key: cord-0274337-woic5p4h
authors: Wojcicki, J.; Escobar, M.; De Castro Mendez, A.; Martinez, S.
title: Household Density and Influences from Family and Friends Shape Vaccine Intent Among Latino Families in the San Francisco Bay Area
date: 2021-09-27
journal: nan
DOI: 10.1101/2021.09.25.21264120
sha: c530168bea1548127f5db783d3d18423a8faf6e0
doc_id: 274337
cord_uid: woic5p4h

Abstract Background: Latino have had higher case counts, hospitalization rates and deaths during the COVID-19 pandemic nationally and in the state of California. Meanwhile, Latino vaccination rates remain lower than those of non-Hispanic Whites. COVID-19 vaccine nonintent, defined as intent not to vaccinate for COVID-19, among Latino individuals continues to be an issue in the state of California. Methods: Families from three Latino longitudinal mother child cohorts previously recruited in the San Francisco Bay Area were surveyed telephonically from February to July 2021 to assess attitudes towards COVID-19 vaccination and prior vaccination, in general, for themselves and their children. Risk for vaccine nonintent was assessed using Mann-Whitney rank sum non-parametric test for continuous predictors and chi-squared tests for categorical ones. Results: Three hundred and eighteen families were surveyed from the Telomere at Birth (TAB), Hispanic Eating and Nutrition (HEN) and Latino Eating and Diabetes Cohort (LEAD). Approximately 36% from TAB and 28% from HEN/LEAD indicated COVID-19 vaccine nonintent for themselves and/or their children. Risk factors for vaccine nonintent included lower maternal age (p=0.01), concern about vaccine side effects (p<0.01) and prior history of a household members being infected with COVID-19 (p<0.01) and indexes of household crowding including number of people sharing a bathroom (p=0.048). Vaccine intent was also associated with receiving vaccine input from friends (p=0.03), family (p<0.01) and/or coworkers (p=0.02) compared with those who were not planning on getting COVID-19 vaccination. It is possible that those with non-intent have received limited input from families, friends and/or coworkers. Discussion: Latino families living in crowded living situations who may not have received any COVID-19 advice family, coworkers or friends are at particular risk for COVID-19 vaccine nonintent. Community based grassroots interventions that focus on trusted individuals with close ties to the community counseling about COVID-19 vaccination could help to boost vaccination rates in this population group.

3 Background

As of September 2021, there have been over 649,3000 deaths from COVID-19 in the United States with high rates among racial and ethnic minority groups (CDC, 2021a) . Over the course of the pandemic, Latinos have been two times more likely to become infected with COVID-19, 2.8 times more likely to have hospitalization, and 2.3 times more likely to die after adjusting for population age structure, in comparison to White, Non-Hispanic groups (CDC, 2021b) . Latinos account for 18.5% of the US population but 27.9% of COVID-19 cases (CDC, 2021b) . The rates of COVID-19 infection and mortality are likely higher among undocumented Latinos but immigration status is not routinely collected by states or counties (Page et. al, 2021) .

Meanwhile, COVID-19 vaccination rates among Latinos are lower than the overall population percentage of Latino in the United States with only 16.2% fully vaccinated (with two doses of Pfizer or Moderna or one of the Janssen/Johnson and Johnson) while Latino represent 17.2% of the eligible population (CDC, 2021c) . In California the disparity between vaccination and eligibility is greater; 29.2% of Latinos have received at least one dosage of COVID-19 vaccination (Covid19.CA.Gov, 2021) . However, this population accounts for 39.2% of the eligible population in the state (Covid19.CA.Gov, 2021) . The burden of disease has also been higher in Latinos compared to other population groups in the state of California. Overall, approximately 30% of vaccines in California have gone to Latino individuals while 63% of cases and 48% of deaths have been among Latinos in California (Kaiser Family Foundation, 2021) .

In the state of California, vaccination is provided to all over the age of 12 without any need for payment or verification of immigration status. Vaccination without need for payment was instituted in California as of April 15, 2021. Vaccination intent is defined as being absolutely certain or likely to be vaccinated (Nguyen et. al, 2021) . Racial and ethnic minorities have a greater rate of nonintent for COVID-19 vaccination (plan not to vaccinate against . A recent review combined data from 13 studies on COVID-19 vaccine intent finding that vaccine hesitancy was much higher for Latinos (41.6%) versus 26.3% for all other adult Americans (Khubchandani and Macias, 2021) . A study focused on vaccine hesitancy among Latino adults found that adults younger than 50 were more likely to be distrustful of the public health authorities than older Latino adults (Kearney et al., 2021) .

