Transtheoretical model stages of change for dietary and physical exercise modification in weight loss management for overweight and obese adults

Nikolaos Mastellos; Laura H Gunn; Lambert M Felix; Josip Car; Azeem Majeed

doi:10.1002/14651858.CD008066.pub3

Transtheoretical model stages of change for dietary and physical exercise modification in weight loss management for overweight and obese adults

Authors' declarations of interest

Version published: 05 February 2014 Version history

https://doi.org/10.1002/14651858.CD008066.pub3

Collapse all Expand all

Abstract

available in

Background

Obesity is a global public health threat. The transtheoretical stages of change (TTM SOC) model has long been considered a useful interventional approach in lifestyle modification programmes, but its effectiveness in producing sustainable weight loss in overweight and obese individuals has been found to vary considerably.

Objectives

To assess the effectiveness of dietary intervention or physical activity interventions, or both, and other interventions based on the transtheoretical model (TTM) stages of change (SOC) to produce sustainable (one year and longer) weight loss in overweight and obese adults.

Search methods

Studies were obtained from searches of multiple electronic bibliographic databases. We searched The Cochrane Library, MEDLINE, EMBASE and PsycINFO. The date of the last search, for all databases, was 17 December 2013.

Selection criteria

Trials were included if they fulfilled the criteria of randomised controlled clinical trials (RCTs) using the TTM SOC as a model, that is a theoretical framework or guideline in designing lifestyle modification strategies, mainly dietary and physical activity interventions, versus a comparison intervention of usual care; one of the outcome measures of the study was weight loss, measured as change in weight or body mass index (BMI); participants were overweight or obese adults only; and the intervention was delivered by healthcare professionals or trained lay people at the hospital and community level, including at home.

Data collection and analysis

Two review authors independently extracted the data, assessed studies for risk of bias and evaluated overall study quality according to GRADE (Grading of Recommendations Assessment, Development and Evaluation). We resolved disagreements by discussion or consultation with a third party. A narrative, descriptive analysis was conducted for the systematic review.

Main results

A total of three studies met the inclusion criteria, allocating 2971 participants to the intervention and control groups. The total number of participants randomised to the intervention groups was 1467, whilst 1504 were randomised to the control groups. The length of intervention was 9, 12 and 24 months in the different trials. The use of TTM SOC in combination with diet or physical activity, or both, and other interventions in the included studies produced inconclusive evidence that TTM SOC interventions led to sustained weight loss (the mean difference between intervention and control groups varied from 2.1 kg to 0.2 kg at 24 months; 2971 participants; 3 trials; low quality evidence). Following application of TTM SOC there were improvements in physical activity and dietary habits, such as increased exercise duration and frequency, reduced dietary fat intake and increased fruit and vegetable consumption (very low quality evidence). Weight gain was reported as an adverse event in one of the included trials. None of the trials reported health‐related quality of life, morbidity, or economic costs as outcomes. The small number of studies and their variable methodological quality limit the applicability of the findings to clinical practice. The main limitations include inadequate reporting of outcomes and the methods for allocation, randomisation and blinding; extensive use of self‐reported measures to estimate the effects of interventions on a number of outcomes, including weight loss, dietary consumption and physical activity levels; and insufficient assessment of sustainability due to lack of post‐intervention assessments.

Authors' conclusions

The evidence to support the use of TTM SOC in weight loss interventions is limited by risk of bias and imprecision, not allowing firm conclusions to be drawn. When combined with diet or physical activity, or both, and other interventions we found very low quality evidence that it might lead to better dietary and physical activity habits. This systematic review highlights the need for well‐designed RCTs that apply the principles of the TTM SOC appropriately to produce conclusive evidence about the effect of TTM SOC lifestyle interventions on weight loss and other health outcomes.

PICOs

Population

Intervention

Comparison

Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

Plain language summary

available in

Behaviour changes for dietary and physical exercise modification in overweight and obese adults

Review question

What are the effects of dietary interventions or physical activity interventions, or both, based on the transtheoretical model (TTM) stages of change (SOC) to produce sustainable (one year and longer) weight loss in overweight and obese adults?

Background

Generally, weight loss programmes tend to involve diet and physical activity interventions. The TTM describes a series of five SOC an individual goes through when changing from an unhealthy behaviour to a healthy one. In this review, we assessed the use of the TTM SOC in weight management programmes for overweight and obese adults especially in terms of the effects on weight loss, dietary habits, physical activity and behaviour changes.

Obesity (body mass index of at least 30 kg/m²) and overweight (body mass index of 25 to less than 30 kg/m²) are increasingly being recognised as important public health issues. Together, they contribute to serious health problems and extensive economic costs worldwide. Body mass index (BMI) is a measure of body fat and is defined as the individual's weight in kilograms divided by the square of the height in metres (kg/m²). The BMI should be considered as a rough guide only because it is mainly used for whole populations and may not correspond to the same degree of fatness in different individuals (like for athletes and physically non‐active individuals).

Study characteristics

We included three studies in our systematic review. Altogether the studies evaluated 2971 participants, with 1467 participants allocated to the intervention groups and 1504 to the control groups. The studies had a length of intervention of 9, 12 and 24 months.

This plain language summary was current as of December 2013.

Key results

The use of the TTM SOC in combination with diet or physical activity, or both, and other interventions in the included studies provided inconclusive evidence about the impact of such interventions on sustainable weight loss (mean difference in favour of the TTM SOC was between 2.1 kg and 0.2 kg at 24 months). However, other positive effects were noted, such as changes in physical activity and dietary habits that included increased exercise duration and frequency, reduced fat intake and increased fruit and vegetable consumption. The studies did not report other important outcomes such as health‐related quality of life, illness (morbidity) and economic costs.

Quality of the evidence

Overall, the quality of the evidence was low or very low. The main limitations included incomplete reporting of outcomes, methodological shortcomings, extensive use of self‐reported measures and insufficient assessment of sustainability due to the lack of long‐term assessments.

Authors' conclusions

Implications for practice

The transtheoretical model (TTM) stages of change (SOC) is widely used as an intervention framework in weight management programmes across community settings, including at home. This review aimed to assess the use of the TTM SOC as a theoretical framework for dietary interventions or physical activity interventions, or both, in weight loss management for overweight and obese adults. The small number of studies and the clinical and methodological heterogeneity among the studies reduce the ability to draw firm conclusions. The included studies provide low quality evidence on the impact of TTM SOC interventions on sustainable weight loss (the mean difference between the intervention and control groups ranged from 2.1 kg to 0.2 kg at 24 months). There is very low quality evidence that TTM SOC and a combination of physical activity or diet, or both, and other interventions (such as stress management and self‐monitoring of blood glucose) can result in significant improvements in dietary (that is reducing dietary fat by 30% and increasing fruit and vegetable servings) and physical activity (that is an increase in mean self‐reported exercise (minutes per week) by around 30 minutes) habits. The review highlights the need for well‐designed randomised controlled trials, applying the principles of the TTM SOC appropriately to produce sustainable health benefits, in order to judge the effectiveness of such interventions. Nevertheless, health managers, administrators and practitioners can use evidence from this review to improve the design and evaluation of TTM SOC lifestyle interventions as well as to better plan, implement and evaluate weight management programmes. In addition, consumers can use the review to enhance their understanding of the effectiveness and limitations of TTM SOC weight loss interventions. Overall, the review may help to improve knowledge, understanding and practice in tackling the important global health challenge of obesity.

Implications for research

Only three randomised controlled trials with 9 to 24 months duration of intervention and follow‐up were included in the review. This may have affected the strength of the evidence. The review may have shown different outcomes, particularly on sustainable weight loss, if all the included trials had applied the principles of the TTM SOC appropriately and assessed weight loss sustainability in follow‐up periods (with at least one‐year intervention and one‐year follow‐up). In addition, the trials were heterogeneous, specifically in terms of interventions and outcomes. It is vital that trials report clear and detailed descriptions of their intervention(s) and the primary and secondary outcome measures to minimise the issue of heterogeneity and to enable meta‐analyses, if appropriate. Some of the trials provided inadequate information on methods of randomisation, allocation concealment and blinding, which affected the methodological quality of the studies (particularly the internal validity). Using a protocol when conducting and reporting research may reduce potential biases and enhance the quality of the study.

There is a need for well‐designed randomised controlled trials, preferably with large sample sizes and long durations of intervention and follow‐up, to evaluate the effectiveness of the TTM SOC for sustainable weight loss in overweight and obese adults. Future trials need to formulate specific and objective outcome measures, especially patient‐important outcomes such as health‐related quality of life, so that appropriate statistical analyses can be conducted to measure their independent impact on sustained weight loss. Finally, a robust systematic review of non‐randomised controlled trials to assess the effectiveness of the TTM SOC for sustainable weight loss in overweight and obese adults may be of value in the near future.

Summary of findings

Open in table viewer

Summary of findings for the main comparison.

