Background

Chronic diseases and mental health disorders are among the most important public health challenges worldwide. We tested whether community gardening, as a multicomponent and synergistic lifestyle intervention, could improve health behaviours and reduce psychosocial risk factors, including stress and anxiety, which are critical for reducing risks for chronic diseases and mental health among age- and ethnically diverse healthy adults.

Methods

We conducted a randomised controlled trial among adults over 18 years old who had not gardened in the past two years. Participants were randomised within waitlists for community gardens in Denver, Colorado. Primary outcomes were diet, physical activity, and anthropometry; secondary outcomes were perceived stress (PSS-10) and anxiety (GAD7). Analysis used intent-to-treat. Difference score models evaluated secondary outcomes.

Findings

From 2017-2019, 291 adults participated in the study. Significant time-by-intervention effects were observed for fibre (p=0·034) and moderate-to-vigorous physical activity (p=0·012). There were no significant time-by-intervention interactions for combined fruit and vegetable intake, healthy eating index, sedentary time, body mass index, and waist circumference (all p>0·04). Difference score models showed greater reductions among intervention participants (T2-T1) in perceived stress (p=0·025) and anxiety (p=0·044).

Interpretation

Community gardening improved fibre intake and moderate-to-vigorous physical activity but not combined fruit and vegetable intake, healthy eating index, sedentary time, body mass index or waist circumference. Moreover, community gardening reduced perceived stress and anxiety. These data, drawing from a robust randomized controlled trial, suggest that gardening can improve key health behaviours and psychosocial health outcomes that are critical for reducing risk for chronic diseases and mental health disorders.

Purpose: Cancer survivors experience high comorbidity and accelerated aging.  A healthier diet and increased physical activity (PA) could resolve these complex problems, and favorably influence other outcomes, e.g., adiposity.  Harvest for Health, a single-blinded, 2-arm, crossover-designed randomized controlled trial (NCT02985411) assessed outcomes of a 1-year, home-based vegetable gardening intervention among 381 cancer survivors across Alabama.

Methods: Harvest for Health was a partnership between the University of Alabama at Birmingham and the Cooperative Extension Service at Auburn University. Survivors of cancers with >60% 5-year survival who self-reported suboptimal vegetable and fruit (V&F) consumption, PA, and physical functioning (PF) were randomly-assigned to waitlist vs. vegetable gardening arms. Trained Extension-certified Master Gardeners were paired with survivors to deliver the intervention, consisting of monthly home visits (with mid-month telephone contact) to oversee the planning, planting, and maintaining of a spring, summer, and fall garden. Supplies included a 4x8’ raised bed or four grow boxes, soil, fertilizer, mulch, seeds and transplants, and gardening tools. Baseline, 1- and 2-year follow-up assessments were conducted (in-person or Zoom-assisted). Primary outcomes included self-reported V&F intake, PA and PF (corroborated by plasma carotenoids, accelerometry and physical performance assessments). Secondary outcomes included body adiposity, stress, sleep, quality-of-life, social support, self-efficacy, gut alpha diversity, and lipid and inflammatory biomarkers. 

Results/Findings: Participants were diagnosed with a wide variety of cancers, with time from diagnosis averaging 7.5 years. Most were female (69%), non-Hispanic white (77%), married (61%), urban dwelling (88%), and retired/unemployed (73%). While few were current smokers, most had BMI’s>25 and ate <2.5 servings of V&Fs/day. At 1-year follow-up, retention averaged 92.9% and adverse events were minimal (no between-arm differences). Improved scores across outcomes were observed in intervention vs. control arms, though statistically significant results were detected only for the 2-minute step test, 30-second chair stand, perceived health, and 3-of-4 microbiome alpha diversity measures (time x arm interaction p-values<0.01). Highly significant within and between-arm differences and time x arm interactions were found for gardening social support (p-values <0.001).

Conclusions: Analyses on Harvest for Health continue; however, results indicate that vegetable gardening interventions are safe, well-accepted, and produce broad benefits.

Purpose: Although school garden programs have been shown to improve dietary behaviors, there has not been a cluster-randomized controlled trial (cRCT) conducted to examine the effects of school garden programs on obesity or metabolic outcomes. The goal of this study was to evaluate the effects of a one-year school-based gardening, nutrition, and cooking intervention (called TX Sprouts) on dietary intake, obesity outcomes, metabolic outcomes, and academic performance in elementary school children.

Methods: This cRCT randomly assigned 16 elementary schools to either TX Sprouts intervention (n=8 schools) or to delayed intervention (control, n=8 schools). The intervention included: a) formation of a Garden Leadership Committee, b) building an outdoor teaching garden, c) 18 student gardening, nutrition and cooking lessons through the school-year taught by trained educators, and d) nine monthly parent lessons. Outcomes were obtained at baseline and post-intervention follow-up (9-months): demographics, anthropometrics (i.e., BMI parameters, waist circumference, and body fat percentage via bioelectrical impedance), blood pressure, dietary intake via diet screener and 24-hr recalls (subsample), standardized test scores, and metabolic outcomes (glucose, lipids, HbA1c) via fasting blood sample (subsample).

Results: A total of 3135 children (7-12 y) were enrolled in study and completed baseline survey and clinical measures; 64% Hispanic, 69% eligible for free and reduced lunch, and 47% male. Eighty seven percent (n=2721) of children completed 9-month follow-up clinical measures. Fasting blood samples were obtained from 1104 children at baseline and 695 children at follow-up. Dietary recalls were obtained in 738 children at baseline and 468 at follow-up. Participants in TX Sprouts compared to control increased vegetable intake (+0.5 vs. 0.01 serv/d, p=0.046), increased unprocessed foods (2.3% vs. -1.8% g, p<0.01), reduced ultra-processed food intake (-2.4% vs. 1.4% g, p=0.04), reduced HbA1c (-0.03 vs. +0.05%, p=0.035), reduced LDL cholesterol (-8.4 vs. -3.5 mg/dL, p=0.05), increased reading scores (+5.3% vs. -1.3%, p=0.047), and increased availability of healthy foods in the home (p=0.01).

Conclusions: Our findings suggest that elementary schools can incorporate garden-based interventions to improve dietary behaviors, academics, and subsequent metabolic parameters in children. Further research on ways to scale and sustain garden, nutrition, and cooking programs in schools are warranted.