Environmental, climate and health-related dietary motivations are associated with higher diet quality in a national sample of US adults with lower incomes

Abstract

Background:

Plant-based diets can have co-benefits for human and planetary health. Associations between environmental, climate and health concerns and dietary intake in US adults are understudied, particularly for underserved populations.

Objectives:

The study objectives were to assess how dietary choices motivated by the environment, climate and health vary by sociodemographic characteristics and how they relate to diet quality and intake frequency of different food groups in US adults with lower incomes.

Design:

The study design is cross-sectional.

Participants/Setting:

A web-based survey was fielded in December 2022 to 1798 US adults with lower incomes (<250% Federal Poverty Guidelines, FPG).

Main outcome measures:

Environmental, climate and health-related dietary motivations and diet quality and dietary food group intake frequency were assessed.

Statistical analyses:

Differences in mean dietary outcomes and dietary motivation ratings by sociodemographic characteristics were evaluated using ANOVA and Kruskall-Wallis tests. Associations between dietary motivations and diet quality scores and dietary intake frequency were examined using generalized linear models adjusted for sociodemographic covariates.

Results:

Younger adults, women and non-binary persons, racially and ethnically minoritized groups, and adults experiencing food insecurity reported higher environmental and climate dietary motivations; older adults, higher-income and food-secure adults reported higher health motivations. Agreeing with environmental (β: 2.28, CI: 1.09–3.47), climate (β: 2.15, CI: 0.90–3.40), and health-related (β: 5.27, CI: 3.98–6.56) dietary motivations were associated with higher diet quality scores, compared to those with neutral rankings. Similarly, agreement with environmental, climate, and health-related dietary motivations was associated with higher intake frequency of fish, fruits and vegetables, and plant proteins, but not with red and processed meat intake frequency. Of several climate mitigation behaviors presented, participants perceived meat reduction as least effective (p<0.001).

Conclusions:

Environment, climate, and health were positive motivators of several healthy dietary choices in US adults with lower incomes. Such motivators did not translate to lower intake frequency of red and processed meat.

Introduction

Global climate change, driven by anthropogenic greenhouse gas emissions (GHGE) is one of the largest and most pervasive threats to human health and well-being today.¹ On average, environmental impacts and GHGE are lower for foods derived from plants than for meat and animal products, particularly beef.^2–5 In addition to promoting planetary health, increasing plant intake and reducing red and processed meat intake may lower risks of cardiovascular disease, type 2 diabetes, and some cancers.^{2, 6}

The Dietary Guidelines for Americans recommend higher intake of vegetables, fruits, and lean proteins, and “lower consumption of red and processed meats” for a healthy dietary pattern.⁷ More than half of Americans report being willing to reduce red meat consumption,⁸ yet red and processed meat intake in the United States (US) persists at a high levels.^{9, 10} US consumer shifts to vegetable equivalents of animal products can greatly reduce food-related GHGE,³ but even single-item substitutions for beef can markedly lower individual carbon footprints while maintaining dietary quality.¹¹ Thus, it is beneficial and feasible to move towards more sustainable and healthy plant-forward diets low in red and processed meats (plant-based diets) in the US. However, to achieve this goal and to ensure that future dietary recommendations are equitable and accessible, it is important to understand what motivates such behaviors in different populations.

Climate change can affect food access and quality, and may disproportionately affect historically marginalized populations, creating greater inequities in health outcomes and access to healthy foods.^{12, 13} The flexibility to shift to plant-based diets in the US may be a future advantage if climate change stresses food systems¹³ and meat costs rise.^{14, 15} Populations subject to social stressors may face unique barriers to adaptability to healthy plant-based diets, thus affecting climate resiliency.^{1, 16} Furthermore groups that have been historically marginalized are reportedly more likely to recognize societal issues as environmental problems,¹⁷ and to value purchasing sustainable food,¹⁸ yet valuation might not always translate into consumption of these foods due to access constraints or cost.⁸ Additionally, due to increasing prevalence of processed and marketed plant-rich foods, low-meat diets may not always translate to healthy plant-based diets,^{19, 20} an issue potentially magnified by limited access to fresh foods. These combined issues highlight the importance of understanding how environmental and climate motivations are associated with diet quality and protein intake in populations that have been socially and economically marginalized.

Attitudes towards meat intake and sustainable diets have been recently reviewed, and both the environment and health have been reported as motivators of meat-reduction behaviors; however, environmental impacts of meat reduction are often under-estimated, and the association between health motivations and meat consumption is inconsistent.^21–24 The inconsistent associations between health motivation and meat consumption may result from the breadth of health concerns and nutritional knowledge in the US. Few studies have evaluated how environmental and health motivators are associated with dietary intake. In a previous observational study, lower red meat intake was associated with sustainability and climate motivations in university students, despite moderate ranking of the mitigating impact of dietary meat reduction on climate change.²⁵ However, health-related motivations did not result in a reduction in red meat intake frequency among university students.²⁵ In another observational study, climate change concern was associated with less frequent red meat intake in a large population-based sample of Italian adults, but not chicken and processed meat.²⁶ Online experimental studies have shown that both environmental and health messaging are motivators for red meat reduction intent in convenience samples of US adults.^{27, 28} Yet, to our knowledge, the associations between environmental, climate, and health-related dietary motivations and overall dietary quality and frequency of intake of multiple specific protein sources has not been previously evaluated in the US, particularly in populations with lower incomes.

