This cohort study investigates food security and all-cause premature mortality and life expectancy across racial and ethnic and sex groups in US adults.
Key Points
Question
What is the association between the entire range of food security and all-cause premature mortality and life expectancy among US adults across racial and ethnic and sex groups?
Findings
In this cohort study of 57 404 adults, lower levels of food security, even marginal food security, were significantly associated with a higher risk of all-cause premature mortality and a shorter life expectancy compared with full food security. Significant racial and ethnic and sex differences were observed in the association of food insecurity with all-cause premature mortality and life expectancy.
Meaning
This research provides a comprehensive picture of the impact of food insecurity on life expectancy and premature mortality in the US, providing scientific evidence for future policies and strategies that ought to prioritize food insecurity as a means to enhance overall health in the population.
Abstract
Importance
Food insecurity has been linked to multiple causes of disease and premature mortality; however, its association with mortality by sex and across racial and ethnic groups remains unknown in the US.
Objective
To investigate the associations of the entire range of food security with all-cause premature mortality and life expectancy across racial and ethnic and sex groups in US adults.
Design, Setting, and Participants
This cohort study included adults (aged ≥18 years) who participated in the National Health and Nutrition Examination Survey from 1999 to 2018, with linkage to the National Death Index through December 31, 2019. Data analysis was performed from August to November 2023.
Exposures
Levels of food security were assessed with the US Department of Agriculture Adult Food Security Survey Module (full, marginal, low, and very low).
Main Outcomes and Measures
All-cause premature mortality (death that occurs before age 80 years) and life expectancy.
Results
The study included 57 404 adults (weighted mean [SE] age, 46.0 [0.19] years; 51.8% female; 12 281 Black individuals [21.4%]; 10 421 Mexican individuals [18.2%]; 4627 Other Hispanic individuals [8.1%]; 24 817 White individuals [43.2%]; and 5258 individuals of other races, including multiracial [9.2%]). During a median (IQR) of 9.3 (5.0-14.3) years of follow-up, 4263 premature deaths were documented. Compared with the full food security group, the adjusted hazard ratios were 1.50 (95% CI, 1.31-1.71), 1.44 (95% CI, 1.24-1.68), and 1.81 (95% CI, 1.56-2.10) across marginal, low, and very low food security groups, respectively (P < .001 for trend). The corresponding life expectancy at age 50 years in each group was 32.5 (95% CI, 32.4-32.6), 29.9 (95% CI, 28.9-30.9), 30.0 (95% CI, 28.9-31.0), and 28.0 (95% CI, 26.8-29.2) years. Equivalently, adults with marginal, low, or very low food security lived on average 2.6 (95% CI, 1.5-3.7), 2.5 (95% CI, 1.4-3.7), or 4.5 (95% CI, 3.2-5.8) fewer years at age 50 years, respectively, compared with those with full food security. The associations appeared to be stronger in women than in men (hazard ratios comparing very low food security with full food security, 2.29 [95% CI, 1.83-2.86] in women and 1.46 [95% CI, 1.19-1.78] in men; P = .009 for interaction) and stronger in White adults than in Black adults (hazard ratios comparing very low food security with full food security, 2.07 [95% CI, 1.70-2.53] in White adults and 1.33 [95% CI, 1.01-1.75] in Black adults; P < .001 for interaction) or in Hispanic adults (hazard ratios comparing very low food security with full food security, 1.06 [95% CI, 0.71-1.58]; P < .001 for interaction).
Conclusions and Relevance
In this cohort study, although the association of food security and life expectancy varied across sex and racial and ethnic groups, overall, lower levels of food security were associated with a higher risk of premature mortality and a shorter life expectancy. The findings of this study highlight the potential importance of improving food security in promoting population health and health equity.
Introduction
Inadequate nutrition is a leading cause of premature death and suboptimal life expectancy in the US and globally.1,2 Food insecurity, defined by the US Department of Agriculture (USDA) as a “lack of consistent access to nutritionally adequate food,”3 is a common cause of inadequate nutrition. In 2021, 10.2% of US households were food insecure at some time during the year.3 A large body of evidence has shown that food insecurity is strongly associated with numerous adverse health outcomes, including diabetes, cardiovascular disease (CVD), death, and others.4,5,6,7,8,9,10 Position papers from the American College of Physicians and the American Heart Association have acknowledged the important role of food security in preventing diet-related chronic diseases and improving health equity.11,12,13
However, major gaps in knowledge exist regarding the association between food insecurity and health. First, evidence is scarce linking food insecurity and life expectancy, the most-used aggregate measure of a population’s overall health. Second, according to the USDA Food Security Survey Module, food security can be divided into 4 levels: full, marginal, low, and very low. Previous studies, including USDA annual food security reports, typically categorized participants as belonging to either the food secure (combined high and marginal food security groups) or food insecure group (combined low and very low food security groups).3,14,15 Evidence has shown that the demographic characteristics of the marginal food security group are more similar to the low food security group than to the full food security group16; thus, the above binary classification approach may lead to misclassification bias and underestimate the association of food insecurity with health outcomes. Third, previous studies have shown that racial and ethnic and sex disparities exist in the prevalence of food insecurity in the US, with Black and Hispanic adults having a higher prevalence of food insecurity than White adults and women having a higher prevalence of food insecurity than men.3,14 However, to our knowledge, no study has prospectively evaluated racial and ethnic and sex differences in the associations of food insecurity with both premature mortality and life expectancy. To fill these research gaps, we investigated the prospective associations of the entire range of food security with all-cause premature mortality and life expectancy across racial and ethnic and sex groups in US adults.
