TITLE: DATE: 9 July 2014 CONTEXT AND POLICY ISSUES

Transcription

1 TITLE: Obesity Management Interventions Delivered in Primary Care for Patients with Hypertension or Cardiovascular Disease: A Review of Clinical Effectiveness DATE: 9 July 2014 CONTEXT AND POLICY ISSUES Obesity is associated with numerous co-morbidities, including cardiovascular disease (CVD), type 2 diabetes, hypertension, and certain cancers. 1 The coexistence of hypertension and obesity greatly enhances cardiovascular risk, and if left untreated can increase the risk of coronary heart disease (CHD), stroke, heart failure, kidney failure, dementia, and other chronic diseases. 2,3 Hypertension and other risk factors for CVD such as dyslipidemia, inflammatory markers, and impaired endothelial function appear to be associated with increases in adiposity and/or reductions in physical activity. 4,5 Adipose tissue is an active endocrine and immune organ whose dysfunction contributes to cardiometabolic diseases. 6 Persistent obesity has been associated with a progressive alteration of vascular tone and reactivity, neo-angiogenesis, and altered patterns of blood flow regulation. 3 It has been reported that body fat correlates with incident atrial fibrillation and that the risk of developing atrial fibrillation increases by 4 to 5% with each one-unit increment in body mass index (BMI). 7 Therefore, it seems intuitive to expect that obesity-induced disease conditions may be amenable to behavioral interventions aimed at influencing dietary and/or physical activity (PA) behavior to promote weight loss. 1,5 The first-line strategy for the treatment of obesity and prevention of cardiometabolic disease is achieving weight loss. 8 Lifestyle measures, which consist of reductions in caloric intake by between 500 to 1000 calories per day, together with increases in physical activity and changes in health behaviors, are the cornerstone of prevention and treatment of obesity. 8,9 However, maintenance of behavioral changes associated with weight loss can be challenging and shortterm improvements commonly fail to be translated into long-term behavioral maintenance. 10 Therefore, adding anti-obesity drugs in obese patients who do not achieve sufficient weight loss or who find it difficult to maintain initial weight loss after lifestyle modifications may be a good strategy for patients who can safely use them. 8,9,11,12 It has been shown in hypertension treatment trials that weight loss, whether alone or in combination with antihypertensive drugs, has a beneficial effect on blood pressure control and Disclaimer: The Rapid Response Service is an information service for those involved in planning and providing health care in Canada. Rapid responses are based on a limited literature search and are not comprehensive, systematic reviews. The intent is to provide a list of sources of the best evidence on the topic that CADTH could identify using all reasonable efforts within the time allowed. Rapid responses should be considered along with other types of information and health care considerations. The information included in this response is not intended to replace professional medical advice, nor should it be construed as a recommendation for or against the use of a particular health technology. Readers are also cautioned that a lack of good quality evidence does not necessarily mean a lack of effectiveness particularly in the case of new and emerging health technologies, for which little information can be found, but which may in future prove to be effective. While CADTH has taken care in the preparation of the report to ensure that its contents are accurate, complete and up to date, CADTH does not make any guarantee to that effect. CADTH is not liable for any loss or damages resulting from use of the information in the report. Copyright: This report contains CADTH copyright material and may contain material in which a third party owns copyright. This report may be used for the purposes of research or private study only. It may not be copied, posted on a web site, redistributed by or stored on an electronic system without the prior written permission of CADTH or applicable copyright owner. Links: This report may contain links to other information available on the websites of third parties on the Internet. CADTH does not have control over the content of such sites. Use of third party sites is governed by the owners own terms and conditions.

2 heart rate and may therefore have a role in the management of hypertension in overweight and obese patients. 3 Weight loss has also been found to ameliorate several other modifiable CVD risk factors such as improving glycemic control, reducing total cholesterol, LDL cholesterol and triglycerides, and increasing HDL cholesterol. 13 In addition, increased PA, reduces the risk for atherosclerotic disease, acute cardiovascular events, stroke, and type 2 diabetes mellitus. 5 Greater weight loss typically confers larger improvements in cardiometabolic risk factors, but even modest weight loss of 5 to 10% has been shown to produce beneficial effects. A reduction in triglycerides and an increase in HDL-cholesterol levels have been reported with 5% and 10% weight reduction. 11,13,14 Though weight loss associated with PA may be minimal compared with interventions that combine reduced caloric intake with exercises, PA has demonstrated favorable changes in body composition, and increased PA is recommended for weight loss maintenance. 14 In view of the abundance of evidence supporting health benefits associated with weight loss, The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation and Treatment of High Blood Pressure (JNC 7) 15 recommends that lifestyle modifications such as weight loss and regular aerobic exercise as part of the initial treatment strategy for lowering high blood pressure. In this respect, the Dietary Approaches to Stop Hypertension (DASH) diet, which is rich in fiber, fruits, vegetables and low-fat dairy products, and low in fat has been recommended by JNC. 16 Because weight loss interventions can be associated with reduced intake of essential nutrients as well as loss of muscle and bone mass, it is important to evaluate strategies for ability to induce desired changes without detrimental impact on vitamins and bone metabolism, and to determine the suitability of available options to targeted patients. 4 The aim of this report is to provide current evidence on the clinical effectiveness of various obesity management interventions that could be used in primary care settings to halt or slow the progression of hypertension or CVD. RESEARCH QUESTIONS 1. What is the clinical effectiveness of long-term primary care obesity management interventions in halting or slowing the progression of hypertension? 2. What is the clinical effectiveness of long-term primary care obesity management interventions in halting or slowing the progression of cardiovascular disease? KEY FINDINGS Greater weight loss correlated positively with greater reductions in blood pressure and cardiovascular risk factors regardless of which weight loss intervention was used. Dietary only interventions achieved greater initial weight loss than physical activity only interventions with the greatest weight loss occurring around 6 months. However, only lifestyle interventions that combine dietary changes and physical activity have been shown to produce significant differences in weight loss beyond 12 months. A lifestyle intervention and comprehensive metabolic risk factor management program resulted in weight loss associated with a substantial reduction in the symptom burden and symptom severity of atrial fibrillation. Motivational Interviewing (MI) was effective in bringing about long-term improvements in walking, and cholesterol levels, and was particularly effective for patients with elevated baseline levels of Obesity Interventions for Hypertension or CVD 2

