Article Text

Prevention and Reduction of Obesity through Active Living (PROACTIVE): rationale, design and methods
  1. R Ross1,2,
  2. S N Blair3,
  3. M Godwin4,
  4. S Hotz5,
  5. P T Katzmarzyk6,
  6. M Lam7,
  7. L Lévesque1,
  8. S MacDonald8
  1. 1
    School of Kinesiology and Health Studies, Queen’s University, Kingston, Ontario, Canada
  2. 2
    Department of Medicine, Division of Endocrinology and Metabolism, Queen’s University, Kingston, Ontario, Canada
  3. 3
    Department of Exercise Science, University of South Carolina, Columbia, South Carolina, USA
  4. 4
    Primary Healthcare Research Unit, Memorial University of Newfoundland, St John’s, Newfoundland, Canada
  5. 5
    Department of Psychology, University of Ottawa, Ottawa, Ontario, Canada
  6. 6
    Pennington Biomedical Research Center, Baton Rouge, Louisiana, USA
  7. 7
    Department of Community Health and Epidemiology, Queen’s University, Kingston, Ontario, Canada
  8. 8
    Centre for Studies in Primary Care, Queen’s University, Kingston, Ontario, Canada
  1. Dr R Ross, School of Kinesiology and Health Studies, 69 Union Street, Queen’s University, Kingston, Ontario, Canada K7L 3N6; rossr{at}


The Prevention and Reduction of Obesity through Active Living (PROACTIVE) is a randomised controlled trial to evaluate the effectiveness of a behaviourally based physical activity and diet composition programme to prevent and reduce obesity and related comorbidities in a primary healthcare setting. 491 abdominally obese men and women 25–75 years of age who were patients of primary care physicians were randomly assigned to either a usual care group (N  =  242) or a behavioural intervention group (N  =  249). Those in usual care received general advice from the physician regarding the merits of physical activity and a healthy diet as a strategy for obesity reduction. Those in the behavioural intervention group received an individually designed counselling programme from a specially trained health educator, with respect to physical activity, diet and obesity reduction. The study was designed to provide 95% power in both men and women to detect a 2% (2 cm) difference in waist circumference and 80% power to identify a 15% reduction in the prevalence of the metabolic syndrome, the two primary outcomes. PROACTIVE is the first behavioural intervention study to assess the effects of physical activity and diet on abdominal obesity and associated metabolic risk factors in a primary healthcare setting, include a generalised sample of men and women and examine long-term (24 months) effects. PROACTIVE has the potential to provide the basis for changing clinical practice (primary care) with respect to the prevention and reduction of obesity and related health risks. The purpose of this report is to present and discuss the rationale, design and methods of PROACTIVE.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

The increase in obesity in adults and children in North America indicates that current interventions to prevent and reduce the prevalence of overweight and obesity are inadequate. In response we developed a novel approach to the prevention and treatment of obesity at the individual level, intrinsic to which are three key elements.

First, primary care physicians remain the cornerstone of healthcare delivery in Canada and are thus an integral component of the programme. Despite acute awareness of the medical importance of reducing obesity and promoting healthy eating and exercise, current data suggest that clinicians do not adequately counsel obese patients.1 Commonly cited barriers include lack of time, lack of counselling expertise and limited staff support and access to educational materials.2 We acknowledge these limitations, and propose in response that effective obesity reduction at the primary care level requires that the clinician work in concert with a specially trained health educator.

Second, the treatment strategy designed to prevent and/or reduce obesity will be achieved through an emphasis on increasing daily physical activity and eating a healthful diet without a specific focus on decreasing energy intake. Several lines of evidence support this position. Current knowledge indicates that a decrease in daily energy expenditure levels in recent decades is a major determinant of the increase in the prevalence of obesity.3 This, combined with the knowledge that cardiorespiratory fitness per se attenuates health risk4 and, that exercise enhances the maintenance of weight loss long term,5 supports a primary role for physical activity in obesity reduction programmes. However, this does not infer that the treatment strategy does not include diet counselling. Advice provided to all participants includes counselling related to proper diet composition (eg, fat intake and emphasis on a plant-based diet and proper portion sizes and avoiding highly processed foods).