Methods
Study Design and Population
This cohort study included data on adults 18 years or older who completed the USDA Adult Food Security Survey Module during the 10 cycles of the National Health and Nutrition Examination Survey (NHANES) from 1999-2000 to 2017-2018. The study design for NHANES has been described in detail previously.17 A total of 57 404 participants were eligible for this study, after excluding participants who had missing values on the Adult Food Security Survey and participants with no linked mortality data (n = 1800). The study protocols were approved by the institutional review board of the National Center for Health Statistics, and written informed consent was obtained from each participant.
Assessments of Exposure
Food insecurity was assessed with the USDA Adult Food Security Survey Module,18 including 10 questions about food security (eTable 1 in Supplement 1). According to the USDA guideline, the severity of food security is classified into 4 levels: full security (0 affirmative responses), marginal security (1 or 2 affirmative responses), low security (3-5 affirmative responses), and very low security (6-10 affirmative responses). The validity of this module has been confirmed in various racial and ethnic groups in several previous studies.19,20
Assessments of Covariates
Based on previous studies,7,8,9,10 the covariates in this study included age, sex, race and ethnicity (non-Hispanic Black [hereafter, Black], Mexican American, Other Hispanic, non-Hispanic White [hereafter, White], and other [other race, including multiracial]), educational attainment (less than high school, high school, and college or above), ratio of family income to poverty (<1.3 [low], ≥1.3 and <3.5 [intermediate], and ≥3.5 [high]), health insurance coverage (yes or no), marital status (single or not), total number of people in the household, and immigration status (born within the US or not). The above information was self-reported and collected by standardized questionnaires. Previous studies have shown that food insecurity is associated with negative health outcomes through nutritional/anthropometric (body mass index [BMI] and diet quality), behavioral (smoking, heavy drinking, and physical inactivity), mental health (access to a mental health professional), access to care (routine place to go for health care), and chronic disease (diabetes, hypertension, CVD, and cancer) pathways21,22; thus, it is not appropriate to adjust for the above factors in the analysis. However, as an exploratory analysis, we further adjusted the above potential mediators in multivariable analyses.
BMI was calculated as weight in kilograms divided by standing height in meters squared. The Alternative Healthy Eating Index diet score (without considering alcohol intake) was calculated to evaluate the diet quality by using the mean values of each dietary component collected from 2 nonconsecutive 24-hour dietary recalls at the baseline. Unhealthy diet was defined as an Alternative Healthy Eating Index score in the bottom 60% of cohort distribution. The information on current smoking, heavy drinking (women: >14 g/d, men: >28 g/d), physical inactivity (<150 minutes of moderate-intensity activity per week or <75 minutes of vigorous-intensity activity per week or an equivalent combination per week), access to a mental health professional (“During the past 12 months, have you seen or talked to a mental health professional such as a psychologist, psychiatrist, psychiatric nurse, or clinical social worker about your health?”, yes or no), and routine place to go for health care (whether the participant had a routine facility to go, other than an emergency department, yes or no) was collected by standardized questionnaires. Diabetes was defined as a self-reported history of diabetes or a fasting plasma glucose level of 126 mg/dL or greater or a hemoglobin A1c level of 6.5% or greater or taking antidiabetic medications. Hypertension was defined as a self-reported history of hypertension or a systolic blood pressure of 140 mm Hg or higher or a diastolic blood pressure of 90 mm Hg or higher or taking antihypertensive medications. CVD was defined as a self-reported history of coronary heart disease, myocardial infarction (referred to as “heart attack” on the questionnaire), or stroke. Cancer was defined as a self-reported history of any kind of cancer or a malignant neoplasm.
Assessments of Death
Information on death and death date was identified through linkage to the National Death Index through December 31, 2019. Deaths that occurred at ages younger than 80 years were defined as premature,23 which was the approximate life expectancy of US adults in 2019. We also evaluated associations of food security with premature death before age 75 years in sensitivity analysis.
Statistical Analysis
To ensure nationally representative estimates, sampling weights were considered in all analyses to account for oversampling of certain subgroups and complex sample design. Multivariable Cox proportional hazards models were used to evaluate the association between the levels of food security and risk of premature mortality; follow-up time was used as the underlying time metric. The proportional hazards assumption was tested by log-minus-log plots, and no violation was found (eFigure 1 in Supplement 1). In multivariable analyses, models were adjusted for age, race and ethnicity, educational attainment, family income, health insurance, marital status, total number of people in the household, and immigration status. Details for the above covariates are described in the eMethods in Supplement 1. To assess the potential sex and racial or ethnic differences, we performed stratified analyses by sex and race and ethnicity groups (Black, Hispanic [only for data after 2007, combined Mexican American and other Hispanic persons into “Hispanic group”], and White).24 To evaluate interactions between food security and sex and race and ethnicity, multiplicative interaction was assessed by adding interaction terms to the Cox models. Missing data on covariates were coded as a missing indicator category for categorical variables. Several sensitivity analyses were performed. First, we imputed missing covariate values with multiple imputation by chained equations and repeated the analyses (educational attainment, family income, health insurance, marital status, and immigration status). Second, to control the potential impact of shifting economic conditions on the results, we adjusted for NHANES cycles in additional sensitivity analysis. Third, we also considered the potential impact of the Supplemental Nutrition Assistance Program (SNAP) on the results.
The life expectancy of participants with different levels of food security (henceforth, exposure groups) was calculated using the life tables.25,26,27 The life tables were constructed starting at age 50 years and ending at age 100 years with the following 3 estimates to calculate the cumulative survival from age 50 years onward: (1) adjusted hazard ratios of all-cause mortality in each exposure group (marginal, low, or very low food security groups) vs the reference (full food security group) from NHANES, (2) the age-specific population prevalence (categorized in 10-year age groups) of each exposure group from NHANES, and (3) the age-specific population all-cause mortality rate in 2019 from the National Center for Health Statistics. The estimated lower survival time (years) associated with lower level of food security was estimated as the difference in the life expectancy at any given age between the reference group and each of the exposure groups. To determine which cause-specific mortality differences were significantly associated with the total change in life expectancy, we also estimated the cause-specific contributions to the life expectancy difference between participants with full food security and those with very low food security with the Arriaga decomposition method.28,29 Details of the methods used for estimating the difference in expected survival time and cause-specific contributions to the life expectancy difference are described in the eMethods in Supplement 1.