3 CVD risk factors. In overweight or obese individuals at increased cardiovascular risk, Mediterranean diets appear to be more effective than low-fat diets in inducing clinically relevant long-term changes in cardiovascular risk factors and inflammatory markers, with evidence from one study demonstrating a lasting benefit of Mediterranean diets for 2 years after the beginning of the diet. METHODS Literature Search Strategy A limited literature search was conducted on key resources including Medline, PsycINFO, PubMed, The Cochrane Library (2014, Issue 5), University of York Centre for Reviews and Dissemination (CRD) databases, Canadian and major international health technology agencies, as well as a focused Internet search. Methodological filters were applied to limit retrieval to health technology assessments, systematic reviews, meta-analyses, randomized controlled trials and non-randomized studies. Where possible, retrieval was limited to the human population. The search was also limited to English language documents published between January 1, 2009 and May 27, Selection Criteria and Methods One researcher screened the citations and abstracts from the literature search; selected articles according to the selection criteria outlined in Table 1, and examined the full-text publications for the final study inclusion for this report. Table 1: Selection Criteria Population Adults with body-mass-index (BMI) of 30 or higher who have hypertension or cardiovascular disease (CVD) or both. Intervention Non-surgical interventions for weight management that can be delivered in primary care or community care, including one or more of the following interventions: pharmacological, lifestyle (diet/healthy eating, exercise), or psychological therapy, applied for a duration of six months or longer. Comparator Hypertension or CVD care that is not specialized on obesity/weight management Outcomes Hypertension status, medication burden (number of and reliance on medication), CVD-related morbidity or mortality; mental health/illness quality of life (QoL); depression severity index scores Study Designs Health Technology Assessments (HTA)/Systematic reviews/metaanalysis; Randomized controlled trials (RCTs), and Non-randomized studies. Exclusion Criteria Articles were excluded if they did not meet the selection criteria in Table 1. In particular, studies that focused on surgical interventions, and interventions targeting hypertension or CVD alone (e.g. antihypertensives) or change in BMI alone were not included. Studies were also excluded if they were published before 2009, if they did not have a comparator group, if they were duplicate publications of already selected study, or included in at least one of the selected HTAs or systematic reviews. Obesity Interventions for Hypertension or CVD 3

4 Critical Appraisal of Individual Studies Two included studies 1,17 were assessed using the SIGN-50 Methodology Checklist 1 for Systematic Reviews and Meta-analysis. 18 The other studies 7,8,10-12,19 were appraised using the SIGN-50 Methodology Checklist 2 for Controlled Trials. 20 The strengths and limitations of the individual studies have been summarized and presented in tabular form in Appendix 3. SUMMARY OF EVIDENCE Quantity of Research Available The literature search yielded 374 citations of which 28 potentially relevant studies were selected upon screening of titles and abstracts. Grey literature searching identified additional 5 papers bringing the total pool of potential articles to 33 of which 8 studies were selected for inclusion in this report. The excluded did not meet the inclusion criteria listed in Table 1. The PRISMA flow chart in Appendix 1 outlines the selection process. Summary of Study Characteristics Characteristics of included studies have been summarized in Appendix 2 Country of origin One systematic review, 1 and one meta-analysis 17 were from the United Kingdom (UK) and Switzerland, respectively; and they both included studies from several different countries. The systematic review included 44 RCTs most of which were conducted in the United States of America (USA) (n = 27). The rest of the studied were from five other countries, Canada (n = 5), UK (n = 5), Australia (n = 4), Finland (n =2), and the Netherlands (n = 1). The meta-analysis 17 included 6 RCTs, three of which were from Italy and one each from Israel, Spain, and USA. Of the six additional randomized controlled trials (RCTs) identified, four were from the USA, 8,11,12,19 and one each from Australia 7 and the UK. 10 Study setting The studies used in the systematic review 1 and meta-analysis 17 were conducted in various settings including primary care, research facilities and university hospitals. All four RCTs from the USA were multi-center studies with one of them (COR-II) 11 conducted at 36 private or institutional practices, another study (CONQUER) 12 had 93 trial sites and a third study (the Look AHEAD Trial) 19 had 16 clinical sites. Another RCTs (SEQUEL) 8 was an extension study of an earlier trial 12 and took place at 36 of the 93 sites used for the initial trial. The 36 sites were selected on the basis of their high initial enrollment numbers and retention rates, and blinding to assigned treatment was maintained through the end of the extension phase study. 8 There were no details given about any of these study sites. The studies from Australia 7 and the UK 10 were both single-center studies with the Australian study 7 drawing participants from a specialized medical centre (Centre for Heart Rhythm Disorders) at a University, while the UK study 10 was conducted in a primary-care setting. Patient population The mean number of participants in studies included in the systematic review was 240 ± 502 (range: 26 to 3,234). The participants were obese adults with mean BMI of 33.1 ± 2.2 kg/m 2 Obesity Interventions for Hypertension or CVD 4