More research is needed to understand how environmental and health concerns and motivations are associated with diet quality and consumption of various proteins in different sociodemographic groups – and structurally-disadvantaged populations in particular.¹³ To address these research gaps, this study aims to evaluate how environmental, climate, and health-related dietary motivations in a sample of US adults with lower incomes relate to overall dietary quality and intake frequency of different food groups, and how these associations vary with sociodemographic characteristics. It was hypothesized that environmental, climate and health-related dietary motivations will be associated with lower red and processed meat intake frequency and higher diet quality scores.

Methods

Study Population

In December 2022, a web-based survey (Qualtrics) was fielded to US adults ages 18 and over with annual household income levels of less than 250% of the 2022 Federal Poverty Guidelines (FPG).²⁹ Previously described,³⁰ the survey assessed dietary intake frequency, environmental, climate and health attitudes, and sociodemographic variables. Recruitment of participants was facilitated by Prime Panels, an online survey recruitment platform designed for academic research that gathers quality samples representative of the US population and reaching underrepresented groups.³¹ The stated purpose of the study was “… to understand factors that affect your diet and health.” Data collection was ongoing from December 9 through December 22, 2022 with a goal of obtaining a sample size which exceeded minimum sample size requirements for an α=0.05, 80% power, and similar effect sizes from prior studies.²⁵ A total of 2898 participants were invited to complete the survey. Participants who did not consent (n=227), failed to answer attention checks correctly (n=231), did not live in the US (n=89), did not meet the income criteria (n=531), or had non-differentiation in responses for multiple sections (n=22) were omitted. Data were further screened for participants who completed answers at an average rate of one second or less per question³¹ and no additional participants were removed, resulting in a final study population of 1798 participants. Participants were compensated with preestablished rates by the Prime Panels research panel through which they were recruited. Participant consent was obtained electronically. The study was determined to be exempt by the University of Michigan Institutional Review Boards.

Diet Quality and Food Group Intake Frequency

Food group intake frequency and diet quality were measured using the 30-day Prime Diet Quality Score (PDQS-30), a previously-validated screener designed to evaluate overall dietary quality by ranking diet according to consumption of healthy and unhealthy food groups.^{32, 33} Briefly, participants were asked about intake frequency of different food groups over the past month. Response options included seven categories: once a month or less, 2–3 times per month, 1–2 times per week, 3–4 times per week, 5–6 times per week, once a day, or 2 or more times per day. Each food group was categorized as healthy or unhealthy. The 14 food groups categorized as healthy included: dark green leafy vegetables; cruciferous vegetables; deep orange vegetables and tubers; other vegetables; citrus fruits; deep orange fruits; other fruits; beans, peas, and soy products; nuts and seeds; poultry; fish (not including shellfish); low fat dairy; whole grains; and liquid oils. The 7 food groups categorized as unhealthy included: white roots and tubers; red meat; processed meat (sum of processed red meats and other processed meats); refined grains and baked potatoes; sugar-sweetened beverages; sweets and ice cream; and fried foods eaten away from home. For each participant, healthy food category responses were assigned a value ranging from 0 (once a month or less) to 6 (2 or more times per day). Unhealthy foods were reverse coded. Thus, possible individual PDQS-30 scores ranged from 0 to 126 with higher scores indicating healthier diets.

To evaluate intake frequency of specific food groups, questions on frequency of intake were converted to times per day. For categorical frequency ranges, the midpoint was used and divided by 7 or 30 to convert weekly or monthly frequency to daily frequency. For processed meats, questions asked about both processed red meats and processed poultry but did not include processed fish. Red and processed meat intake frequency was calculated by summing daily intake frequency of red meat, processed red meat, and processed poultry. Fruit and vegetable intake frequency was calculated by summing daily intake frequency of all fruits and vegetables scored in the PDQS-30 score. Plant protein intake frequency included combining daily intake frequency of nuts and seeds with intake frequency of beans, peas, and soy products.

Environment, Climate, and Health Attitudes

Motivations for food and beverage choices.

Questions were asked about the extent to which food and beverage choices were motivated by concern for (1) the environment, (2) climate change, and (3) health. These questions have been used in prior research²⁵ and asked “How much do you agree with these statements? I try to make food and beverage choices that…”: “Are good for the environment,” “Reduce my impact on climate change,” and “Are good for my health,” with a 7-point response scale ranging from 1= “Strongly Disagree” to 7= “Strongly Agree” for each item. Responses were classified as disagree (response of 1–3), neutral (response of 4, reference group) or agree (response of 5–7).

Perceptions of climate-friendly behaviors.

Items assessing perceived effectiveness of behaviors for combatting climate change have been previously described,²⁵ and were modified from a scale developed by de Boer and colleagues.³⁴ The questionnaire asked, “For each of the lifestyle changes, indicate whether you think this is an effective way of combatting climate change” (in random order): “Eat local, seasonal foods”, “Eat less meat”, “Drive less”, “Eat organic foods”, “Use less plastic”, “Save energy at home”, “Recycle.” Response options ranged from 1=“Not effective at all,” to 4=“Highly effective,” and included “Don’t know.” Responses of “Don’t know” (ranging from n=90 to n=226 per response) were excluded from statistical analyses.

