Federal, State, and Local Nutrition Policies for Cancer Prevention: Perceived Impact and Feasibility, United States, 2018

Abstract

Objectives. To assess stakeholder perceptions of the impact and feasibility of 21 national, state, and local nutrition policies for cancer prevention across 5 domains in the United States.

Methods. We conducted an online survey from October through December 2018. Participants were invited to take the survey via direct e-mail contact or an organizational e-newsletter.

Results. Federal or state Medicare/Medicaid coverage of nutrition counseling and federal or state subsidies on fruits, vegetables, and whole grains for participants in the Supplemental Nutrition Assistance Program were the policies rated as having the highest perceived impact and feasibility. Overall, the 170 respondents rated policy impact higher than policy feasibility. Polices at the federal or state level had a higher perceived impact, whereas local policies had higher perceived feasibility.

Conclusions. Our findings might guide future research and advocacy that can ultimately motivate and target policy actions to reduce cancer burdens and disparities in the United States.

Suboptimal diets are a major preventable cause of cancer.¹ Although we can promote optimal diets to motivated individuals, individual-based approaches are often costly and poorly sustained. Conversely, population-based policy strategies such as taxation, food subsidies, food labeling, and institutional nutrition guidelines or standards could have a broader, more powerful, and more sustainable impact on dietary changes.²

With limited public health budgets and political opportunities, policymakers are often required to prioritize the implementation of interventions that are most effective and feasible. However, the perceived impact and feasibility of policy strategies to improve diets and reduce cancer burden have not been evaluated. Using a survey, we assessed the impact and feasibility of selected nutrition policies for cancer prevention in the United States as perceived by stakeholders including elected government officials and staff, civil or public service employees, advocates, researchers, and health care professionals.

METHODS

We administered an online survey to federal-, state-, or local-level stakeholders from 10 organizations and groups in the United States (Table A, available as a supplement to the online version of this article at http://www.ajph.org). To administer the survey, we sent 4 rounds of emails comprising an introductory e-mail that described the purpose of the survey and 3 subsequent reminder emails for survey completion. When direct e-mail contact was not feasible, we asked organizations to send the survey link through their internal organizational e-mail system or an e-newsletter for 1 or 2 rounds. Between October and December 2018, 170 individuals provided complete responses and were included in our analysis.

Survey Instrument

A 5-point Likert scale was used to collect data from respondents on their perceptions regarding the impact and feasibility of 21 policies in 5 domains:

1. food taxes and subsidies,

2. food labeling,

3. nutrition quality standards,

4. nutrition education, and

5. nutrition promotion in health care (Table B, available as a supplement to the online version of this article at http://www.ajph.org).

The question on impact asked “To what extent would the following policies, if enacted, be effective in reducing the cancer burden?” (5 = very effective, 4 = somewhat effective, 3 = neither effective nor ineffective, 2 = somewhat ineffective, 1 = not effective at all). The question on feasibility asked “How feasible is it to enact each of the following policies in the next 5 years, at the specified level?” (5 = very feasible, 4 = somewhat feasible, 3 = neither feasible nor infeasible, 2 = somewhat infeasible, 1 = not feasible at all). Both questions included an additional “don’t know” category (rated as 6).

Statistical Analysis

We estimated the means and standard deviations of the impact and feasibility scores for each policy according to level of implementation (federal, state, or local). To assess policy impact jointly with policy feasibility, we plotted mean impact scores as the x-axis and mean feasibility scores as the y-axis. For each policy, we ranked the percentages of respondents who rated the policy favorably as very effective or somewhat effective for impact and as very feasible or somewhat feasible for feasibility. We then compared the percentages of policies rated favorably by policy domain. All statistical analyses were 2 sided, and significance was considered at an alpha level of .05. We used SAS version 9.4 (SAS Institute Inc, Cary, NC) in conducting all of our statistical analyses.