5 (range: 30.1 to 38.8 kg/m 2 ) and a mean age of 55.0 ± 6.8 years. The meta-analysis involved 6 RCTs with a total of 2,650 adult participants with age ranging from 35 to 68 years, and mean body mass index (BMI) from 29 to 35 kg/m 2 with at least one additional cardiovascular risk factor. One of the RCTs 7 involved 150 overweight or obese participants age 21 to 75 years with BMI greater than 27 kg/m 2 who had symptomatic paroxysmal or persistent atrial fibrillation. Another study 11 included 1,496 participants 18 to 65 years of age with a BMI 30 to 45 kg/m 2, or a BMI 27 to 45 kg/m 2 and controlled hypertension and/or dyslipidemia. The COR-II study 11 included 1,496 obese (BMI 30 to 45 kg/m 2 ) or overweight (27 to 45 kg/m 2 ) participants with dyslipidemia and/or hypertension. The CONQUER study 12 enrolled participants 18 to 70 years of age who had BMI of 27 to 45 kg/m 2, and had 2 weight-related co-morbidities. Hypertension was defined as systolic blood pressure 140 and 160 mm Hg (or 130 and 160 mm Hg if diabetic), diastolic blood pressure 90 and 100 mm Hg (or 85 and 100 mm Hg if diabetic), or the use of 2 antihypertensive medications. Another study (SEQUEL) 8 which was an extension of the CONQUER study 12 had similar inclusion criteria like the latter and included 676 patients with 2 or more weight-related comorbidities. The participants were eligible to enroll in the study if they completed the CONQUER study 12 on treatment and complied with protocol requirements, were BMI 22 kg/m 2 at the completion of the CONQUER study, and if they did not stop taking the study drug continuously for 4 weeks at the completion of the CONQUER study. 12 One study 10 included 334 participants aged 18 to 65 years with BMI of 28 kg/m 2 or more, who exhibited at least one of the following CVD risk factors; hypertension (SBP/DBP at least 150/90 mmhg), or hypercholesterolemia. The Look AHEAD trial 19 enrolled 5,145 participants with type 2 diabetes, aged 45 to 76 years, who were overweight or obese (BMI > 25 kg/m 2 ). Interventions and comparators The systematic review 1 and the meta-analysis 17 involved studies that compared lifestyle interventions for weight loss in overweight/obese patients with risk factors for cardiometabolic risk factors. Lifestyle interventions included diet, physical activity (PA), and behavioral changes to achieve weight loss. The systematic review 1 classified interventions as diet only (D-only), PA only (PA-only), or diet and PA (D-PA). These interventions were either compared with each other or to usual care. Usual care was not described. Details of specific lifestyle interventions in the studies included in the systematic review 1 were not provided and the authors stated that dietary and PA behaviors were reported using a variety of different measurements. The meta-analysis specifically included studies that compared Mediterranean diets to low fat diets. Mediterranean diets were defined as diets with moderate fat intake (where the main sources of added fat were olive oil and nuts), rich in vegetables, and low in red meat (with poultry and fish replacing beef and lamb). 8 Low-fat diets were defined as diets aiming at an energy intake with 30% of calories from fat. 8 Three of the remaining studies 8,11,12 evaluated pharmacological interventions for weight loss while three others 7,10,19 investigated lifestyle interventions for weight loss. Pharmacological studies One of the pharmacological intervention studies 11 randomized 1,496 participants via a centrally administered interactive voice response system in a 2:1 ratio, stratified by study site, to receive Obesity Interventions for Hypertension or CVD 5