Factors that may impact vaccine intent in Latino communities include lack of legal status to minimal prior history with US healthcare services, wariness of authorities, and cultural and linguistic barriers (Strully et al., 2021) . In particular, there has not been sufficient inclusion of the Latino community or collaboration with Latino community resources during public health planning for COVID-19 vaccination roll-out (Strully et al., 2021) .

With overall vaccination rates rising in the USA (as of early September 62% have received one dose (CDC, 2021d)), COVID-19 infection and hospitalization rates are primarily among the unvaccinated (Griffin et al., 2021) . As such, there is increased urgency to better understand COVID-19 vaccine nonintent among Latino communities. The present study examines vaccine nonintent and the intention for vaccine uptake among three pre-existing longitudinal cohorts of Latinos residing in the Greater San Francisco Bay Area.

In Spring 2021, in response to the availability of COVID-19 vaccines in the United States, we surveyed three existing Latino mother-child, family based longitudinal cohorts in the San Francisco Bay Area, the Hispanic Eating and Nutrition (HEN), Latino Eating and Diabetes (LEAD) study and the Telomeres at Birth (TAB) cohort. In short, these Latino cohorts were recruited initially as mother-child birth cohorts at UCSF Benioff and Zuckerberg San Francisco General (ZSFG) hospitals to evaluate early life risk factors for childhood obesity. Details about these cohorts has been previously described including methods of recruitment and demographics (Wojcicki et al., 2011; Ville et al., 2017) . The HEN and LEAD cohorts consist primarily of foreign born mothers (from Mexico and Central American countries) and were recruited primarily from ZSFG whereas TAB includes a more heterogenous mix of Latino families recruited primarily from UCSF Benioff. We re-contacted these cohorts telephonically in 2020 during an early period in the pandemic (May to September 2020) to assess food insecurity, housing conditions and risk for COVID-19 infection (Escobar et al., 2021; Mendez et al., in press ).

Subsequently, we re-contacted families again to assess attitudes towards COVID-19 vaccination in February-June 2021 once vaccinations had received emergency US Food and Drug Administration (FDA) approval (Pfizer-BioNTech, Moderna and Johnson & Johnson Janssen) in the United States. At this time, we surveyed participants about vaccination history for themselves and their children, COVID-19 vaccination intent and intent to vaccinate their children (once approval was given for their children's age group for COVID-19 vaccination) using a short 10-minute questionnaire administered telephonically. We additionally surveyed participants regarding those individuals and institutions including media sources that had possibly influenced their attitudes toward COVID-19 vaccination.

We combined the HEN and LEAD cohorts for analyses purposes due to similar inclusion criteria (Latina mothers anticipating a healthy newborn) and methods of recruitment primarily at the ZSFG hospital were similar (Wojcicki et al., 2011; Ville et al., 2017) . HEN and LEAD are primarily Spanish speaking, low-income households and the majority of families participated in the Special Supplementation Program of Women, Infants and Children (WIC). The TAB cohort was analyzed separately as women and children were recruited primarily at the UCSF Benioff hospital and population demographics were more heterogenous including more highly educated parents, US-born parents and English speakers. Demographic factors including age, household size and household density and employment status as well as past vaccination history were assessed in relationship to self-disclosed COVID-19 vaccination nonintent for self and/or child. We also assessed differences in influences on COVID-19 vaccination intent between those who intended or had already vaccinated and those who were not planning to be vaccinated using chisquared tests and Wilcoxon rank sum tests. Continuous predictors were tested for normality using the Shapiro Wilks test for normality and all predictors followed non-parametric . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted September 27, 2021. ; https://doi.org/10.1101/2021.09.25.21264120 doi: medRxiv preprint distributions. All statistical analyses were conducted using Stata 15.0. This study was approved by the UCSF Institutional Review Board (IRB) and participants gave oral consent for their and their children's participation.

We had 211 participants from the TAB cohort who completed the COVID-19 vaccine survey and 108 from the HEN/LEAD cohorts. Demographics including language use and household socioeconomic status are presented by cohort in Table 1 . Average household size was 5.11±1.82 in HEN/LEAD and slightly smaller at 4.7±1.9 in TAB. Both cohorts reported a mean of 2 to 3 children in the household (Table 1) . Almost all HEN/TAB participants reported Spanish as the primary language (93.5%) in contrast with TAB where 32.1% reported Spanish as their primary language (64.6% reported English as their primary language) (Table 1). Maternal age was slightly higher in HEN/LEAD (37.4±6.8) versus 35.4±5.6 in TAB. A higher percentage of TAB reported being married or living with partner (91.4%) versus 48.6% in HEN/LEAD and similarly a high percentage of TAB had a high school degree or higher (81.8%), 83.2% of HEN/LEAD had a high school diploma or less (Table 1 ).