Application of the transtheoretical model stages of change (TTM SOC) compared with usual advice on diet or exercise, or both, for overweight and obesity
Population: adults with overweight and obesity Settings: hospital and community Intervention: TTM SOC on diet or physical activity, or both Comparison: usual advice on diet or physical activity, or both
Outcomes	Mean differences (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
Weight loss [kg] a) Follow‐up: 12 months b) Follow‐up: 24 months	a) 0.7 b1) Objectively measured: ‐0.2 (‐1.0 to 0.9) b2) Self measured: ‐2.1	2971 (3)	⊕⊕⊝⊝ low^a	a) TTM SOC subgroup ‐ 1.4 kg, control ‐ 0.7 kg (difference not statistically significant) b1) P = 0.50 b2) P < 0.05
Health‐related quality of life	See comment	See comment	See comment	No study reported this outcome
Adverse events	See comment	See comment	See comment	No adverse events were reported
Physical activity, self‐reported exercise [minutes per week] Follow‐up: 24 months	32 (8 to 55)	665 (1)	⊕⊝⊝⊝ very low^b	P = 0.008 (from 6 to 24 months) in favour of TTM SOC
Dietary habits a) Diet with <30% fat [%] Follow‐up: 12 months b) Servings of fruit (vegetables) per day [n] Follow‐up: 12 months	a) 0.9 b) 0.2 (+ 0.1)	a) 665 (1)	a) ⊕⊝⊝⊝ very low^b	a) TTM SOC 35.2%, control 36.1% (P = 0.004) b) TTM SOC 1.9 (+ 2.2), control 1.7 (+ 2.1); (P = 0.016, P = 0.011)
Costs	See comment	See comment	See comment	No study evaluated this outcome
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
^aDowngraded by two levels because of imprecise results (confidence interval includes null effect and benefit or harm) and high risk of performance, detection and recall bias ^bDowngraded by three levels because of few participants and one trial only, high risk of performance, detection and recall bias and indirectness (surrogate outcome parameter)

Background

Description of the condition

In this review, overweight and obesity are defined as abnormal or excessive fat accumulation that may impair health. 'Overweight' refers to a body mass index (BMI) equal to or more than 25 to less than 30, and 'obesity' to a BMI equal to or more than 30. BMI is a common measure used in classifying overweight and obesity in adults, and conforms to the World Health Organization (WHO) standard. It is defined as the weight in kilograms divided by the square of the height in metres. It provides the most useful population‐level measure of overweight and obesity for both sexes and for all ages of adults. Nevertheless, it must be considered as a rough guide only because it may not correspond to the same degree of fatness in different individuals.

Obesity is a major global public health threat due to increasing trends in overweight and obesity among adults and children in many developed and developing countries. The WHO projected that approximately 1.5 billion adults (age 20 years and above) would be overweight in 2008, 500 million of whom would be obese; while at least 43 million children under the age of five years were overweight or obese in 2010 (World Health Organization 2012). In the United States (US), obesity is reaching alarming rates. According to the most recent (2009 to 2010) estimates of the National Health and Nutrition Examination Survey (NHANES), 78 million US adults aged 20 years and over (35.7%) and 12.5 million US children and adolescents between 2 and 19 years of age (16.9%) are obese (Ogden 2012a), while 33% and 31.8% of adults and children, respectively, are overweight (Ogden 2012b). In the United Kingdom (UK), obesity figures are slightly lower. According to data from the English Health Survey 2009 to 2010, 23% of adults and 15.5% of children are classified as obese, while approximately 38.5% of adults and 14.5% of children (aged between 2 and 15 years) are overweight (NHS Information Centre 2011).

Obesity results in significant impairment of health and longevity. Obesity also increases individuals’ risk of illness and reduces their life expectancy (London Health Observatory 2011). Overweight and obesity are major risk factors for serious chronic diseases, such as type 2 diabetes mellitus, cardiovascular disease, hypertension, stroke and some forms of cancer (World Health Organization 2012). Osteoarthritis is also more commonly seen among overweight and obese individuals. Obesity reduces quality‐adjusted life expectancy by about three years in males and six years in females (Brønnum‐Hansen 2007; Pryke 2008). In the US, adult obesity and overweight were associated with 111,909 and 33,746 excess deaths, respectively, in 2000 (Flegal 2005). In the UK, the number of deaths as a result of excess weight was estimated to be 8.7% of the total number of deaths (Banegas 2003), with severely obese individuals on average dying 11 years earlier than non‐obese people (London Health Observatory 2011). Furthermore, obesity has huge economic implications for a country from direct treatment costs and from indirect costs (such as sickness absence). For example, in England the disease burden for obesity alone was estimated at GBP 2.7 billion (approximately EUR 3.2 billion, May 2013 conversion rate) in 2007, with a projection to rise to GBP 3.9 billion (approximately EUR 4.6 billion) by 2015, while the NHS expenditure for both obese and overweight individuals was approximately GBP 4.2 billion (approximately EUR 4.9 billion) in 2007 and is expected to rise as high as GBP 6.3 billion (approximately EUR 7.4 billion) by 2015 (Foresight 2007). In the US, the medical care costs of obesity were approximately USD 147 billion (approximately EUR 113 billion, May 2013 conversion rate) in 2008 (Finkelstein 2009).

Description of the intervention

The transtheoretical model (TTM) describes the sequential behaviour change in an individual from an unhealthy behaviour to a healthy one. It is a model of intentional change predicting the possible outcomes during the adaptation process of the 'new' acquired behaviour. The TTM has proven successful as an interventional approach in smoking reduction amongst adults (Velicer 1998) but its effectiveness for producing weight reduction in obesity is unclear. Studies have shown that the TTM stages of change (SOC) can be used to plan dietary interventions for short‐term weight loss amongst overweight and obese individuals over a minimum of three months (Shaw 2006). The effectiveness of TTM dietary interventions beyond one year is inconsistent (Curry 1992; Greene 1999; Johnson 2006; Johnson 2008; Laforge 1994; Prochaska 2008a; Vallis 2003; Wee 2005). One study found that the TTM algorithm was insensitive and most individuals failed to meet the behavioural criteria of the model stages (Greene 1994), but other studies identified stages of change for uptake of low‐fat diet in adults (Auld 1997; Lamb 1996; Read 1996; Steptoe 1996).

The TTM provides a conceptual explanation of the processes that individuals go through when modifying a problem behaviour or acquiring a positive behaviour, in this case changing dietary intake or physical activity, or both, in order to achieve a sustainable weight loss. The SOC is the main construct of the TTM and illustrates the sequential progress and series of stages that individuals will progress through for a specific behaviour transformation (Velicer 1998). The series of five stages of change are pre‐contemplation, contemplation, preparation, action and maintenance; which an individual will go through in adopting a healthy behaviour or quitting the unhealthy one (as shown in Appendix 1) (Prochaska 1992; Prochaska 1997; Prochaska 2008b). The model's two main underlying assumptions are firstly that the majority of people are not ready to change their behaviour and will therefore not be helped by traditional action‐oriented prevention programs. Secondly, behavioural change is complex and may unfold in a sequence of stages. Individuals typically adapt these different processes of change according to the progress they have made towards changing their behaviour (DiClemente 1985).

Adverse effects of the intervention

The potential main adverse effects of the intervention include relapse into unhealthy behaviour; weight gain over a specific period of time; and economic costs.

How the intervention might work

The intervention might work by providing information on stage‐related strategies that can be applied to individuals' weight loss management programs. The proposed strategies are intended to change both the dietary and physical activity behaviours of participants to achieve a sustainable proportion of weight loss among overweight and obese adults. The hypothesis is that the TTM truly reflects human behaviour in the process of change (DiClemente 1985). The intervention also enables predictions on which strategies are suitable for the individuals at certain stages; therefore, weight loss strategies are targeted and tailored to meet the participants' needs.

Dietary strategies based on the TTM SOC might work by meeting individuals' needs according to the predictions of the TTM; as a result, there will be a change in the dietary habits (such as reduction in daily calories and fatty food consumption) which is repeatable (as the behaviour change takes place), leading to sustainable weight loss. Similarly, physical activity strategies tailored according to the model possibly work by increasing the level of exercise and physical activity, occurring in a continuous and sustainable manner, resulting in the targeted outcome. The significance of such an approach is that the behaviour change takes place voluntarily and is highly self‐driven, which may contribute to a sustainable desired behaviour change.

A study among overweight or obese adults (1277 participants with a BMI of 25 to 39.9) claimed that TTM‐based tailored feedback can improve healthy eating, exercise, emotional distress management, and weight of the study population. The results showed a significant increase in fruit and vegetable intake and individuals tended to progress to action and maintenance at 24 months (Johnson 2008). However, a review done on the TTM application found that it is difficult to apply the model when looking at dietary change because most studies demonstrated differences in terms of the aspect of diet being examined, as well as the staging algorithms and dietary assessment methodology (Ni Mhurchu 1997).

The TTM is a useful theoretical model in guiding interventions and predicting outcomes for dietary management among adults, as shown in some of the studies above. The studies with a rigorous design have shown statistically significant results that link stages of the TTM with the primary measured outcomes, particularly for large sample studies with longer follow‐up periods. It is potentially plausible to apply the TTM to other settings and it may be applicable in measuring other outcomes such as physical exercise modification and weight loss. The two common primary outcomes measured in dietary modification using the TTM as the guidelines are reduction in fat consumption and increase in healthy food intake (that is increase in fruit and vegetable consumption) (Di Noia 2008; Greene 1994; Johnson 2008; Laforge 1994).

Why it is important to do this review

Obesity drugs, dietary modification and physical activity are common interventions used in the management of obesity among overweight and obese individuals in primary care (or community) and clinical settings. A large systematic review (44 clinical trials) of long‐term (more than two years) weight loss studies in overweight and obese individuals (19,273 adults with a BMI of at least 25 kg/m²) from 1966 to 2003 investigated dietary and 'lifestyle', drug therapy (orlistat or sibutramine) and surgical (for example gastric bypass) methods resulting in modest weight loss, and potentially improving markers of cardiovascular risk factors (Douketis 2005). Dietary and lifestyle therapy provided less than 5 kg weight loss after two to four years, drug therapy provided 5 to 10 kg weight loss after one to two years, and surgical therapy provided 25 to 75 kg weight loss after two to four years. The review, however, reported methodological limitations in the included studies that restricted the applicability of findings to overweight and obese individuals in other settings (Douketis 2005). There are few systematic reviews and no clear evidence exists of the effectiveness of such interventions in producing sustainable weight loss beyond one year after intervention among overweight and obese individuals (Douketis 2005; Jain 2005; Nield 2007; Shaw 2006).