6 a combined oral formulation of 32 mg/ day naltrexone SR plus 360 mg/day bupropion SR (NB32), or matching placebo, administered in divided doses, twice daily for a 56 week study including up to 4 weeks over which doses were titrated to the maximum. Participants received instructions at baseline, as well as at weeks 12, 24, 36, and 48, to follow a hypocaloric diet (500 kcal/day deficit) and increase physical activity, and behavioral modification advice. If participants in the NB32 group achieved less than 5% weight loss at visits between weeks 28 and 44 inclusive, they were re-randomized 1:1 in a double-blind fashion to continue receiving NB32 or escalate to NB48 (48 mg/day naltrexone SR plus 360 mg/day bupropion SR) for the remainder of the study participants. The CONQUER trial 12 randomized a total of 2,487 participants to either of two combinations of phentermine (PHEN) and topiramate extended-release (TPM ER) or placebo. Thus patients received either PHEN 7.5 mg/tpm ER 46 mg, or PHEN 15 mg/tpm ER 92 mg, or placebo. All participants received standardized diet and lifestyle modification counseling at each study visit, including a 500 kcal/day reduction in caloric intake. 12 The third study, SEQUEL 8 was an extension of the CONQUER study with the participants continuing with the original treatment to which they were randomly assigned during CONQUER for an additional 52 weeks (a total of 108 weeks from the beginning of the CONQUER trial). Lifestyle interventions studies In one of the lifestyle intervention studies, patients were randomized to either a physicianled weight loss program (intervention group) or to self-directed general lifestyle measures (control group). Both groups underwent intensive management of cardiometabolic risk factors. Another study 10 involving 334 participants had a motivational interview (MI) as the intervention compared to what was described as minimal intervention. 10 Patients randomized to MI received standard exercise and nutrition information plus up to five 20 to 30 minutes face-toface MI sessions delivered by a physical activity specialist and registered dietician for 6 months. The MI sessions focused on exploring ambivalence and eliciting self-directed change. The counselors adopted strategies such as agenda setting, exploration of advantages and disadvantages of the weight-loss, and the importance of the program and its rules in order to build motivation in participants who were ambivalent about behavior change. For those who were sufficiently motivated to change, strategies employed included strengthening commitment to change and negotiating a change plan. 10 Patients randomized to the minimal intervention received standard information only without any MI counselling sessions. All participants in both the MI and minimal intervention groups received a standard leaflet that provided lifestyle guidelines information on exercise and nutrition. The guidelines included recommendation to be physically active for 30 minutes, at least five times a week; to consume five portions of fruit and vegetables per day, and recommended fat intake. 10 Physiological and psychological benefits of increased physical activity were also listed in the leaflet, as well as a food and physical activity quiz and advice depending upon scores. 10 In the Look AHEAD trial, 19 5,145 participants were randomized to an intensive behavioral intervention referred to as Intensive Lifestyle Intervention (ILI) or standard diabetes support education (DSE). The Look AHEAD trial was a 4-year study, with follow-up visits past Year 4 intended primarily for collection of data on cardiovascular disease (CVD) events and those measures that potentially mediate the effect of the intervention on CVD events (weight, blood pressure, lipids, and HbA1c). All participants continued on medical care provided by their Obesity Interventions for Hypertension or CVD 6

7 personal physicians, including changes in medication, except temporary adjustment to avoid hypoglycemia during periods of intensive weight loss for patients in the ILI arm. 19 Specific components of the ILI included daily caloric goal of between 1200 and 1800 calories depending on initial body weight, and a home-based exercise plan involving 175 minutes or more per week of physical activities similar in intensity to brisk walking. The caloric intake plan required less than 30% of total calories from fat (with less than 10% from saturated fat), and a minimum of 15% of total calories from protein. Participants were provided with pedometers to motivate them and they could count any moderate activity toward the exercise goal provided it was at least 10 minutes in duration and of similar intensity to brisk walking. 19 Furthermore, patients in the ILI group were seen weekly for the first 6 months and three times per month for the next 6 months, either individually or in a group. Subsequently, they were seen individually at least once a month, contacted by phone or once a month, and given the opportunity to participate in a variety of approved group classes. 19. Participants in the DSE group used standardized protocols focused on diet, physical activities and social support. Though they were offered three group sessions per year, information on behavioral strategies was not provided and patients were not weighed. 19 Outcome Measures This report focuses on outcomes related to blood pressure and lipid parameters. Another Rapid Response report discusses details of outcomes that impact diabetes status, plasma glucose, and glycemic control. 21 Outcome of interest in the systematic review 1 were change in behavior, weight, and/ or risk factors for cardiovascular and metabolic morbidity. Change in behavior was measured using altered kilocalorie and fat intake as a proxy while other outcomes were measured using standard laboratory and clinical procedures. 1 Most included studies reported outcomes at 6 and 12 months. 1 Outcomes of interest considered by the meta-analysis 17 included cardiovascular risk factors at 2 years of follow-up, in terms of mean differences in body weight, body mass index, systolic and diastolic blood pressure, total high-density lipoprotein (HDL) cholesterol, and low-density lipoprotein (LDL) cholesterol. Pharmacological studies One pharmacological intervention study 11 had two co-primary efficacy endpoints: the percent change in weight and the proportion of participants with 5% or more weight loss at week 28. The secondary endpoints of the study 11 were analyses of these measures at week 56. Additional secondary endpoints included the proportion of participants with 10% or more weight loss and changes in markers of cardiometabolic risk, participant-reported measures of food cravings and control of eating, and weight-related quality of life at week 28. Quality of Life was assessed using the Impact of Weight on Quality of Life questionnaire (IWQOL)-Lite. In the CONQUER trial, 12 the primary end points were percentage weight loss and the proportion of participants achieving 5% or more weight loss. Additional end points were changes in lipid variables and blood pressure in the dyslipidemia and hypertensive populations, respectively, stratified by treatment and magnitude of weight loss. The SEQUEL study 8 had two co-primary endpoints: mean percentage weight loss and percentage of subjects achieving 5% weight loss from baseline to week 108 (beginning with CONQUER 12 ). Secondary endpoints included weight Obesity Interventions for Hypertension or CVD 7