In addition to differences in demographics of the cohorts, participants from HEN/LEAD versus TAB overall had lower socioeconomic status as measured by a number of indicators. Specifically, HEN/LEAD cohorts had a higher number sharing a bathroom and 64.5% of the cohorts experienced food insecurity during the COVID-19 lockdowns versus 34.1% of the TAB cohort (Table 1) . Similarly, 55.4% of HEN/LEAD cohorts had household member working during the COVID-19 lockdown in California (3-7/2020) compared with 23.2% of those in the TAB cohort. A high percentage of HEN/LEAD participants are foreign born (93.5%) and of Mexican origin (55.9%). Place of birth was not collected for TAB participants and lower percentage report Mexican origin (31.0%). A higher percentage of HEN/LEAD participants reported someone in the household had tested positive for COVID-19 compared with a household member testing positive in TAB (38.0% versus 21.4%) ( Table 1) .

A high percentage from both cohorts reported up-to-date vaccination with 94.8% of TAB and 95.4% of HEN/LEAD having vaccination up-to-date for self and 94.3% and 97.2% having them up to date for their children (Table 2 ). Eighty percent of TAB versus a higher 84.3% of HEN/LEAD reported they planned or had already received the COVID-19 vaccination with a lower 66.7% of TAB and 74.1% of HEN/LEAD reporting that they planned to have their children receive the COVID-19 vaccination (if and when available) ( Table 2) . Thirty-five point two percent of TAB participants and 27.8% of HEN/LEAD stated COVID vaccination nonintent for themselves or their children. The most significant reasons for not wanting a COVID-19 vaccine for self or children were 1) the idea that the vaccine had not been around for long enough and/or 2) concern about side effects associated with the vaccine (Table 2 ). Concern about side effects were in general greater for children than adults (Table 2 ). Both groups reported that news information informed decisions the most about COVID-19 vaccination with friends ranking second for HEN/LEAD while family was the second biggest influence for TAB (Table 2) .

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted September 27, 2021. ; https://doi.org/10.1101/2021.09.25.21264120 doi: medRxiv preprint

Comparing COVID-19 vaccination intent with nonintent, we found that mothers who were planning on getting vaccinated were slightly older in TAB (36.2±5.1 versus 34.1±6.3 years) and for both cohorts, households with greater household person density were at greater risk for COVID nonintent as measured by number of persons sharing bathrooms (4.67±1.79 versus 3.94±.48; p=0.048 in HEN/LEAD and 3.8±1.8 versus 3.3±1.6; p=0.13 in TAB) (Table 3 ). In TAB, there were also a greater number of total individuals in the house 3.8±1.8 versus 3.3±1.6 (p=0.09) in families with nonintent versus intent to vaccinate against COVID-19.

Families who had someone in the household previously test positive for COVID-19 were more likely to have COVID-19 vaccination nonintent in TAB (31.1% versus 16.2%; p<0.01) although there was no association in HEN/LEAD with prior exposure to COVID-19. For both cohorts, greater concerns about side effects were associated with vaccine nonintent for adults and children (Table 3) . Spanish language use in TAB neared statistical significance (p=0.1) as nonintent was higher in the Spanish-speaking participants versus English speaking ones (39.2% versus 28.2%). When the data was analyzed just for adult intent to vaccinate themselves in relation to language use the results were highly significant (34.3% versus 13.3%; p<0.01). For HEN/LEAD, those that stated they were born in the USA were more likely to have COVID-19 vaccination nonintent, although the results neared statistical significance (p=0.07; Table 3 ). Those that planned to get vaccinated against COVID-19 in the TAB cohort were more likely to attribute significant influences from family, friends, co-workers and news sources compared to those with nonintent for COVID-vaccination. Influences from family were statistically significant in HEN/LEAD to predict COVID-19 vaccination intent (Table 3) .

We found no association between Latino ethnicity, country of origin or food insecurity and risk for COVID-19 vaccine nonintent in HEN/LEAD and TAB cohorts (Table 3) .

For those that answered they would not vaccinate themselves or their children against COVID-19, we asked a question about what would make them more comfortable to vaccinate against COVID-19. Many suggested more information, more time or more assurance form their doctor/health provider that there would be no side effects or that it was safe/needed for their personal health. As a number of our participants were women of childbearing age, there was a significant amount of concern surrounding potential risk form the COVID-19 vaccine while pregnant, breastfeeding or during the postpartum period and women wanted more assurance and more information that the COVID-19 vaccine was safe.