A large Cochrane systematic review of 41 randomised controlled trials (from the US, Netherlands, Canada, Australia and UK, with a total of 3476 participants) assessed exercise as a means of achieving weight loss and demonstrated that exercise had a positive effect on body weight in adults who were overweight or obese (Shaw 2006). Exercise alone resulted in a small weight loss compared with no treatment. However, exercise combined with diet resulted in a greater weight reduction than diet alone (mean difference (MD) ‐1.0 kg) and increasing exercise intensity increased the magnitude of weight loss (MD ‐ 1.5 kg). The major limitation of the review was the lack of long‐term trials included in the analyses (Shaw 2006).

Another Cochrane systematic review of 18 randomised controlled trials examined the effects of the type and frequency of dietary advice given to adults with type 2 diabetes mellitus (1467 participants who were either overweight or had normal weight) (Nield 2007). They reported that dietary advice plus exercise was associated with a statistically significant mean decrease in the glycosylated haemoglobin A1c (HbA1c) levels of 0.9% at six months and 0.1% at 12 months. The study found no statistically significant results in relation to weight loss. There were insufficient data for a meta‐analysis, so conclusions on the effects of low‐fat or other weight reducing diets were limited (Nield 2007).

A systematic clinical literature review found that dietary and exercise treatments for obese adults produced moderate weight loss (about 3 kg to 5 kg) compared with no treatment or usual care (Jain 2005). Meanwhile, weight loss from drugs used with diet or exercise programs also produced 3 kg to 5 kg of weight loss, but the effects did not last after the drug was stopped. The reported weight loss can be statistically significant but it may not be clinically sufficient to improve patients' health or quality of life. There was a tendency for weight regain or relapse as shown by most studies with long‐term follow‐up in the review (Jain 2005).

This review collated evidence and allowed rigorous appraisal of how and to what extent the TTM works as a theoretical and pragmatic ('real life tested') framework for lifestyle modification (with diet or physical activity, or both) resulting in sustained weight loss among the target population. The outcomes of this review are relevant for patients and practitioners trying to understand strategies and treatment regimes for overweight and obese people in the hospital and primary care (or community, including at home) settings. The findings of this review are also useful for planning and implementing obesity management programs as well as for policy makers.

This is an updated version of the original Cochrane systematic review (Tuah 2011). In this update we have carefully reviewed the studies included in the original publication, in response to feedback on their eligibility, and searched for new studies. Inconsistencies in the original review have been resolved to provide an accurate assessment of the use of the TTM SOC as a model, theoretical framework or guideline in designing lifestyle modification interventions in overweight and obese adults.

Objectives

To assess the effects of dietary interventions or physical activity interventions, or both, and other interventions based on the transtheoretical model (TTM) stages of change (SOC) to produce sustainable (one year and longer) weight loss in overweight and obese adults.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled clinical trials.

Types of participants

Adults, age 18 years and over, who were overweight or obese according to any standard parameters used by the WHO (for example body mass index (BMI), waist measurement, waist‐to‐hip ratio) and the criteria valid in the country at the time of the start of the trial. Overweight is defined as a BMI of 25 to 29.9 and obesity as a BMI of at least 30.

Participants with co‐morbidities, such as diabetes, heart diseases and hypertension, were included in the review.

Types of interventions

Intervention

Application of the transtheoretical model (TTM) stages of change (SOC) combined with lifestyle modification strategies, consisting of dietary intervention or physical activity intervention, or both, which was tailored to an individual who was overweight or obese.

The included studies had to describe the intervention as using the TTM as a model, theoretical framework or guideline in designing lifestyle modification strategies, as stated above. The intervention needed to fulfil the criteria of the TTM SOC including pre‐contemplation, contemplation, preparation, action, and maintenance (Appendix 1) as described by Prochaska and DiClemente (Prochaska 1992). The intervention must be delivered by healthcare professionals or trained lay people and targeted for overweight and obese adults at the hospital or community level, such as at community health centres, general practice clinics, community centres, schools and homes. All studies with an intervention duration from one to 12 months and above were included in the review.

Control

Usual advice on diet or advice on physical activity, or both.

Types of outcome measures

Primary outcomes

Sustained weight loss (changes in weight or BMI at one to five years and above).
Short‐term weight loss (changes in weight or BMI at less than 12 months).
Health‐related quality of life.

Secondary outcomes

The first two outcomes below comprise the main types of secondary outcomes, although additional secondary outcomes are given.

(1) Change in self‐reported or measured dietary consumption, defined as

A reduction in the daily number of calories.
A reduction in fatty food intake.
An increase in daily fruit and vegetable consumption.

(2) Change in self‐reported or measured physical activity, referring to an increase in any form of physical activity (in terms of intensity, frequency, duration and type) that was non‐prescribed or prescribed by health professionals.

Uptake or increase in physical activity.

(3) Change in other weight loss measures (skin fold measurement, waist measurement, and waist‐to‐hip ratio).

(4) Change in the SOC progression.

Adverse events

There are three main adverse events measured. These include

Relapse into unhealthy behaviour and weight gain.
Morbidity.
Economic costs.

Covariates, effect modifiers and confounders

Underlying chronic diseases such as cancer, diabetes, and respiratory disease that may cause weight loss.
Compliance.
Pharmaceutical interventions.
Bariatric surgery.

Timing of outcome measurement

At one month, three months, six months, nine months, one year and, if available, beyond one year, as stated by each trial.

Search methods for identification of studies

Electronic searches

We used the following sources, from inception until specified, for the identification of trials:

The Cochrane Library (17 December 2013).
MEDLINE (17 December 2013).
EMBASE (17 December 2013).
PsycINFO (17 December 2013).

We also searched databases of ongoing trials including the metaRegister of Controlled Trials (www.controlled‐trials.com/mrct/).

For detailed search strategies, see Appendix 2.

Additional key words that were of relevance could have been detected during any of the electronic or other searches. If this had been the case, we would have modified the electronic search strategies to incorporate these terms. Studies published in any language were included.

Searching other resources

We tried to identify additional studies by searching the reference lists of included trials and (systematic) reviews, meta‐analyses and health technology assessment reports.

Potential missing and unpublished studies were sought by contacting experts in the field. We used the library resources at Imperial and the British Library if potentially relevant articles were cited but not available via databases or websites.

Data collection and analysis

Selection of studies

To determine the studies to be assessed further, two review authors (NM, LF) independently scanned the abstract, title, or both sections of every record retrieved. All potentially relevant articles were investigated as the full text. Inter‐rater agreement for selection of potentially relevant studies was measured using the kappa statistic (Cohen 1960) and the value was 0.81, which showed that the strength of agreement between assessors was very good. Where differences in opinion existed, they were resolved by a third party (LG). If resolving disagreement was not possible, the article was added to those 'awaiting assessment' and study authors were contacted for clarification. An adapted PRISMA (preferred reporting items for systematic reviews and meta‐analyses) flow chart of the study selection is presented in Figure 1 (Liberati 2009).

Figure 1

Study flow diagram.

Data extraction and management

For studies that fulfilled the inclusion criteria, two review authors (NM, LF) independently abstracted the relevant population and intervention characteristics using standard data extraction templates (for details see Characteristics of included studies; Table 1; Appendix 3; Appendix 4; Appendix 5; Appendix 6; Appendix 7; Appendix 8) with any disagreements resolved by discussion, or if required by a third party (LG).

Open in table viewer

Table 1. Overview of study populations

Characteristic Study ID	Intervention(s) and control(s)	Sample size^a	Screened [N]	Randomised [N]	ITT [N]	Complete data or finishing study [N]	Randomised with complete data or finishing study [%]	Follow‐up^b
Johnson 2008^c	I: SOC + diet, physical activities + stress management	‐	4290	628		335	53.7	24 mo
	C: usual care	‐	4290	649		426	66.7	24 mo
			total:	1277	1277	761
Jones 2003^d	I1: PTC	‐	‐	250		‐	‐	12 mo
	I2: PTC + blood test strips			260
	C1: usual diabetes treatment			250
	C2: usual diabetes treatment + blood test strips			269
			total:	1029	1029
Logue 2005	I: TM‐CD	540 (90% power to detect a difference of 4.5 kg; α = 0.05; 20% drop‐out rate))	‐	329		266	79.2 (62.2)^e	24 mo
	C: augmented usual care		‐	336		271	82.4 (68.7)^e	24 mo
			total:	665	665	537
*Grand total*	*All interventions*			*1467*
	*All comparators*			*1504*
	*All interventions and comparators*			*2971*

"‐" denotes not reported

^aAccording to power calculation in study publication or report
^bDuration of intervention or follow‐up, or both, under randomised conditions until end of study
^cMinor mismatch between N with complete data and % as reported in figure 2 of the publication (Johnson 2008)
^dData on drop‐outs, losses to follow‐up and missing were not reported
^eValues in parentheses indicate measured weight (versus measured weight or weight abstracted from chart)

C: control; I: intervention; ITT: intention‐to‐treat; mo: months; PTC: pathways to change; SOC: stages of change; T: total; TM‐CD: transtheoretical model‐chronic disease

We sent an e‐mail to all authors of included studies to enquire whether they were willing to answer questions regarding their trials. We present the results of this e‐mail survey in Appendix 9. Thereafter, we sought relevant missing information on the trial from the primary author(s) of the article, if required.

Assessment of risk of bias in included studies

Two review authors (NM, LF) assessed each trial independently. We resolved possible disagreements by consensus, or through consultation with a third party (LG). In cases of disagreement, the rest of the authors were consulted and a judgement was made based on consensus.

We assessed the risk of bias using the Cochrane Collaboration's tool (Higgins 2011a; Higgins 2011b). We used the following criteria.