All statistical analyses were conducted using SAS, version 9.4 (SAS Institute), and R, version 3.6.1 (R Foundation for Statistical Computing). Monte Carlo simulation with 1000 runs was used to calculate the CIs of the life expectancy estimation with R package boot. All statistical tests were 2-sided, and we considered P < .05 to be statistically significant. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Results
Basic Characteristics of Study Participants According to Levels of Food Security
A total of 57 404 adults were included in the study (weighted mean [SE] age, 46.0 [0.19] years; 51.8% female; 12 281 Black individuals [21.4%]; 10 421 Mexican individuals [18.2%]; 4627 Other Hispanic individuals [8.1%]; 24 817 White individuals [43.2%]; and 5258 individuals of other races, including multiracial [9.2%]). Characteristics of the sample with weighted population numbers are presented in Table 1. Of the participants, 78.4% had full food security; 8.5% of participants had marginal food security; 7.4% of participants had low food security; and 5.6% of participants had very low food security. Participants with lower levels of food security were younger; more likely to be single; more likely to be Black, Mexican American, or Other Hispanic; more likely to have a lower level of family income, a lower level of educational attainment, and a larger household size; and less likely to be covered by health insurance. Participants with a lower level of food security were also more likely to have an unhealthy lifestyle (unhealthy diet, current smoking, heavy drinking, and physical inactivity) and a higher prevalence of obesity, hypertension, diabetes, CVD, and cancer. Moreover, they were more likely to have underlying psychological problems (access to a mental health professional) and lack regular health care. The SNAP participation rate increased significantly as food security levels decreased. Notably, we found that the overall demographic characteristics of the marginal food security group were more similar to the low food security group than the full food security group.
Table 1. Characteristics of Participants at Baseline by Different Levels of Food Security.
| Characteristic | Weighted No. (%), millions | |||
|---|---|---|---|---|
| Food security | ||||
| Full | Marginal | Low | Very low | |
| Prevalence | 171.0 (78.4) | 18.7 (8.5) | 16.2 (7.4) | 12.3 (5.6) |
| Participants, No. | 41 177 | 6329 | 5900 | 3998 |
| Age, mean (SE), y | 47.3 (0.21) | 41.4 (0.33) | 41.1 (0.35) | 40.7 (0.39) |
| Sex | ||||
| Female | 88.1 (51.3) | 10.3 (55.0) | 8.6 (53.3) | 6.5 (53.0) |
| Male | 83.7 (48.7) | 8.4 (45.0) | 7.6 (46.7) | 5.8 (47.0) |
| Single | 57.6 (33.5) | 8.0 (42.6) | 7.3 (45.3) | 6.4 (51.6) |
| Race and ethnicity | ||||
| Black | 16.2 (9.4) | 3.6 (19.3) | 2.9 (18.0) | 2.2 (18.1) |
| Mexican American | 10.0 (5.8) | 3.1 (16.8) | 3.3 (20.6) | 1.6 (12.8) |
| Other Hispanic | 7.3 (4.3) | 1.9 (10.3) | 1.8 (11.0) | 1.2 (10.1) |
| White | 126.0 (73.5) | 8.6 (46.1) | 7.1 (43.5) | 6.5 (52.4) |
| Othera | 12.0 (7.0) | 1.4 (7.5) | 1.1 (6.9) | 0.8 (6.6) |
| Educational attainment | ||||
| Less than high school | 22.7 (13.2) | 4.9 (26.3) | 5.6 (34.8) | 3.7 (29.9) |
| High school | 38.0 (22.1) | 5.4 (28.7) | 4.0 (24.7) | 3.5 (28.5) |
| Some college or above | 105.6 (61.5) | 7.5 (40.3) | 5.9 (36.1) | 4.5 (36.9) |
| Family income | ||||
| Low | 23.1 (13.4) | 7.5 (40.1) | 8.1 (49.6) | 7.4 (60.0) |
| Intermediate | 55.7 (32.4) | 8.0 (42.6) | 5.8 (35.6) | 3.9 (31.8) |
| High | 82.5 (48.0) | 2.0 (10.6) | 1.1 (6.5) | 0.5 (4.3) |
| Immigration status (not born in the US) | 24.8 (14.4) | 4.9 (25.9) | 4.8 (29.8) | 2.2 (18.2) |
| Total No. of people in the household, mean (SE) | 2.9 (0.02) | 3.5 (0.04) | 3.6 (0.05) | 3.4 (0.07) |
| Unhealthy diet | 81.0 (47.1) | 10.5 (56.1) | 9.3 (57.4) | 7.7 (62.9) |
| Obesity | 56.1 (32.7) | 7.4 (39.5) | 6.7 (41.4) | 5.0 (40.4) |
| Current smoking | 30.3 (17.6) | 5.3 (28.1) | 4.9 (30.5) | 5.3 (43.1) |
| Heavy alcohol drinking | 17.4 (10.2) | 1.2 (6.6) | 1.1 (6.9) | 0.8 (6.4) |
| Physical inactivity | 100.5 (58.5) | 13.0 (69.2) | 11.8 (72.8) | 8.8 (71.1) |
| No health insurance coverage | 22.6 (13.2) | 5.9 (31.5) | 6.0 (36.7) | 4.3 (35.2) |
| Access to a mental health professional | 12.5 (7.3) | 2.3 (10.5) | 1.9 (12.0) | 2.2 (18.2) |
| No regular health care facility | 24.1 (14.0) | 4.4 (23.6) | 4.4 (26.9) | 3.2 (25.7) |
| Hypertension | 62.1 (36.2) | 6.2 (33.3) | 5.5 (34.0) | 4.5 (36.9) |
| Diabetes | 18.9 (11.0) | 2.6 (13.9) | 2.3 (14.0) | 1.8 (14.7) |
| Cancer | 17.1 (9.9) | 1.2 (6.2) | 0.9 (5.4) | 0.9 (7.5) |
| Cardiovascular disease | 11.5 (6.7) | 1.2 (6.4) | 1.3 (7.9) | 1.1 (8.6) |
| SNAP participants | 11.7 (6.8) | 5.6 (30.1) | 5.9 (36.7) | 5.8 (46.8) |
Abbreviation: SNAP, Supplemental Nutrition Assistance Program.