8 loss, percentage of patients achieving 10%, 15%, or 20% weight loss, and changes from baseline to week 108 in blood pressure, serum lipid variables, and concomitant medications for weight-related comorbidities. The 9-item Patient Health Questionnaire self-reported assessment for depressive symptoms and the 11-item, clinician-administered the C-SSRS were used to assess participants mental health and the presence of suicidal ideation or behavior. Lifestyle interventions studies In one of the lifestyle intervention studies, 7 the primary outcome was atrial fibrillation symptom burden, measured by the Atrial Fibrillation Severity Scale (AFSS). The AFSS is a validated scale that encompasses 3 domains of atrial fibrillation: event frequency (scored 1-10), duration (scored ), and global episode severity (scored 1-10). 7 The scale can range from 3.25 representing a single minimally symptomatic episode to 30 for a continuous highly symptomatic episode lasting longer 48hours. 7 In addition, the AFSS has a symptom-specific continuous subscale to assess symptom severity ranging from 0 for no symptoms to 35 for severe symptomatology. The secondary outcomes were 7-day Holter-derived atrial fibrillation episode and duration burden, echocardiographic left atrial area, and left ventricular wall thickness. 7 Another study 10 had categorized outcomes as behavioral or biomedical. Behavioral outcomes consisted of total physical activity, walking, moderate physical activity, vigorous physical activity, fat intake, fruit and vegetable intake), while biomedical outcomes comprised changes in BMI, bodyweight, SBP, DBP, cholesterol, LDL, HDL, and triglycerides. Weight, height, systolic, and diastolic blood pressure, and fasting cholesterol, were assessed using standard protocols. Fat intake was assessed using a scale from the Dietary Instrument for Nutrition Education (DINE), which is a food frequency questionnaire of 19 groups of food that account for around 70% of the fat and fibre in the typical UK diet. 10 Fruit and vegetable consumption was assessed using the five-a-day Community Evaluation Tool questionnaire (FACET) which has been shown to be correlated with consumption recorded via food diaries and is considered sufficient for analyzing group consumption patterns necessary for evaluating community interventions. 10 The major outcomes examined by the Look AHEAD study 19 were changes weight, physical fitness, and cardiovascular risk factors. Weight and height were measured in duplicate using digital scale and stadiometer. A submaximal exercise test was given at Years 1 and 4 (and for a subset of participants at Year 2) following a maximal graded exercise test at baseline. Cardiovascular risk factors were evaluated through changes in blood pressure and lipid parameters from baseline, using automated device and standardized laboratory procedures, respectively. In addition, participants health related quality of life (HRQOL) status was measured using the SF-36 instrument. Summary of Critical Appraisal Appendix 3 provides further details of the critical appraisal of individual studies. Systematic Review and Meta-Analysis The systematic review 1 and the meta-analysis 17 had clearly defined research questions relevant to the goals of this report and they were based on published randomized controlled trials. In both studies, 1,17 extensive literature searches covering multiple electronic databases were conducted. The systematic review 1 reported that in addition to electronic databases, three Obesity Interventions for Hypertension or CVD 8

9 journals were searched for relevant studies to include. The method used to search the journals was not specified. While the meta-analysis clearly stated that no language restrictions were imposed in the literature search, the systematic review did not mention language restrictions. A search of grey literature was not done in either synthesis. The systematic review 1 stated that two researches independently screened the first 200 references of RCTs and resolved differences by discussion, with one of them completing identification of the remaining studies. Of the 44 included studies, data from the initial three were extracted by three researchers (it was not described whether this was done independently or not), with one researcher extracting data from the remaining studies. These selection and data extraction practices are uncommon and it is unclear whether they could increase the risk of bias. In the meta-analysis, two researchers independently identified trials to be included, assessed their quality, and extracted data. The systematic review 1 and meta-analysis 17 used standard criteria to appraise the methodological quality of included RCTs. 8,12 However, the systematic review 1 did not assess included studies for publication bias, and although researchers in the meta-analysis reported not detecting any publication bias, they could not entirely rule it out because of the relatively small number of included studies. 17 In pooling the results from the various included studies, both the systematic review and meta-analysis acknowledged the presence of heterogeneity and reported outcomes with statistically determined levels of heterogeneity. Though researchers in the metaanalysis reported robust results across various sensitivity analyses accounting for differences in trial quality, population studied, and co-interventions, the entire study was based on only six trials, with three trials published by the same group of authors. It is unclear how this more limited perspective might impact the results. Researchers in both the systematic review and meta-analysis reported that they had no relationship with industry or financial associations that might pose a conflict of interest in their studies. 1,17 In general, the systematic review and meta-analysis were well-done and of acceptable quality. However, the applicability of their findings in primary-care settings in Canada have to be assessed based on their limitations, including the heterogeneity of pooled results. Pharmacological interventions studies All the pharmacological weight loss intervention studies 8,11,12 were randomized controlled trials. In one study, 11 randomization was done through a centrally administered interactive voice response system, while the other studies 8,12 randomly assigned participants using a computer generated algorithm. In all the studies 8,11,12 blinding was established through the use of matching placebo. One study 12 analyzed data using the intent-to-treat (ITT) population, while two others 8,11 used the modified intent-to-treat (mitt) population defined as randomized participants with a baseline and at least one post-baseline measurements of body weight while on the study drug. Baseline characteristics were generally well balanced in all the three studies. The majority of participants were women and each study had more than 84% (range: 84.6 to 86.6%) Caucasian participants. 8,11,12 The prevalence of obesity in Canada is 19.7% in Caucasians compared with 18.7% and 9.5% in people of African and South Asian descent, respectively. 22 Thus there seems to be an overrepresentation or Caucasians in the study populations, compared to what may be expected in the Canadian setting. However, it is unclear how this will affect the generalizability of the findings to a Canadian context. Obesity Interventions for Hypertension or CVD 9

