Food insecurity remains a serious public health problem. In 2019, 11 % of U.S. households suffered from food insecurity(Reference Coleman-Jensen, Rabbitt and Gregory1), defined as the limited and uncertain acquisition of nutritionally adequate foods through socially acceptable ways(Reference Bickel, Nord and Price2). The burden of household food insecurity has been disproportionately borne by racial/ethnic minority groups, with food insecurity prevalence for Black and Hispanic households hovering at 22 % and 17 % in 2020, respectively(Reference Coleman-Jensen, Rabbitt and Gregory1,Reference Coleman-Jensen, Rabbitt and Gregory3) .
Little is known about the prevalence of food insecurity among Asian Americans (AA) whose population grew by 35·5 % over the past decade(Reference Jones, Marks and Ramirez4). Data on AA are usually aggregated, ignoring the diversity of experiences of AA subgroups and potentially masking ethnic subgroup differences(Reference Holland and Palaniappan5). The few published studies of food insecurity among AA report varying prevalence among ethnic subgroups, consistent with the heterogeneity of income among Asian ethnic subgroups, which is greater than that of other racial/ethnic groups(Reference Becerra, Mshigeni and Becerra6). Pooling data from the California Health Interview Survey (CHIS) over 10 years, Becerra et al. examined ethnic subgroup differences in food insecurity among AA adults and found wide variation in prevalence, ranging from 2·3 % among Japanese to 16·4 % among Vietnamese(Reference Becerra, Mshigeni and Becerra6). Not speaking English at home was associated with increased food insecurity risk among Chinese, Koreans and Vietnamese but not Filipinos or Japanese. Adults aged 45+ years had significantly higher prevalence of food insecurity compared with younger adults (18–44 years) among Chinese, South Asian, Korean and Vietnamese subgroups. In another study, Louie et al. studied a convenience sample of sixty-eight California-residing Asians and Pacific Islanders and found 60 % were food insecure but only 30 % had ever applied for CalFresh (California’s Supplemental Nutrition Assistance Program) benefits. Shame and pride and lack of knowledge about eligibility requirements were cited as primary reasons(Reference Louie, Kim and Chan7).
These existing studies examining food insecurity among AA do not report results specific to older adults (ages 60+ years). Older adults are uniquely vulnerable to food insecurity given a high prevalence of chronic disease, physical and cognitive limitations and fixed income(Reference Boersma, Black and Ward8). Having multiple chronic diseases is associated with higher risk of food insecurity; strained household budgets from increased healthcare expenses may partially explain this observation(Reference Jih, Stijacic-Cenzer and Seligman9). Ageing also increases the risk for physical and cognitive limitations, which may impact an older adult’s ability to perform daily living activities, such as purchasing and preparing food, increasing food insecurity risk(Reference Petersen, Brooks and Titus10). Older adults are more likely to live on fixed incomes, limiting how much they can spend on food when prices increase with inflation.
This study uses CHIS data to compare food insecurity prevalence across age groups among AA ethnic subgroups residing in California(11). The goal is to inform the implementation of food assistance programs and policies especially during the recovery years of the pandemic, while also filling a gap in the literature on food insecurity in older adult AA, a group reported by the media to have been seeking food assistance and the subject of anti-Asian attacks during the pandemic(Reference Yee12,Reference Yee13) .
Methods
Data source and study population
CHIS is the nation’s largest state health survey(11). Starting in 2001 as a biennial phone survey conducted in multiple languages, CHIS used a dual-frame random-digit-dial sampling technique prior to 2019(14). Currently, it uses an address-based sampling frame and is a phone and web-based survey of over 20 000 households per year that is conducted on a continuous basis. The survey is conducted in six languages (English, Spanish, Chinese (Mandarin and Cantonese dialects), Vietnamese, Korean and Tagalog). We used 2-year public use data files from 2011 to 2018 for adults and AA subgroups with sample sizes that allowed for stable statistical estimates of food insecurity rates, specifically, Chinese, Korean, Filipino and Vietnamese.
Measurements
Food insecurity
Food insecurity was assessed using the six-item USDA food security survey module which used Likert scale or yes/no responses to assess agreement with statements such as ‘The food that (I/we) bought just didn’t last, and (I/we) didn’t have money to get more’ ‘Often true’ and ‘sometimes true’ or ‘yes’ were deemed affirmative responses(15). Food security was operationalised as having no more than one affirmative response while food insecurity was operationalised as having two or more affirmative responses. The food security module was administered only to households with income ≤ 200 % of the federal poverty level (FPL) or unknown income. Food insecurity prevalence was calculated as the percent of all respondents (including those with income > 200 % FPL) who were assessed as food insecure.