Random sequence generation (selection bias).
Allocation concealment (selection bias).
Blinding, separated for blinding of participants and personnel (performance bias) and blinding of outcome assessment (detection bias).
Incomplete outcome data (attrition bias).
Selective reporting (reporting bias).
Other bias.

We judged the risk of bias criteria as 'low risk', 'high risk' or 'unclear risk' and evaluated individual bias items as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). We present a 'Risk of bias' graph and a 'Risk of bias summary' figure (Figure 2; Figure 3).

Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

We assessed the impact of the individual bias domains on study results at the endpoint and study levels.

For performance bias (blinding of participants and personnel), detection bias (blinding of outcome assessors) and attrition bias (incomplete outcome data), we evaluated risk of bias separately for subjective and objective outcomes.

We defined the following outcomes as objective outcomes.

Sustained weight loss (measured using weight or BMI at one year and above).
Short‐term weight loss (measured using weight or BMI at less than 12 months).
Measured change in dietary consumption.
Measured uptake or change in physical activity.
Change in other weight loss measures (skin fold measurement, waist measurement and waist‐to‐hip ratio).
Relapse into unhealthy behaviour and weight gain.
Morbidity.
Economic costs.

We defined the following endpoints as subjective outcomes.

Health‐related quality of life.
Self‐reported change in dietary consumption.
Self‐reported uptake or change in physical activity.
Progression through the SOC.
Adverse events.

Measures of treatment effect

Dichotomous data were expressed as odds ratios (ORs) or risk ratios (RRs) with 95% confidence intervals (CIs). Continuous data were expressed as differences in means (MDs) with 95% CIs.

Unit of analysis issues

We took into account the level at which randomisation occurred, such as with cross‐over trials, cluster‐randomised trials and multiple observations for the same outcome.

We attempted to obtain baseline and follow‐up weight and height (or other weight measures used in the trials) from the authors if not reported. For cluster‐randomised and cross‐over trials, the focus of analysis was on the weight loss value, both absolute and relative, as defined by each study. Different units of analysis (for example OR and RR) were planned to be subjected to a sensitivity analysis.

In a cluster‐randomised trial, individuals are randomised in groups (that is the group is randomised, not the person). For example, in a TTM study the patients in one general practice may be randomised as a group to receive either the TTM or usual care. Had cluster‐randomised trials been included, we had planned to use appropriate statistical analyses of cluster‐randomised trials with the intra‐cluster correlation coefficient and design effect playing an important role in these analyses (Campbell 2004). In a cross‐over trial, individuals are randomised to a sequence of interventions and each person is his or her own control. Had there been any cross‐over trials among the included studies, we would have compared the intervention(s) and control for each patient to assess the effect of the TTM within each patient. Furthermore, we would have examined any potential sources of bias as a result of the cross‐over design (for example any carry‐over effects that could bias the results). For example, did patients begin the second period (that is the intervention if the first period consisted of the control, or vice versa) in a similar state to how they began the first period, or have the patients’ characteristics changed throughout the course of the first period? Analysis of any cross‐over trials would have followed guidelines outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). Our initial searches suggested that there would be few if any cross‐over trials in the area of the TTM and behaviour modification with respect to obesity.

Dealing with missing data

We contacted the authors to obtain relevant missing data (Appendix 9), if feasible. We carefully performed evaluation of important numerical data, as necessary, with either screened, randomised patients, intention‐to‐treat (ITT) populations, as‐treated or per protocol (PP) populations.

Assessment of heterogeneity

Since substantial clinical and methodological heterogeneity was present across the included trials, study results were not reported as pooled effect estimates from meta‐analyses. Had a meta‐analysis been appropriate, we would have identified any statistical heterogeneity by visually inspecting the forest plots and using a standard Chi² test with a significance level of α = 0.1, in view of the low power of this test. We would have specifically examined heterogeneity with the I² statistic, quantifying inconsistency across studies to assess the impact of heterogeneity on the meta‐analysis (Higgins 2002; Higgins 2003), where an I² statistic of 75% or more would have been considered a considerable level of inconsistency (Higgins 2009).

Furthermore, if a meta‐analysis had been conducted and statistical heterogeneity had been found, then we would have attempted to determine potential explanations for this heterogeneity by exploring individual study and subgroup characteristics.

Assessment of reporting biases

We had planned to use funnel plots to assess the potential existence of small study bias, in the case where we could include 10 studies or more investigating a particular outcome. Several explanations can be offered for the asymmetry of a funnel plot, including true heterogeneity of effect with respect to trial size, poor methodological design (and hence bias of small trials) and publication bias (Sterne 2011).

Data synthesis

We had planned to summarise data statistically if the data were available, sufficiently similar and of sufficient quality (Higgins 2011a). We would have performed analyses according to the statistical guidelines contained in the latest version of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a).

Subgroup analysis and investigation of heterogeneity

In order to avoid a hypothesis‐generating situation, we had planned to carry out subgroup analyses only if the primary outcome parameter demonstrated statistically significant differences between the intervention and control groups.

We had planned the following subgroup analyses and to investigate interaction.

Overweight and obese groups.
With co‐morbidities and without co‐morbidities groups.
Age groups.
Gender.

Sensitivity analysis

We had planned to perform sensitivity analyses in order to explore the influence of the following factors on effect size.

Restricting the analysis to published studies.
Restricting the analysis to take into account risk of bias, as specified in Assessment of risk of bias in included studies.
Restricting the analysis to very long or large studies to establish how much they dominated the results.
Restricting the analysis to studies using the following filters: diagnostic criteria, language of publication, source of funding (industry versus other), country.

We had also planned to test the robustness of the results by repeating the analysis using different measures of effect size (RR, OR etc.) and different statistical models (fixed‐effect and random‐effects models).

Results

Description of studies

See ‘Characteristics of included studies’ table and ‘Characteristics of excluded studies’ table.

Results of the search

The original search strategy (from inception to January 2011) identified 2557 records and the updated search (from January 2011 to December 2013) added another 2403 reports. Following initial screening of titles and abstracts, 186 potentially eligible studies (96 and 90 for the original and updated searches, respectively) were identified for full text review. A total of three studies met the inclusion criteria and were included in the updated review. See Figure 1 for the flow chart of the study selection.

Included studies

The details of the studies are described in the table ‘Characteristics of included studies’. A total of three studies were included in the review. Two trials (Johnson 2008; Logue 2005) were of parallel design with a one to one randomisation ratio and one trial (Jones 2003) was of factorial design. The transtheoretical model stages of change (TTM SOC) was used as a framework for intervention and assessment in all included studies. Dietary modification and physical activity were common interventions for weight loss. The trials were published between 2003 and 2008, and the study sample size varied from 665 to 1277 participants. Trial durations across the included studies ranged from 9 to 24 months.

Participants and setting

There were a total of 2971 participants across the three trials, 1467 of which were randomised to intervention groups and 1504 to control groups. Two trials (Johnson 2008; Logue 2005) reported the data for participants in the intervention and control groups finishing each study, whilst one trial did not provide this information (Jones 2003). The percentage of participants who completed the studies ranged from 53.7% to 79.2% in the intervention groups and from 66.7% to 82.4% in the control groups. All participants in the three included trials were analysed on the basis of intention to treat (ITT).

An overview of the populations in the included studies is shown in Table 1. The trials were conducted with the participation of overweight and obese adults only. One trial (Logue 2005) included more female participants, whereas two trials (Johnson 2008; Jones 2003) comprised more male than female participants. Two trials reported age as a mean value (Johnson 2008; Jones 2003), whereas one trial (Logue 2005) reported age as a range of values (with included participants ranging between 40 and 69 years). The included trials used a variety of weight entry criteria: two studies used BMI measures only (BMI cut‐off points and BMI range) (Johnson 2008; Jones 2003), whilst one trial used BMI with waist‐to‐hip ratio (WHR) (Logue 2005). Of the two trials that used BMI measures only, one used a BMI cut‐off point of 27 kg/m² (Jones 2003) while one trial applied a BMI range from 25 to 39.9 (Johnson 2008). Another trial (Logue 2005) used a BMI of at least 25 kg/m² alongside WHR for men and women. Overall, the studies included participants within the BMI range of 25 to 39.9.

One trial (Johnson 2008) did not report whether it included participants with co‐morbidities, while two trials included participants with one or more co‐morbidities, such as type 2 diabetes mellitus (Jones 2003) and hypercholesterolaemia (Logue 2005). Two trials included participants on a variety of medications, such as psychotropic drugs (Logue 2005) and insulin (Jones 2003), whilst one study provided no information on medications (Johnson 2008). Participants in the included trials were mostly white or Caucasian and African American. Two studies were conducted in the US (Johnson 2008; Logue 2005) and one in Canada (Jones 2003). The baseline characteristics of the included trials are stated in Appendix 5 and Appendix 6.

Interventions

The TTM SOC was used as a framework for intervention and assessment of participants' stages of change in the included studies (Johnson 2008; Jones 2003; Logue 2005) (Appendix 3).

The TTM SOC was used with physical activity or dietary modification, or both, with other interventions. One trial evaluated dietary modification (by dietary assessment and telephone counselling) plus other interventions (such as information on self‐help diabetes care and blood test strips) compared with usual treatment (blood test strips only) and showed significantly greater weight loss (measured as a direct weight reduction) for those progressing to the action stage compared to those who remained in a pre‐action stage (that is contemplation or pre‐contemplation) (Jones 2003). Another trial evaluated a combination of physical activity, diet and other interventions such as stress management strategies (by giving individualised feedback) compared to usual care and showed significant sustainable weight loss (measured as a direct weight reduction), particularly at 24 months (Johnson 2008). In addition, a trial involving assessment, advice and ‘prescription’ of dietary changes and physical activity (alongside anthropometric evaluation) combined with monetary rewards for completing each assessment, compared to augmented usual care, resulted in early weight loss (measured as direct weight reduction) at six months only, but no sustainable weight loss at 12 or 24 months (Logue 2005). The descriptions of interventions for the included trials are shown in Appendix 3.

