adults who tended not to trust federal statistics (OR: 2.16, 95% CI: 1.84–2.52, p < 0.001; aOR: 2.11, 95% CI: 1.79–2.49, p < 0.001), had very little confidence in the US Census Bureau (OR: 2.23, 95% CI: 1.53–3.26, p < 0.001; aOR: 1.69, 95% CI: 1.14–2.51, p = 0.009), or had very little confidence in US statistical agencies (OR: 2.28, 95% CI: 1.74–2.99, p < 0.001; aOR: 1.87, 95% CI: 1.40–2.48, p < 0.001). By February 2025 (Table 4), these associations persisted: tending not to trust federal statistics (OR: 1.64, 95% CI: 1.38–1.96, p < 0.001; aOR: 1.63, 95% CI: 1.35–1.97, p < 0.001), very little confidence in the US Census Bureau (OR: 1.99, 95% CI: 1.34–2.95, p < 0.001; aOR: 1.65, 95% CI: 1.13–2.39, p = 0.009), and very little confidence in US statistical agencies (OR: 1.88, 95% CI: 1.39–2.53, p < 0.001; aOR: 1.63, 95% CI: 1.22–2.19, p = 0.001) were each linked to higher odds of not receiving the influenza vaccine. Discussion Principal’s findings and broader implication Consistent with the hypothesis, lower trust in federal statistical agencies was significantly associated with not taking the COVID-19 or influenza vaccine, and this relationship persisted after adjusting for sociodemographic factors. Additionally, individuals with very little confidence in the US Census Bureau or US statistical agencies had a higher likelihood of not receiving the COVID-19 vaccine compared to individuals with at least some confidence in these institutions. Similarly, individuals who tend not to trust federal statistics had higher odds of not receiving the influenza vaccine than those who tend to trust federal statistics. Furthermore, people with very little confidence in the US Census Bureau or US statistical agencies also had increased odds of not receiving the influenza vaccination. While weighted prevalence estimates indicated declines in influenza and COVID-19 vaccine uptake across waves, regression results showed trust in federal statistical agencies continued to be significantly associated with vaccination behavior, even after accounting for sociodemographic factors. Influenza vaccination prevalence fluctuated between 2014 and 2021, and during our survey waves we observed further declines in both influenza and COVID-19 uptake [20]. These patterns underscore the importance of understanding the factors shaping vaccine uptake rather than viewing coverage as static. In our analysis, trust emerged as a significant factor associated with vaccination uptake, with lower confidence in federal statistical agencies associated with reduced likelihood of vaccination. This aligns with prior work showing institutional trust plays a vital yet nuanced role in vaccine behavior, where the nature of trust may be critically associated with vaccine uptake [21–24]. The role of institutional trust, however, is complex and context dependent. For example, low confidence in governmental institutions has been shown to negatively affect vaccine trust [21–24]. Our findings similarly indicate lower trust in federal agencies is associated with reduced uptake. One potential explanation for these findings is that federal statistical agencies play a central role in generating and communicating population health information used to guide public health recommendations [7,11]. Distrust in these institutions may therefore reduce confidence in the validity of vaccine-related data, perceived disease risk, or the credibility of public health messaging more broadly [21–24]. This may be associated with how individuals interpret vaccination recommendations and engage with preventive health interventions [21–24]. While many studies document an association between greater trust in healthcare providers and higher vaccine acceptance, recent research suggests skepticism toward official narratives can, in some contexts, motivate preventive health behaviors [23,24]. The timing of our data collection, particularly around significant political events, may also have introduced dynamics that may have been associated with trust and vaccine uptake differently than in other settings. Further research is needed to disentangle how trust, risk perception, and sociopolitical context interact to shape vaccination behavior [25–32]. Strengths, limitations, and future research directions A strength of the Household Pulse Survey is it is nationally representative and conducted by the US Census Bureau. This recent data provides a unique perspective over other studies that relied on data from 2020 to 2021 [20]. However, the limitations of self-reported data in vaccination studies are well documented in the literature [33–35]. Self-reported data may introduce information bias because participants may report false information on their vaccination status. Additionally, because the Household Pulse Survey was administered by the U.S. Census Bureau, responses regarding trust in federal statistics and statistical agencies may have been influenced by survey source effects or social desirability bias. Potential participants were informed during recruitment that the survey was conducted by the U.S. Census Bureau, and individuals with lower levels of institutional trust may therefore have been less likely to participate, introducing possible selection or nonresponse bias into trust-related estimates. These potential biases may affect the accuracy and representativeness of the findings and should be considered when interpreting the results. Additionally, voluntary participation in the survey may have contributed to nonresponse bias despite the application of survey weighting procedures. Furthermore, the cross-sectional design limits causal inference, as the data provide only a snapshot in time and make it difficult to assess how institutional trust and vaccination behavior evolve over time [35]. Because vaccine uptake was relatively common in some subgroups, odds ratios may somewhat overestimate the magnitude of association relative to prevalence ratios. Residual confounding is also possible because some factors associated with both institutional trust and vaccine uptake, such as political orientation, household income, and urban/rural residence, were not included in the adjusted analyses. Future research could extend these findings by linking electronic health record data on vaccine uptake and tracking changes in institutional trust and vaccination behaviors over time. New survey waves may also help monitor evolving trends in COVID-19 and influenza vaccination. Longitudinal studies are especially important for assessing causal relationships and understanding how trust in federal statistical agencies may be related to vaccine-related decision-making. Future work should additionally incorporate qualitative and mixed-methods approaches to better understand the drivers of distrust, including historical experiences, perceptions of institutional credibility, communication strategies, and community perspectives. Incorporating these perspectives may support the development of more effective public health outreach, trust-building interventions, and vaccination promotion efforts while reducing information bias through the inclusion of objective health behavior measures and a broader range of vaccination determinants. Conclusions The significant associations between trust in statistical agencies, federal statistics, and the U.S. Census Bureau and vaccination suggest that institutional trust is an important factor in public health initiatives. Individuals with lower trust in agencies such as the U.S. Census Bureau had higher odds of not receiving COVID-19 and influenza vaccines. Weighted analyses also indicated declining COVID-19 and influenza vaccine uptake across survey waves relative to October 2024. Taken together, these findings highlight the association between institutional trust and vaccination behaviors. Supplementary Material Funding This work was supported by a grant from the U.S. National Library of Medicine [Grant number: T15LM012502]. The views expressed in this publication are those of the author and do not necessarily reflect the position or policy of the National Library of Medicine or the United States government. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.jvacx.2026.100847. Footnotes CRediT authorship contribution statement Nathaniel J. Maxey: Writing – original draft, Software, Methodology, Formal analysis, Conceptualization. Jemar R. Bather: Writing – review & editing, Supervision, Software, Methodology, Conceptualization. Taliyah S. Griffin: Writing – review & editing. Plamena P. Powla: Writing – review & editing. Felix M. Pabon-Rodriguez: Writing – review & editing, Software, Investigation. Ethics statement The analysis conducted in this study was reviewed by the Indiana University Human Research Protection Program (HRPP) staff, who determined that it does not involve human subjects research (IRB #29297). Thus, the study did not require further Institutional Review Board (IRB) review before conducting the analysis. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Data availability statement The data used in this study are publicly available on the US Census Bureau website (www.census.gov/data/experimental-data-products/household-pulse-survey.html). References Associated Data Supplementary Materials Data Availability Statement The data used in this study are publicly available on the US Census Bureau website (www.census.gov/data/experimental-data-products/household-pulse-survey.html).