Other includes the response option of other, including multiracial.
Association of Food Security and Risk of Premature Mortality
During a median (IQR) follow-up of 9.3 (5.0-14.3) years, a total of 4263 all-cause premature deaths were documented. In the total population, after adjustment for age, sex, race and ethnicity, educational attainment, family income, health insurance, marital status, total number of people in the household, and immigration status, lower levels of food security were significantly associated with a higher risk of all-cause premature mortality. Notably, we found the hazard ratio of premature mortality was already higher in the marginal food security group compared with full food security group; the adjusted hazard ratios were 1 (reference), 1.50 (95% CI, 1.31-1.71), 1.44 (95% CI, 1.24-1.68), and 1.81 (95% CI, 1.56-2.10) across groups with full food security, marginal food security, low food security, and very low food security, respectively (P < .001 for trend) (Table 2). These results remained robust in sensitivity analyses using younger than 75 years old to define premature death or using multiple imputation to impute data for missing covariates (eTables 2 and 3 in Supplement 1). The results also did not change appreciably after further adjustment for NHANES cycles or SNAP (eTables 4 and 5 in Supplement 1). As an exploratory analysis, further adjustment for potential mediators (eg, unhealthy diet, obesity) attenuated the observed associations, but the results remained significant at all levels of food security (Table 2).
Table 2. Hazard Ratios for Categories of Food Security and Premature Mortality.
| Model | Food security category, hazard ratio (95% CI) | P value for trend | |||
|---|---|---|---|---|---|
| Full | Marginal | Low | Very | ||
| All | |||||
| Cases, No./total No.a | 3009/41 177 | 435/6329 | 448/5900 | 371/3998 | NA |
| Age, sex, race and ethnicity adjustedb | 1 [Reference] | 1.94 (1.70-2.21) | 2.01 (1.74-2.32) | 2.78 (2.43-3.18) | <.001 |
| Multivariable adjustedb | 1 [Reference] | 1.50 (1.31-1.71) | 1.44 (1.24-1.68) | 1.81 (1.56-2.10) | <.001 |
| Further adjusted mediatorsb,c | 1 [Reference] | 1.26 (1.10-1.45) | 1.19 (1.02-1.38) | 1.35 (1.17-1.57) | <.001 |
| Women | |||||
| Cases, No./total No.a | 1173/21 053 | 183/3462 | 184/3111 | 167/2085 | NA |
| Age, sex, race and ethnicity adjustedb | 1 [Reference] | 2.00 (1.62-2.46) | 2.05 (1.67-2.53) | 3.33 (2.75-4.05) | <.001 |
| Multivariable-adjustedb | 1 [Reference] | 1.62 (1.32-2.00) | 1.55 (1.24-1.95) | 2.29 (1.83-2.86) | <.001 |
| Further adjusted mediatorsb,c | 1 [Reference] | 1.34 (1.09-1.65) | 1.24 (0.98-1.57) | 1.61 (1.28-2.02) | <.001 |
| Men | |||||
| Cases, No./total No.a | 1836/20 124 | 252/2867 | 264/2789 | 204/1913 | NA |
| Age, sex, race and ethnicity adjustedb | 1 [Reference] | 1.91 (1.60-2.28) | 1.98 (1.65-2.38) | 2.38 (1.96-2.89) | <.001 |
| Multivariable adjustedb | 1 [Reference] | 1.42 (1.18-1.70) | 1.37 (1.13-1.66) | 1.46 (1.19-1.78) | <.001 |
| Further adjusted mediatorsb,c | 1 [Reference] | 1.23 (1.02-1.48) | 1.15 (0.95-1.40) | 1.14 (0.91-1.42) | .06 |
| Non-Hispanic Black individuals | |||||
| Cases, No./total No.a | 811/8141 | 124/1689 | 143/1410 | 108/1041 | NA |
| Age, sex, race and ethnicity adjustedb | 1 [Reference] | 1.19 (0.98-1.46) | 1.57 (1.26-1.96) | 1.79 (1.40-2.29) | <.001 |
| Multivariable-adjustedb | 1 [Reference] | 1.02 (0.83-1.25) | 1.25 (01.01-1.56) | 1.33(1.01-1.75) | .01 |
| Further adjusted mediatorsb,c | 1 [Reference] | 0.95 (0.77-1.16) | 1.12 (0.90-1.40) | 1.08 (0.81-1.46) | .43 |
| Hispanic individuals | |||||
| Cases, No./total No.a | 202/5031 | 46/1551 | 49/1639 | 37/971 | NA |
| Age, sex, race and ethnicity adjustedb | 1 [Reference] | 1.15 (0.78-1.68) | 0.98 (0.69-1.41) | 1.34 (0.91-1.98) | .26 |
| Multivariable adjustedb | 1 [Reference] | 1.05 (0.72-1.53) | 0.85 (0.56-1.27) | 1.06 (0.71-1.58) | .84 |
| Further adjusted mediatorsb,c | 1 [Reference] | 1.04 (0.70-1.55) | 0.86 (0.57-1.30) | 1.00 (0.67-1.49) | .70 |
| Non-Hispanic White individuals | |||||
| Cases, No./total No.a | 1522/20 294 | 167/1675 | 149/1472 | 165/1376 | NA |
| Age and sex adjustedb | 1 [Reference] | 2.33 (1.94-2.80) | 2.41 (1.98-2.95) | 3.48 (2.90-4.17) | <.001 |
| Multivariable adjustedb | 1 [Reference] | 1.68 (1.39-2.03) | 1.58 (1.27-1.96) | 2.07 (1.70-2.53) | <.001 |
| Further adjusted mediatorsb,c | 1 [Reference] | 1.35 (1.11-1.66) | 1.25 (0.99-1.57) | 1.48 (1.21-1.80) | <.001 |
Abbreviation: NA, not applicable.
