The effect of different monitoring strategies on weight, health, and eating behaviour

Abstract

A variety of dietary and exercise approaches appear to be effective for weight management in the short-term, however these need to be accompanied by behavioural strategies to make them sustainable long-term. One of the strongest predictors of success is adherence to monitoring progress. However, although multiple monitoring options exist, they have typically been included as part of a suite of behaviour strategies, making it difficult to determine which monitoring components are the most effective.

This thesis presents findings from the SWIFT (Support strategies for Whole-food diets, Intermittent Fasting, and Training) randomised parallel study, which determined the effectiveness of four different monitoring strategies (face-to-face monitoring, self-monitoring dietary intake, self-monitoring daily body weight, or self-monitoring hunger) on weight loss, body composition, blood markers, and psychosocial indices in overweight and obese adults undertaking a 12-month weight loss programme. We first needed to ascertain the feasibility of the hunger training arm, as adherence had never been reported and was of concern. Hunger training teaches individuals to connect their physical symptoms of hunger with their blood glucose levels and to eat only when blood glucose is below a set target. We conducted a two-week feasibility study with 30 participants. The first 20 participants were asked to only eat if their blood glucose was 4.7 mmol/L or less prior to eating (protocol A), and the remaining 9 participants used an individualised cut-off based on their fasting glucose concentrations (protocol B). All but one participant completed the study (97% retention compared to feasibility criterion of 85%) and participants measured their blood glucose before 94% (95% CI 91, 98) of eating occasions (criterion 80%). However, participants following protocol A adhered to eating when blood glucose was below the prescribed level 66% of the time compared with 84% in those following protocol B (p = 0.010, within-person criterion of 75%).

Our individualised approach to hunger training (protocol B) was therefore used in the wider SWIFT study. 250 overweight or obese adults were randomised to control, brief monthly individual consults (brief support), daily self-monitoring of body weight, self-monitoring of dietary intake using MyFitnessPal, or hunger training. All groups received diet and exercise advice and 171 participants remained at 12 months (68.4% retention). All groups lost weight over the course of the intervention (typically 3.9-6.8kg) with no evidence of a difference between the control group and each of the intervention groups (all p≥0.084). However, participants in the hunger training group lost significantly more weight at 12 months than those in brief support (2.9kg, 95% CI 0.8-5.1kg, p=0.008) or MyFitnessPal (3.2kg, 0.1-6.4kg, p=0.046) groups. Few significant differences were observed in eating behaviour (all p≥0.111), although brief support and hunger training reported more favourable effects on depression and anxiety at 12 months than control participants. Adherence to the monitoring strategies (% recommended days) ranged from 29.6% for hunger training to 63.6% for attendance at the monthly brief support sessions.

Lastly, we conducted post-hoc analyses of how adherence to hunger training influenced weight, eating and dieting behaviour, given the potential promise of this approach. Adherence was analysed according to blood glucose measurement, and completion of a booklet, throughout the study period. Participants who completed at least 60 days (recommended 63 days) lost 6.3 percentage points (95% CI 3.1, 9.5; p<0.001) more weight than those with lower completion rates. Compared to participants who completed less than 30 days of hunger training, participants who completed at least 30 had a decrease in emotional eating (p=0.004) and an increase in dietary restraint (p=0.004) and those who completed at least 60 days of hunger training had an increase in their intuitive eating score (p=0.042), and a decrease in both emotional eating (p<0.001) and external eating (p=0.029), compared to those that completed less than 30 days.

Collectively, the results from these three studies highlight that when monitoring strategies are used in a public health intervention with minimal contact, they lead to a modest amount of weight loss, with no difference between groups. However, when monitoring strategies are compared to each other, hunger training appears to be the most effective for weight loss. A high adherence to hunger training is necessary to yield clinically significant weight loss and improvements in eating behaviour, including a decrease in emotion and external eating, and an increase in intuitive eating. Overall, these results demonstrate that there is little difference between self-monitoring diet, weight, and face-to-face monitoring compared with a control group on weight or health markers, and hunger training may be a worthwhile strategy to continue to investigate.