Diet, physical activity, and behavioural interventions for the treatment of overweight or obesity in preschool children up to the age of 6 years
Abstract
Background
Child overweight and obesity has increased globally, and can be associated with short‐ and long‐term health consequences.
Objectives
To assess the effects of diet, physical activity, and behavioural interventions for the treatment of overweight or obesity in preschool children up to the age of 6 years.
Search methods
We performed a systematic literature search in the databases Cochrane Library, MEDLINE, EMBASE, PsycINFO, CINAHL, and LILACS, as well as in the trial registers ClinicalTrials.gov and ICTRP Search Portal. We also checked references of identified trials and systematic reviews. We applied no language restrictions. The date of the last search was March 2015 for all databases.
Selection criteria
We selected randomised controlled trials (RCTs) of diet, physical activity, and behavioural interventions for treating overweight or obesity in preschool children aged 0 to 6 years.
Data collection and analysis
Two review authors independently assessed risk of bias, evaluated the overall quality of the evidence using the GRADE instrument, and extracted data following the Cochrane Handbook for Systematic Reviews of Interventions. We contacted trial authors for additional information.
Main results
We included 7 RCTs with a total of 923 participants: 529 randomised to an intervention and 394 to a comparator. The number of participants per trial ranged from 18 to 475. Six trials were parallel RCTs, and one was a cluster RCT. Two trials were three‐arm trials, each comparing two interventions with a control group. The interventions and comparators in the trials varied. We categorised the comparisons into two groups: multicomponent interventions and dietary interventions. The overall quality of the evidence was low or very low, and six trials had a high risk of bias on individual 'Risk of bias' criteria. The children in the included trials were followed up for between six months and three years.
In trials comparing a multicomponent intervention with usual care, enhanced usual care, or information control, we found a greater reduction in body mass index (BMI) z score in the intervention groups at the end of the intervention (6 to 12 months): mean difference (MD) ‐0.3 units (95% confidence interval (CI) ‐0.4 to ‐0.2); P < 0.00001; 210 participants; 4 trials; low‐quality evidence, at 12 to 18 months' follow‐up: MD ‐0.4 units (95% CI ‐0.6 to ‐0.2); P = 0.0001; 202 participants; 4 trials; low‐quality evidence, and at 2 years' follow‐up: MD ‐0.3 units (95% CI ‐0.4 to ‐0.1); 96 participants; 1 trial; low‐quality evidence.
One trial stated that no adverse events were reported; the other trials did not report on adverse events. Three trials reported health‐related quality of life and found improvements in some, but not all, aspects. Other outcomes, such as behaviour change and parent‐child relationship, were inconsistently measured.
One three‐arm trial of very low‐quality evidence comparing two types of diet with control found that both the dairy‐rich diet (BMI z score change MD ‐0.1 units (95% CI ‐0.11 to ‐0.09); P < 0.0001; 59 participants) and energy‐restricted diet (BMI z score change MD ‐0.1 units (95% CI ‐0.11 to ‐0.09); P < 0.0001; 57 participants) resulted in greater reduction in BMI than the comparator at the end of the intervention period, but only the dairy‐rich diet maintained this at 36 months' follow‐up (BMI z score change in MD ‐0.7 units (95% CI ‐0.71 to ‐0.69); P < 0.0001; 52 participants). The energy‐restricted diet had a worse BMI outcome than control at this follow‐up (BMI z score change MD 0.1 units (95% CI 0.09 to 0.11); P < 0.0001; 47 participants). There was no substantial difference in mean daily energy expenditure between groups. Health‐related quality of life, adverse effects, participant views, and parenting were not measured.
No trial reported on all‐cause mortality, morbidity, or socioeconomic effects.
All results should be interpreted cautiously due to their low quality and heterogeneous interventions and comparators.
Authors' conclusions
Muticomponent interventions appear to be an effective treatment option for overweight or obese preschool children up to the age of 6 years. However, the current evidence is limited, and most trials had a high risk of bias. Most trials did not measure adverse events. We have identified four ongoing trials that we will include in future updates of this review.
The role of dietary interventions is more equivocal, with one trial suggesting that dairy interventions may be effective in the longer term, but not energy‐restricted diets. This trial also had a high risk of bias.
PICOs
Plain language summary
Diet, physical activity, and behavioural interventions for the treatment of overweight or obesity in preschool children up to the age of 6 years
Review question
How effective are diet, physical activity, and behavioural interventions in reducing the weight of overweight and obese preschool children?