Socio-demographic characteristics
Socio-demographic characteristics included the following: whether the respondent was born in the USA, citizenship status; language spoken at home, income expressed as percent of the FPL, household size, housing, educational attainment, employment status, health insurance, participation in SNAP/CalFresh and receiving supplemental security income or social security disability insurance (SSI/SSDI).
Statistical analysis
We examined three age groups: 18–39, 40–59 and 60+ years. Unweighted counts and weighted percentages were generated from PROC SURVEYFREQ. Weighted means were generated from PROC SURVEYMEANS. Associations between food insecurity and socio-demographic characteristics were assessed using the χ 2 test for each ethnic and age group. To examine trends in food insecurity prevalence over time, rolling averages were calculated for each ethnic and age group across four-year periods to allow for adequate sample sizes in each ethnic/age group (2011–2014, 2013–2016 and 2015–2018). To assess the precision of each point estimate, we calculated the CV, which is the se divided by the point estimate(16). We identified any point estimates with a CV > 0·3 in the results as potentially statistically unstable, as recommended by CHIS(17). All analyses were conducted using SAS, version 9.4 (SAS Institute Inc.) and took account of the complex sampling design and sample weights of CHIS. Statistical significance was set at P < 0·05 (two-sided). This study was determined to be exempt from review by the University of California, Los Angeles Institutional Review Board.
Results
Socio-demographic characteristics
Ethnic subgroup differences in socio-demographic characteristics are presented in Table 1. Education levels were similar among Chinese, Korean and Filipino with over half reporting having a college degree or higher but lower for Vietnamese with 52 % reporting having only a high school diploma or lower. They were also generally lowest among older adults. The majority were born outside of the USA; older respondents were more likely to be born outside the USA than younger respondents. A high proportion of Chinese (43 %), Korean (45 %) and Vietnamese (58 %) respondents reported they spoke only their native language at home, with the oldest age group reporting the highest proportions. About a quarter to a third of Filipino, Chinese and Korean and over half of Vietnamese respondents were classified as having household income ≤ 200 % FPL. Poverty levels were highest among older adults in all ethnic subgroups.
Study population characteristics by age and Asian subgroup, California Health Interview Survey 2011–2018 †
* Indicates statistically unstable estimates with CV > 0·3.
† Analyses conducted using survey weights provided by California Health Interview (CHIS). Frequencies are unweighted and proportions are weighted. Means and se are weighted.
‡ Food insecurity was only assessed among those with incomes </= 200 % FPL; not applicable indicates those above 200 % FPL.
§ Supplemental Nutrition Assistance Program (SNAP) enrollment was only assessed among those with incomes </= 300 % FPL.
|| Supplemental Security Income (SSI)/Supplemental Security Disability Insurance (SSDI) receipt was only assessed among those with incomes </= 300 % FPL.
Among all adults, food insecurity prevalence was highest among Vietnamese (16 %) followed by Filipino (10 %) with lower levels among Chinese (7 %) and Korean (7 %) (Table 1). For every ethnic subgroup, food insecurity prevalence was highest among older adults with especially high levels among Chinese (17 %) and Vietnamese (26 %). Among all adults, food insecurity prevalence was significantly higher among those with lower education, born outside of the USA, without U.S. citizenship, who spoke a language other than English at home, with the lowest incomes (0–99 % FPL), who were unemployed, and without health insurance (Table 2). Dose–response effects were observed for education and language spoken at home for all age/ethnic subgroups. U.S. citizenship, poverty level and health insurance were not associated with food insecurity among older adults.
Association between prevalence of food insecurity and study population characteristics among Asian Americans by age, California Health Interview Survey 2011–2018 †
CHIS, California Health Interview Survey; FPL, federal poverty level.
Bolded values indicate P < 0·05.
‡Food insecurity was only assessed among those with incomes </= 200 % FPL; not applicable indicates those above 200 % FPL.
* Indicates statistically unstable estimates with CV > 0·3.
† Analyses conducted using χ 2 test with survey weights provided by CHIS.
§ SNAP enrollment was only assessed among those with incomes </= 300 % FPL.