Sociodemographic Variables

Sociodemographic data collected included self-reported age, gender, race and ethnicity, education, annual household income, and state of residence. For analysis, those self-identifying as female and transgender woman were classified as woman, male and transgender man were classified as man, and those responding as neither male nor female or “none of these” were classified as gender non-conforming. Due to small numbers in some response categories, race and ethnicity were reclassified as non-Hispanic Black, Hispanic, non-Hispanic White (includes Middle Eastern/North African as currently defined by US Census³⁵), and a fourth group encompassing participants self-identifying as Asian, Native Hawaiian or Other Pacific Islander, American Indian or Alaska Native, or Multiracial/Multiethnic. Food security status was assessed using the US Household Food Security Survey Module and scores were categorized according to USDA guidelines.³⁶ Food insecurity refers to individuals in the low or very low food security groups.

Statistical Analysis

First, descriptive statistics were used to examine participants’ sociodemographic characteristics. Then, means and standard deviations (SD) were calculated for PDQS-30 scores, red and processed meat daily intake frequency, and environmental, climate and health motivations by participants’ sociodemographic characteristics. Differences in means by sociodemographic characteristics were compared using one-way analysis of variance for PDQS-30 scores and motivations, and Kruskall-Wallis nonparametric tests for differences in means for red and processed meat intake frequency (due to the skewed distribution of red and processed meat intake frequency). Associations between environmental, climate and health motivations and dietary intake frequencies were examined using generalized linear models with a log link (to allow for interpretation of coefficients as percent difference), and gamma distribution (to account for the skewed distribution of dietary intake frequency). Frequency ratios (FR) were calculated by exponentiating the model coefficients; [FR-1] *100 represents percent higher frequency of consumption for each point on the perception scales. A generalized linear model with an identity link and normal distribution was used to evaluate associations between motivations and PDQS-30 scores, due to the normal distribution of the PDQS-30 scores. Models were adjusted for age, gender, race and ethnicity, education, %FPG, education, and environmental or health motivations, due to the potential influence of these variables on both dietary outcomes and environmental or health motivations.^{18, 25, 37, 38} Interactions between sociodemographic variables and red/processed meat intake frequency were examined for the outcome of PDQS-30. Wilcoxon signed-rank nonparametric tests were used to evaluate within-individual differences in mean ranking of the effectiveness of climate mitigation behaviors. Statistical significance was considered at p < 0.05. Statistical analyses were performed using SAS version 9.4.³⁹

Results

Mean PDQS-30 score and red and processed meat intake frequency (times/day) are shown in Table 1. Mean diet quality score for this population (50.2) was less than half of the maximum score (126). Mean diet quality scores differed significantly by age (p=0.01), race and ethnicity (p<.001), education (p<.0001), %FPG (p=0.03), food security (p=0.004) and geographic region (p=0.03). Means were lowest for adults age 40–49, people self-identifying as non-Hispanic White, and groups with lowest education levels, lowest income, and food insecurity. Participants living in the Midwestern or Southern US had lower PDQS-30 scores than those living in the Northeastern or Western US. Mean intake frequencies of red and processed meat also differed by age (p<.0001), race and ethnicity (p=0.01), education (p=0.001), food security (p=0.04), and geographic region (p<.001). Red and processed meat intake frequency (times/day) was generally highest for younger age groups, with highest mean intake frequency observed for ages 30–39. People who identified as Hispanic had highest mean red and processed meat intake frequency, followed by those identifying as non-Hispanic Black, non-Hispanic White, and those falling into the racial and ethnic group including people identifying as Asian, Native Hawaiian or Other Pacific Islander, American Indian or Alaska Native, or Multiracial/Multiethnic. Red and processed meat intake frequency was highest for those with the lowest education level and those with food insecurity. Those living in the Southern and Midwestern US had the highest red and processed meat intake frequency, while those living in the Northeastern and Western US had the lowest intake frequency.

Table 1.

Sociodemographic distribution and mean with standard deviation (SD) diet quality score and dietary intake frequency for US adults with lower incomes (n= 1798) surveyed in December 2022.

