Small Steps for Big Changes - Recipe for Health

The study is a one-group repeated measures design and the primary outcome is hemoglobin A1c% (HbA1c%). There is no randomization of individuals to conditions. HbA1c will be measured using point-of-care test at eight timepoints. Secondary outcomes include diet (caloric, carbohydrate and fibre intake), anthropometrics (weight and waist circumference) and eating motives.

Objectives:

Given the importance of diet for HbA1c% improvement among those at risk of developing type 2 diabetes (T2D), the purpose of this study is to assess caloric, carbohydrate and fibre intake in real-life settings and evaluate whether these dietary factors predict HbA1c levels over 24 months. Second, the researchers aim to investigate whether dietary changes predict weight and waist circumference among the study participants. Further, food choices are determined by many factors, including but not limited to biological, physiological, psychological and socio-cultural factors. Subsequently, the investigators aim to explore whether participants' motives for eating affect their dietary intake, and whether such motives change after partaking in SSBC.

Sample Size:

A total sample size of 214 is required to detect an HbA1c change between the baseline and the three follow-up segments. Linear trajectory was assumed between baseline and the 9-12-month segment, then a stable trajectory to the 21-24 month segment. The trial is powered to detect a clinically significant change across all segments. As such, the study is powered to detect the smallest expected change in HbA1c, which occurred at the 3-6 month segment based on the Phase IV effectiveness findings (d = .23) based on a previous RCT. Using G*Power, an initial sample size of N = 178 was calculated assuming two-sided type-I error probability of 5% and a power of 80%, d = .23 change in HbA1c at 3-6 months, and three covariates (age, sex, gender). Informed by Phase IV trial statistics, the investigators increased the sample to account for a <17.2% loss to follow-up (178/.828 = 214). The community partner agreed to deliver SSBC program to up to 250 individuals. Accordingly, 250 patients will be recruited.

Recruitment:

Participants will be primarily recruited through the Valley Medical Laboratories (Kelowna, BC). Potentially eligible participants with an HbA1c of 5.7-6.4% will receive a flyer about the program and contact the study team if they are interested. Community members may also contact our research team directly if they are interested in participating.

Trial Design:

This study is a treatment-only repeated measures design. The investigators aim to recruit adults with prediabetes to take part in the SSBC program with baseline measures taken at (-3) months before and at the beginning of the program, immediate post-intervention (after session 6) and follow-up measurement time-points at 3-, 6-, 9-, 12-, 21-, and 24- months post-intervention. The primary outcome is glycated hemoglobin (HbA1c). The secondary outcomes are caloric, carbohydrate and fibre intake as well as weight and waist circumference. Eating motives are also the secondary outcomes and predictor variables which will be dichotomized into physiological (hungry, tired) and psychological motives (bored, stressed, sad, social, eating time).

Diet and Exercise Intervention (SSBC program) at Community Site:

SSBC program will be delivered by partnering organization, the YMCA of Southern Interior. Participants will receive diet and exercise counselling aimed at setting personalized goals, improving self-monitoring skills, and self-regulatory efficacy. These consults are designed to increase participants' adherence to healthy diet and regular exercise. Coaches highlight healthy dietary strategies (reducing added sugar and refined/processed carbohydrate intake, increasing vegetable consumption, reading food labels, Canadian Food Guide) that are linked to reduced risk of T2D. Participants and coaches create detailed plans for how to meet their goals and check in to discuss progress at each visit. Counselling sessions last approximately 20-30 minutes. Following counselling sessions, exercise sessions with a trained coach occur over the course of the 6-week. Participants are supervised while they complete their chosen form of aerobic exercise: 1) traditional continuous exercise that involves exercising at a constant pace for 20-30 minutes (moderate-intensity continuous training, MICT); or 2) interval exercise that involves alternating between 30 seconds and 1-minute of brisk exercise and 1-minute of easy exercise for ~15 min (high-intensity interval training, HIIT). Both exercise forms could be completed on treadmill, elliptical or stationary bike. The trained coaches apply the principles of gradual progression. A combination of continuous MICT and HIIT are used to increase enjoyment and variety, as well as encourage progression.

