The Healthy Bodies Healthy Hearts (HBHH) study was a randomized, controlled trial delivered in coordination with primary care and intended to increase leisure physical activity (PA) levels and decrease weight in overweight and obese, inactive, middle-aged women. Primary outcomes were changes in PA level and weight at 3 and 12 months. Secondary outcomes were BMI, waist circumference, and blood pressure. The protocol was approved by University of Pittsburgh institutional review board (IRB) and registered with ClinicalTrials.gov (NCT01099007).
Assessments and the interventionist-led group intervention were conducted at the largest and most centrally located of three primary care practices affiliated with University of Pittsburgh Medical Center (UPMC). Two of the primary care practices were general medicine academic practices with faculty and resident providers; the last was a primary care practice for mid-life women with both obstetrics and gynecology (OB-GYN) and general medicine faculty providers.
Recruitment and Baseline Screening
Recruitment began in June 2009 and was completed in February, 2010. We recruited women aged 45–65 years, with a BMI ≥ 25 kg/m2 who were relatively physically inactive (less than one hour of PA per week), and receiving primary care at one of three UPMC practices. Exclusions included: unstable cardiac or pulmonary disease (e.g., recent myocardial infarction), poorly controlled hypertension (i.e., systolic blood pressure [SBP] ≥ 180), primary care physician (PCP) unwilling to allow moderate PA, and participant unable to perform moderate PA (i.e., unable to walk due to severe pain).
Our primary recruitment tool was a best-practice alert in the electronic health record that notified the primary care PCP of a woman who potentially qualified for the study. If the woman expressed interest, the PCP would enter a referral order and endorse that moderate PA was safe for the patient. After referral, a screening call was made to the woman by study staff to assess eligibility and interest. If the woman met criteria and was interested, a 60–90 min, in-person baseline assessment was scheduled.
Baseline screening started with a review of the consent form and the participation requirements of the study by study staff. Following the signed consent, height, weight, blood pressure, and waist circumference were obtained by trained study personnel. The Modified Activity Questionnaire (MAQ) was administered by a trained research staff member.15
Once the participant had been screened, the consent signed, and the initial baseline assessments completed, randomization occurred in a 1:1 allocation. Each woman was allowed to draw a sealed envelope that contained a designation assignment, either interventionist-led (IL) or self-guided (SG).
IL groups were comprised of 12 weekly group sessions and were conducted in a conference room in the largest of the three primary care practices. There were 12 designated topics for the sessions (Online Appendix Table A), led by the study physician (MBC), who was also a provider at the practice, or one of two PhD-level interventionists (KLS, KCS) not affiliated with the practice. The content focused on physical activity, diet, and stress relief adapted from a lifestyle intervention for premenopausal and postmenopausal women13
17 as well as portions of the Diabetes Prevention Program (DPP).18 Participants were given PA and dietary goals similar to those given in the DPP, with target calorie and fat gram goals based on starting weight and a PA goal of at least 150 min of moderate PA per week. Participants were given a calorie-counter book, a pedometer, and sheets for tracking PA and diet; tracking sheets were not collected or reviewed by the interventionist. The sessions included 30 min discussions followed by 30 min of group-based, moderate-intensity PA.
Mindfulness concepts were formally introduced through mindful eating and mindful PA sessions, and were integrated throughout the 12 week intervention.19 The participants practiced bringing their awareness and attention to the present moment with intention and without judgment of thoughts, emotions, body sensations, sensory perceptions, and behavioral actions. They were instructed to be an impartial witness to themselves during the experience of eating and exercise.
SG participants received a 12-week, self-guided manual based on the American Heart Association’s Choose to Move program20 at randomization, as well as a calorie-counter book and a pedometer. There was no interventionist contact with the SG group, but they had the contact information of study personnel for any questions or concerns.
In-person assessments for outcomes were performed at 3 and 12 months after baseline. Participants were not compensated for the 3-month visit, but did receive a $25 payment for completing the 12-month visit either in person or over the phone. If it was not possible for a participant to come in for a 12-month, in-person assessment, a brief version phone visit was conducted, focusing on PA data and a self-reported weight. Referring providers received letters after the 3 and 12 month outcome assessment visits informing them of their patients’ current weights and physical activity statuses.
Weight was measured by a trained staff member in clinic using a standard balance beam scale (SECA Medichoice) and following a written protocol.13
Physical Activity was measured using the one-month version of the Modifiable Activity Questionnaire (MAQ), administered by a trained, certified staff member. We focused on leisure activity estimates, because our intervention was designed to target leisure PA.13 PA levels were calculated as the product of the duration and frequency of each activity (in hours per week), weighted by an estimate of the metabolic equivalent (MET) of that activity and summed for all activities performed. PA data were expressed as metabolic equivalent hours per week (MET-hour/week). The MAQ has been shown to be both reliable and valid, and has been widely used in numerous populations and intervention studies.15
Clinically relevant secondary outcomes, typically lowered by a physical activity and weight loss intervention, include:
BMI, calculated using a standard equation.
Waist Circumference, measured using a non-stretch tape measure marked in centimeters.
Blood pressure, measured averaging three readings collected using a written protocol13 by trained staff using manual blood pressure cuffs (Welch Allyn).
Psychosocial outcomes included Physical Activity Enjoyment Scale (PACES),21 Beck Depression Inventory (BDI),22 Pittsburgh Sleep Quality Index (PSQI),23 Short Form-12 Health Survey (SF-12),24 and Exercise Self-Efficacy.25
27 Information about these measures and the results of these analyses are found in the online Exploratory Psychosocial Measures Appendix (Online Appendix B; Online Appendix Tables B1/B2).
Race, education, marital status, and medical comorbidities were assessed using self-report. The first three were assessed at baseline, and the last at 12 months (n = 83).
Descriptive statistics characterized the outcome variables at baseline, 3 months and 12 months. To assess differences between groups at baseline and 3-month and baseline and 12-month follow-up, two sample t-tests (continuous, normally distributed outcomes) or Wilcoxon’s Rank Sum (continuous, non-normally distributed) were conducted. Our primary outcome analysis compared groups according to the intention-to-treat approach. We used mixed models with fixed effects of group, time (3 and 12 months) and group x time interaction, controlling for baseline measure and BMI for primary outcomes and only for baseline measurements for secondary outcomes. For models, square root transformation was used for the physical activity outcome variable. The mixed models take into account the correlation among measurements within a person over time. We also conducted analyses stratified by race. Analyses were conducted using SAS version 9.3 statistical software (SAS Institute Inc., Cary, NC, USA).