Among those with income ≤ 200 % FPL, Vietnamese respondents aged 40–59 years reported the highest enrollment in SNAP at 19 %, but the majority of respondents across all subgroups were not enrolled in SNAP (Table 1). SNAP enrollment was associated with higher levels of food insecurity (Table 2).
Age group differences in food insecurity prevalence varied over the three time periods examined, among ethnic subgroups (Figure 1). The largest gap in food insecurity rates between age groups (higher among older adults) was observed among Chinese followed by Koreans and Filipinos. This gap appears to have increased over time among all three groups due to increasing food insecurity prevalence among the older adults well as decreasing food insecurity rates among most of the younger age groups. Between 2011–2014 and 2015–2018, food insecurity rates among older adults increased by 45 % for Vietnamese, 25 % for Chinese and about 20 % for Filipinos and Koreans. In contrast, they decreased among the two younger age groups for all ethnic subgroups except Vietnamese.
Prevalence of Food Insecurity from 2011 to 2018 by Asian Subgroup and Age, California Health Interview Survey 2011–2018a. a Food insecurity over time calculated using 4-year rolling averages from 2011 to 2018.
Discussion
Our study examined differences in food insecurity by age group among AA ethnic subgroups. Determining age group differences in food insecurity has implications for the allocation of limited resources. We found older adults had the highest prevalence of food insecurity and that in some ethnic subgroups, older and younger adults exhibited contrasting trends in food insecurity. The difference in food insecurity prevalence between older adults and young adults was most stark for Chinese. Older adult Chinese had a food insecurity prevalence that was quadruple that for young adult Chinese. Between 2011–2014 and 2015–2018, food insecurity prevalence increased among older adults for every ethnic subgroup, with Vietnamese experiencing a 45 % increase and Chinese, a 25 % increase. In contrast, except for 40- to 59-year-old Vietnamese, food insecurity prevalence decreased over time in each younger age group.
For all adults, the socio-demographic factors associated with increased food insecurity risk in most if not all age/ethnic subgroups were lower education, being foreign born, not being a US citizen, not speaking English at home, being unemployed, not having health insurance and being enrolled in CalFresh (SNAP). This is consistent with the findings from Beccera et al except that they did not include U.S. citizenship, health insurance status and SNAP enrollment in their analysis(Reference Becerra, Mshigeni and Becerra6). We found that not having U.S. citizenship and not having health insurance were associated with increased food insecurity among the younger age groups but not older adult AA. Poverty level was not associated with food insecurity, likely due to the fact that the food security questions were only asked of those with incomes <200 % of poverty.
To our knowledge, this is the first study to compare food insecurity rates between older and younger adults for AA ethnic subgroups and report climbing rates of food insecurity among older adult AA in the last decade. However, this study has limitations. First, we had to pool several years of data (to obtain adequate samples), which may mask granular trends in food insecurity. Additionally, due to sample size limitations, we were unable to examine food insecurity among Pacific Islanders who often aggregated with AA; studies suggest food insecurity may be more prevalent in Pacific Islanders than AA(Reference Long, Rowland and McElfish18). Second, CHIS data are only representative for California, so our findings may not be generalisable to other states. Lastly, the data are cross-sectional, limiting our ability to determine causal relationships.
The fastest growing racial group in the country, AA have been overlooked in studies of food insecurity. Our finding of AA older adults having higher rates of food insecurity are corroborated by media reports of AA seniors seeking assistance at food distribution events during the pandemic(Reference Yee13) and reports of rising poverty rates among older adult AA(Reference Tran19). We also find that those who are not U.S. citizens and those who speak a language other an English at home are more likely to be food insecure and that SNAP, the largest food assistance program in the country, does not reach all who may benefit from the program. Further studies of SNAP/CalFresh enrollment among older adult AA during the COVID-19 pandemic to provide insights into the impact of pandemic-related food policy provisions, which allowed for waivers and flexibilities in the operations of food assistance programs, may be helpful(20). More complete collection and reporting of disaggregated health-related data for AA and Pacific Islanders is needed. Older adult AA and Pacific Islanders are often left out of the conversation about the public health needs and well-being of seniors and of AA and Pacific Islanders. Community-engaged research to develop culturally appropriate strategies for increasing the reach and services of SNAP and senior nutrition programs and access to culturally appropriate foods for AA and Pacific Islander older adults is recommended(Reference Nhan and Wang21).
Open access