Sociodemographic Variable	n (%)	PDQS-30a			Red & Processed Meat Intake Frequency (times/day)
		Mean	SD	p-valueb	Mean	SD	p-valuec
All participants	1798 (100)	50.2	11.3		0.77	0.74
Age (years)
18–29	328 (18.4)	50.2	11.3	0.01	0.89	0.79	<0.0001
30–39	358 (19.9)	49.5	11.5		0.91	0.85
40–49	330 (18.4)	49.3	11.4		0.86	0.86
50–59	307 (17.1)	49.5	11.4		0.67	0.63
≥60	475 (26.4)	51.7	10.9		0.58	0.53
Gender
Woman	924 (51.4)	50.4	11.6	0.44	0.73	0.70	0.12
Man	858 (47.7)	49.9	11.0		0.81	0.79
Gender non-conforming	16 (0.89)	52.9	9.4		0.63	0.61
Race and Ethnicity
Non-Hispanic Black	279 (15.5)	52.4	11.4	<0.001	0.81	0.82	0.01
Hispanic	171 (9.5)	51.7	11.7		0.89	0.91
Non-Hispanic Whited	1237 (68.8)	49.4	11.2		0.75	0.69
A/NHPI/AIAN/M/Oe	111 (6.2)	50.8	10.6		0.63	0.76
Education
High school diploma/GED or less	716 (39.8)	48.3	11.0	<0.0001	0.86	0.85	0.001
Some college	546 (30.4)	49.7	11.1		0.74	0.69
College degree or higher	536 (29.8)	53.1	11.2		0.67	0.62
% Federal Poverty Guidelines
<100%	622 (34.6)	49.4	11.9	0.03	0.80	0.84	0.66
100 to <200%	795 (44.2)	50.3	10.6		0.76	0.68
200 to <250%	381 (21.2)	51.3	11.6		0.74	0.71
Food Securityf
High/marginal food security	868 (48.8)	51.0	11.6	0.004	0.73	0.68	0.04
Low/very low food security	912 (51.2)	49.4	10.9		0.80	0.79
Geographic Region
Northeast	260 (14.5)	50.7	11.0	0.03	0.64	0.60	<0.001
Midwest	410 (22.8)	49.6	11.1		0.80	0.78
South	818 (45.5)	49.7	11.1		0.84	0.82
West	310 (17.2)	51.7	12.1		0.65	0.53

^a30 day Prime Diet Quality Score (PDQS-30), possible scores range from 0–126.

^bp-value from one-way ANOVA

^cp-value from Kruskall-Wallis nonparametric test for differences in means

^dIncludes participants identifying as Middle Eastern/North African

^eIncludes participants identifying as Asian (A, n=26), Native Hawaiian/Other Pacific Islander (NHPI, n=4), American Indian/Alaska Native (AIAN, n=19), Multiracial/Multiethnic (M, n=44), and other races (O, n=18) as self-described.

^fn=18 missing due to self-reported unknown food security status

In the study population, 41% of adults agreed that they make dietary choices that are good for the environment, and these choices differed by age (p<.001), gender (p=0.01), race and ethnicity (p=0.001), education (p=0.002) and food security (p=0.03) (Table 2). Younger adults, women and gender-nonconforming persons, minoritized racial and ethnic groups, and those with higher education and with food insecurity were more likely to agree that they make dietary choices for the environment. Similarly, 35% of participants agreed that they make dietary choices to reduce their impact on climate change; this agreement level varied by age (p<.0001), gender (p=0.03), race and ethnicity (p<.0001), education (p=0.01), food security (p=0.03), and geographic region (p=0.046). Climate was a stronger motivator amongst young adults, women and gender-nonconforming persons, minoritized racial and ethnic groups, people with higher education levels, those with food insecurity, and adults living in the Northeastern US. About 64% of participants agreed that they make dietary choices that were good for health, and agreement level differed significantly by age (p=0.003), education (p<.0001), %FPG (p=0.04), and food security (p=0.002). Adults with older age, greater educational attainment, higher income, and food security more highly ranked health as a motivator for dietary choices.

Table 2.

Mean (SD) environmental, climate, and health-related motivation scores, stratified by sociodemographic characteristics, for 1798 US adults with lower incomes surveyed in December 2022.

Sociodemographic Variable	n	I try to make food and beverage choices that… a
		Are good for the environment				Reduce my impact on climate change				Are good for my health
		% Agreeb	Mean	SD	P c	% Agreeb	Mean	SD	P c	% Agreeb	Mean	SD	P c
All participants	1798	40.5	4.19	1.71		34.7	3.91	1.76		64.1	4.98	1.58
Age (years)
18–29	328	45.1	4.48	1.70	<0.001	39.6	4.23	1.69	<.0001	62.5	4.94	1.57	0.003
30–39	358	43.0	4.31	1.71		39.9	4.09	1.76		60.6	4.98	1.54
40–49	330	40.3	4.23	1.70		33.9	3.96	1.76		61.5	4.83	1.66
50–59	307	35.5	3.97	1.71		27.7	3.64	1.78		61.2	4.83	1.65
≥60	475	38.7	4.03	1.71		32.2	3.69	1.77		71.6	5.21	1.50
Gender
Woman	924	41.7	4.30	1.67	0.01	36.4	4.00	1.72	0.03	65.5	5.04	1.57	0.21
Man	858	39.2	4.07	1.75		32.4	3.80	1.80		62.8	4.91	1.60
Gender non-conforming	16	43.8	4.31	1.45		56.3	4.44	1.55		56.3	5.25	1.34
Race and Ethnicity
Non-Hispanic Black	279	44.4	4.51	1.75	0.001	39.8	4.28	1.76	<.0001	67.4	5.21	1.57	0.05
Hispanic	171	41.5	4.28	1.80		40.9	4.13	1.77		62.0	5.01	1.60
Non-Hispanic Whited	1237	39.2	4.09	1.69		32.6	3.78	1.75		63.2	4.92	1.58
A/NHPI/AIAN/Me	111	43.2	4.41	1.58		35.1	4.09	1.72		69.4	4.99	1.55
Education
HS Diploma/GED or less	716	39.5	4.18	1.71	0.002	34.2	3.90	1.74	0.01	59.2	4.80	1.62	<.0001
Some college	546	36.5	4.02	1.72		31.3	3.75	1.75		62.3	4.88	1.60
College degree or higher	536	45.9	4.38	1.70		38.6	4.07	1.79		72.6	5.32	1.46
% Federal Poverty Guidelines
<100%	622	41.0	4.23	1.77	0.76	35.2	3.96	1.82	0.63	59.8	4.86	1.67	0.04
100 to <200%	795	39.9	4.19	1.70		34.3	3.88	1.76		65.8	5.01	1.57
200 to <250%	381	40.9	4.14	1.64		34.4	3.87	1.68		67.7	5.11	1.45
Food Securityf
High/marginal food security	868	39.6	4.11	1.75	0.03	33.4	3.82	1.84	0.03	67.6	5.10	1.60	0.002
Low/very low food security	912	41.8	4.28	1.68		36.2	3.99	1.70		61.1	4.86	1.56
Geographic Region
Northeast	260	41.2	4.34	1.70	0.38	39.2	4.18	1.76	0.046	69.2	5.11	1.59	0.53
Midwest	410	40.5	4.21	1.69		34.4	3.93	1.72		62.4	4.93	1.60
South	818	40.0	4.17	1.73		34.0	3.85	1.78		63.7	4.97	1.59
West	310	41.3	4.10	1.71		32.9	3.80	1.77		63.2	4.95	1.54