10 One of the studies 11 indicated that its sample size would provide 99% power to detect a statistically significant difference in mean percent weight loss but there was no indication that it was powered to address outcomes related to hypertension or cardiovascular disease risk factors. The other studies 8,12 had no discussion on sample calculation and power determination for weight, hypertension, or cardiovascular disease. All three pharmacological intervention studies 8,11,12 measured pre-specified outcomes with reliable tools and applied rigorous statistical analysis. Though each of the trials 8,11,12 was multi-centered, they each presented a composite result for all sites and comparability of outcomes from individual sites could not be determined. One of the studies 11 reported a high dropout rate with 54% of randomized participants in each treatment group completing the trial. More intervention-treated participants discontinued because of an adverse event, whereas more placebo-treated participants discontinued because of insufficient weight loss. 11 In the CONQUER trial 12 completion rates for the study arms are not discussed clearly, though the overall discontinuations in the intervention (PHEN/TPM ER) groups were reported to be between 12% to 20% and discontinuation due to adverse events was reported to be dose-related. The SEQUEL trial, 8 which is an extension of the CONQUER trial, also reported a high retention rate (84 % in both treatment arms), which may be expected because only participants from select study sites who completed the CONQUER trial while still using the study drug were eligible, suggesting that participants with positive treatment outcomes were more likely to enroll in the extension study. 8 The last observation carried forward imputation approach was used for missing data in all the studies. 8,11,12 Since patients who are more compliant are more likely to derive greater benefit from the intervention, it is probable that LOCF is a conservative approach which underestimates the overall effect size. In two of the studies 8,12 co-morbidities like hyperglycemia, high blood pressure, and dyslipidemia were actively managed on the basis of treatment guidelines including medication use, resulting in the confounding impact of medication changes on hypertension and cardiometabolic variables. Therefore it is uncertain whether actual treatment effects may be reproducible with altered background medication/treatment options to manage these comorbidities. Lifestyle interventions All three lifestyle interventions in overweight/obese studies 7,10,19 were randomized controlled trials. Details about methods or randomization were not provided in any of the studies, so it is unclear whether appropriate measures were used or the degree of risk of selection bias. In one study 7 counsellors could not be blinded to patients allocation and it was not reported whether assessment was done by independent assessors blinded to the allocation. Another study 10 indicated that the assessors were blinded to treatment allocation. The Look AHEAD study 19 did not provide any details about blinding. In each study, the baseline characteristics of participants were well-balanced across treatment groups and all relevant outcomes were measured in a standard, valid, and reliable way. 7,10,19 One study 7 had a sample size sufficient to provide at least 80% power to meet the study s primary end point, weight loss, but there was no indication that it was powered to detect differences in secondary outcomes such as hypertension and cardiovascular events across treatment groups. Another study 10 had power determination to detect differences in treatment groups with regards to weight, and blood pressure, but it is uncertain whether it was powered to detect clinically important changes in cardiovascular disease or disease risk factors. However, the proportion of patients in the intervention group who attended the counselling sessions that differentiated intervention from control was low, with 32%, 4%, 14%, 38%, and 12% attending 0, 1, 2, 3, and 4 or 5 consultations, respectively. The investigators attributed this low and variable Obesity Interventions for Hypertension or CVD 10

11 attendance to lack of resources to motivate patients to attend or to allow for an ongoing check on counselling attendance. The Look AHEAD study 19 had a large sample size (n=5,145) with at least 80% power to detect an 18% difference in major cardiovascular disease events between its two study groups. The dropout rate in one study 7 was high but similar in the treatment groups at all time points (44% in intervention group and 48% in the control at 15 months). The most common reasons for discontinuation were co-morbidities or medical/surgical procedure pre-specified as reason to censor participants. Ablation of atrial fibrillation was the single most frequent reason why patients prematurely discontinued the study. 7 According to the investigators of one study, 7 the use of the AFSS to assess the atrial fibrillation symptom burden has the potential to underestimate the true arrhythmia burden, and periodic assessment with 7-day continuous Holter monitoring may miss episodes during the periods patients were not monitored. Another study 10 did not discuss retention/dropout rates; however 38% and 29% of randomized participants in the intervention and control group, respectively, had been lost to follow-up at 6 months. The Look AHEAD trial 19 reported that 94.1% of participants in the intervention arm (ILI) and 93.1% in the comparator arm (DSE) completed the 4-year assessment. 19 Two studies 7,10 handled missing data using last observation carried forward imputation approach. Given the high dropout rate in one study 7 and the significant proportion lost to follow-up in the other, 10 it is uncertain how the LOCF approach might have affected the reported findings. The Look AHEAD study does not provide information about how missing data were handled. Though patients characteristics in all the studies generally seem reasonably similar to what might pertain in Canada, it is not completely certain how generalizable findings from these studies will be in primary care setting in Canada in view of a number of factors. First, all the studies were conducted in foreign countries and apart from a study in the UK which took place at a primary-care setting, trials were either conducted in private research centers or specialized places which may differ considerable from primary care setting in Canada, or there was not enough information about the study settings. Secondly, some trials motivated patient participation with free replacement meals 19 or involved staff with specialties 7,10 that may not be readily available in primary care settings. Summary of Findings Rapid Response reports are organized so that the evidence for each research question is presented separately. Further details on individual study findings and authors conclusions have been provided in Appendix What is the clinical effectiveness of long-term primary care obesity management interventions in halting or slowing the progression of hypertension? This report uses SBP and DBP as surrogates of hypertension where the latter has not been specified separately as an outcome since persistently elevated blood pressure is a major symptom of hypertension. The systematic review 1 reported that compared to usual care, diet and physical activity (PA) interventions led to greater behavioral changes that promote weight loss and reduce cardiometabolic risk factors. Twenty included studies in the systematic review 1 comparing combined diet and PA interventions to waiting list control/usual showed significant Obesity Interventions for Hypertension or CVD 11

