Step Monitoring to Improve ARTERial Health

BACKGROUND: High walking levels reduce myocardial infarction, stroke, and mortality rates in overweight/obese patients with diabetes and/or hypertension, but our own work, led by Nominated Principal Applicant K. Dasgupta, indicates low daily step counts in these patients, at approximately 5,000 steps/day on average with a further 15% reduction during fall and winter. A meta-analysis of physical activity programs indicates that pedometer-based monitoring programs can lead to higher daily step counts, but evidence for impact on arterial health is limited, as is evidence for the effectiveness of a pedometer-based strategy specifically implemented into the usual clinical care of patients with diabetes and/or hypertension. The SMARTER trial will address these knowledge gaps.

PRIMARY RESEARCH QUESTION: Among sedentary overweight/obese adults with diabetes and/or hypertension do physician-delivered step count prescriptions integrated into usual care reduce arterial stiffness more than usual care alone, over a one-year period? Arterial stiffness (primary outcome), a summative indicator of arterial health, is more precise and reliable than individual risk factors. An analysis of the Framingham Heart Study demonstrated that, even after adjustment for traditional risk factors, increased arterial stiffness was independently associated with a 48% increase in vascular disease risk. Co-Principal Applicant S. Daskalopoulou is an expert in the noninvasive assessment of arterial stiffness and has a well-equipped Vascular Lab funded through a CFI grant.

STUDY DESIGN: Randomized, allocation concealed, single-blind (outcome assessors), intervention allocation ratio 1:1, multisite clinical trial. This design will allow for the level A evidence necessary to justify widespread change in clinical practice.

TRIAL SYNOPSIS: Given that the majority of diabetes and hypertension patients are managed in primary care settings, the SMARTER trial interventions will be delivered through the large network of primary care clinics accessible to Co-Principal Applicant E. Rosenberg as well as diabetes and internal medicine clinics throughout Montreal where patients may receive their primary diabetes and hypertension follow-up. Twenty-four collaborating physicians have been identified. Physicians/clinic staff will obtain assent from candidates within their practice for contact by the SMARTER coordinator. The number of collaborating physicians continues to be increased, including physicians at primary care, diabetes, hypertension, internal medicine, and endocrinology clinics. Eligibility: Candidates will be adults with 25≤BMI<40 kg/m2 followed for diabetes and/or hypertension and sedentary to somewhat active. Evaluations: Formal trial evaluations, conducted at baseline and 12 months, will include assessments of arterial stiffness (carotid femoral pulse wave velocity measured noninvasively with applanation tonometry); step counts (pedometer with concealed window) and physical activity (accelerometer) worn for one week; fitness (exercise stress test; ˙VO2max); anthropometric parameters; and individual vascular risk factors. Intervention arm: The physician gives the active trial participants a pedometer, log book, and a step count prescription based on the baseline daily step count. The time frame for a > 3,000 steps/day net increase is 10 months for sedentary participants (<5,000 steps/day), 7 months for low active participants (5,000-7,499 steps/day), and 5 months for somewhat active participants (7,500-9,999 steps/day). There will be four clinic visits over one year. Control arm: Same visit frequency with advice to engage in 30-60 minutes of activity on most days of the week. Sample Size: Allowing for a loss to follow-up of up to 17% based on our previous studies, investigators will require a sample size of 364 individuals (i.e. 182 per arm) to detect a 10% difference in change in arterial stiffness between our active and control arms to an accuracy of +/- 5% over a one-year period. Analysis: Intention-to-treat. Between-arm differences in 'after minus before changes' with 95% CIs for main analysis.

Addedum to stress testing: Due to timeline limitations, we were obliged to forego stress testing assessments as of October 20, 2014. This does not impact our primary outcome. Moreover, stress testing is not required when engaging in a walking program in type 2 diabetes. We will be able to assess impacts of the intervention on fitness (secondary outcome) in a subgroup of patients (i.e., those who completed assessments before October 20, 2014).

IMPORTANCE: With increasing numbers of diabetes and hypertension patients, there is a pressing need for effective and efficient clinical practice strategies to help physicians support their patients to achieve the arterial health benefits of higher physical activity levels. The SMARTER trial seeks to provide such a tool. If effectiveness is demonstrated, all efforts will be made for the inclusion of our approach in Clinical Practice Guidelines for diabetes and hypertension, and investigators will develop training tools (manuals, websites, CD-ROMs) to allow maximal uptake of our proposed strategy.

AN OBSERVATIONAL SUBSTUDY: Novel Real-Time Measurement of Physical Activity Patterns in Type 2 Diabetes and Hypertension Through GPS Monitoring and Accelerometry

In addition to the main clinical trial, we are conducting additional measurements among consenting type 2 diabetes patients in order to examine the effects of the walkability of their home neighbourhood on their baseline step count and time at different physical activity intensities (accelerometer measurement already being performed through SMARTER). The additional measurements include wearing a Geographical Positioning Systems (GPS) device for the 7-day period that they wear the pedometer with concealed viewing window and accelerometer. The GPS device collects time-stamped location information such that X,Y coordinates are collected. These are used to determine the times that they are within or outside neighbourhood buffer zones.

For the assessment of neighbourhood walkability,the parameters assessed include population density, pedestrian-friendly design and diversity of destinations - commonly referred to in the urban planning literature as the 3D's. The variables that best capture density, design, and diversity include residential density, street connectivity and land use mix. Residential density is defined as the number of residences per square kilometre of residential land area. Street connectivity is defined as the number of ≥3-way intersections per square kilometre in neighbourhood, where a greater number of intersections facilitates movement between origins (e.g., residences) and destinations (e.g., shops and parks).Land-use mix is a measure of the number of different land uses located within a neighbourhood.Land use mix is assessed via an entropy score - a value between zero and one that captures the degree of heterogeneity of land uses in a neighbourhood. A subcomponent of land use mix that may be a particularly important for encouraging individuals to walk within their neighbourhood and that is easily incorporated into the design of new neighbourhoods is greenspace/recreational land area.

We are using Geographical Information System (GIS) mapping (computer-based assessment of neighbourhood characteristics derived from existing data sources that have some spatially referenced identification, such as a home address) to measure these facets of neighbourhood walkability.In brief, each of the variables will be derived by geocoding participants six-digit home postal codes, constructing 1-kilometre polygonal buffers zones around each participants home address (i.e., a geographical zone around the centroid of the postal code area) and calculating the measures of interest for each neighbourhood using tools within a GIS software package (ArcGIS) and publically available shape files.

Means and standard deviations will be used to describe the number of steps per day occurring specifically in home neighbourhoods (i.e., as determined through GPS) and overall. Multiple linear regression analyses will be used to assess the relationship between 1) home neighbourhood environments and the number of steps taken per day in the home neighbourhood and 2) home neighbourhood environments and the number of steps taken per day taken in any location. These analyses will be repeated with time at moderate to vigorous activity in lieu of steps as the outcome variable. Several variables measured through SMARTER will be considered for exclusion in models (e.g., age, sex, educational level, BMI).

This observational substudy is partly funded by an operating grant from the Heart and Stroke Foundation (Quebec) awarded to K. Dasgupta (Principal Investigator) and Nancy Ross (Co-Principal Investigator on substudy) and is being led by Samantha Hajna, their doctoral candidate student.