We conducted a systematic evidence review of the benefits and harms of behavioral counseling interventions to prevent cardiovascular disease (CVD) in persons with established risk factors to assist the U.S. Preventive Services Task Force (USPSTF) in updating its previous recommendation statements.
We searched MEDLINE, PsycInfo, the Database of Abstracts of Reviews of Effects, and the Cochrane Central Register of Controlled Trials from 2001 through October 2013 to locate relevant trials for all key questions published since the previous reviews in support of prior recommendations. We supplemented our searches with reference lists from relevant existing systematic reviews, suggestions from experts, and information from Clinicaltrials.gov to identify ongoing trials.
Two investigators independently reviewed 7,218 abstracts and 553 articles against a set of a priori inclusion criteria. Investigators also independently critically appraised each study using design-specific quality criteria based on USPSTF methods. We included fair- or good-quality trials evaluating behavioral counseling interventions to promote a healthy diet, physical activity, or both in persons with CVD risk factors, including hypertension, dyslipidemia, metabolic syndrome, and impaired fasting glucose or glucose tolerance. We resolved discrepancies by consensus.
One investigator abstracted data from 74 included studies into evidence tables and a second reviewer checked these data. We conducted meta-analyses on 57 of the 71 trials that provided necessary data to estimate the effect size of counseling on intermediate health outcomes (lipids, blood pressure, weight measures, and glucose measures). We qualitatively summarized the evidence for effects on health outcomes, behavioral outcomes, and harms.
Only a subset of trials (k=16) reported measures of patient health outcomes, including CVD events (k=5) and self-reported measures of quality of life (QOL) or depression symptoms (k=11). In general, intensive interventions that combined lifestyle interventions did not reduce CVD events or mortality at up to 10 years of followup, although event rates were generally low. In one early good-quality trial, a high-intensity behavioral counseling intervention in conjunction with a protocol to start medication reduced CVD events at 6.6 years compared with usual care (relative risk [RR], 0.71 [95% CI, 0.51 to 0.99]). This study was conducted in Swedish men at high risk for CVD (which included persons with diabetes and known CVD). Overall, combined lifestyle interventions did not appear to improve self-reported depression symptoms (k=4) in persons with impaired fasting glucose or glucose tolerance at 6 to 12 months. Findings that showed a benefit on self-reported QOL measures were mixed. While three combined lifestyle counseling trials showed improvement on selected QOL measures, two combined lifestyle counseling trials and two physical activity–only counseling trials showed no benefit on self-reported QOL at 6 to 12 months. Medium- (31 to 360 minutes) to high-intensity (>360 minutes) combined lifestyle counseling in persons selected for CVD risk factors reduces total cholesterol, low-density lipoprotein (LDL) cholesterol, blood pressure, fasting glucose, diabetes incidence, and weight outcomes. Overall, at 12 to 24 months, behavioral counseling appears to reduce total cholesterol (k=34) by an average of 4.48 mg/dL (95% CI, 6.36 to 2.59), LDL cholesterol (k=25) by 3.43 mg/dL (95% CI, 5.37 to 1.49), systolic blood pressure (k=31) by 2.03 mm Hg (95% CI, 2.91 to 1.15), diastolic blood pressure (k=24) by 1.38 mm Hg (95% CI, 1.92 to 0.84), fasting glucose (k=22) by 2.08 mg/dL (95% CI, 3.29 to 0.88), diabetes incidence (k=8) by an RR of 0.58 (95% CI, 0.37 to 0.89), and weight outcomes (k=34) by a pooled mean difference of 0.26, using standardized units (95% CI, 0.35 to 0.16). There was substantial statistical heterogeneity for weight outcomes. High-intensity combined lifestyle counseling in persons with impaired fasting glucose or glucose tolerance (k=5) can reduce diabetes incidence in the long term (RR, 0.55 [95% CI, 0.45 to 0.67]). Intensive diet-only counseling interventions primarily in persons with dyslipidemia who are not yet taking medications can also modestly lower total (k=8) and LDL (k=7) cholesterol at 12 to 24 months. In contrast, medium-intensity (k=8) physical activity–only counseling interventions (k=10) did not appear to improve intermediate CVD outcomes at 12 to 24 months. Findings from trials that could not be included in quantitative analyses were generally consistent with pooled findings. Overall, objectively measured and self-reported changes in dietary intake and physical activity were concordant with intermediate outcome findings. Only three of the 61 trials that reported behavioral outcomes did not also report intermediate health outcomes. In selected trials conducted in persons who were already taking medications to lower cholesterol or blood pressure, counseling interventions appeared to improve dietary intake and physical activity despite a lack of benefit on lipid or blood pressure outcomes. Many physical activity–only counseling trials (k=9) had less than 12 months of followup. Four of five trials reporting behavioral outcomes at 12 to 24 months found statistically significant improvements in self-reported physical activity (i.e., number of persons meeting the recommended 150 minutes of moderate activity per week, minutes per week of total or moderate physical activity). We examined all included counseling trials for harms, including any paradoxical change in outcomes. While we searched for additional studies examining harms of healthy lifestyle counseling interventions, we did not find any. Overall, harms (or lack thereof) were not commonly reported (k=10). In general, included interventions did not have significant adverse effects, except for two persons who had serious events resulting from physical activity in one trial targeting older adults. Reported increases in carbohydrate intake (k=8) were accompanied by dietary improvements in fat, saturated fat, fiber, or fruits and vegetables, without an overall increase in sugar or total calories consumed.
Only a small subset of trials reported patient health outcomes, longer-term followup of intermediate and behavioral outcomes, and harms. We were unable to identify important contributors to statistical heterogeneity other than gross categorizations of type of population or intervention.
Medium- and high-intensity diet and physical activity behavioral counseling in overweight or obese persons with CVD risk factors resulted in consistent improvements across a variety of important cardiovascular intermediate health outcomes up to 2 years. High-intensity combined lifestyle counseling reduced diabetes incidence in the longer term. The applicability of these findings depends largely on the availability of intensive counseling in practice and real-world fidelity and adherence to these interventions.