Unweighted number of participants.
Sampling weights were considered in analyses. Multivariable adjusted for age, sex, race and ethnicity, educational attainment, family income, health insurance, marital status, total number of people in the household, and immigration status. Women compared with men: P = .009 for interaction; White individuals compared with Black individuals: P < .001 for interaction; White individuals compared with Hispanic individuals: P < .001 for interaction.
Mediators included current smoking, heavy drinking, unhealthy diet, physical inactivity, underlying psychological problems, routine place to go for health care, obesity, diabetes, hypertension, cardiovascular disease, and cancer.
Moreover, we found the association between levels of food security with premature mortality appeared to be stronger in women than in men (hazard ratios comparing very low food security with full food security, 2.29 [95% CI, 1.83-2.86] in women and 1.46 [95% CI, 1.19-1.78] in men; P = .009 for interaction) and stronger in White adults than in Black adults (hazard ratios comparing very low food security with full food security, 1.33 [95% CI, 1.01-1.75] in Black adults and 2.07 [95% CI, 1.70-2.53] in White adults; P < .001 for interaction) or in Hispanic adults (hazard ratios comparing very low food security with full food security, 1.06 [95% CI, 0.71-1.58]; P < .001 for interaction). The levels of food security were not associated with premature mortality in Hispanic adults (Table 2).
Association of Food Security With Estimated Life Expectancy
Lower levels of food security were associated with a shorter life expectancy in the total population. At age 50 years, the estimated total life expectancy was 32.5 (95% CI, 32.4-32.6), 29.9 (95% CI, 28.9-31.0), 30.0 (95% CI, 28.9-31.0), and 28.0 (95% CI, 26.8-29.2) years in adults with full food security, marginal food security, low food security, and very low food security, respectively (Figure 1A). Equivalently, adults with marginal, low, or very low food security lived on average 2.6 (95% CI, 1.5-3.7), 2.5 (95% CI, 1.4-3.7), or 4.5 (95% CI, 3.2-5.8) years shorter at age 50 years, respectively, compared with those with full food security. In the cause-specific decomposition analysis of the life expectancy differences between full food security and very low food insecurity, 27.1% of estimated lost life expectancy at age 50 years was attributable to CVD death, 23.8% was attributable to cancer death, and 49.1% was attributable to other causes (Figure 1A).
Figure 1. Estimates of Cumulative Survival Time From Age 50 Years Onward Among Adults With Different Levels of Food Security in Total Population, Women, and Men.
CVD indicates cardiovascular disease; LE, life expectancy; YLL, years of life lost.
Women with marginal, low, or very low food security lived on average 2.7 (95% CI, 1.3-4.1), 2.5 (95% CI, 1.0-4.0), or 5.8 (95% CI, 3.9-7.4) years shorter at age 50 years, respectively, compared with those with full food security (Figure 1B). The corresponding differences in men were 2.5 (95% CI, 0.9-4.2), 2.5 (95% CI, 0.9-4.3), or 3.0 (95% CI, 1.0-4.8), respectively (Figure 1C). In women, 22.0% of the lost life expectancy at age 50 years due to very low food security was attributable to CVD death, and 23.6% was attributable to cancer death (Figure 1B). The corresponding percentages were 36.5% and 25.6%, respectively, in men (Figure 1C).
We also performed stratified analyses by racial and ethnic groups (Figure 2). We observed that lower levels of food security were associated with a shorter estimated life expectancy in Black and White adults. In White adults, compared with adults with full food security, adults with marginal, low, or very low food security lived on average 3.2 (95% CI, 1.7-4.7), 3.5 (95% CI, 1.9-5.3), or 6.2 (95% CI, 4.4-8.0) years shorter at age 50 years, respectively (Figure 2A). Compared with adults with full food security, 27.4% of the lost life expectancy due to very low food security was attributable to CVD death, and 23.9% was attributable to cancer death (Figure 2A). Black adults with very low food security lived on average 2.3 (95% CI, 0.3-4.7) years shorter at age 50 years compared with Black adults with full food security (Figure 2B). The levels of food security were not associated with life expectancy in Hispanic adults.
Figure 2. Estimates of Cumulative Survival Time From Age 50 Years Onward Among White and Black Adults With Different Levels of Food Security.
CVD indicates cardiovascular disease; LE, life expectancy; YLL, years of life lost.
Moreover, the differences in life expectancy between food security groups decrease with age (eFigure 2 in Supplement 1). Notably, until age 70 years, the estimated life expectancy remained significantly lower in the very low food security group than in the full food security group in all participants and each subgroup (eTable 6 in Supplement 1).
Discussion
In a nationally representative cohort of US adults, we found that lower levels of food security, even marginal food security, were significantly associated with a higher risk of all-cause premature mortality and a shorter life expectancy compared with full food security. Moreover, we found significant racial and ethnic and sex differences in the association of food insecurity with all-cause premature mortality and life expectancy.