RESULTS

More than half of the 170 survey respondents were elected government officials and their staff or civil service employees (54.7%); 18.2% were advocates or employees of not-for-profit organizations, 11.2% were scientists or research experts from academic institutions, 10.6% were health care professionals, and 5.3% were classified as “other.” The respondents’ expertise related to different levels of government, mostly the state level (58.2%) followed by the local or community level (23.5%) and the federal or national level (18.2%). State-level respondents came from 36 US states, with 18 (18.2%) from the Northeast, 29 (29.3%) from the Midwest, 19 (19.1%) from the South, and 30 (30.3%) from the West (Table C, available as a supplement to the online version of this article at http://www.ajph.org).

For the 21 policies being evaluated, mean policy impact scores ranged from 2.90 to 4.25 (with 1 being least effective and 5 being most effective), and mean policy feasibility scores ranged from 1.86 to 4.13 (with 1 being least feasible and 5 being most feasible) (Table D, available as a supplement to the online version of this article at http://www.ajph.org). Overall, the 21 policies received higher scores for impact than for feasibility. Federal- or state-level policies were perceived to have a higher impact, whereas local policies were perceived to have higher feasibility (Table E, available as a supplement to the online version of this article at http://www.ajph.org).

When scores were evaluated jointly for each individual policy, the policies with both high impact and high feasibility scores were Medicare or Medicaid coverage of nutrition counseling for people with chronic conditions (at the federal and state levels) and subsidies on fruits, vegetables, and whole grains among SNAP (Supplemental Nutrition Assistance Program) participants (at the federal and state levels). The policies with both low impact and low feasibility scores were taxes on processed meat and taxes on junk food at each implementation level (Figure 1). Across the 5 policy domains (food taxes or subsidies, food labeling, nutrition quality standards, nutrition education, nutrition promotion in health care policies), impact was perceived as highest for food subsidy policies and feasibility was perceived as highest for food labeling and nutrition education policies (Table F and Figure A, available as supplements to the online version of this article at http://www.ajph.org).

FIGURE 1—

Perceived Impact Versus Perceived Feasibility of Federal, State, and Local Nutrition Policies: United States, 2018

Note. SNAP = Supplemental Nutrition Assistance Program.

DISCUSSION

Stakeholders perceived federal and state Medicare/Medicaid coverage of nutrition counseling and federal and state subsidies on fruits, vegetables, and whole grains for SNAP participants to have the highest impact on and feasibility for cancer prevention. Currently Medicare covers outpatient nutrition counseling but only for highly selective diagnoses (e.g., diabetes or chronic kidney disease); cancer patients and survivors are excluded from this benefit. Although more than 80% of cancer care occurs at outpatient clinics,³ 70% of cancer patients have never received nutrition advice from their providers during or after treatment.⁴ This represents a missed opportunity. Lessons can be drawn from the expanded model of the Medicare Diabetes Prevention Program, through which Medicare beneficiaries with prediabetes receive a Centers for Disease Control and Prevention clinical intervention focusing on weight management.⁵

Disparities in diet quality have persisted or worsened in the past 15 to 20 years.⁶ The 2018 Farm Bill significantly expanded the Food Insecurity Nutrition Incentive program, which provides financial incentives to SNAP participants for purchasing fruits and vegetables.⁷ A 30-cent incentive for every SNAP dollar spent on fruits and vegetables was shown to increase intake among SNAP participants by 26%.⁸ This supports the impact of subsidizing fruits and vegetables for low-income populations and paves the way for further program expansions focused on reducing disparities in the diet-associated cancer burden in the United States.

Despite the promising cost effectiveness of food tax policies, the 2 tax policies included in our study, taxes on processed meats and taxes on junk food, were perceived to have the least impact and feasibility at all implementation levels. This may reflect the complexity of defining the categories of processed meat or junk food to be taxed and the appropriate mechanisms for implementing the tax^9,10; it may also reflect the widespread unpopularity of food taxes in the United States.¹¹

PUBLIC HEALTH IMPLICATIONS

Government nutrition policies can be cost-effective strategies for reducing cancer burdens and disparities in the United States. In our analysis, federal and state Medicare/Medicaid coverage of nutrition counseling and federal and state subsidies on fruits, vegetables, and whole grains among SNAP participants had the highest perceived impact and feasibility, whereas taxes on processed meats or junk food were perceived to have the lowest impact and feasibility. Our findings have implications for prioritizing government policies to improve nutrition and reduce cancer burdens and disparities.