The third key component of the proposed trial is that change in waist circumference and not body mass index (BMI) or body weight is the primary obesity-related outcome. It is now firmly established that waist circumference predicts morbidity and mortality independent of age and BMI.6 7 It is also established that abdominal obesity (subcutaneous and visceral fat) is preferentially reduced in response to weight loss, in particular, exercise-induced weight loss.8 Furthermore, increasing evidence suggests that exercise is associated with a marked reduction in abdominal obesity (waist circumference) despite minimal or no weight loss (ie, no change in BMI).9 Together, these observations reinforce the importance of waist circumference as a clinical tool to assess the effectiveness of obesity prevention and reduction and thus our decision to employ waist circumference as a primary outcome variable.


The primary objective of the Prevention and Reduction of Obesity through Active Living (PROACTIVE) trial is to assess the effectiveness of a 24-month behaviourally based physical activity and diet programme to prevent and reduce obesity and related comorbid conditions in abdominally obese men and women. We hypothesise that an individualised education and behaviour counselling programme (intervention) delivered to patients by a health educator in a primary care setting will result in a reduction in waist circumference and a reduction in the metabolic syndrome and progression to the metabolic syndrome compared with usual care. A secondary objective is to evaluate the dose–response relationships between increases in physical activity and corresponding changes in waist circumference and the metabolic syndrome.

There are two primary outcome variables: waist circumference and the metabolic syndrome. Waist circumference was chosen as the indicator of abdominal obesity because of its independent association with morbidity and mortality. The metabolic syndrome is defined as the co-occurrence of a constellation of atherogenic, diabetic and hypertension risk factors within an individual, which substantially increases the risk of premature morbidity and mortality.10 11


Description and selection criteria of participants

PROACTIVE participants include 491 sedentary, community dwelling adult men and women aged 25–75 years who are overweight (BMI 25–39.9 kg/m2) and abdominally obese (waist circumference >102 or >88 cm in men and women, respectively) recruited from family medicine clinics in Kingston, Ontario, Canada. The inclusion and exclusion criteria for PROACTIVE are listed in table 1 and baseline characteristics of the subjects are presented in table 2. Because this is a pragmatic trial designed for general implementation in a clinical setting, patients identified with dyslipidaemia, type 2 diabetes mellitus or hypertension were not excluded from participation. All participants were willing and able to increase daily physical activity and provided informed consent.

Table 1 Inclusion and exclusion criteria for PROACTIVE
Table 2 Baseline characteristics of the intervention and control groups by gender

Clinical setting, recruitment and screening

Participating physicians in PROACTIVE work within one of three family practice clinics in Kingston. Approximately 12 physicians within each clinic are participating in the trial. Each physician supplied his or her patient list to the PROACTIVE project coordinator, who subsequently created an information letter for each potential participant. Before mailing, each physician reviewed letters addressed to his or her patients and removed those for patients known to be ineligible or unable to participate in PROACTIVE for other reasons. Participant flow from screening to randomisation is illustrated in fig 1.

Figure 1 Participant flow from screening to randomisation.


Eligible participants were randomly assigned by the data manager based on a computer-automated randomisation sequence, following acquisition of all primary outcome data to prevent potential enrolment data bias. Although randomisation theoretically leads to an equality of all factors in both intervention and usual care groups, we employed a stratification procedure to ensure equal numbers of participants with specific and potentially confounding characteristics in both groups. Randomisation was stratified based on sex, age and waist circumference.

Blinding, contamination, bias

It was not possible to conceal the group assignment from patients or physicians in this study. Nonetheless, strategies were employed to decrease the likelihood of contamination and bias; these included the following: (1) all baseline data were collected before randomisation, which ensured that measurement staff were “blinded” to group assignment; (2) study participants were asked not to disclose their group assignment with measurement staff at subsequent visits; (3) the blood pressure measurements were performed by a semi-automated blood pressure machine (BP TRU) greatly reducing human error, inter-tester reliability and possible bias; (4) the remaining measurements used to determine the presence of the metabolic syndrome are automated laboratory tests (fasting glucose, triglycerides, high-density lipoprotein (HDL) cholesterol) and are unlikely to be affected by human bias; (5) physicians were explicitly informed of the importance of not making changes to their usual approach to lifestyle counselling during the course of the study.