12 improvements in SBP and DBP at 6 months. Five studies in which diet only interventions were compared to waiting list control/usual care showed no significant differences in SBP and DBP between treatment groups at 6 months. Similarly, six studies which compared PA-only interventions to waiting list control / usual care could not detect a significance difference between treatment arms with respect to blood pressure. The meta-analysis 17 showed that individuals assigned to a Mediterranean diet had more favorable changes than participants on low-fat diet in weighted mean differences of systolic blood pressure ( 1.7 mm Hg; 95% confidence interval [CI], 3.3 to 0.05) and diastolic blood pressure ( 1.5 mm Hg; 95% CI, 2.1 to 0.8) after 2 years of follow-up. In one of the pharmacological intervention studies, 11 the intervention (NB32) and placebo arms demonstrated similar categorical changes in blood pressure and pulse rate, with the mean systolic and diastolic blood pressure tending to remain within approximately 1 mm Hg of baseline values in both arms throughout the study. The reduction in blood pressure was observed in patients who lost weight and the extent of blood pressure reduction correlated with how much weight was lost, irrespective of the treatment group to which the patient was assigned. In a sub-study, 24-hour systolic and diastolic blood pressure and heart rate patterns were similar between NB and placebo-treated participants at baseline, week 24, and week Another study (CONQUER) 12 demonstrated that dyslipidemia and/or hypertensive patients receiving the intervention (PHEN/TPM ER) achieved significantly more weight loss of 5%, 10%, and 15% compared with those receiving placebo. As in a previously reported study 11, greater weight loss resulted in greater improvements in blood pressure, regardless of treatment group assignment, with no significant differences between treatment groups in blood pressure reduction after adjusting for degree of weight loss. An extension study(sequel) 8 to the CONQUER trial 12 found that though diastolic and systolic blood pressure showed equal reductions in the placebo and intervention groups after 2 years of therapy; participants in the intervention showed a net decrease in concomitant antihypertensive medication use, whereas antihypertensive medications were increased in the placebo group. There was no analysis done to examine how the background antihypertensive medication use influenced the reported blood pressure levels. The most common adverse events in CONQUER and SEQUEL trials 8,12 were dry mouth, paresthesia, constipation, upper respiratory tract infection, and nasopharyngitis. One of the lifestyle intervention studies 7 showed a greater reduction in the numbers of patients with elevated blood pressure among the participants in the intervention group (from 64 to 16 patients) compared with the control group (67 to 44 patients). At 15 months, patients in the intervention group experienced a statistically significant reduction in systolic and diastolic blood pressure from baseline (136 to 133 mm Hg systolic and 82 to 80 mm Hg diastolic, respectively; P <0.001 for both). The change in the control group was relatively smaller, though statistically significant (137 to 136 mm Hg systolic, P = 0.05; and 84 to 83 mm Hg diastolic P = 0.05). The difference between the treatment arms was statistically significant (P < 0.001) but the clinical significance of the reported levels of blood pressure reduction is not clear. Furthermore, while patients in the intervention group experienced a decrease in the use of antihypertensive agents from baseline to 15 months, patients in the control group increased antihypertensive medication use in the same time period. 7 Another lifestyle intervention study 10 did not find a sustained reduction in blood pressure in either the intervention (MI) or control group though the intervention led to a significant improvements in walking. 10 Obesity Interventions for Hypertension or CVD 12