The trials varied in length of intervention (from 9 to 24 months). One trial (Logue 2005) had an intervention duration of 24 months, another trial lasted for 12 months (Jones 2003), while the length of intervention in one study was nine months (Johnson 2008). Two trials reported final assessments but did not follow up participants after the end of the intervention (Jones 2003; Logue 2005), whilst one trial followed participants at intervals after the intervention (Johnson 2008).

All studies were community‐based and were conducted in primary care practices (Logue 2005), health centres (Jones 2003) and homes (Johnson 2008). The majority of interventions were delivered by health professionals, including weight loss advisors, dieticians and counsellors. One trial did not clearly state which personnel delivered the intervention (Johnson 2008).

Outcomes

In our systematic review, the primary outcomes measured were sustained weight loss maintenance (at one to five years and above), short‐term weight loss (at 1 to 12 months) (both measured using weight or BMI) and health‐related quality of life. One trial reported weight loss at 12 months (Jones 2003) and two trials at 24 months (Logue 2005; Johnson 2008). None of the included studies measured or reported health‐related quality of life. The secondary outcomes that were measured were self‐reported changes in calorie or energy intake or expenditure (Jones 2003; Johnson 2008; Logue 2005), fruit and vegetable consumption (Johnson 2008; Jones 2003), physical activity frequency (Johnson 2008) and duration (Logue 2005), as well as measured change in waist circumference (Logue 2005) and progression through the SOC (Johnson 2008; Jones 2003). Weight gain was reported as an adverse event in one study (Logue 2005). No other adverse intervention effects were reported. Details of the outcomes are stated in the 'Characteristics of included studies' table and 'Effects of interventions' section of this review.

Excluded studies

In total, there were 183 excluded studies in this review: 93 in the original review and 90 in the updated review. The details of those studies are shown in the 'Characteristics of excluded studies' table. The main reasons for excluding these studies were that other theoretical frameworks, such as cognitive behaviour therapy, self‐efficacy theory, social action theory and social cognitive theory, were used in designing the intervention (whether in combination with the TTM SOC or not); study participants were children or adolescents or had a normal body weight (BMI less than 25); or the study design did not meet the criteria for a randomised controlled trial. For instance, one study was excluded as it employed a non‐random method (that is alternate allocation) to allocate participants to the intervention and control groups (Vermunt 2011).

Risk of bias in included studies

The risk of bias of the included studies is described in 'Characteristics of included studies'.

All trials had some methodological weaknesses according to the criteria applied. No trial reported adequate methods for randomisation and allocation. None of the included studies reported methods for blinding participants and personnel. Two studies used an intention‐to‐treat (ITT) analysis to deal with missing objective data (Jones 2003; Logue 2005) while one study used subjective outcome measures only (ITT was used for missing weight values in a subsample of participants who provided objective data) (Johnson 2008). All studies reported adequate imputation techniques for subjective outcomes. However, they were subjected to reporting bias as they inadequately reported weight loss and other outcome data (Johnson 2008; Jones 2003; Logue 2005). The assessment for each domain is explained below and shown in the 'Risk of bias' graph (Figure 2) and 'Risk of bias' summary (Figure 3).

Allocation

Random sequence generation (selection bias)

One trial reported the method of randomisation (Logue 2005) and was therefore categorised as 'low risk' for selection bias. The other trials stated that participants were randomised and no further explanation was given (Johnson 2008; Jones 2003). These studies were graded ‘unclear' with regards to selection bias (Johnson 2008; Jones 2003).

Allocation concealment (selection bias)

One trial reported that allocation to the groups was concealed (Logue 2005) but it was unclear whether concealment was done appropriately. The other two studies did not describe allocation concealment (Johnson 2008; Jones 2003).

Blinding

Blinding (performance bias): objective and subjective outcomes

One trial reported that participants and staff were blinded while obtaining baseline measures; however, it was unclear whether they were blinded during the study (Logue 2005). The rest of the included trials did not explain whether investigators or participants, or both, were blinded during the study (Johnson 2008; Jones 2003). All included trials were considered ‘unclear’ in terms of blinding for objective outcomes. For subjective outcomes there was a high risk of performance bias in all trials.

Blinding (detection bias): objective outcomes

One study did not use any objective measures (that is weight loss was self‐reported) (Johnson 2008), therefore it was classified as ‘unclear’ with regards to this domain (Johnson 2008). The rest of the included trials did not provide adequate information on blinding during assessment (Jones 2003; Logue 2005). However, it was unlikely that the objective outcomes (that is weight loss measures) were affected by unblinded outcome assessors and therefore these studies were designated 'low risk' for detection bias (Jones 2003; Logue 2005).

Blinding (detection bias): subjective outcomes

The included studies did not explain whether outcome assessors (that is participants in this case) were blinded during assessment (Johnson 2008; Jones 2003; Logue 2005). They were thus considered ‘high risk’ in terms of detection bias.

Incomplete outcome data

Incomplete outcome data (attrition bias): objective outcomes

Two trials addressed incomplete data for objective outcomes by incorporating imputation techniques (Jones 2003; Logue 2005). One trial did not employ any objective measures and was categorised as ‘unclear risk’ in terms of attrition bias (Johnson 2008).

Incomplete outcome data (attrition bias): subjective outcomes

The included trials addressed the incomplete data for subjective outcomes by incorporating imputation techniques (Johnson 2008; Jones 2003; Logue 2005). However, one trial had high attrition rates (in total, only 54% in the intervention group and 67% in the control group finished the trial) and was characterised 'unclear risk' with regards to this domain (Johnson 2008). The other two trials were categorised ‘low risk’ in terms of attrition bias (Jones 2003; Logue 2005).

Selective reporting

Although the study protocols were not available, it appears that the included studies reported all expected outcomes (that is weight loss, diet and physical activity outcomes, SOC progression). However, the included trials did not provide complete weight loss data (that is short‐term weight loss, at least 5% weight loss, weight loss at 12 months, energy intake and expenditure) and were thus subjected to a ‘high risk’ of reporting bias (Johnson 2008; Jones 2003; Logue 2005).

Other potential sources of bias

All trials used valid measures (Johnson 2008; Jones 2003; Logue 2005). However, they also used self‐reported instruments to measure subjective outcomes (that is self‐reported changes in weight, physical activity, dietary intake and SOC), which has subjected these trials to risk for recall bias. One study performed objective measures of weight, physical activity and food intake in a subsample of participants (n = 202/1277) to try to protect against recall bias (Johnson 2008).

Effects of interventions

See: Summary of findings for the main comparison

For details on primary and secondary outcome measures see Appendix 8. All reported outcomes refer to the comparison of the application of the TTM SOC with 'usual' care.

Meta‐analysis for sustained or short‐term weight loss, as well as other outcomes, was not appropriate, primarily because of the clinical and methodological heterogeneity across the study interventions. In particular, interventions varied in content, frequency and intention (see Characteristics of included studies). There were also variations in the timing of the outcome measurement and the types of outcomes presented (dichotomous versus continuous, objective versus self‐reported) in the included trials. Last but not least, the included studies were methodologically weak with regards to reporting bias and thus meta‐analysis could be misleading, if feasible at all.

Primary outcomes

Weight loss

The application of the TTM SOC as a theoretical framework for dietary or physical activity interventions, or both, as well as combined with monetary rewards or stress management interventions, resulted in statistically significant, sustainable (one year and longer) weight loss in one trial (Johnson 2008) while one trial indicated non‐significant sustainable weight loss (Logue 2005). Another trial reported significant weight loss for participants in the intervention group in the action stage compared to those in a pre‐action stage at 12 months, but no comparison was reported between the intervention and control groups (Jones 2003). None of the trials reported short‐term weight loss results (less than 12 months), although short‐term weight change was measured in two studies (Johnson 2008; Logue 2005). All trials used a direct measure of weight (kg) as the outcome measure (Johnson 2008; Jones 2003; Logue 2005). Two studies reported objective weight loss outcomes measured at a diabetes centre (Jones 2003) and primary care physician offices using calibrated weight scales (Logue 2005) while one trial used self‐reported measures of weight, which were found to correlate well (0.99) with objective measures that were conducted in a subgroup (n = 202) of participants using beam scales (Johnson 2008).

All trials had some methodological weaknesses (see 'Risk of bias in included studies'). In the methodologically strongest trial, the TTM SOC used as a framework for an intervention including assessment, advice and ‘prescription’ of dietary changes and physical activity (alongside anthropometric evaluation) compared to usual care resulted in early (at 12 months) mean weight loss for both the intervention and control groups; however, the mean difference of ‐0.5 kg (95% CI ‐1.3 to 0.3) between the groups at 12 months was not statistically significant (Logue 2005). Furthermore, at the end of the intervention (24 months) 60% of all participants regained the lost weight and returned to their baseline weight. The mean weight change was slightly higher in the intervention group (‐0.4 kg; 95% CI ‐1.1 to 0.4) compared to the control group (‐ 0.2 kg; 95% CI ‐1.0 to 0.7) at 24 months, though this difference was statistically non‐significant (P = 0.17). The weight change between the intervention and control groups was ‐0.2 kg (95% CI ‐1.0 to 0.9; P = 0.50) at 24 months of the trial, indicating no statistically significant effect on sustainable weight loss at either 12 or 24 months (Logue 2005). Although weight change was measured at 6, 12, 18 and 24 months, short‐term (six months) results were not provided for this outcome (results for 6 and 12 months were combined, as were those for 18 and 24 months). This trial had a high risk of selective reporting and other bias.

