Does joining a CSA change how you eat?

In community-supported agriculture (CSA), you pay a farm up front for a share of its harvest and collect produce, usually weekly, through the growing season^4,7. The question is whether that changes what ends up on your plate.

Who joins. Members are typically women who are white, highly educated and affluent, drawn by a wish for local, high-quality, organic produce; most are satisfied, but turnover is often high⁸. Low-income households that apply to a subsidized CSA already report better knowledge, habits and diets than similar households that do not⁶. So a member's good diet may be a reason for joining, not a result.

What members report. In a matched study at three Minnesota employers, 324 CSA buyers reported more vegetables at home, more family meals and less eating out, but health assessments showed no significant change in dietary intake, health status or BMI⁹. In a rural New York county, members reported higher intake of 11 vegetables during the season, and over half tried a new vegetable¹⁰.

What trials show. In a Massachusetts trial, adults with excess weight given a subsidized share scored 4.3 points higher on the Healthy Eating Index, a 100-point diet-quality score, than adults given the same amount in cash¹. In a four-state trial of half-price shares plus nutrition classes, caregivers ate about 1.1 cups more produce a day than comparison caregivers, but children did not change and the gains faded after the season⁴. A 50-woman pilot found more produce at home but no significant rise in consumption⁵.

The practical side. In the Massachusetts trial a full share cost $690 for the June-to-November season and a small share $480, with six or nine items a week chosen from 15 to 20 kinds¹. Low-income members collected shares in a median 88% of weeks¹¹, more where farms offered several share sizes². Families found unfamiliar foods and quick spoilage hard to plan around³, and running the four-state program cost about $2,439 per household a year⁷.

What it means. For a household that will cook what arrives, a CSA can make vegetables easier to eat for a season. It is not a guarantee, and children's diets may need more than a box.

The randomized trials

Berkowitz and colleagues (2019) enrolled 128 adults with a BMI above 25 who were seen at a community health center in central Massachusetts or lived nearby; 122 were analyzed (56 CSA, 66 comparison) and 82% completed follow-up. The CSA group received $300 per growing season to spend on a share running June to November; the comparison group received healthy eating information and an equal cash payment. On the Healthy Eating Index, the CSA group scored 60.2 and the comparison group 55.9 (difference 4.3 points, 95% CI 0.5 to 8.1, P = 0.03; 3 points counts as clinically meaningful), with gains in total vegetables, total fruit and whole fruit. Food insecurity was lower (RR 0.68, 95% CI 0.48 to 0.96). Members collected a median 79% of weekly shares. Limits: one site, mostly white women, participants could not be blinded, no data on produce used or wasted, and too few people to judge secondary outcomes¹.

The Farm Fresh Foods for Healthy Kids trial assigned 305 caregiver-child pairs in New York, North Carolina, Vermont and Washington State to a half-price CSA plus nutrition classes or to a delayed start. After one season, caregivers' produce intake was 1.10 cups a day higher than the comparison group's (P = 0.02, adjusted for baseline), skin carotenoids were higher (P = 0.01) and households were more likely to be food secure (odds ratio 1.67). Children's diet, BMI and activity showed no significant net change. A year after baseline, caregivers' attitudes and cooking confidence were still better, but diet and food security were no longer better than the comparison group's. More weeks of share pick-up, not more classes attended, went with larger gains⁴.

A feasibility trial randomized 50 low-income women with children to 16 weeks of produce boxes plus five education sessions or to nothing. Fruit and vegetable stock at home rose significantly; the rise in consumption did not reach significance. Women collected boxes in 9.2 of 16 weeks, held back by transport and work schedules⁵.

Surveys and matched comparisons

In the Minnesota employer study, 324 buyers were matched to non-buyers by age, sex, employer and occupation; participants reported changes, but health assessments found no significant change in dietary intake, health status or BMI⁹. In New York, 151 questionnaires from four groups of members (new or returning, full-paying or subsidized) showed higher intake of 11 vegetables during the season, a sustained rise in some winter vegetables and new vegetables tried by more than half, though the authors caution that people who join may be more open to new vegetables¹⁰. Among cost-offset CSA applicants, children in households of 46 purchasers ate 4.14 cups of produce a day against 1.83 in 18 non-purchasers' households, a gap the randomized trial did not reproduce for children^6,4. In 148 caregivers from the intervention group of the four-state trial, preparing vegetables for children's snacks and dinners, and leafy greens, was still more frequent a year later (107 followed), but that analysis had no comparison group¹².

Practical lessons

Across ten communities, share pick-up ranged from 33% to 89% of weeks. Where farms offered several share sizes (69% of farms), members collected shares in 77% of weeks against 58%; community pick-up sites went with lower participation (65% against 78%)². The process evaluation found a mean season of 21 weeks, a median pick-up of 88% of weeks, and 82% of caregivers saying their household ate most or all of the produce¹¹. In focus groups of 53 caregivers, quantity was abundant and choosing items helped, but payment timing was a barrier for some, and unfamiliar foods and quick spoilage hindered meal planning, with waste concerns likely sharpest for families with little money³. The program cost $2,439 per household a year ($1,884 to run and $555 in participants' own costs), or $1,507 to $2,439 for each extra cup of caregiver produce intake and $2,271 to $3,137 for each household moved out of food insecurity⁷.

Where the evidence is thin

Three randomized trials of 50 to 305 participants, all in the United States and mostly with subsidized low-income participants, followed people for one season or one year, measured diet by recall or screener (skin carotenoids aside) and none measured disease. For unsubsidized members the evidence is surveys and one matched study; a narrative review found a dearth of experimental research⁸. Emerging evidence (2 of 5). See also farmers markets and diet quality and fresh vs frozen produce for the off-season.