To our knowledge, this is the first study evaluating the associations of the entire range of food security with premature mortality and life expectancy in US adults. Consistent with our study, a study from the Canadian Community Health Survey found a significant graded association between levels of food security and premature mortality. Notably, they also found that the risk of premature death was significantly higher (about 10%) even in the marginal food security group compared to the full food security group.9 Another study from the National Health Interview Survey has evaluated the association between levels of food security and 10 common chronic diseases and found that the risk of hypertension, diabetes, and chronic obstructive pulmonary disease was 20%, 59%, and 158% higher, respectively, in the marginal food security group compared to the full food security group.4 In the current study, we found that even marginal food security, which was typically classified as food security in previous studies, was associated with a 50% higher risk of premature mortality and a 2.6-year shorter life expectancy at age 50 years compared with those with full food security. These results suggest that we should consider all 4 categories of food security in future analyses, rather than a simple binary combination.
Notably, we found that associations of levels of food security with premature mortality and life expectancy significantly varied by racial and ethnic groups. The association was stronger in White adults than in Black adults, and no significant association was found in Hispanic adults. These findings were somewhat surprising given the greater prevalence of food insecurity in Black and Hispanic adults than in White adults in the US.3,14 It is widely accepted that food insecurity is associated with negative health outcomes largely through its effects on diet quality,21,22 whereas this pathway may vary widely across racial and ethnic groups. Two previous studies from NHANES found that significant associations of food insecurity with poor-quality diet were observed in White individuals but not in Black and Hispanic individuals.30,31 Another study found that lower levels of food security were significantly associated with lower diet quality in both Black and White individuals; however, the association appeared to be much stronger in White individuals than Black individuals.32 Notably, we cannot rule out that the observed racial and ethnic disparity in the association between food insecurity and premature mortality and life expectancy is due to the smaller sample size for Black and Hispanic individuals than White individuals. Future studies are needed to explore its potential mechanisms.
Moreover, we found the association of food security with premature mortality and life expectancy appeared to be stronger in women than in men. This finding was supported by previous studies assessing the associations between food insecurity and obesity, CVD, and cardiovascular mortality, with stronger associations observed in women.7,10,33,34 In most cases, women have a higher level of decision-making power than men when it comes to food purchases and distribution within the family.35 This role in the family may cause more psychological burden in women when facing food insecurity, and some women may also sacrifice their own nutritional intake to meet the nutritional intake of the family as much as possible.36,37,38
The present findings may have a great public health implication. Life expectancy estimates provide policymakers and the public with a more intuitive picture than traditional mortality risk estimates because they provide an absolute quantitative assessment. Our results suggest that in addition to encouraging people to improve their lifestyle and cardiovascular health,25,27,39 improving food security may also be a way to curb the stagnant life expectancy of US residents.40 Moreover, our study provides a comprehensive picture of the association of food insecurity with life expectancy and premature mortality among US adults with different sex and race and ethnicity, providing scientific evidence for future policies and strategies.
Limitations
This study had several limitations. First, because information on food security was only available at baseline, we were unable to analyze the impact of changes in food security levels on premature mortality and life expectancy. Notably, several previous studies have explored the association between changes in food security status and BMI (an important mediator in the association of food insecurity with mortality), but none of them found that persistent food insecurity was more strongly associated with the outcomes than transient food insecurity.41,42,43,44 Second, we could not exclude the possibility that food insecurity is a marker for other low socioeconomic factors, and residual confounding might exist even though we adjusted for a wide range of potential confounders. Third, Asian and multiracial individuals were not included in the stratified analyses by race and ethnicity due to their limited sample size. Fourth, although we carefully selected potential confounders in our analysis based on previous literature,7,8,9,10 some potential confounders were not included due to incomplete information (eg, disability status, sexual orientation, and number of children). Fifth, the lack of data on nutrition security is another limitation in the study. Different from food security, nutrition security focuses on assessing the uncertainty of access to nutritious food.13 Future research should focus on the association between nutrition security and health outcomes.
Conclusions
In this cohort study, although the association of food security with life expectancy varied across sex and racial and ethnic groups, overall, lower levels of food security were associated with a higher risk of premature mortality and a shorter life expectancy. These results call for a higher prioritization in tackling food insecurity as a means to improve population health and reduce health inequalities.
eMethods.