Follow-up examinations

Follow-up data collection visits occur at 6-month intervals after randomisation. As with baseline measurements, all follow-up visits occur within the clinical laboratory of the principal investigator (RR). At these visits, data to determine the primary outcomes are obtained as well as data on participation in physical activity using the 7-day physical activity recall (PAR) questionnaire. Information on adverse events and medication use is also obtained each visit.

Outcome measures

The primary outcome variables are waist circumference and the metabolic syndrome. Secondary outcome variables include other measures of body composition, total serum cholesterol, low-density lipoprotein cholesterol, physical activity, barriers to physical activity and other psychosocial variables. Other variables of interest include medical history and medication. Reliability data for the acquisition of all anthropometric measures were obtained for each staff member. Primary outcome data for all patients at each time point were reported to the respective family physician.

Anthropometric measurements

Waist circumference is measured at the superior edge of the iliac crest according to established procedures.12 Other circumferences include proximal thigh, hip and biceps. Standing height is obtained using a wall-mounted stadiometer and body mass is measured on a balance scale.

Metabolic syndrome

The third report of the National Cholesterol Education Program Adult Treatment Panel13 criteria are used to diagnose the metabolic syndrome. Clinical data are collected at all measurement times by the study nurse. All testing is performed in the morning after an overnight fast. Fasting glucose, HDL-cholesterol and plasma triglycerides are measured using standard procedures. Arterial blood pressure is obtained with the subject in a seated position on the right arm, using a semi-automated system (BpTRU).14

Physical activity

Physical activity is assessed as energy expenditure (kcal/kg per day) estimated by the PAR questionnaire, which has been widely used and validated.15 16 The PAR is a structured interviewer-administered interview in which the participant is asked to recall the amount of time spent in each 24-h period during the past 7 days in six categories of activities classified according to their intensity expressed as metabolic equivalents (MET, or kcal/kg per hour). The six categories are sleep, sit, and light, moderate, hard and very hard physical activity. The amount of time spent in each category is multiplied by that category’s average MET value and the results are assumed to obtain an estimate of energy expenditure (kcal/kg per day). Time in light activity is obtained by subtracting hours in other activities and in sleep from 24 h. Light activity is partitioned further by considering hours spent sitting and other light activities separately.

Additional measures related to physical activity are collected by self-administered questionnaire, and include benefits and barriers to being physically active,17 exercise self-efficacy,18 processes of change,19 perceived environment related to physical activity (eg, neighbourhood safety, walkability, green space and parks)20 and social support from family and friends.21

Psychosocial measures

Information on a variety of psychosocial measures is collected by self-administered, validated questionnaires at baseline and each subsequent follow-up. Measures include depression,22 perceived stress,23 stressful life events,24 quality of life,25 self-determination,26 mastery,27 self-esteem,28 self-motivation,29 personality traits,30 and weight goals31 and a “weight loss data collection tool” comprising items on weight cycling and body image.

Safety monitoring

A data safety monitoring board was formed to ensure that the proper procedures were in place to protect the safety of participants while ensuring the validity and integrity of the study. The primary safety concerns in PROACTIVE are cardiovascular and musculoskeletal events that may be associated with engaging in physical activity. All serious adverse events that are brought to the attention of the health educator are reported immediately to the family physician, PROACTIVE medical director and principal investigator (RR) and follow-up is tracked. The data safety monitoring board receives notification of all serious adverse events. In addition to adverse events reported to the health educator, the measurement staff complete an adverse event questionnaire with each participant at each follow-up visit.

Sample size and statistical power

The required sample size was determined for the two primary outcome variables, waist circumference and the metabolic syndrome. We believe that clinically relevant changes are a 5% reduction in waist circumference and a 15% reduction in the prevalence of the metabolic syndrome. Given that the metabolic syndrome is expressed as a proportion and waist circumference is continuous and preliminary calculations indicated that a larger sample size would be required to detect differences in the metabolic syndrome, the sample size derivation was based on the metabolic syndrome, and the power for the sample to detect changes in waist circumference was calculated.