^aParticipants were asked how much they agree with the statement “I try to make food and beverage choices that…are good for the environment, reduce my impact on climate change, are good for my health.” Responses were on a scale of 1–7, strongly disagree (1) to strongly agree (7).

^bPercent of participants reporting 5, 6, or 7 on scale of 1–7.

^cP-value for differences in means, results from one-way ANOVA

^dIncludes participants identifying as Middle Eastern/North African

^eIncludes participants identifying as Asian (A), Native Hawaiian/Other Pacific Islander (NHPI), American Indian/Alaska Native (AIAN), Multiracial/Multiethnic (M), and other races as self-described.

^fn=18 missing due to self-reported unknown food security status

Table 3 presents the association between environmental, climate, and health-related dietary motivations and dietary outcomes. In multivariate models, compared to those with neutral motivations, PDQS-30 scores were higher for participants who agreed that they make food and beverage choices that are good for the environment (β=2.28, CI=1.09, 3.47, p<.001), climate (β =2.15, CI=0.90, 3.40, p<.001), or health (β =5.27, CI=3.98, 6.56, p<.0001). PDQS-30 scores were lower for participants who disagreed that they make dietary choices for climate change (β =−1.34, CI=−2.57, −0.10, p=0.03) or health (β =−1.99, CI=−3.66, −0.31, p=0.02) when compared to those with neutral attitudes.

Table 3.

Associations between environmental, climate, and health-related dietary motivators and dietary intake frequency for 1798 US adults with lower incomes surveyed in December 2022.

Dietary outcome	Motivation Classification (reference = neutral)	I try to make food and beverage choices that… a
		Are good for the environmentb			Reduce my impact on climate changeb			Are good for my healthc
		βd	95% CI	P	β d	95% CI	P	β d	95% CI	P
PDQS-30e	Agree	2.28	1.09, 3.47	<.001	2.15	0.90, 3.40	<.001	5.27	3.98, 6.56	<.0001
	Disagree	−0.93	−2.24, 0.38	0.16	−1.34	−2.57, −0.10	0.03	−1.99	−3.66, −0.31	0.02
Dietary Intake Frequency (times/day)		FR f	95% CI	P	FR f	95% CI	P	FR f	95% CI	P
Red meat (unprocessed)	Agree	0.89	0.79, 1.00	0.05	0.94	0.84, 1.07	0.35	0.98	0.86, 1.11	0.72
	Disagree	0.90	0.79, 1.02	0.09	1.01	0.90, 1.14	0.82	1.01	0.86, 1.19	0.87
Processed red meat	Agree	0.98	0.86, 1.11	0.77	1.03	0.90, 1.18	0.66	0.91	0.79, 1.04	0.18
	Disagree	0.97	0.84, 1.11	0.65	0.99	0.87, 1.13	0.89	0.95	0.80, 1.14	0.60
Poultry (unprocessed)	Agree	1.03	0.93, 1.15	0.54	0.96	0.85, 1.08	0.49	1.09	0.97, 1.23	0.15
	Disagree	0.99	0.88, 1.12	0.90	0.95	0.85, 1.06	0.38	0.93	0.79, 1.08	0.33
Processed poultry	Agree	1.15	1.01, 1.31	0.04	1.21	1.06, 1.39	0.006	0.81	0.70, 0.94	0.004
	Disagree	0.95	0.82, 1.09	0.44	0.96	0.84, 1.10	0.59	1.02	0.85, 1.23	0.82
Fish	Agree	1.41	1.24, 1.61	<.0001	1.28	1.12, 1.48	<.001	1.19	1.03, 1.37	0.02
	Disagree	0.70	0.61, 0.81	<.0001	0.79	0.69, 0.90	<.001	0.98	0.82, 1.18	0.82
Fruits and vegetables	Agree	1.14	1.05, 1.25	0.003	1.14	1.04, 1.25	0.006	1.31	1.19, 1.44	<.0001
	Disagree	0.91	0.83, 1.01	0.07	0.94	0.85, 1.03	0.15	0.96	0.85, 1.09	0.54
Plant Proteinsg	Agree	1.24	1.10, 1.40	<.001	1.19	1.05, 1.34	0.006	1.24	1.09, 1.41	<.001
	Disagree	0.95	0.84, 1.09	0.48	0.85	0.75, 0.96	0.009	1.05	0.89, 1.24	0.58