Background
Across the world more children are becoming overweight and obese. These children are more likely to suffer from health problems, both while as children and in later life. More information is needed about what works best for treating this problem.
Study characteristics
We found 7 randomised controlled trials (clinical studies where people are randomly put into one of two or more treatment groups) comparing diet, physical activity, and behavioural (where habits are changed or improved) treatments (interventions) to a variety of control groups (who did not receive treatment) delivered to 923 overweight or obese preschool children up to the age of 6 years. We grouped the studies by the type of intervention. Our systematic review reported on the effects of multicomponent interventions and dietary interventions compared with no intervention, 'usual care', enhanced usual care, or some other therapy if it was also delivered in the intervention arm. The children in the included studies were monitored (called follow‐up) for between six months and three years.
Key results
Most studies reported the body mass index (BMI) z score: BMI is a measure of body fat and is calculated by dividing weight (in kilograms) by the square of the body height measured in metres (kg/m²). In children, BMI is often measured in a way that takes into account sex, weight, and height as children grow older (BMI z score). We summarised the results of 4 trials in 202 children reporting the BMI z score, which on average was 0.4 units lower in the multicomponent intervention groups compared with the control groups. Lower units indicate more weight loss. For example, a 5‐year‐old girl with a body height of 110 cm and a body weight of 32 kg has a BMI of 26.4 and a BMI z score of 2.99. If this girl loses 2 kg weight within a year (and gained 1 cm in height), she would have reduced her BMI z score by approx. 0.4 units (her BMI would be 24.3 and her BMI z score 2.58). Accordingly, the average change in weight in the multicomponent interventions was 2.8 kg lower than in the control groups. Other effects of the interventions, such as improvements in health‐related quality of life or evaluation of the parent‐child relationship, were less clear, and most studies did not measure adverse events. No study investigated death from any cause, morbidity, or socioeconomic effects. One study found that BMI z score reduction was greater at the end of both dairy‐rich and energy‐restricted dietary interventions compared with a healthy lifestyle education only. However, only the dairy‐rich diet continued to show benefits two to three years later, whereas the energy‐restricted diet group had a greater increase in BMI z score than the control group.
This evidence is up to date as of March 2015.
Quality of the evidence
The overall quality of the evidence was low or very low, mainly because there were just a few studies per outcome measurement or the number of the included children was small. In addition, many children left the studies before they had finished.
Authors' conclusions
Summary of findings
| Diet, physical activity, and behavioural interventions for the treatment of overweight or obesity in preschool children aged 0 to 6 years | ||||||
| Population: preschool children (aged 0 to 6 years) with overweight or obesity Settings: various Intervention: multicomponent interventions Comparison: usual care/enhanced usual care/information control/wait‐list control | ||||||
| Outcomes | Control | Multicomponent intervention | Relative effect | No of participants | Quality of the evidence | Comments |
| Changes in BMI and body weight a. BMI z scorea [units] Follow‐up: 12 to 18 months b. Weight [kg] Follow‐up: 12 to 18 months | a. The mean change in BMI z score ranged across control groups from ‐0.3 units to +0.4 units b. The mean change in weight ranged across control groups from +3.1 kg to +5.2 kg | b. The mean change in BMI z score in the intervention groups was 0.4 units lower (0.6 to 0.2 lower) b. The mean change in weight in the intervention group was 2.8 kg lower (4.4 to 1.2 lower) | ‐ | a. 202 (4) b. 202 (4) | a. ⊕⊕⊝⊝ b. ⊕⊕⊝⊝ | Lower units indicate more weight loss |
| Adverse events Follow‐up: 24 months | See comment | See comment | See comment | 88 (1) | ⊕⊝⊝⊝ | Only 1 trial (abstract only) reported on adverse events, stating no adverse events were observed |