eTable 1. US Department of Agriculture Adult Food Security Survey Module Questions
eTable 2. Hazard ratios and 95% confidence interval for category of food security with the hazard of premature mortality (Deaths that occurred at ages younger than 75)
eTable 3. Hazard ratios and 95% confidence interval for category of food security with the hazard of premature mortality after multiple imputation for covariates
eTable 4. Hazard ratios and 95% confidence intervals for category of food security and premature mortality (further adjusted for cycles)
eTable 5. Hazard ratios and 95% confidence interval for category of food security with the hazard of premature mortality (further adjustment for SNAP)
eTable 6. Life expectancy at age 60 or 70 years by food security level
eFigure 1. The proportional hazards assumption was tested by log-minus-log plots
eFigure 2. Years of life expectancy lost at different ages (50-100 years old) due to marginal, low or very low levels of food security compared to full food security group
Data Sharing Statement
References
- 1.Murray CJ, Atkinson C, Bhalla K, et al. ; US Burden of Disease Collaborators . The state of US health, 1990-2010: burden of diseases, injuries, and risk factors. JAMA. 2013;310(6):591-608. doi: 10.1001/jama.2013.13805 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Wang DD, Li Y, Afshin A, et al. Global improvement in dietary quality could lead to substantial reduction in premature death. J Nutr. 2019;149(6):1065-1074. doi: 10.1093/jn/nxz010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Coleman-Jensen A, Rabbitt MP, Gregory CA, Singh A. Household food security in the United States in 2021. Accessed December 12, 2023. https://www.ers.usda.gov/webdocs/publications/104656/err-309.pdf
- 4.Christian VJ, Miller KR, Martindale RG. Food insecurity, malnutrition, and the microbiome. Curr Nutr Rep. 2020;9(4):356-360. doi: 10.1007/s13668-020-00342-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gundersen C, Ziliak JP. Food insecurity and health outcomes. Health Aff (Millwood). 2015;34(11):1830-1839. doi: 10.1377/hlthaff.2015.0645 [DOI] [PubMed] [Google Scholar]
- 6.Gundersen C, Tarasuk V, Cheng J, de Oliveira C, Kurdyak P. Food insecurity status and mortality among adults in Ontario, Canada. PLoS One. 2018;13(8):e0202642. doi: 10.1371/journal.pone.0202642 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sun Y, Liu B, Rong S, et al. Food insecurity is associated with cardiovascular and all-cause mortality among adults in the United States. J Am Heart Assoc. 2020;9(19):e014629. doi: 10.1161/JAHA.119.014629 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Walker RJ, Chawla A, Garacci E, et al. Assessing the relationship between food insecurity and mortality among US adults. Ann Epidemiol. 2019;32:43-48. doi: 10.1016/j.annepidem.2019.01.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Men F, Gundersen C, Urquia ML, Tarasuk V. Association between household food insecurity and mortality in Canada: a population-based retrospective cohort study. CMAJ. 2020;192(3):E53-E60. doi: 10.1503/cmaj.190385 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Vercammen KA, Moran AJ, McClain AC, Thorndike AN, Fulay AP, Rimm EB. Food security and 10-year cardiovascular disease risk among US adults. Am J Prev Med. 2019;56(5):689-697. doi: 10.1016/j.amepre.2018.11.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Serchen J, Atiq O, Hilden D; Health and Public Policy Committee of the American College of Physicians . Strengthening food and nutrition security to promote public health in the United States: a position paper from the American College of Physicians. Ann Intern Med. 2022;175(8):1170-1171. doi: 10.7326/M22-0390 [DOI] [PubMed] [Google Scholar]
- 12.Thorndike AN, Gardner CD, Kendrick KB, et al. ; American Heart Association Advocacy Coordinating Committee . Strengthening US food policies and programs to promote equity in nutrition security: a policy statement from the American Heart Association. Circulation. 2022;145(24):e1077-e1093. doi: 10.1161/CIR.0000000000001072 [DOI] [PubMed] [Google Scholar]
- 13.Brandt EJ, Mozaffarian D, Leung CW, Berkowitz SA, Murthy VL. Diet and food and nutrition insecurity and cardiometabolic disease. Circ Res. 2023;132(12):1692-1706. doi: 10.1161/CIRCRESAHA.123.322065 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Brandt EJ, Chang T, Leung C, Ayanian JZ, Nallamothu BK. Food insecurity among individuals with cardiovascular disease and cardiometabolic risk factors across race and ethnicity in 1999-2018. JAMA Cardiol. 2022;7(12):1218-1226. doi: 10.1001/jamacardio.2022.3729 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Seligman HK, Laraia BA, Kushel MB. Food insecurity is associated with chronic disease among low-income NHANES participants. J Nutr. 2010;140(2):304-310. doi: 10.3945/jn.109.112573 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Jyoti DF, Frongillo EA, Jones SJ. Food insecurity affects school children’s academic performance, weight gain, and social skills. J Nutr. 2005;135(12):2831-2839. doi: 10.1093/jn/135.12.2831 [DOI] [PubMed] [Google Scholar]
- 17.National Center for Health Statistics . About the National Health and Nutrition Examination Survey. Accessed December 12, 2023. https://www.cdc.gov/nchs/nhanes/about_nhanes.htm
- 18.Bickel G, Nord M, Price C, Hamilton W, Cook J. Guide to measuring household food security, 2000. Accessed December 12, 2023. https://nhis.ipums.org/nhis/resources/FSGuide.pdf
- 19.Derrickson JP, Fisher AG, Anderson JE. The core food security module scale measure is valid and reliable when used with Asians and Pacific Islanders. J Nutr. 2000;130(11):2666-2674. doi: 10.1093/jn/130.11.2666 [DOI] [PubMed] [Google Scholar]
- 20.Hamilton WL, Cook JT, Thompson WW, et al. Household food security in the United States in 1995: summary report of the Food Security Measurement Project. Accessed December 12, 2023. https://www.fns.usda.gov/household-food-security-united-states-1995-summary-report-food-security-measurement-project
- 21.Palakshappa D, Ip EH, Berkowitz SA, et al. Pathways by which food insecurity is associated with atherosclerotic cardiovascular disease risk. J Am Heart Assoc. 2021;10(22):e021901. doi: 10.1161/JAHA.121.021901 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chang R, Javed Z, Taha M, et al. Food insecurity and cardiovascular disease: current trends and future directions. Am J Prev Cardiol. 2021;9:100303. doi: 10.1016/j.ajpc.2021.100303 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Johnson NB, Hayes LD, Brown K, Hoo EC, Ethier KA; Centers for Disease Control and Prevention (CDC) . CDC national health report: leading causes of morbidity and mortality and associated behavioral risk and protective factors—United States, 2005-2013. MMWR Suppl. 2014;63(4):3-27. [PubMed] [Google Scholar]