Based on analysis of the Canada Heart Health Surveys (CHHS), the prevalence of the metabolic syndrome in Canada is 17.0% in men and 13.2% in women.32 However, the prevalence of the metabolic syndrome in Canadians with the inclusion criteria in the proposed study (waist circumference >102 or 88, BMI <35 kg/m2) is 55.3% (men 61%; women 51%) (unpublished analysis of CHHS database). The higher prevalence in this group is to be expected as all participants have at least one risk factor (high waist circumference). A 15% absolute reduction in the prevalence of the metabolic syndrome would thus decrease it from 55.3% to 40.3%. In the Finnish Diabetes Prevention Study, the change in waist circumference after one year of lifestyle intervention was −4.3 (5.2 SD)%, or −4.4 (5.2 SD) cm.33 Therefore 5.2 cm was used in our calculations for the variance in change in waist circumference. Given that the mean waist circumference of participants in the CHHS with the inclusion criteria for the proposed study was 101.3 cm (men 108.3 cm; women 100.6 cm), we believe that a 5% reduction in waist circumference is of considerable clinical importance, as this translates into a reduction of 5.1 cm (men 5.4 cm; women 5.0 cm).

Assuming an alpha of 0.05, a beta of 0.2, and the 55.3% prevalence of the metabolic syndrome, 174 subjects were required for each arm to be able to identify a 15% reduction in prevalence.34 This sample size provides a power of over 99.9% to detect a 5% difference (5.1 cm) in waist circumference and a power of 94.6% to detect a 2% (2 cm) difference in waist circumference. We anticipate a 20% attrition rate for dropouts based on our previous experience, thus we were required to recruit a minimum of 218 subjects for each arm (a total of 436 subjects). We adopted a conservative approach to ensure an adequate sample size and power and thus aimed to recruit 250 subjects per group.

Statistical analyses

A two-tailed t test (or its non-parametric equivalent such as the Mann–Whitney U test) for continuous variables and a χ2 for categorical variables will be used to analyse potential differences between the usual care and behavioural intervention groups at baseline and to examine differences between drop-outs and participants who remained in the study. Intent-to-treat analysis35 will be applied. Missing data due to dropouts or non-compliance will be handled using multiple imputation methods. Sensitivity analysis will be conducted to study the impact of missing data. For each outcome, longitudinal multilevel modelling36 will be the primary analysis. The model includes the random subject effect and the following fixed effects: intervention, age, gender, baseline BMI, baseline waist circumference and the metabolic syndrome. The covariates age, gender, baseline BMI, waist circumference and metabolic syndrome included in the model are used to account for the baseline difference between the two groups. An alpha of 0.05 will be used to determine statistical significance.



This 2-year multiphase intervention to promote physical activity involvement among physically inactive abdominally obese adults is being conducted within the primary care setting. Intervention goals include a reduction in abdominal adiposity as indicated by decreases in waist circumference (waist loss), increased physical activity adoption and maintenance and improved diet quality. The theoretically based behavioural intervention designed to build self-regulatory skills for physical activity and diet is delivered by trained health educators. Our decision to use health educators to implement the behavioural intervention addresses some of the shortcomings noted in previous interventions in primary care, including lack of time that medical personnel and clinic staff have to counsel patients and lack of personnel with adequate training to counsel patients for changing physical activity and dietary habits. Each health educator holds a degree in kinesiology and undergoes comprehensive training before working with PROACTIVE participants.

A full-time health educator is assigned to each family practice clinic and delivers the programme sessions on-site within a private office equipped with a treadmill and all of the intervention materials. This setting offers a private and comfortable environment conducive to learning and discussion and conveys professionalism, credibility and a sense of security to participants by virtue of its location within a healthcare context.


Waist circumference reduction goal

The long-term goal is to sustain a reduction in waist circumference that ranges between 5% and 10% (∼4–15 cm) of the initial circumference. A reduction within this range is feasible and is associated with substantial reductions in metabolic risk factors.37 The short-term goal is to achieve a reduction of 5% in waist circumference within the first 6 months of the programme.

Specific behavioural goals

The intervention phase of the programme is based on the principle that participants will achieve negative energy balance. We attempt to ensure that the majority of the negative energy balance is induced by an increase in energy expenditure (physical activity). The rationale for this approach is twofold. First, it provides the flexibility required to accommodate individual differences and tolerance to physical activity. Second, it ensures that the majority of the change in waist circumference is induced by an increase in physical activity, which is consistent with the knowledge that exercise enhances long-term maintenance of weight loss.5 Study participants are also counselled to adjust their diet composition to meet current recommendations for a balanced diet following the Mediterranean food pattern.