| HrQoL and self esteem a. DUX 25 (Dutch Child AZL TNO Quality‐of‐Life tool: total score and 4 domains; scale 0 to 100; higher score indicates better HrQoL) Follow‐up: 12 months b. CHQ‐PF50 (Dutch edition of the Child Health Questionnaire Parent Form: 15 items; score 0 to 100; higher score indicates better HrQoL) Follow‐up: 12 months c. PedsQL (Pediatric Quality of Life Inventory, physical functioning subscale; higher score indicates better HrQoL) Follow‐up: 6 months/12 months d. PedsQL (total score) Follow‐up: 12 months | See comment | See comment | See comment | a. 40 (1) b. 40 (1) c. 17 (1) d. 16 (1) | a/b/c/d | No trials reported self esteem a. Change in median of the total score: +5 in the intervention group versus ‐5 in the control group; change in median of 1 of 4 domains (physical functioning): +8 in the intervention group versus ‐4 in the control group b. No statistically significant differences in any of the 15 items c. 6 months' change in mean: +9.5 units in the intervention group versus ‐1.7 units in the control group, data not reported for total score and 3 other subscales; 12 months' change in mean +13.8 units in the intervention group versus ‐2.7 units in the control group, data not reported for total score and 3 other subscales d. No substantial differences between multicomponent intervention and control group |
| All‐cause mortality | See comment | See comment | See comment | See comment | See comment | No trials reported all‐cause mortality |
| Morbidity | See comment | See comment | See comment | See comment | See comment | No trials reported morbidity |
| Parent‐child relationship or assessment of parenting | See comment | See comment | See comment | 44 (2) | ⊕⊝⊝⊝ | Limited data were reported, no substantial differences between intervention and control groups |
| Socioeconomic effects | See comment | See comment | See comment | See comment | See comment | No trials reported socioeconomic effects |
| BMI: body mass index; CI: confidence interval; HrQoL : health‐related quality of life | ||||||
| GRADE Working Group grades of evidence | ||||||
| a"A BMI z score or standard deviation score indicates how many units (of the standard deviation) a child's BMI is above or below the average BMI value for their age group and sex. For instance, a z score of 1.5 indicates that a child is 1.5 standard deviations above the average value, and a z score of ‐1.5 indicates a child is 1.5 standard deviations below the average value" (NOO NHS 2011). | ||||||
| Diet, physical activity, and behavioural interventions for the treatment of overweight or obesity in preschool children aged 0 to 6 years | ||||||
| Patient or population: preschool children (aged 0 to 6 years) with overweight or obesity Settings: obesity research clinic Intervention: dietary interventions + healthy lifestyle education Comparison: healthy lifestyle education | ||||||
| Outcomes | Healthy lifestyle education | Dietary intervention + healthy lifestyle education | Relative effect | No of participants | Quality of the evidence | Comments |
| Changes in BMI and body weight 1. Dairy‐rich diet a. BMI z score [units]a Follow‐up: 6 months b. BMI z score [units] Follow‐up: 36 months 2. Energy‐restricted diet a. BMI z score [units] Follow‐up: 6 months b. BMI z score [kg/m²] Follow‐up: 36 months | 1. Dairy‐rich diet a. The mean change in BMI z score was ‐0.5 units in the control group b. The mean change in BMI z score was +0.6 units in the control group 2. Energy‐restricted diet a. The mean change in BMI z score was ‐0.5 units in the control group b. The mean change in BMI z score was +0.6 units in the control group | 1. Dairy‐rich diet a. The mean change in BMI z score in the intervention group was b. The mean change in BMI z score in the intervention group was 2. Energy‐restricted diet a. The mean change in BMI z score in the intervention group was b. The mean change in BMI z score in the intervention group was | ‐ | 1. Dairy‐rich diet a. 59 (1) b. 52 (1) 2. Energy‐restricted diet a. 57 (1) b. 47 (1) | 1. Dairy‐rich diet a/b ⊕⊝⊝⊝ 2. Energy‐restricted diet a/b ⊕⊝⊝⊝ | Lower units indicate more weight loss 2 dietary interventions and 1 control compared in one 3‐arm randomised controlled trial (the number of participants in the control group was halved for the analysis and is shown here) |
| Adverse events | See comment | See comment | See comment | See comment | See comment | Not reported |
| Health‐related quality of life and self esteem | See comment | See comment | See comment | See comment | See comment | Not reported |
| All‐cause mortality | See comment | See comment | See comment | See comment | See comment | Not reported |
| Morbidity | See comment | See comment | See comment | See comment | See comment | Not reported |
| Parent‐child relationship or assessment of parenting | See comment | See comment | See comment | See comment | See comment | No trials reported parent‐child relationship or assessment of parenting |
| Socioeconomic effects | See comment | See comment | See comment | See comment | See comment | Not reported |
| *The basis for the assumed risk (e.g. the median control group risk across trials) 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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a"A BMI z score or standard deviation score indicates how many units (of the standard deviation) a child's BMI is above or below the average BMI value for their age group and sex. For instance, a z score of 1.5 indicates that a child is 1.5 standard deviations above the average value, and a z score of ‐1.5 indicates a child is 1.5 standard deviations below the average value" (NOO NHS 2011). bDowngraded by three levels because of reporting bias, indirectness, and imprecision (one trial only with small number of participants); see Appendix 9. | ||||||