- 24.Johnson CL, Paulose-Ram R, Ogden CL, et al. National Health and Nutrition Examination Survey: analytic guidelines, 1999-2010. Vital Health Stat 2. 2013;(161):1-24. [PubMed] [Google Scholar]
- 25.Ma H, Wang X, Xue Q, et al. Cardiovascular health and life expectancy among adults in the United States. Circulation. 2023;147(15):1137-1146. doi: 10.1161/CIRCULATIONAHA.122.062457 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ma H, Xue Q, Wang X, et al. Adding salt to foods and hazard of premature mortality. Eur Heart J. 2022;43(30):2878-2888. doi: 10.1093/eurheartj/ehac208 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Li Y, Pan A, Wang DD, et al. Impact of healthy lifestyle factors on life expectancies in the US population. Circulation. 2018;138(4):345-355. doi: 10.1161/CIRCULATIONAHA.117.032047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Arriaga EE. Measuring and explaining the change in life expectancies. Demography. 1984;21(1):83-96. doi: 10.2307/2061029 [DOI] [PubMed] [Google Scholar]
- 29.Auger N, Feuillet P, Martel S, Lo E, Barry AD, Harper S. Mortality inequality in populations with equal life expectancy: Arriaga’s decomposition method in SAS, Stata, and Excel. Ann Epidemiol. 2014;24(8):575-580, 580.e1. doi: 10.1016/j.annepidem.2014.05.006 [DOI] [PubMed] [Google Scholar]
- 30.Leung CW, Tester JM. The association between food insecurity and diet quality varies by race/ethnicity: an analysis of National Health and Nutrition Examination Survey 2011-2014 results. J Acad Nutr Diet. 2019;119(10):1676-1686. doi: 10.1016/j.jand.2018.10.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Kardashian A, Dodge JL, Terrault NA. Racial and ethnic differences in diet quality and food insecurity among adults with fatty liver and significant fibrosis: a US population-based study. Aliment Pharmacol Ther. 2022;56(9):1383-1393. doi: 10.1111/apt.17219 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Allen AJ, Kuczmarski MF, Evans MK, Zonderman AB, Waldstein SR. Race differences in diet quality of urban food-insecure Blacks and Whites reveals resiliency in Blacks. J Racial Ethn Health Disparities. 2016;3(4):706-712. doi: 10.1007/s40615-015-0189-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Franklin B, Jones A, Love D, Puckett S, Macklin J, White-Means S. Exploring mediators of food insecurity and obesity: a review of recent literature. J Community Health. 2012;37(1):253-264. doi: 10.1007/s10900-011-9420-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Jung NM, de Bairros FS, Pattussi MP, Pauli S, Neutzling MB. Gender differences in the prevalence of household food insecurity: a systematic review and meta-analysis. Public Health Nutr. 2017;20(5):902-916. doi: 10.1017/S1368980016002925 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Martin MA, Lippert AM. Feeding her children, but risking her health: the intersection of gender, household food insecurity and obesity. Soc Sci Med. 2012;74(11):1754-1764. doi: 10.1016/j.socscimed.2011.11.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Ivers LC, Cullen KA. Food insecurity: special considerations for women. Am J Clin Nutr. 2011;94(6):1740S-1744S. doi: 10.3945/ajcn.111.012617 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Stevens CA. Exploring food insecurity among young mothers (15-24 years). J Spec Pediatr Nurs. 2010;15(2):163-171. doi: 10.1111/j.1744-6155.2010.00235.x [DOI] [PubMed] [Google Scholar]
- 38.McIntyre L, Glanville NT, Raine KD, Dayle JB, Anderson B, Battaglia N. Do low-income lone mothers compromise their nutrition to feed their children? CMAJ. 2003;168(6):686-691. [PMC free article] [PubMed] [Google Scholar]
- 39.Wang X, Ma H, Li X, et al. Association of cardiovascular health with life expectancy free of cardiovascular disease, diabetes, cancer, and dementia in UK adults. JAMA Intern Med. 2023;183(4):340-349. doi: 10.1001/jamainternmed.2023.0015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Ho JY, Hendi AS. Recent trends in life expectancy across high income countries: retrospective observational study. BMJ. 2018;362:k2562. doi: 10.1136/bmj.k2562 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Vercammen KA, Moran AJ, Carnethon MR, et al. Longitudinal analysis of food insufficiency and cardiovascular disease risk factors in the CARDIA study. Am J Prev Med. 2022;62(1):65-76. doi: 10.1016/j.amepre.2021.06.020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Olson CM, Strawderman MS. The relationship between food insecurity and obesity in rural childbearing women. J Rural Health. 2008;24(1):60-66. doi: 10.1111/j.1748-0361.2008.00138.x [DOI] [PubMed] [Google Scholar]
- 43.Whitaker RC, Sarin A. Change in food security status and change in weight are not associated in urban women with preschool children. J Nutr. 2007;137(9):2134-2139. doi: 10.1093/jn/137.9.2134 [DOI] [PubMed] [Google Scholar]
- 44.Jones SJ, Frongillo EA. The modifying effects of Food Stamp Program participation on the relation between food insecurity and weight change in women. J Nutr. 2006;136(4):1091-1094. doi: 10.1093/jn/136.4.1091 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods.
eTable 1. US Department of Agriculture Adult Food Security Survey Module Questions
eTable 2. Hazard ratios and 95% confidence interval for category of food security with the hazard of premature mortality (Deaths that occurred at ages younger than 75)
eTable 3. Hazard ratios and 95% confidence interval for category of food security with the hazard of premature mortality after multiple imputation for covariates
eTable 4. Hazard ratios and 95% confidence intervals for category of food security and premature mortality (further adjusted for cycles)
eTable 5. Hazard ratios and 95% confidence interval for category of food security with the hazard of premature mortality (further adjustment for SNAP)
eTable 6. Life expectancy at age 60 or 70 years by food security level
eFigure 1. The proportional hazards assumption was tested by log-minus-log plots
eFigure 2. Years of life expectancy lost at different ages (50-100 years old) due to marginal, low or very low levels of food security compared to full food security group
Data Sharing Statement