^aParticipants were asked how much they agree with the statement “I try to make food and beverage choices that…are good for the environment, reduce my impact on climate change, are good for my health.” Responses were on a scale of 1–7, strongly disagree (1) to strongly agree (7). Participants were classified as agree (response of 5–7), neutral (response of 4), or disagree (response of 1–3), with neutral as the reference group.

^bModel adjusted for age, gender, race and ethnicity, education, % Federal Poverty Guidelines, food security, and health motivations

^cModel adjusted for age, gender, race and ethnicity, education, % Federal Poverty Guidelines, food security, and environmental motivations

^dResults from a generalized linear model with an identity link and normal distribution, to account for normal distribution of PDQS-30

^e30 day Prime Diet Quality Score (PDQS-30), possible scores range from 0–126.

^fResults from a generalized liner model with a log link and gamma distribution, to account for positive skewed distribution of dietary intake frequency variables. FR = Frequency Ratio. 100(FR-1) = % higher intake compared to neutral response.

^gSum of nuts/seeds and beans/peas/soy intake frequency

Participants who agreed with the statement “I try to make food and beverage choices that are good for the environment” had 15% higher intake frequency of processed poultry (p=0.04), 41% higher intake frequency of fish (p<.0001), 14% higher intake frequency of fruits and vegetables (p=0.003), and 24% higher intake frequency of plant proteins (p<.001) compared to those with neutral responses (Table 3). Those motivated by climate had 21% higher processed poultry intake frequency (p=0.006), 28% higher fish intake frequency (p<.001), 14% higher fruit and vegetable intake frequency (p=0.006), and 19% higher plant protein intake frequency (p=0.006) compared with those with neutral responses. Those who disagreed that they make choices for the environment had 30% lower intake frequency of fish (p<.0001), while those who disagreed that they make choices for climate had 21% lower intake frequency of fish (p<.001) and 15% lower intake frequency of plant proteins (p=0.009) than participants with neutral responses. Participants who agreed with the statement “I try to make food and beverage choices that are good for my health” had 19% lower intake frequency of processed poultry (p=0.004), 19% higher intake frequency of fish (p=0.02), 31% higher intake frequency of fruits and vegetables (p<.0001), and 24% higher intake frequency of plant proteins (p<.001) than those with neutral responses.

Mean ranking of the effectiveness of different climate mitigation behaviors in this population is displayed in Table 4. Recycling was rated by participants as the most effective behavior (mean=3.4), followed by “use less plastic” and “save energy at home” (means = 3.3). Eating less meat was ranked as the lowest (mean=2.4) compared to all other climate mitigation behaviors (p’s<0.0001). Diet-related climate mitigation behaviors had the highest percentage of “don’t know” responses (11–13%).

Table 4.

Mean ranking of the effectiveness of different climate mitigation behaviors for a national sample of US adults with lower incomes, surveyed in December 2022^a

Climate mitigation behavior	Meanb	SD	n	p-valuec	Don’t knowd (%)
Eat less meat	2.4	1.1	1600	ref	11.0
Eat organic foods	2.5	1.1	1572	<.0001	12.6
Eat local, seasonal foods	2.9	1.0	1600	<.0001	11.0
Drive less	3.1	1.0	1677	<.0001	6.7
Use less plastic	3.3	1.0	1697	<.0001	5.6
Save energy at home	3.3	0.9	1694	<.0001	5.8
Recycle	3.4	0.9	1708	<.0001	5.0

^aParticipants were asked to indicate whether they think each behavior is an effective way of combatting climate change (response ranged from 1=not effective, to 4= highly effective, 5= don’t know). Response options were displayed in random order.

^bResponses of “don’t know” were excluded from statistical analyses.

^cP-value for Wilcoxon signed-rank nonparametric test for within-individual comparison

^dPercentage of participants responding as “don’t know” for each mitigation behavior

Discussion

American adults with lower incomes reported making food and beverage choices that are good for the environment (41%), reduce impact on climate change (35%), or are good for health (64%). Agreement with environmental, climate or health-related dietary motivations was associated with higher diet quality scores when compared with neutral ranking. Compared with neutral rating, agreement with environmental, climate or health motivations was associated with higher intake frequencies of fish, fruits and vegetables, and plant proteins, but not with lower intake frequencies of red and processed meats. Additionally, processed poultry intake frequency was significantly higher for those motivated by the environment and climate change. Of several possible climate mitigation behaviors presented, participants perceived meat reduction as least effective. These results suggest that dietary choices of individuals with lower incomes are in part motivated by environment, climate, and health concerns, yet the role of red and processed meat in these outcomes may be underestimated.