Background
The prevalence of overweight and obese children and adolescents has increased throughout the world, presenting a global public health crisis (Ng 2014; WHO 2015a). Although once considered to be a condition affecting only high‐income countries, rates of paediatric overweight and obesity have recently started to rise dramatically in some low‐ and middle‐income countries (Wang 2012). Using the International Obesity Task Force (IOTF) standard definition, the age‐standardised prevalence of overweight and obesity in children and adolescents has increased in low‐, middle‐, and high‐income countries over the last 30 years (Cole 2000). In 2013, the prevalence of overweight and obese children and adolescents in high‐income countries was estimated at 23.8% (95% confidence interval (CI) 22.9 to 24.7) for boys and 22.6% (95% CI 21.7 to 23.6) for girls. In low‐ and middle‐income countries, the prevalence was estimated as 12.9% (95% CI 12.3 to 13.5) of boys and 13.4% (95% CI 13 to 13.9) of girls (Ng 2014). Very young children are also affected. In 2010, de Onis 2010 used the World Health Organization growth standards to estimate that over 42 million children under 5 years of age were overweight or obese, with approximately 35 million of these children living in low‐ and middle‐income countries (WHO 2015b).
Inequalities in overweight and obesity prevalence have also been documented. Generally, socioeconomically disadvantaged children in high‐income countries (Knai 2012; NCB 2015; Shrewsbury 2008), and children of higher socioeconomic status in low‐ and middle‐income countries (Lobstein 2004; Wang 2012), are at greater risk of becoming overweight. However, this relationship may vary by population demographics (for example age, gender, ethnicity), and environment (for example country, urbanisation) (Wang 2012). The prevalence of obesity has been shown to vary by ethnicity, with large data sets showing substantial ethnic variation in English (HSCIC 2015), American (Freedman 2006; Skinner 2014), and New Zealand (Rajput 2014) child populations.
Whilst there is some evidence that the rate of increase in paediatric obesity may be slowing in some high‐income countries, current levels remain too high, and continue to rise in many low‐ and middle‐income countries (Olds 2011; Rokholm 2010). However, an additional concern in some high‐income countries such as the USA, in Kelly 2013 and Skinner 2014, and England, in CMO 2015 and Ells 2015, is the rise in severe paediatric obesity. Whilst the IOTF published an international definition for severe paediatric (morbid) obesity in 2012 (Cole 2012), often severe obesity prevalence is reported using country‐specific cut points, making international comparisons difficult. However, data from the USA, in Skinner 2014, and England, in Ells 2015, has shown that the prevalence of severe paediatric obesity varies by socioeconomic status and ethnicity, and may result in a greater risk of adverse cardiometabolic events and severe obesity in adulthood (Kelly 2013).
Description of the condition
Childhood overweight and obesity results from an accumulation of excess body fat, and can increase the risk of both short‐ and longer‐term health consequences. Numerous obesity‐related comorbidities can develop during childhood, which include muscular skeletal complaints (Paulis 2014); cardiovascular risk factors such as hypertension, insulin resistance, and hyperlipidaemia (Reilly 2003), even in very young children (Bocca 2013); motor and developmental delays (Cataldo 2015); and conditions such as such as sleep apnoea (Narang 2012), asthma (Egan 2013), liver disease, and type 2 diabetes (Daniels 2009; Lobstein 2004). The condition can also affect psychosocial well‐being, with obese young people susceptible to reduced self esteem and quality of life (Griffiths 2010), as well as stigmatisation (Puhl 2007; Tang‐Peronard 2008). Evidence also shows that childhood obesity can track into adulthood (Parsons 1999; Singh 2008; Whitaker 1997), and is therefore associated with an increased risk of ill health later in life (Reilly 2011).
Description of the intervention
Given the serious implications associated with childhood and adolescent obesity, effective treatment is imperative. Whilst the fundamental principles of weight management in children and adolescents are the same as in adults (that is reduced energy intake and increased energy expenditure), the primary aim of treatment (that is weight reduction or deceleration of weight gain) and the most suitable intervention approach vary, and are dependent on the child’s age and degree of excess weight, among other considerations. Family‐based interventions combining dietary, physical activity, and behavioural components have been shown to be effective and are considered the current best practice in the treatment of childhood obesity in children under 12 years of age (Oude Luttikhuis 2009).