13 In the Look AHEAD trial, 19 participants in ILI had greater improvements than the control (DSE) in systolic blood pressure (SBP, 5.33 vs mmhg, P < ), diastolic blood pressure (DBP, 2.92 vs 2.48 mmhg, P < 0.012) averaged across four years of follow-up. 2. What is the clinical effectiveness of long-term primary care obesity management interventions in halting or slowing the progression of cardiovascular disease? The systematic review 1 reported that 20 of its included studies demonstrated that diet and PA produced an overall significant improvement in all risk factors with the exception of HDL cholesterol, though SBP and triglycerides showed the most consistent improvements. 1 At 6 months, diet and exercise showed significant improvements in SBP, DBP, fasting blood glucose (FBG), total cholesterol, and triglycerides. Five studies comparing diet only interventions to waiting list control/usual care showed no significant differences in SBP, DBP, and FBG between treatment arms at 6 months, though improvements were reported in total cholesterol, triglycerides, and SBP at 3 months. 1 Another six included studies in the systematic review 1 showed that changes in risk factors due to PA-only interventions were less consistent and produced no significant detectable effects when compared with a combined diet and PA, and D- only interventions. 1 The meta-analysis 17 found that after 2 years of follow-up total cholesterol and triglyceride values changed more favorably in participants randomized to Mediterranean diets than in those randomized to low-fat diets. There were no statistically significant differences in low-density lipoprotein cholesterol or HDL cholesterol. Individuals assigned to a Mediterranean diet also had more favorable changes in weighted mean differences of body weight, body mass index, systolic blood pressure, diastolic blood pressure, fasting plasma glucose, and high-sensitivity C- reactive protein. One of the pharmacological intervention studies 11 found that at week 28, patients in the intervention group experienced greater improvements than those in the placebo group in various cardiometabolic parameters, including, triglycerides, LDL, HDL as well as reduced fasting insulin. Most of these improvements were maintained at week In another study (CONQUER) 12 participants with dyslipidemia who received the interventions (PHEN/TPM ER) experienced significantly greater improvements compared with placebo in serum triglyceride, HDL, and non-hdl levels. Improvements in these lipid variables and in inflammatory biomarkers were greater with increasing degrees of weight loss with no clinically meaningful or statistically significant differences among treatment groups in lipid parameters after adjusting for degree of weight loss. A greater percentage of participants in the intervention groups demonstrated a trend toward a net reduction in using lipid-lowering medications compared with the placebo group. The SEQUEL study 8 which as an extension the CONQUER, 11 demonstrated that the greater improvements in cardiovascular and metabolic variables in participants in the intervention (PHEN/TPM ER) groups at 56 weeks were maintained at 108 weeks with treatment leading to progressively greater reductions in triglycerides and greater increases in HDL cholesterol than did placebo. LDL cholesterol decreased in all treatment arms, with the greatest reduction in the placebo group, whereas reduction in non HDL cholesterol was similar in all groups. The placebo group had a markedly greater net increase in the number of lipid lowering medications used compared with the PHEN/TPM CR groups. More participants in the intervention group had a decrease in lipid-lowering medications than those receiving placebo. Obesity Interventions for Hypertension or CVD 13

14 In one of the lifestyle intervention studies, 7 there was a decline in atrial fibrillation symptom burden and symptom severity scores in both the intervention and the control groups with the intervention group showing significantly greater decline in both domains from 6 months to 15 months (P < 0.001, both domains). The intervention group also demonstrated significantly greater decline in secondary analyses for AFSS symptom frequency, duration, and global episode severity scores compared with the control group (P < in all domains). At 15 months, patients in the intervention group experienced significantly greater reduction from baseline in the mean number of atrial fibrillation episodes (3.3 to 0.62) compared with the control group (2.8 to 2.0). Furthermore, while the mean duration of atrial fibrillation declined from baseline in the intervention group (1,176 to 491 minutes), the control group experienced increased duration of atrial fibrillation (1,393 to 1,546 minutes). The difference between the 2 groups was significant in all measures (P < 0.001). While levels of insulin and CRP decreased more in the intervention group than in the control group, serum triglycerides decreased only in the intervention group. 7 The intervention group showed a greater reduction compared with the control group in the numbers of patients with elevated blood pressure, and elevated lipid levels. More people in the intervention group reduced alcohol consumption than control group. 7 Another study 10 found that participants allocated to the intervention (MI) group exhibited significantly greater improvements in cholesterol than the comparison group. While the MI intervention demonstrated improvements in behavioral outcomes at both 6 and 18 months in the area of walking, the minimal intervention had no significant effect on walking scores over time. The study 10 showed a significant decrease from baseline in dietary fat intake at both the 6 and 18 month follow-up period for the minimal intervention group, whereas there was no difference in fat intake in the MI intervention group. 10 The Look AHEAD trial 19 showed that averaged across four years of follow-up, the ILI achieved greater improvements than the control (DSE) in blood pressure, HDL-cholesterol, and triglycerides. However, nominal reductions in LDL-C were greater in DSE than ILI, though the difference between the two groups disappeared when adjustments were made for medication use. Although the greatest benefits were often seen at 1 year, ILI participants maintained greater improvements than DSE in weight, fitness, HbA1c, SBP, and HDL-C at 4 years. 19 Limitations Limitations of included studies have been summarized in Appendix 3. According to the authors of the systematic review, 1 the quality limitation included a limited number of studies available for both short-term and long-term outcomes of interest, the use of a variety of different measurements to report outcomes, and unverifiable self-reported outcomes. Also, there was considerable heterogeneity among studies which can reduce the clinical relevance of the findings. In addition, the majority of studies did not report important quality measures such as method of randomization, reasons for participant dropout, and blinding of outcome assessors. 1 The meta-analysis 17 is based on six studies of which 3 trials were published by the same group of authors. According to the authors, although they tested and found no indication of publication bias, the relatively small number of trials affords low power of any test to detect publication bias, therefore such bias cannot be definitely ruled out. Obesity Interventions for Hypertension or CVD 14