Another trial applied the TTM SOC in combination with diet, physical activity and stress management interventions (Johnson 2008). Participants in the treatment group were categorised into three subgroups for three types of behaviour: healthy eating (that is reducing dietary fat to 30% of calories and calorie reduction of 500 calories per day), physical exercise (that is at least 30 minutes of moderate exercise per day, five days per week or more) and addressing emotional stress (that is using healthy strategies rather than eating to cope with the reduction in food intake). The study showed a statistically significant sustained self‐reported weight loss in the treatment group among those who were in a pre‐action stage for both healthy eating and exercise (n = 617) compared to the control group (mean difference (MD) ‐2.1 kg; P < 0.05) at 24 months (Johnson 2008). For those in the healthy eating behaviour group, a weight loss of at least 5% of one’s body weight was more frequent amongst participants in the treatment group (27.4%) compared to those in the control group (20.3%) with a statistically significant overall effect at 24 months (OR 1.22 (95% CI 1.01 to 1.48); P < 0.05). Similarly, for the exercise behaviour intervention, a weight loss of 5% or more was more frequent in the treatment group (28.8%) than in the control group (19.4%) with a trend toward significantly increasing differences over time (OR 1.32 (95% CI 0.99 to 1.75); P = 0.05). In the intervention that combined healthy eating and exercise behaviours, a weight loss of 5% or more was more frequent amongst participants in the intervention group (30%) compared to those in the control group (18.6%) at 24 months. The overall intervention effect appeared to increase over time (OR 1.35 (95% CI 1.01 to 1.81); P = 0.05). Although weight loss was self‐reported, the investigators collected objective weight measures from a subgroup of 202 participants, which were found to correlate well (0.99) with self‐reported data. In addition, despite the fact that participants in both groups completed assessments at baseline, 6, 12 and 24 months, short‐term (six months) or minimally sustainable (12 months) weight loss results were not presented in the report as the scope of the report focused on sustained weight loss only (Johnson 2008). This trial was associated with a high risk of performance and detection bias for this outcome and also had a high risk of selective reporting and other bias.

Finally, a trial used the TTM SOC with diet and blood testing strip interventions to assess participants' readiness to change their behaviour with regards to self‐monitoring of blood glucose (SMBG), healthy eating and smoking cessation (Jones 2003). Participants received one or more of the above interventions (SMBG, healthy eating, smoking cessation). The study reported greater sustained weight reduction amongst participants in the healthy eating group who progressed to an action stage (that is individuals who had changed their behaviour) than participants who remained in a pre‐action stage (that is individuals in the pre‐contemplation, contemplation or preparation stages) for the intervention group at 12 months (1.4 kg versus 0.7 kg), but this did not reach statistical significance (Jones 2003). There was a statistically significant weight loss amongst participants progressing to an action stage compared to those remaining in a pre‐action stage for the intervention group (1.8 kg versus 0.3 kg; P < 0.01) in both the SMBG and healthy eating groups at 12 months. Assessments were obtained at 3, 6, 9 and 12 months for the intervention groups and at baseline and 12 months for participants in the control groups. However, the data for the outcomes measured, such as short‐term weight loss at three, six and nine months in the healthy eating and both healthy eating and SMBG combined intervention groups, were not reported. In addition, no comparison between the intervention and control groups was reported (Jones 2003). This trial had a high risk of selective reporting and other bias.

Health‐related quality of life

Health‐related quality of life was one of the primary outcomes identified in this review but was not reported in any of the included trials.

Secondary outcomes

Self‐reported change in dietary consumption and measured change in dietary consumption

The TTM SOC combined with diet or physical activity, or both, and other interventions often resulted in statistically significant self‐reported changes in dietary consumption (measured as change in daily energy or calorie intake or change in daily fruit and vegetable intake, or both) as reported in the included trials (Johnson 2008; Jones 2003; Logue 2005). Dietary consumption was measured using validated Food Frequency Questionnaires (FFQ). All studies employed self‐administered retrospective data collection methods to measure the effect of TTM SOC interventions on dietary consumption and thus the results need to be interpreted with caution due to the possibility of recall, performance and detection bias.

Change in daily calorie intake

A trial using the TTM SOC combined with diet and blood testing strip interventions reported significantly lower calorie intake from fat (that is consuming a diet with < 30% fat) amongst participants in the intervention group compared to the control group (35.2% versus 36.1%; P = 0.004) for healthy eating at 12 months (Jones 2003).

In addition, in another trial the TTM SOC application in combination with diet, physical activity and stress management interventions showed a statistically significant overall change (OR 1.61 (95% CI 1.33 to 1.94); P < 0.001) in the number of participants who progressed to the action or maintenance stages in the intervention group compared to the control group for healthy eating behaviour at 6 (43.9% versus 31.3%), 12 (43.1% versus 35.2%) and 24 months (47.5% versus 34.3%) (Johnson 2008). Healthy eating was defined as reducing dietary fat to 30% of calories as well as a reduction of 500 calories per day. The term 'progress to action or maintenance stage' refers to an individual's readiness to engage in a healthy behaviour (Johnson 2008). Dietary calorie intake data for the intervention and control groups at 6 and 12 months were not explicitly reported. In addition, no overall comparison between the TTM SOC and control groups was reported (Johnson 2008).

Finally, a trial that combined the TTM SOC with diet, physical activity and monetary reward interventions reported no statistically significant mean change in energy intake per day in the intervention group compared to the control (P = 0.69) at 24 months (Logue 2005). There was a statistically significant reduction in the mean energy intake per day for both groups combined (‐ 250 kcal/day; P < 0.0001) throughout the 6 to 24 months follow‐up. The mean energy intake values for the intervention and control groups were not provided (Logue 2005).

Change in fruit and vegetable consumption

Two trials reported statistically significant changes in fruit and vegetable consumption at 6, 12 and 24 months. In one trial, when the TTM SOC was applied with a diet intervention a resulting significant (P = 0.016) between‐groups change in servings of fruits per day was observed at 12 months, with an average of 1.9 servings for the intervention group compared to 1.7 servings for the control group (Jones 2003). There was also a significant (P = 0.011) increase in vegetable servings in the intervention group (+2.2 servings per day) compared to the control group (+2.1 servings per day) at 12 months (Jones 2003).

In addition, a trial that used the TTM SOC in combination with diet, physical activity and stress management interventions showed a statistically significant overall change (OR 1.63 (95% CI 1.34 to 1.97); P < 0.0001) in fruit and vegetable consumption for participants progressing to the action and maintenance stage in the intervention group compared to those in the control group at 6 (44.0% versus 31.4%), 12 (45.3% versus 39.6%) and 24 months (48.5% versus 39.0%) (Johnson 2008). Short‐term (up to 12 months) fruit and vegetable consumption data for the intervention and control groups were not explicitly reported. Again, no overall comparison between the intervention and control groups was provided (Johnson 2008).

Meta‐analysis for this outcome was not appropriate mainly due to the variability in the types of outcomes used (dichotomous versus continuous data).

Self‐reported uptake of physical activity and measured change in physical activity

There were two trials (Johnson 2008; Logue 2005) reporting uptake of physical activity using the TTM SOC in combination with diet, physical activity and stress management interventions at 24 months. The outcomes reported were changes in physical activity frequency (that is per cent increased uptake per week) (Johnson 2008) as well as changes in exercise duration (that is minutes per week) and energy expenditure (kcal/kg per day) (Logue 2005). Both studies used validated questionnaires (Johnson 2008; Logue 2005). However, the use of self‐reported retrospective data subjected these studies to risk for recall bias, performance and detection bias.

Change in physical activity frequency

In one trial, the TTM SOC in combination with diet, physical activity and stress management interventions showed an increase in exercise habits (that is 30 minutes per day, 5 to 7 days per week) amongst participants progressing to the action and maintenance stage in the intervention group compared to the control group at 6 (43.0% versus 34.6%), 12 (37.7% versus 35.9%) and 24 months (44.9% versus 38.1%) (Johnson 2008). There was a significant group effect among those in the pre‐action stages for exercise at baseline, beginning at six months, which was maintained at all time points (OR 1.27 (95% CI 1.03 to 1.57); P < 0.05) (Johnson 2008). No overall comparison between the TTM SOC and usual care groups was provided (Johnson 2008).

Change in physical activity duration

The TTM SOC combined with diet, physical activity and monetary reward interventions resulted in a statistically significant increase in the mean self‐reported exercise (minutes per week) in the intervention group versus the control group (P = 0.008) from 6 to 24 months; the mean difference between the intervention and the usual care groups over the total duration of the study (between 6 and 24 months) was 31.5 minutes (95% CI 7.98 to 55.02) (Logue 2005).

Change in energy expenditure

Energy expenditure was measured using weekly physical activity recalls in one study (Logue 2005). The findings showed a significant increase in mean energy expenditure per day (+2 kcal/kg per day; P = 0.04) for both groups combined (Logue 2005). However, the difference in energy expenditure for the intervention group versus the control group was not statistically significant (P = 0.31) at 24 months (Logue 2005). The data on energy expenditure at 6, 12 and 18 months were not explicitly reported. Moreover, the mean energy expenditure values for the intervention and control groups were not provided (Logue 2005).

Change in weight loss measures

Change in other weight loss measures (for example skin fold measurement, waist measurement and waist‐to‐hip ratio) was reported in one study (Logue 2005). In this study, the TTM SOC in combination with diet, physical activity and monetary rewards interventions showed no significant mean waist girth change in the intervention group compared to the control group (P = 0.57); however, the effects for both groups combined showed a significant decrease in mean waist girth (1.7 cm (95% CI 0.9 to 2.5 cm); P = 0.0001) at 24 months (Logue 2005).