Adverse effects of the intervention
It is not anticipated that diet, physical activity, and behavioural interventions will lead to adverse outcomes. However, as with all obesity treatment interventions in children and young people, potential adverse effects should be considered, including effects on linear growth, eating disorders and psychological well‐being.
How the intervention might work
The home environment is important in the aetiology of childhood obesity, with parents playing a large role in food choice and physical activity for their children. In surveys in the USA, Wansink estimated that the ‘nutritional gatekeeper’ (who buys and cooks the food) controls 72% of the food eaten by children, both within and outside the home (Wansink 2006). A systematic review by Clark et al. showed that a high level of parental restriction of snack foods is associated with increased energy intake and weight gain in children (Clark 2007). In contrast, ‘covert’ control of children’s food intake by controlling the home eating environment to limit exposure to unhealthy foods (that is not buying unhealthy foods) is shown to lower the intake of unhealthy snacks when compared with ‘overt’ control (that is buying the snacks but not allowing access) (Ogden 2006). In terms of physical activity, a systematic review showed that parental support is strongly associated with physical activity levels in children, albeit the influence of parental modelling by being physically active themselves was inconsistent (Gustafson 2006).
Poor family functioning, such as poor communication and high levels of conflict, is also associated with higher risk of obesity in children (Halliday 2014). Authoritative parenting style is associated with lower risk of obesity in children, when compared with other parenting styles (Sleddens 2011). Due to the importance of the role of parents in the home environment and the importance of parenting styles and skills, parents have been defined as the ‘agents of change’ in the family for intervening with children under 12 years of age who are obese (Golan 2004). In addition, young children themselves are receptive to early and fact‐based health education (Baxter 2015). Qualitative research suggests that interventions for preschool‐aged children should aim to promote parental modelling of positive behaviours, create home and preschool environments that promote healthy diets, and simultaneously target factors at the family and preschool/childcare levels (Paes 2015).
Why it is important to do this review
The first version of this systematic review was published in 2003 and included analysis of childhood obesity treatment trials published up until July 2001 (Summerbell 2003). The second version was published in 2009, updating the 2003 review (Oude Luttikhuis 2009).
To reflect the rapid growth in this field, the third update to this review has been split across six reviews focusing on the following treatment approaches: surgery; drugs; parent‐only interventions; diet, physical activity, and behavioural interventions for young children aged 0 to 6 years; schoolchildren aged 6 to 11 years; and adolescents aged 12 to 17 years.
The current review examines the effectiveness of interventions for preschool children aged up to 6 years. Previous systematic reviews identified an absence of randomised controlled trials assessing interventions for preschool‐aged children (Bluford 2007; Bond 2009; Bond 2011; Oude Luttikhuis 2009), however a number of trials have since been published.
The results of this current review and other systematic reviews in this series will provide information on which to underpin clinical guidelines and health policy on the treatment of childhood obesity.
Objectives
To assess the effects of diet, physical activity, and behavioural interventions for the treatment of overweight or obesity in preschool children up to the age of 6 years.
Methods
Criteria for considering studies for this review
Types of studies
We included randomised controlled clinical trials with at least six months of follow‐up.
Types of participants
We included trials of overweight or obese children with a mean trial age of 0 to 6 years at the commencement of the intervention. We excluded the critically ill, or children with a syndromic cause for their obesity (for example Prader‐Willi).
Types of interventions
We planned to investigate the following comparisons of intervention versus control/comparator.
Intervention
Any form of lifestyle intervention with a primary aim to treat overweight or obesity in children (any form of dietary, physical activity and/or behavioural therapy delivered as single‐ or multicomponent interventions).
Comparator
The comparison could be no intervention, usual care (however defined), or an alternative concomitant therapy providing it is delivered in the intervention arm.
Concomitant interventions had to be the same in the intervention and comparator groups to establish fair comparisons.
Types of outcome measures
Primary outcomes
-
Changes in body mass index (BMI) and body weight.
-
Adverse events.
Secondary outcomes
-
Health‐related quality of life and self esteem.
-
All‐cause mortality.
-
Morbidity.
-
Anthropometric measures other than BMI.
-
Behaviour change.
-
Participant views of the intervention.
-
Parent‐child relationship or assessment of parenting.
-
Socioeconomic effects by validated measures.
Method and timing of outcome measurement
-
Changes in BMI (kg/m²) and body weight (kg): measured at baseline and at least at six months.