Data Collection:

1. HbA1c: HbA1c% scores will be collected using the point-of-care HbA1c test at DPRG lab. It is a simple finger-prick test that provides the patient's HbA1c value within minutes. The HbA1c test indicates the sugar levels in the participant's blood without fasting. Point-of-care assessment will be conducted using the World Health Organization (WHO) guidelines for capillary sampling.
2. Diet \& eating motives: Dietary outcomes and eating motives will be assessed through Keenoa image based food tracking application. Keenoa is a smartphone imaged-based dietary assessment application that recognizes and identifies food items using artificial intelligence. To capture free-living diets, participants will be asked to take pictures of their foods, snacks, and beverages for 3 days at each time point on consecutive 2 weekdays and 1 weekend day. Participants can log packaged and processed foods using the barcode scanner feature of the app. In addition to the food images, participants will be asked to answer a multiple-choice question about their eating motives through the app prior to each meal. They will be given the option to choose one of the following reasons for eating: 'bored', 'stressed', 'sad', 'social', 'it's time', 'tired', or 'hungry'. Participants will complete the dietary assessment at home. Keenoa sources its nutrient data from the Canadian Nutrient File, United States Department of Agriculture, and Hong Kong Food Database, which makes it convenient to use in a study conducted in Canada. To date, two validation studies comparing Keenoa with other dietary assessment tools have been published. In addition to the acceptable validity, a qualitative study found that Keenoa is user friendly and time saving.
3. Weight \& waist circumference: Weight will be measured on a stadiometer. Waist circumference will be measured at the top of the iliac crest (hip bone) using a tape.

Statistical Analyses:

An interrupted time-series design will be used to test client-level outcomes on an intention-to-treat basis. Interrupted time-series consists of repeated outcome observations across time, with a break in series corresponding to the introduction of an intervention. The average of multiple baseline "control phase" assessments are compared to "intervention phase" outcome trends to determine effectiveness. HbA1c will be assessed at 8 primary time points: 2 control phase assessments (3-months before, and at baseline [month "0"]) and 6 support phase follow-up time points (3-, 6-, 9-, 12-, 21-, and 24-months post-intervention). The 3-month periodization is clinically meaningful as the primary outcome of HbA1c reflects average blood glucose over ~3 months. Two baseline assessments account for 6 months of pre-intervention HbA1c measurement (i.e., 6-3 and 3-0 months pre-baseline), strengthening our ability to detect meaningful change. Interrupted time series HbA1c data will be analyzed using a piecewise "segmented" regression with four segments: | baseline | 3, 6 | 9, 12 | 21, 24 | to indicate changes from baseline (0, -3 months) to early adoption (3-, 6-months post-intervention), medium (9- and 12-months post-intervention), and long-term (21-, 24-months post-intervention) adherence. The researchers will adjust for time trends and covariates (age and sex) by allowing for adjustment of both within-subject and within-site correlations. A sensitivity analysis with missing data imputation will be conducted to rule-out the potential impact of non-adherence on study outcomes.

A repeated measures ANOVA will be performed to determine whether there are any differences in dietary intake [caloric (kcal), carbohydrate (grams and %) and fibre (grams) intake] in participants over time. A multiple regression will be performed to determine the relationship between dietary intake [caloric (kcal), carbohydrate (grams and %) and fibre (grams) intake] and HbA1c levels while controlling for confounding factors (i.e., age and sex). The relationship between dietary behaviours, weight and waist circumference will be analyzed using a multiple regression. Eating motives are predictor variables and will be dichotomized into two groups: physiological (i.e., hungry, tired) and psychological (i.e., bored, stressed, sad, social and eating time) eating motives. A binomial logistic regression will be performed to evaluate whether eating motives are predictive of dietary intake.