Progression through the SOC

Two trials reported progression through the SOC as an outcome (Johnson 2008; Jones 2003). The term 'progress to action stage' refers to individuals who have changed behaviour within the last six months, whereas 'maintenance stage' refers to individuals who have maintained the behaviour change for at least six months. One trial using the TTM SOC in combination with diet and blood testing strip interventions reported that more participants in the SMBG intervention groups (43.4% for 'pathway to change' plus strips and 30.5% for 'pathway to change' only) progressed to the action or maintenance stages in comparison to the control group (27.0% for treatment as usual plus strips and 18.4% for treatment as usual only) at 12 months (P < 0.001) (Jones 2003). Similarly, there was a greater proportion of participants who moved to the action or maintenance stages in the intervention (32.5%) versus control (25.8%) groups for healthy eating behaviour (P < 0.001). Some of the information for the outcomes measured in the intervention and control groups was not complete, specifically sample size and proportions of no events (Jones 2003).

In addition, in another trial the TTM SOC used in combination with diet, physical activity and stress management interventions showed that a greater proportion of participants progressed to action or maintenance stages (individuals' readiness to engage in healthy behaviour) in the intervention group than the control group for healthy eating behaviour at 6 (43.9% versus 31.3%), 12 (43.1% versus 35.2%) and 24 months (47.5% versus 34.3%) (Johnson 2008). The overall group effect for all time points was statistically significant (OR 1.61 (95% CI 1.33 to 1.94); P < 0.001). With regards to exercise outcomes, participants in the intervention group were more likely to move to action or maintenance stages compared to the control group at 6 (43.0% versus 34.6%), 12 (37.7% versus 35.9%) and 24 months (44.9% versus 38.1%). There was a significant group effect at six months that was maintained at all time points (OR 1.27 (95% CI 1.03 to 1.57); P < 0.05). In addition, in comparison with the control group, more participants in the intervention group moved to action or maintenance stages in the fruit and vegetable outcome at 6 (44.0% versus 31.4%), 12 (45.3% versus 39.6%) and 24 months (48.5% versus 39.0%). Based on the overall group effect, the association between the intervention and outcome was statistically significant at all time points (OR 1.63 (95% CI 1.34 to 1.97); P < 0.0001). Data on this outcome measure were not adequately reported for the intervention and control groups (specifically, values for no event and sample size) (Johnson 2008).

Adverse events

Morbidity

Morbidity as an adverse event outcome was not reported by the included trials.

Weight gain

There was significant weight gain for both the intervention and control groups combined after 12 months (from 12 to 24 months) in one trial (P < 0.0001), but there were no other data reported on the given outcome (Logue 2005). Most participants lost weight during the first 6 to 12 months but then relapsed.

Costs

None of the included trials reported any kind of economic cost as an outcome.

Reporting bias

In this review, it was not possible to assess reporting bias using funnel plots because there were only three trials included, the types of outcomes varied and the estimated effect measures used in each trial differed.

Discussion

Summary of main results

Three trials were identified that met the criteria for the review and were relatively recent (published in the last 10 years). The trials were heterogeneous, particularly in terms of interventions and outcomes and had different sample sizes (from 665 to 1277 participants), with 2971 participants evaluated in total. They were mostly conducted in primary care settings (apart from one intervention which was delivered at home), were mainly delivered by health professionals, and had short to medium term follow‐up (24 months or less). Weight in kilograms was the primary body weight measure used in the trials. Waist girth was also used in one trial. The TTM SOC was used as a framework for intervention to assess participants' stage of change. All included trials were performed on an intention‐to‐treat (ITT) basis.

When looking at the available evidence on sustained weight loss one study using self‐reported measures reported a statistically significant difference of 2.1 kg (P < 0.05) in favour of the intervention group (Johnson 2008), while the other study using objective measures found a non‐significant difference of 0.2 kg (P = 0.17) when comparing the intervention to the control group (Logue 2005). However, due to the small number of studies and their variable methodological quality (especially with regards to selective reporting), it is challenging to draw solid conclusions about the effectiveness of dietary, physical activity and other TTM SOC‐based interventions in producing sustainable (one year and longer) weight loss in overweight and obese adults. Specifically, sustainability of weight loss was not sufficiently assessed in the included studies. Only one study measured weight loss at 12 months follow‐up (Johnson 2008), while the other two studies did not include a post‐intervention assessment to measure weight loss beyond the end of the intervention (Jones 2003; Logue 2005). In addition, selective reporting was observed with some weight loss and other outcome data not adequately reported. One study provided 12‐month weight loss data for the intervention groups only (Jones 2003); another trial just reported post‐intervention weight loss data (Johnson 2008); while only one trial provided results at both 12 and 24 months (Logue 2005). Heterogeneity in the measures employed to estimate weight loss was also evident. One study used self‐reported weight loss measures with objective measures applied to a subsample of participants (Johnson 2008), while the rest of the included trials used objective measures only (Jones 2003; Logue 2005). The varying levels of weight loss in combination with the limited number and the heterogeneity of studies do not allow firm conclusions about sustainable weight loss.

The review also shows that the TTM SOC combined with physical activity or diet, or both, and other interventions (for example stress management) can produce statistically significant effects on other outcome measures, particularly changes in dietary consumption and physical activity levels. One study reported significant treatment effects for calorie intake (47.5% versus 34.3%; P < 0.001), fruit and vegetable consumption (48.5% versus 39.0%; P < 0.001) and physical activity frequency (44.9% versus 38.1%; P < 0.05) in the intervention group progressing to action and maintenance stages compared to control at 24 months (Johnson 2008); another study found significant between‐group differences for calorie intake from fat (35.2% versus 36.1%; P = 0.004), fruit consumption (1.9 versus 1.7 servings; P = 0.016) and vegetable consumption (2.2 versus 2.1 servings; P = 0.011) at 12 months (Jones 2003); while one trial found non‐significant reductions in energy intake (P = 0.69) and expenditure (P = 0.31) but significant differences in exercise duration (31.5 minutes; 95% CI 7.98 to 55.02; P = 0.008) between the intervention and control groups at 24 months (Logue 2005). Although the small number of studies and the use of self‐reported measures in these studies did not permit us to collate conclusive evidence about the impact of TTM SOC interventions on dietary and physical activity changes, the available evidence demonstrates some significant improvements in dietary and exercise habits among those receiving a TTM SOC intervention, which is worth noting.

Overall completeness and applicability of evidence

Relevance of the evidence

The TTM SOC is a useful theoretical and pragmatic intervention framework for some aspects of lifestyle modification in overweight and obese individuals. This review aimed to demonstrate the effectiveness of TTM SOC diet or physical activity, or both, and other interventions in producing sustained weight loss. The included studies do not provide strong evidence to judge whether such interventions can lead to sustainable weight loss. The small number of studies and the clinical and methodological heterogeneity among the studies reduce the likelihood of drawing conclusive evidence. Nevertheless, there are some early signs of positive effects of TTM SOC lifestyle modification interventions on diet and physical activity, with participants in the intervention groups significantly reducing calorie intake from fat while increasing fruit and vegetable consumption as well as exercise levels at 12 and 24 months.

External validity

All trials included male and female adult participants from diverse backgrounds and were conducted in community settings. From this aspect, the findings of the review are generalizable to overweight and obese adults who are undergoing lifestyle modification programmes for weight loss, specifically programmes which are delivered in community settings, including at home. However, the small number of studies, the methodological weaknesses among those studies (for example use of self‐reported outcome measures) and the limited contextual heterogeneity (studies were conducted in Canada and the US) hinder the replicability and applicability of findings in other settings. In addition, one trial recruited participants from a selected population, namely people with diabetes, which might affect its generalizability to other settings (Jones 2003). Another trial included obese participants with a number of co‐morbidities, such as hypertension, hypercholesterolaemia, arthritis and diabetes (Logue 2005). The two trials that contained participants with co‐morbidities (Jones 2003; Logue 2005) did not present analyses by co‐morbidity subgroups; therefore, it is unclear whether there would be any significant differences in outcomes with co‐morbidities versus without co‐morbidities.

Relevance to review’s objectives

The included studies provide insufficient information to examine the effectiveness of dietary or physical activity interventions, or both, in some instances also combined with other interventions, based on the TTM SOC for weight loss in overweight and obese adults. Two studies reported between‐group differences in weight loss at 24 months and their results varied (Johnson 2008; Logue 2005). Another study provided 12‐month weight loss data for the intervention group only (Jones 2003). The review may benefit from studies with at least 12 months duration of intervention and 12 months follow‐up to assess the sustainability of weight loss, particularly at 24 months and beyond. The relevant points in the inclusion criteria were investigated and presented in the results, including additional and adverse outcomes, a summary of outcomes and potential bias.

Relevance to current practice

Obesity is one of the world’s fastest growing health threats, and commissioning and developing obesity management programmes is a priority for policy makers, clinicians and administrators in health systems across the world. This review can be used to improve the design and evaluation of TTM SOC lifestyle interventions by informing those involved in such programmes about the current limitations in the planning, implementation and evaluation of TTM SOC lifestyle modification programmes and the need for well‐designed interventions that apply the principles of the model appropriately to produce sustainable health benefits. It also informs practitioners on existing evidence and expected outcomes (such as weight loss, change in physical activity and dietary intake) when using the TTM SOC with weight management programmes. Finally, it can also serve to inform and enhance patients' understanding of the effectiveness and limitations of TTM SOC weight loss programmes.

The TTM SOC is a promising model of behaviour change. It can lead to improvements in dietary and physical activity habits when combined with diet, physical activity and other interventions. However, there is weak evidence on the impact of TTM SOC interventions on weight loss.