Application of our approach will help ensure that the participants attain the goal of a 5% reduction in waist circumference within the initial 6 months. Furthermore, we calculate that for most individuals the exercise fraction of the negative energy balance will be achieved by accumulating approximately 45–60 minutes of daily, moderate (brisk walking) physical activity. This is a feasible target and consistent with current public health recommendations4 and with physical activity levels attained in earlier studies that have employed an intensive lifestyle modification approach (Diabetes Prevention Program and Finnish Diabetes Prevention Study).

Usual care group

Participants randomly assigned to the usual care group receive general advice from their physicians about the merits of physical activity as a strategy for obesity reduction, and continue to meet with their physician according to their usual schedule (typically once a year). At the end of the intervention, all usual care participants will be invited to attend a workshop on strategies to integrate physical activity and healthy eating into their lifestyle and will receive a selection of materials used in the intervention.

Behavioural intervention group

Participants randomly assigned to the intervention group receive individually tailored counselling based on the transtheoretical model38 39 and social cognitive theory.40 Given that people at different stages of change are considered to have correspondingly different needs to motivate stage progression,38 39 individual stage-based tailoring was chosen over group-based counselling. Physical activity interventions designed to encourage people to use the cognitive and behavioural strategies that correspond to their stage of readiness for change (from precontemplation to maintenance) have been shown to be successful in moving people forward through the stages of change.41 42

Although there is disagreement in the literature concerning the applicability of all transtheoretical model constructs to physical activity behaviour change, the central idea that people differ in terms of their readiness to commit to, adopt and maintain behaviour change provided a valuable heuristic upon which to tailor the counselling sessions. In particular, we believed that unless the intervention was “stage-matched”, there would be a risk that the counselling would not meet participants’ needs; this would reduce the face validity of the intervention and reduce the power of the intervention. Consequently, the health educators were taught to assess: readiness to change concerning the relevant behavioural targets; the degree of importance that participants placed on change and participants' confidence (self-efficacy) in modifying their behaviour. The health educators were taught to use their findings to guide the counselling process. Motivational interviewing43 served as the counselling model because it shares the perspective that readiness is strongly influenced by subjective importance and confidence and because it specifies a set of counselling skills that are believed to facilitate participants’ intrinsic motivation.

Phase 1: Intensive phase (0–6 months, 15 sessions, 15 h)

During the first 20 weeks the health educator works interactively one-on-one with participants to provide the knowledge and skills to increase daily physical activity based on his/her current level of psychological and physical readiness to change. Based on our previous work, we can assume that the majority of participants will be in the contemplation or preparation stage of change for exercise. A stage-based curriculum is implemented, based on materials adapted from those developed for Project Active, ACT and PRIME.44 45

During the first 6 weeks of the intervention, participants attend eight one-on-one hour-long sessions at the health educator’s office. Following an assessment at the first session, an individualised exercise plan is designed to allow participants to progress to approximately 45 minutes of moderate (eg, brisk walking) daily exercise. This is developed by the participants themselves, through behavioural tools such as goal-setting, addressing barriers to activity and self-monitoring.

During the subsequent 14 weeks, participants meet with the health educator once every 2 weeks for 60 minutes. During each session, the health educator reviews the participant’s physical activity self-monitoring log and food diary. Behavioural goals are reviewed and a different “theme” is covered at each meeting, during which specific behaviour change skill-building exercises are practised with the health educator. In addition to learning how to exercise, participants learn to set realistic goals, self-monitor activity (eg, use a pedometer or track minutes per day), problem solve, access social support and prevent relapse with the guidance of the healthy lifestyle educators.

At the sixth session, 3-day food diaries are reviewed by the health educator as an introduction to the Mediterranean food pattern. This pattern promotes the daily consumption of whole-grain foods, fruits, vegetables, legumes, nuts and seeds, healthy fats (olive and canola oil) and low fat dairy products. Changes in diet composition are implemented by the participant with the guidance of the health educator.