Evidence of higher dietary quality with higher environmental valuation has been observed in previous studies. Valuation of sustainable foods and sustainable food practices has been associated with better diet quality in young adults,^{40, 41} and environmental and climate motivations have been associated with lower red meat intake frequency among university students.²⁵ A large survey recently conducted in Italy determined that participants who reported that climate change is an urgent environmental problem had lower intake frequency of beef and pork, but not chicken or processed meat.²⁶ In the current study, agreement with making dietary choices for the environment, climate, or health was associated with higher intake frequency of plant proteins, possibly reflecting increasing accessibility, affordability, or acceptability of these protein types.¹⁹

Counter to the study hypothesis, none of the dietary motivations were significantly associated with lower frequency of red and processed meat intake in this study. This may be due, in part, to the low perceived effectiveness of eating less meat for combating climate change relative to other behaviors such as recycling, whereas research shows that eating a completely plant-based diet to be four times more effective than recycling.⁴² Consistent with the current findings, it is well documented that consumers underestimate the environmental impact of food,²³ and meat consumption in particular.^{21, 23, 25, 43} Additionally, there are mixed messages in the US about the health benefits and risks of red meat intake⁴⁴ and saturated fat.⁴⁵ It also may be engrained in many Americans that meat is essential to a healthy diet. In a previous study, primary reasons for consuming meat were beliefs that a healthy diet includes meat and that a meal is incomplete without meat.²⁴ Furthermore, the popularity of meat-heavy diets such as the Keto and Paleo diets⁴⁶ may confuse consumers as to the health impacts of red and processed meat. Findings from the current study suggest that, like the population overall, US adults with lower incomes may in part be confused about the health and environmental impacts of red and processed meat consumption, but it is also likely that other factors, such as cost and availability, food and marketing environments, food preferences, and cultural norms around the importance of red and processed meat consumption affect choices in this population.^{21, 24}

A previous review of the literature concluded that sustainability is not highly considered when making food choices,²³ yet the current study demonstrates that the environment, climate change, and health can motivate dietary choices of adults with lower incomes. The changing environment affects food availability and helps shape attitudes towards sustainability; as climate change and other human-environment disruptions affect food systems, environmental perceptions and dietary motivations may shift.^{13, 47} For example, there is recent evidence that the COVID-19 pandemic, possibly due to changes in food system functioning, corresponded with increased interest in sustainable diets in Europe.⁴⁸

Previously, the environment, climate and health were also shown to motivate dietary choices in young adults.²⁵ Younger adults in the current study more highly ranked the importance of the environment and climate in dietary choices. Climate anxiety persists in young people around the world;⁴⁹ therefore efforts to promote sustainable dietary choices among younger adults might be more effective and result in long-term behavior change. The current results also underscore that, given underestimation of the importance of red meat in the climate impact of dietary choices, there is a role for public health messaging and education around this topic to support such behavior change.

Sociodemographic differences in environmental and climate dietary motivations track with concerns of environmental health risks in the US. Non-Hispanic Black and other minoritized races, females, and higher educated groups in the US showed stronger perceptions of environmental health risks.⁵⁰ Similarly, Hispanic/Latino and African American groups were more likely to be alarmed or concerned about climate change than White groups.⁵¹ Greater environmental concern may reflect perceptions of vulnerability⁵² and personal experiences, such as inequities in environmental risks;^{53, 54} furthermore, social problems might be observed through an environmental justice lens in some groups that have been economically and socially marginalized.¹⁷ Lastly, environmental and climate change concerns of lower income and historically marginalized racial and ethnic groups may be underestimated.⁵⁵ This study reiterates the need to ensure that policies and programs for sustainable eating are inclusive of historically underrecognized populations.

While herein environmental, climate and health-based dietary motivations are associated with some components of healthy diets, it is possible that these associations may be underestimated as barriers may prevent translation of environmental and dietary motivations to behaviors.^{56, 57} Both external factors (economic, social, cultural, lack of infrastructure, convenience) and internal factors (motivation, education, values, emotional involvement) can be barriers to pro-environmental behavior,^{56, 58} and dietary behavior,^59–62 as well as drivers of food choice, possibly affecting sustainable eating and the associations between motivations and dietary intake. Consumers often view sustainable diets as inaccessible, inconvenient and expensive,²³ and this may be especially true for lower-income households.⁸ Neighborhoods where economically disadvantaged populations reside reportedly have a greater prevalence of fast food and convenience stores, which promote unhealthy diets^{63, 64} and may affect access to healthy plant-based foods.

Secondly, actions may not always track to the most effective choices for health or the environment. This may in part be to lack of consumer knowledge and information regarding the most healthful and environmentally conscious choices (as may be the case with red and processed meat intake), but also perceptions about healthfulness,^{65, 66} environmental claims of products,⁶⁷ and the resulting “health halo” stemming from corporate social responsibility (CSR) marketing.⁶⁸ For example, in a recent study, positive framing of climate-friendly fast food menu choices influenced perceived healthfulness of choices.⁶⁹ Such health halos and greenwashing – the act of conveying misleading information regarding environmental practices⁷⁰ – may confuse consumers, particularly when combined with perceived effectiveness of climate reduction behaviors. These effects can reach beyond individual impacts on climate change and potentially impact public support for climate-friendly policies. Additionally, consumer perceptions and purchasing power may reinforce CSR marketing and contribute to the effectiveness of corporate greenwashing.⁷⁰

Lastly, a plant-based diet or lower red meat intake does not necessarily equate to a healthy diet or health benefits.²⁰ For example, previous studies have shown that benefits of plant-based diets were observed with healthy plant-based diets (e.g., high fruits and vegetables, nuts, and whole grains, but low in refined grains, added sugars, and potatoes) but not unhealthy plant-based diets.⁶ This is important to consider as populations shift to plant-based diets, particularly in under-resourced communities where access to healthy plant-based foods may be limited.

