Efeitos em longo prazo de intervenção nutricional coletiva sobre o consumo de frutas e hortaliças

Abstract

Introduction: The World Health Organization (WHO) recommends an intake of 400g/day of fruit and vegetables (FV), but intake is low worldwide. This context requires effective and sustainable interventions, for which Primary Health Care is an ideal scenario. This study aims to: 1) to validate the data on FV intake obtained in telephone interviews; 2) to assess the effects of a group-based nutrition intervention on FV intake at 12, 36, and 48 months of follow-up; 3) to assess whether the intervention was associated with maintaining positive changes in FV intake at 36 and 48 months. Methods: we analyzed data from baseline, 12, 36, and 48 months of follow-up of a randomized controlled community trial carried out in a representative sample of the Health Academy Program (PAS) in Belo Horizonte, Minas Gerais, Brazil, with participants aged 20 and older. The control group (CG) received usual care, and the intervention group (IG) received usual care plus a group-based intervention to promote FV intake, based on the Transtheoretical Model (TTM) and the problematizing approach of Paulo Freire, during seven months. At baseline and at 12 months we collected data on demographics, anthropometric measurements, health, FV intake estimated with a validated instrument, and eating behavior. At 36 and 48 months we collected data on FV intake using telephone interviews. In the 1st manuscript we validated FV intake from telephone interviews using one sample t tests, concordance correlation coefficients (CCC), and Bland-Altman plots. In the 2nd manuscript we assessed intervention effects over time using difference-in-differences models, with baseline data as a reference. In the 3rd manuscript we analyzed the maintenance of positive changes in FV intake with the use of Generalized Estimating Equations (GEE) models, comparing the CG and IG on FV intake from 12 to 36 months and from 36 to 48 months. Longitudinal data were analyzed according to intention-to-treat, and missing data underwent multiple imputation. Results: we collected data from 3414 PAS participants at baseline. At 12 months, 2241 baseline respondents (65.5%) participated in the first follow-up. At 36 and 48 months, 2371 (69,4%) and 2112 (61,9%) participants were assessed, respectively. The validation sample was made up of 441 participants, and agreement between data from face to face and telephone interviews was substantial for fruit intake (CCC=0,629) and moderate for vegetables (CCC=0,419). At 12 months the intervention was effective in increasing fruit intake in the overall sample, and fruit and FV intake among participants with low baseline FV intake. Among the overall sample, fruit intake at 12 months went from 164,1g/day to 181,1g/day in the IG, and from 172,3g/day to 177,4g/day in the CG, which indicates a significant change in the difference between groups (p<0,05). At 12 months, IG participants who had a low FV intake at baseline increased their fruit intake from 103,2g/day to 162,7g/day, while in the CG it went from 109,1g/day to 150,1g/day; FV intake in the IG went from 243,4g/day to 339,0g/day, while in the CG it went from 252,0g/day to 327,9g/day (p<0,05). Regarding the maintenance of intervention effects, there was a decrease in FV intake of 92.9g/day at 36 months (-95,3g/day among the CG and -89,8g/day among the IG, p>0,05), and an increase of 30.3g/day at 48 months (+32,0g/ day among the CG and +28,1g/day among the IG, p>0,05), with no association with the intervention at either 36 months (p=0,23) or 48 months (p=0,34). Conclusion: data from telephone interviews were valid for estimating FV intake, allowing for faster and less costly monitoring. The intervention promoted a favorable evolution of fruit intake at 12 and 36 months and FV intake at 12 months, but was not associated with the maintenance of positive results at 36 or 48 months. Fruit intake evolved differently from vegetable intake over time, which reinforces the need to separately assess the intake of these food groups. Considering these results, we suggest periodically repeating interventions to sustain positive results in the long-term.