Phase 2: Months 7–12 (6 sessions, 3–6 h)

When a participant attains the waist loss goal of a 5% reduction in waist circumference, he or she enters a maintenance phase. For most participants, this phase begins at session 16 (month 6). At this point, participants are encouraged by the health educator to continue the behaviour programme as per the fundamental principles (45–60 minutes of activity per day and healthy eating patterns). Should a participant continue to lose weight (within the acceptable limit of 0.5 kg per week), this will serve to lend further support to our intervention.

What is already known on this topic

Although much is understood about the efficacy of physical activity as a strategy for obesity reduction, little is known about the effectiveness of physical activity as a strategy for obesity reduction in clinical and public health settings.

What this study adds

  • This study is the first behavioural intervention study to assess the effects of physical activity and diet on abdominal obesity and associated metabolic risk factors in a primary healthcare setting, include a representative sample of men and women and examine long-term (24 months) effects.

  • This study describes a comprehensive approach to delivering behavioural interventions for physical activity in a primary care setting.

Phase 3: Months 13–24 (12 sessions, 6–12 h)

During the second year of the intervention, contact with health educators continues, with the duration of sessions based on their waist circumference values and the adoption of physical activity. Those achieving the study targets meet the health educator bimonthly for 30-minute sessions (see fig 2). Maintenance issues of concern to the participant are discussed. Those participants who have not achieved the waist and physical activity goals see the health educator monthly, for 60-minute sessions to continue building physical activity and healthy eating behaviours into their routine.

Figure 2 Timeline of 2-year intervention including success criteria and time spent with health educator. aDuring phase 1 (baseline—6 months), participants have 15 1-h sessions with their health educator (HE). There are eight sessions during the first 6 weeks of the intervention. Subsequently, sessions are scheduled at biweekly intervals. bDuring phase 2 (7–12 months), participants have six sessions with their health educator at monthly intervals. The duration of the sessions is 30–60 minutes, depending on whether participants have achieved their short-term goals or not. cDuring phase 3 (13–24 months), participants have 12 sessions with their health educator at monthly intervals. As with phase 2, the sessions are 30–60 minutes in duration. PA, physical activity; WC, waist circumference.

Tracking system

The health educator maintains a record of each participant’s progress, tracking the average number of pedometer steps or physical activity minutes completed by the participant between sessions. Participants also receive regular newsletters, which highlight new opportunities for physical activity in Kingston, seasonal recipes and updated information on the study’s progress.

Intervention quality control

Adherence to intervention protocols is of great concern because non-adherence seriously jeopardises study outcomes, including generalisability, treatment effects and evaluating feasibility and effectiveness. We promote adherence to the intervention in a number of ways. First, the specified frequency of contacts allows the detection of non-adherence early, which identifies early the need to implement strategies to help participants resolve problems. Participants are encouraged to complete activity logs and other types of self-monitoring exercises. The health educator telephones participants after missed meetings to reschedule and/or discuss the reasons for not attending. Second, participants have a primary health educator with whom they will develop a relationship to allow for one-on-one problem solving. If necessary, to stay on track with the goal to deliver the intensive phase material by the end of month 5, then the health educator may combine the content of two sessions into one visit. If material is not completed in one session (due to time constraints) the health educator carries it forward to complete it in the next visit.


The dramatic increase in adult obesity in North America clearly indicates that current interventions designed to reduce the prevalence of overweight and obesity are inadequate. We designed a novel approach to investigate the effectiveness of physical activity combined with a healthful diet in the prevention and treatment of obesity and related comorbid conditions at the individual level. It is anticipated that the results of the PROACTIVE trial will provide the basis for changing healthcare delivery with respect to the prevention and treatment of obesity and associated health risk in adults in primary healthcare settings.


Supplementary materials


  • The extended version of this paper is published online only at

  • Funding: This study is sponsored by the Canadian Institutes of Health Research Grant OHN-63277 from the Institute of Nutrition, Diabetes and Metabolism.

  • Competing interests: Declared. RR has received honoraria from the following pharmaceutical companies as a consultant and/or lecturer: Sanofi-Aventis, Theratechnologies and receives book royalties from Human Kinetics. SNB receives book royalties from Human Kinetics.