Plant-based diets, particularly those low in red and processed meat, are recommended for human and environmental health^{2, 7}, but rising costs of meats with climate change⁷¹ may provide an opportunity to address affordable and sustainable healthy diets in higher-income countries such as the US.⁷² It has been predicted that even in developed countries, increasing food prices due to climate change may necessitate shifts to lower cost foods, particularly among groups with lower incomes.⁷³ There is a need to support such communities vulnerable to the double-burden of climate change and negative food environments. Educational programming may increase knowledge of environmentally friendly choices while improving dietary quality,⁷⁴ but more research is needed in economically and socially marginalized populations to understand the multi-factorial and structural barriers to adopting more sustainable diets. Individual and community-based intervention programs have been effective at improving diet quality in populations with lower incomes,⁵⁹ and such programs could expand to sustainable eating. Food labeling may also be effective in supporting sustainable dietary choices, but must be aligned with health goals.⁶⁹ Sustainability-focused food labels must also be understandable to consumers, and convey information without causing confusion or information overload.^{75, 76} Ultimately governing bodies, food processors, retailers, and servicers should be involved to create change.² Institutions such as public schools, hospitals, and government buildings have potential to implement sustainable food practices,⁷⁷ environmental sustainability should be included in federal and individual dietary guidelines⁷⁸ and institutional procurement policies, and plant-based foods should be subsidized while taxing⁷⁷ or lifting subsidies on foods with high GHGE. To the extent possible, policies that improve the sustainability of the food offered (e.g., shifting the default options or increasing the ratio of plant-based entrees to red meat entrees) should be prioritized. Doing so will help facilitate more sustainable food choices without placing the burden on the consumer to make the more sustainable choice. Lastly, there should be continued focus on improving the food environment for socioeconomically disadvantaged populations so that healthy plant-based foods are convenient and accessible.

This study has several limitations. It is a cross-sectional study, and therefore causal inferences cannot be made. Data were self-reported, which may be subject to recall bias and social desirability bias.⁷⁹ However, participants were blinded to the research hypothesis which may minimize social desirability bias. Selection bias may have occurred if panel members differed from non-responders with respect to beliefs and actions. It is possible that due to the method of recruitment, this study population might be more tech savvy or highly educated,⁸⁰ and thus the population might not be representative of the lower-income US population in all aspects. The survey was also presented only in English, which may have limited participation of non-English speaking persons. Data collected via online participant panels may be prone to providing quick “good enough” answers rather than thoughtful answers.⁸⁰ It is likely that such bias in this study would underestimate the observed effects. Furthermore, this limitation was minimized through use of attention checks and screening for unrealistically-short survey completion times. Small numbers of participants falling into some groups, such as gender non-conforming persons and people identifying as races other than Black, Hispanic, and White did not allow for meaningful interpretation of associations within these groups; future studies of diverse populations are important to further evaluate the observed associations. A dietary screener was used to capture dietary patterns which may not capture all nuances of a healthy diet; it may be helpful for future studies to consider additional dietary indices that more directly align with the Dietary Guidelines for Americans,⁷ such as the Healthy Eating Index.⁸¹ Lastly, there are many potential factors influencing food choices such as taste preferences, cultural norms, culinary skills, access, and price that were not measured in this study.^{82, 83} These influences may serve as potential confounders, or may result in disconnect between motivations and dietary intake and are thus an important area of future research.

Conclusions

Among US adults with lower incomes, environmental, climate, and health-related dietary motivations were associated with higher diet quality scores and higher intake frequency of some healthy dietary components after controlling for sociodemographic variables. Counter to the hypothesis, these motivations were not associated with intake frequency of red and processed red meat. Environmental and climate motivations were higher for women and people identifying as gender-nonconforming, people from historically minoritized racial and ethnic groups, those with higher education, and people with low or very low food security. Understanding sociodemographic differences in dietary motivations and how such motivations are associated with dietary food group intake frequency lays the foundation for supporting sustainable dietary behaviors in underserved populations. Recognizing and cultivating such motivations in these populations presents opportunities to support healthy diets, improve health, and reduce environmental impacts while advancing transformative adaptation to climate change.

Research Snapshot.

Research Question:

How are environmental perceptions and dietary motivations associated with dietary intake frequency of different food groups in US adults with lower incomes?

Key Findings:

In this cross-sectional study of 1,798 US adults with lower-incomes, those who reported that their dietary choices were motived by the environment, climate, or health had significantly higher diet quality scores, and higher intake frequency of fish, fruits and vegetables, and plant proteins compared to those with neutral motivations. Motivations were not associated with red and processed meat intake frequency, and reducing meat intake was perceived as minimally effective as a climate mitigation strategy.