We found a high rate of COVD-19 vaccine nonintent in our primarily Latino family cohorts in the San Francisco Bay Area (35.2% in the TAB cohort and 27.8% in HEN/LEAD) comparable with other studies that have focused on Latino communities in the United States (Khubchandani and Macias, 2021). These cohorts represent a diverse group of Latino families in the San Francisco Bay Area including a high percentage of foreign-born parents in the HEN/LEAD cohorts as well as socioeconomic heterogeneity in the TAB cohort. Families in HEN/LEAD reported lower socioeconomic status as indicated by maternal education level, employment status and food insecurity compared with those in TAB. At the time of recruitment, HEN/LEAD . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 27, 2021. ; https://doi.org/10.1101/2021.09.25.21264120 doi: medRxiv preprint families participated uniformly in the Special Supplemental Program for Women, Infants and Children (WIC) (Ville et al.,2018; Wojcicki et al., 2011) 

California compared with other population groups and the greater burden of disease in Latinos (CDC 2021b; Kaiser Family Foundation, 2021) , it is necessary to further untangle reasons for COVID-19 vaccination nonintent in this population group.

Similar to other studies that have found a high rate of vaccine nonintent among younger Latino individuals, we found an overall high rate of vaccine nonintent in all three cohorts (27.8% in HEN/LEAD and 35.2% in TAB). Furthermore, those mothers that were younger in age were more likely to have vaccine nonintent compared with older mothers, specifically in TAB but also in HEN/LEAD although the differences were not statistically significant for HEN/LEAD. It is possible that there is a generational effect and younger Latinos similar to younger other population groups do not feel as vulnerable or at high risk compared with older populations. Alternatively, as other studies have demonstrated there is less trust of public health institutions and government by younger Latinos than older ones (Kearney et al., 2021) .

We also found increased risk for vaccine nonintent among those individuals in with higher population density in households but no associations with employment status or food insecurity or other markers of lower socioeconomic status. Previous surveys have found that higher socioeconomic status individuals are more likely to vaccinate using metrics focused on education, income and insurance status (Allen et al., 2021) or income alone (Reiter et al., 2020) . In our low-income cohorts of primarily foreign-born Latina women (HEN/LEAD), we did not find any association with educational status or employment but rather household density was more predictive of vaccination nonintent. Similarly, among the more economically heterogeneous TAB cohort, we also did not find any associations with education or income but crowding indices were also associated with vaccination intent. Paradoxically, increased housing density elevates risk for COVID-19 infection (Mendez et al., in press) , and our findings suggest that these same high-risk individuals may be less likely to get access to COVID-19 vaccination. Crowding indices may reflect immigration status, which was not collected in our survey, as previous studies have found that undocumented immigrants live in more crowded and less secure housing situations (Hall and Greenman, 2013) . In TAB, the more socioeconomically heterogenous of the two cohorts, Spanish language use was higher among those not intending to vaccinate, particularly for adult vaccination compared with English language speakers. There was no difference in language use for HEN/LEAD as the population was relatively homogenous linguistically with 93.5% reporting primary Spanish language use.

Greater concern with side effects from COVID-19 vaccination were significantly associated with COVID-19 vaccine nonintent for all cohorts that were surveyed as similar studies have shown (Nguyen et al., 2021) . Future outreach and community-based initiatives need to focus on the minimal risks from vaccination side effects compared with the much larger risk associated with non-vaccination and must be partnered with community approaches and outreach versus marketing campaigns (Jimenez et al., 2021) . Interestingly, those participants that had more . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this this version posted September 27, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 direct exposure with COVID-19 infection in the household were less likely to have COVID-19 vaccination intent although this finding was only apparent in the TAB cohort and not HEN/LEAD. It is possible that Latino families have misconceptions about prior exposure to COVID-19 and need for vaccination. Alternatively, other studies have had participants who voiced desire for physiological immunity as an option to immunity through vaccination (Taylor et al., 2020) .

For all three cohorts, those that individuals that have COVID-19 vaccine nonintent cite fewer influences from family, friends, co-workers and news sources. A large nationally representative Kaiser Family Foundation study found that family and friends were instrumental in persuading individuals to get COVID-19 vaccination after initial nonintent (Kirzinger et al., 2021) . Another study with adult women found that exposure to social media may play a positive role in encouraging vaccine uptake (Berenson et al., 2021) and news media exposure may differentially impact Republican versus Democrat-aligned individuals (Fridman et al., 2021) . To our knowledge, this is the first study focused exclusively on the Latino population in California and suggests the importance of social networks and connections to change COVID-19 nonintent to intent to vaccinate.

