Insufficient physical activity is a major risk factor for several common diseases, for example heart disease, type 2 diabetes, and cancer. Conversely, regular physical activity reduces the risk of these diseases. Only about a third of Norwegian adults manage to accumulate enough weekly physical activity to achieve health benefits. Cognitive therapies have documented effects in a range of disorders, e.g. depression, anxiety, insomnia, and chronic pain, but we do not know if they are useful in promoting physical activity in adults.
The term cognitive therapies is used in this report and is understood to include cognitive behavioural therapies, which is a common use in Norway.
We carried out this systematic review to answer the question “What is the effect of cognitive therapies on physical activity in adults ≥ 18 years, compared to no intervention, usual care or another intervention?”
We searched systematically in five electronic databases. In addition, we searched the reference lists of included studies. Two persons independently screened titles and abstracts, selected studies from full text publications, and assessed risk of bias in the included studies. One person extracted data from the studies and another person verified the data extraction. We summarized the results by random-effects meta-analyses and presented standardized mean differences and 95% confidence intervals. We carried out analyses to explore whether characteristics of the populations or the interventions, or the quality of the studies influenced the results. We rated our confidence in the effect estimates using GRADE (Grading of Recommendations Assessment, Development and Evaluation) and presented the results in summary of findings tables. In the GRADE system, high quality means that we are very confident that the true effect is close to that of the estimate of the effect. Moderate quality means that the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low quality means that the true effect may be substantially different from the estimate of the effect. Very low quality means that the true effect is likely to be substantially different from the estimate of effect.
We did not find any systematic reviews that could answer our question. We found 21 randomized controlled trials with 25 comparisons that fulfilled our inclusion criteria. Most of the studies included different patient groups, such as persons with cancer, type 2 diabetes, or coronary heart disease, while a third of the studies included sedentary but otherwise healthy persons. All studies but three were judged to have an unclear risk of bias, most commonly due to insufficient information on how allocation of participants to study groups was concealed.
We found a moderate effect of cognitive therapies on physical activity compared to no intervention or usual care, including receiving advice. The standardized mean change difference was 0.47 with a 95% confidence interval of 0.19 to 0.74. According to GRADE, we rated our confidence in the effect estimate as moderate. Other studies that compared cognitive therapies with no intervention or usual care reported their results as mean differences. Our confidence in the effect estimate was very low and we considered these data to be too uncertain to consider whether there was an effect or not.
We found a small to moderate effect of cognitive therapies on physical activity when added to exercise or rehabilitation programs compared to these programs alone. The standardized mean difference was 0.42 with a 95% confidence interval of 0.15 to 0.69. Our confidence in this effect estimate was moderate. The documentation is too sparse to conclude about the effect of cognitive therapies directly compared to exercise or rehabilitation.
Finally, we found that cognitive therapies might have a small effect similar to health education. The standardized mean difference was 0.14 with a 95% confidence interval of -0.14 to 0.42. Our confidence in this effect estimate was low.
In sensitivity analyses, the characteristics of the populations or interventions, or quality of the studies did not systematically influence the results across the comparisons.
The studies included in this systematic review covered a range of adult populations, including both patients and sedentary but otherwise healthy persons. The interventions included basic elements of cognitive therapies such as goal-setting, development of skills to identify and handle maladaptive thoughts and behaviours, and relapse prevention. The interventions were delivered by several different categories of health professionals. There was great variation in the duration and frequency of the intervention given. These factors, explored in sensitivity analyses, did not seem to influence the results systematically, which supports the applicability of the results.
However, applicability may be limited by the fact that most studies used strict exclusion criteria meaning that, for example, persons with co-morbidities or mental health problems were excluded from participation. Thus, the results may not capture the effectiveness of cognitive therapies, i.e. if they work well in routine clinical practice. Another possible limitation pertains to the measurement of physical activity. All methods to measure physical activity have their strengths and limitations. The most serious limitation related to the present results is the unknown ability of self-report questionnaires to detect change in physical activity from one point in time to another in a reliable way. Another limitation is that many different methods were used to measure physical activity across the studies. The consequence of this is that results that go into meta-analyses needed to be standardized, and the connection with the original measurement scales such as minutes per day, steps per day, or calories used is lost. We can only describe effects in terms of direction (to the advantage of the intervention or control group, or showing little or no difference) and magnitude (small, moderate, or large). This makes it difficult to interpret the results. The lack of direct comparison of cognitive therapies with exercise or rehabilitation programs may limit the applicability of the results. We do not know the effect on physical activity of participating in either a cognitive therapy program or an exercise or rehabilitation program. Finally, few studies had follow-up beyond six months after the end of the intervention.
Research gaps include standardization of primary outcome measurements in physical activity, preferably organised as a task for international cooperation; follow-up results for at least 12 months after the end of the intervention; direct comparison of cognitive therapies with exercise or rehabilitation programs.
Cognitive therapies probably lead to a moderate increase in physical activity when they are compared to no intervention or usual care. When cognitive therapies are added to exercise or rehabilitation programs they probably lead to a small to moderate additional increase in physical activity compared to exercise or rehabilitation only. The follow-up times are short, up to six and nine months, respectively. Cognitive therapies may have a similar small effect on physical activity as health education, up to nine months post intervention.