Quality of the evidence

Three randomised controlled trials (including a total of 2971 participants) were evaluated in this systematic review following the use of a set of inclusion and exclusion criteria. Meta‐analysis was not appropriate because there were different types of outcomes presented in the trials (dichotomous versus continuous) and the data for the intervention and control groups for each outcome were not completely reported in the published reports. There were also variations in the intervention content, frequency and duration, as well as timing of outcome measurement, in the included trials. The summary of findings Table for the main comparison provides an assessment of the quality of evidence for weight loss (low quality), physical activity (very low quality evidence) and dietary habits (very low quality evidence). Although the table includes six outcomes, we have been unable to undertake an assessment of quality of life, adverse events and cost data due to inadequate information.

One of the key methodological limitations in the included trials was the selective reporting of weight loss data, hindering assessment of the effectiveness of TTM SOC lifestyle interventions in producing short and long‐term weight loss. Selective data disclosure was also observed in the reporting of diet and physical activity outcomes. We considered imprecision to affect the quality of evidence across all the outcomes. Two of the three interventions did not include a post‐intervention assessment and therefore it was challenging to assess long‐term weight loss sustainability, prompting us to downgrade the outcome for indirectness. There was also inadequate reporting of information on methods of randomisation, allocation concealment and blinding in most trials, to the extent that most studies were categorised as ‘unclear’ in terms of bias. Other potential sources of bias were also identified (for example recall bias due to self‐reported data gathering). The trials were performed on an intention‐to‐treat basis but the aforementioned bias issues affected the internal validity of the results, leading us to downgrade the evidence due to risk of bias across the outcomes reported in the summary of findings Table for the main comparison.

Potential biases in the review process

The ways in which potential biases in the review process were minimised include well‐defined inclusion and exclusion criteria; independent data extraction by two assessors; and use of the Cochrane risk of bias tool (Higgins 2009). Though not necessarily a limitation, conclusions on sustainable weight loss beyond two years cannot be made since only a small number of studies met the inclusion criteria and these studies measured weight loss up to two years. It was not possible to assess reporting bias using funnel plots primarily because the types of outcomes and the estimated effect measures used in each trial were different. Furthermore, due to the heterogeneity across the studies and the lack of ability to combine the trials into a meta‐analysis, it is challenging to make a firm judgement about sustainable weight loss up to two years when one study reports significant effects, one finds no significant sustainable weight loss, and another trial does not report an overall between‐group comparison.

Agreements and disagreements with other studies or reviews

The included studies do not provide conclusive evidence about the effectiveness of the TTM SOC in combination with physical activity or diet, or both, and other interventions in producing sustainable weight loss (the mean difference in the included studies was between 2.1 kg and 0.2 kg at 24 months). Several reviews support this finding, although they did not look specifically at the TTM SOC as a theoretical framework, and emphasize the need for well‐designed clinical trials to assess the effectiveness of such interventions. A large systematic review of dietary and lifestyle therapy interventions showed a mean weight loss of 3.5 (SD 2.4) kg after two to three years amongst overweight and obese individuals, which was increased to 3.6 (SD 2.6) kg after four to seven years (Douketis 2005). However, the authors concluded that the methodological limitations in the included studies hinder the applicability of the findings to obese participants assessed in everyday clinical practice and that additional research is needed to assess the effectiveness and clinical significance of such interventions. Similarly, a systematic clinical review found moderate weight loss (about 3 to 5 kg) for dietary and exercise interventions amongst obese adults compared with usual care (Jain 2005). Again, the review author pointed out potential methodological flaws in the included studies (for example unclear randomisation, lack of blinding, high attrition rates) that can bias the results and the author highlighted the need for high quality clinical trials that fulfil the requirements of evidence‐based medicine. Another large systematic review argued that exercise combined with diet resulted in a greater weight reduction than diet alone or physical activity alone (MD ‐1.0 kg) (Shaw 2006). However, the authors recognised that no conclusive evidence can be drawn from the included trials as their duration varied from 3 to 12 months, hindering evaluation of weight loss sustainability.

In addition, this review provides early evidence that the TTM SOC combined with diet or physical activity, or both, and other interventions can improve exercise and dietary habits, in particular reduced fatty food intake and increased fruit and vegetable consumption, which were sustainable over relatively long periods (12 to 24 months). This finding contrasts with the results from an earlier TTM SOC and diet review which reported mixed evidence on dietary change amongst overweight and obese adults. However, the authors of the review argued that most included studies differed in terms of the aspect of diet being examined, staging algorithms and dietary assessment methodology used. Therefore, there were significant variations in results which made it difficult to interpret the results of the studies (Ni Mhurchu 1997). The studies included in the review were not specifically randomised controlled trials and the limited number of trials at that time, combined with the use of a less robust review methodology, may have affected the results.

Figure 1

Study flow diagram.

Navigate to figure in ReviewOpen in new tab

Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Navigate to figure in ReviewOpen in new tab

Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Navigate to figure in ReviewOpen in new tab

Application of the transtheoretical model stages of change (TTM SOC) compared with usual advice on diet or exercise, or both, for overweight and obesity
Population: adults with overweight and obesity Settings: hospital and community Intervention: TTM SOC on diet or physical activity, or both Comparison: usual advice on diet or physical activity, or both
Outcomes	Mean differences (95% CI)	No of participants (studies)	Quality of the evidence (GRADE)	Comments
Weight loss [kg] a) Follow‐up: 12 months b) Follow‐up: 24 months	a) 0.7 b1) Objectively measured: ‐0.2 (‐1.0 to 0.9) b2) Self measured: ‐2.1	2971 (3)	⊕⊕⊝⊝ low^a	a) TTM SOC subgroup ‐ 1.4 kg, control ‐ 0.7 kg (difference not statistically significant) b1) P = 0.50 b2) P < 0.05
Health‐related quality of life	See comment	See comment	See comment	No study reported this outcome
Adverse events	See comment	See comment	See comment	No adverse events were reported
Physical activity, self‐reported exercise [minutes per week] Follow‐up: 24 months	32 (8 to 55)	665 (1)	⊕⊝⊝⊝ very low^b	P = 0.008 (from 6 to 24 months) in favour of TTM SOC
Dietary habits a) Diet with <30% fat [%] Follow‐up: 12 months b) Servings of fruit (vegetables) per day [n] Follow‐up: 12 months	a) 0.9 b) 0.2 (+ 0.1)	a) 665 (1)	a) ⊕⊝⊝⊝ very low^b	a) TTM SOC 35.2%, control 36.1% (P = 0.004) b) TTM SOC 1.9 (+ 2.2), control 1.7 (+ 2.1); (P = 0.016, P = 0.011)
Costs	See comment	See comment	See comment	No study evaluated this outcome
The basis for the assumed risk* (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.
^aDowngraded by two levels because of imprecise results (confidence interval includes null effect and benefit or harm) and high risk of performance, detection and recall bias ^bDowngraded by three levels because of few participants and one trial only, high risk of performance, detection and recall bias and indirectness (surrogate outcome parameter)

Navigate to table in Review

Table 1. Overview of study populations

Characteristic Study ID	Intervention(s) and control(s)	Sample size^a	Screened [N]	Randomised [N]	ITT [N]	Complete data or finishing study [N]	Randomised with complete data or finishing study [%]	Follow‐up^b
Johnson 2008^c	I: SOC + diet, physical activities + stress management	‐	4290	628		335	53.7	24 mo
	C: usual care	‐	4290	649		426	66.7	24 mo
			total:	1277	1277	761
Jones 2003^d	I1: PTC	‐	‐	250		‐	‐	12 mo
	I2: PTC + blood test strips			260
	C1: usual diabetes treatment			250
	C2: usual diabetes treatment + blood test strips			269
			total:	1029	1029
Logue 2005	I: TM‐CD	540 (90% power to detect a difference of 4.5 kg; α = 0.05; 20% drop‐out rate))	‐	329		266	79.2 (62.2)^e	24 mo
	C: augmented usual care		‐	336		271	82.4 (68.7)^e	24 mo
			total:	665	665	537
*Grand total*	*All interventions*			*1467*
	*All comparators*			*1504*
	*All interventions and comparators*			*2971*
"‐" denotes not reported ^aAccording to power calculation in study publication or report ^bDuration of intervention or follow‐up, or both, under randomised conditions until end of study ^cMinor mismatch between N with complete data and % as reported in figure 2 of the publication (Johnson 2008) ^dData on drop‐outs, losses to follow‐up and missing were not reported ^eValues in parentheses indicate measured weight (versus measured weight or weight abstracted from chart) C: control; I: intervention; ITT: intention‐to‐treat; mo: months; PTC: pathways to change; SOC: stages of change; T: total; TM‐CD: transtheoretical model‐chronic disease

Table 1. Overview of study populations

Navigate to table in Review

Cochrane Review language

Website language

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

PICOs

PICOs

Population

Intervention

Comparison

Outcome

Plain language summary

Behaviour changes for dietary and physical exercise modification in overweight and obese adults

Visual summary

Authors' conclusions

Implications for practice

Implications for research

Summary of findings

Background

Description of the condition

Description of the intervention

Adverse effects of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Intervention

Control

Types of outcome measures

Primary outcomes

Secondary outcomes

Adverse events

Covariates, effect modifiers and confounders

Timing of outcome measurement

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

Results

Description of studies

Results of the search

Included studies

Participants and setting

Interventions

Outcomes

Excluded studies

Risk of bias in included studies

Allocation

Random sequence generation (selection bias)

Allocation concealment (selection bias)

Blinding

Blinding (performance bias): objective and subjective outcomes

Blinding (detection bias): objective outcomes

Blinding (detection bias): subjective outcomes

Incomplete outcome data

Incomplete outcome data (attrition bias): objective outcomes

Incomplete outcome data (attrition bias): subjective outcomes

Selective reporting

Other potential sources of bias

Effects of interventions