Our study has demonstrated the critical importance of social networks for COVID-19 vaccination intent including the importance of family and friends as well as community with increased risk potential among those with higher density living conditions with possibly lacking legal status. Key community stakeholders and clinics need to be involved at the grassroots level to dispel myths concerning vaccination and as our study has shown concern, in particular, surrounding the side effects of vaccination. The US Surgeon general has made a similar recommendation that local "trusted messengers" a key to COVID-19 vaccination (NPR, 2021). As described below, since the collection of data for this study, there have been increased efforts to publicize the notion that immigrant status will not play into access to vaccination or enforcement. Community clinics need to better circulate this message as well to ensure that local stakeholders can disseminate this information widely.

The participants in the present study are part of three, pre-existing, longitudinal cohorts in the Greater San Francisco Bay Area. While our study's demographics provide insight into reasons for vaccination nonintent among the Latino community, our sample may not be representative of the individual beliefs, values, and behaviors of all United States Latinos. Furthermore, developments in 2021 may have altered some perceptions concerning vaccinations. Over the time period that our study was conducted the United States government and the United States Department of Homeland Security made clear that vaccines will be available free of charge to all individuals in the United States and the immigrant status will not play into vaccination nor will enforcement be conducted near clinic sites or sites of vaccination distribution (United States Department of Homeland Security, 2021). Furthermore, the Biden administration has made a greater effort to facilitate partnership with Community Health Centers to promote equal access to vaccines among minority population groups (Corallo and Tolbert, 2021) . These political changes may have altered attitudes and intent concerning COVID-19 vaccination.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted September 27, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 Fridman A, Gershon R, Gneezy A. COVID-19 and Vaccine Hesitancy: A Longitudinal Study. PLOS One April 2021. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0250123, accessed September 5, 2021.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted September 27, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted September 27, 2021. ; https://doi.org/10.1101/2021.09.25.21264120 doi: medRxiv preprint Concern about COVID-19 Vaccine Side effects^ . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) preprint

The copyright holder for this this version posted September 27, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 

Factors associated with the intention to obtain a COVID-19 vaccine among a racially/ethnically diverse sample of women in the USA

Intent to get vaccinated against COVID-19 among reproductive-aged women in Texas

COVID Data Tracker: Demographic Characteristics of People Receiving COVID -19 Vaccinations in the United States

Biden Wants to Partner with Health Centers to Promote More Equitable Access to COVID-19 Vaccines. Kaiser Family Foundation

Vaccine Progress Data

High food insecurity in Latino families and associated COVID-19 infection in the Greater Bay Area

SARS-CoV-2 Infections and Hospitalizations Among Persons Aged >=16 Years, by Vaccination Status

Housing and neighborhood quality among undocumented Mexican and Central American immigrants

Black and Latino Community Perspectives on COVID-19 Mitigation Behaviors, Testing, and Vaccines

Latest Data on COVID-19 Vaccinations by Race

Vaccine Hesitancy Among Hispanic Adults. Kaiser Family Foundation

COVID-19 Vaccination Hesitancy in Hispanics and African-Americans: A Review and Recommendation for Practice

KFF COVID-19 Vaccine Monitor

Overcrowding and Exposure to Secondhand Smoke Increase Risk for COVID-19 Infection Among Latino Families in the Greater San Francisco Bay Area, In press Tobacco Induced Diseases. National Public Radio (NPR)

COVID-19 Vaccination Intent, Perceptions, and Reasons for Not Vaccinating Among Groups Prioritized for Early Vaccination -United States

Acceptability of a COVID-19 vaccine among adults in the United States: How many people would get vaccinated?

Strategies to Address Vaccine Hesitancy and Mitigate Health Disparities in Minority Populations

A Proactive Approach for Managing COVID-19: The Importance of Understanding the Motivational Roots of Vaccination Hesitancy for SARS-COV2

DHS Statement on Equal Access to COVID-19 Vaccines and Vaccine Distribution Sites

Early Antibiotic Exposure and Risk of Childhood Obesity in Latinos

Chronic maternal depression is associated with reduced weight gain in latino infants from birth to 2 years of age

This project was supported by the generosity of the Eric and Wendy Schmidt by recommendation of the Schmidt Futures program, through Covid Catalyst at the Center of Emerging and Neglected Diseases. Additional funding came from NIH NIDDK 080825 097458 as well as the Hellman Family Foundation and Allen Foundation and Marc and Lynn Benioff.

(3/26) *Higher scores are indicative of greater food insecurity ^Higher scores indicate greater concern on a scale from 1-10.