-
Adverse events: defined as an adverse outcome that occurred during or after the intervention but is not necessarily caused by it, and measured at baseline and at least at six months.
-
Health‐related quality of life: evaluated by a validated instrument such as Pediatric Quality of Life Inventory and measured at baseline and at least six months.
-
All‐cause mortality: defined as any death that occurred during or after the intervention and measured at six months or later.
-
Morbidity: defined as illness or harm associated with the intervention and measured at baseline and six months or later.
-
Anthropometric measures other than change in BMI: defined by the use of validated tools such as waist circumference, skin fold thickness, waist‐to‐hip ratio, dual X‐ray absorptiometry, or bioelectrical impedance analysis and measured at baseline and at least at six months.
-
Behaviour change: defined as validated measures of diet and physical activity and measured at baseline and at least at six months.
-
Participant views of the intervention: defined as documented accounts from participant feedback and measured at baseline and at least at six months.
-
Parent‐child relationship or assessment of parenting: evaluated by a validated instrument and measured at baseline and at least at six months.
-
Socioeconomic effects: defined as a validated measure of socioeconomic status such as parental income or educational status and measured at baseline and at least at six months.
Summary of findings
We present a 'Summary of findings' table in which we reported the following outcomes, listed according to priority.
-
Changes in BMI and body weight.
-
Adverse events.
-
Health‐related quality of life.
-
All‐cause mortality.
-
Morbidity.
-
Parent‐child relationship or assessment of parenting.
-
Socioeconomic effects.
Search methods for identification of studies
Electronic searches
We searched the following sources from inception of each database to 10 March 2015 and placed no restrictions on the language of publication.
-
Cochrane Library
-
Cochrane Database of Systematic Reviews (Issue 3, 2015)
-
Cochrane Central Register of Controlled Trials (CENTRAL) (Issue 2, 2015)
-
Database of Abstracts of Reviews of Effects (DARE) (Issue 1, 2014)
-
Health Technology Assessment (HTA) Database (Issue 1, 2014)
-
-
Ovid MEDLINE(R) In‐Process & Other Non‐Indexed Citations and Ovid MEDLINE(R) <1946 to Present>
-
EMBASE <1974 to 2015 March 09>
-
PsycINFO <1806 to March Week 1 2015>
-
CINAHL
-
LILACS
-
World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (http://apps.who.int/trialsearch/), including:
-
Australian New Zealand Clinical Trials Registry (2 March 2015)
-
Chinese Clinical Trial Registry (2 March 2015)
-
EU Clinical Trials Register (EU‐CTR) (2 March 2015)
-
ISRCTN (2 March 2015)
-
The Netherlands National Trial Register (2 March 2015)
-
Brazilian Clinical Trials Registry (2 February 2015)
-
Clinical Trials Registry ‐ India (2 March 2015)
-
Clinical Research Information Service ‐ Republic of Korea (3 March 2015)
-
Cuban Public Registry of Clinical Trials (3 March 2015)
-
German Clinical Trials Register (3 March 2015)
-
Iranian Registry of Clinical Trials (3 March 2015)
-
Japan Primary Registries Network (3 March 2015)
-
Pan African Clinical Trial Registry (3 March 2015)
-
Sri Lanka Clinical Trials Registry (2 March 2015)
-
Thai Clinical Trials Register (3 March 2015)
-
We continuously applied a MEDLINE (via Ovid SP) email alert service established by the Cochrane Metabolic and Endocrine Disorders (CMED) Group to identify newly published trials using the same search strategy as described for MEDLINE (for details on search strategies, see Appendix 1). If we identified new trials for inclusion, we evaluated these, incorporated the findings into our review, and resubmitted another review draft (Beller 2013).
If we detected additional relevant key words during any of the electronic or other searches, we modified the electronic search strategies to incorporate these terms and documented the changes.
Searching other resources
We attempted to identify other potentially eligible trials or ancillary publications by searching the reference lists of retrieved included trials, (systematic) reviews, meta‐analyses, and health technology assessment reports. We also contacted authors of included trials to identify any further trials that we may have missed.
Data collection and analysis
Selection of studies
Two review authors (two of KR, JC, EL, COM, LA, LAl‐K, EM, LE) independently scanned the abstract, title, or both, of every record retrieved, to determine which trials we should assess further. We investigated all potentially relevant articles as full text. We resolved any discrepancies through consensus or through recourse to a third review author (KR, JC, EL, LAl‐K). Where resolution of a disagreement was not possible, we added the article to those 'awaiting assessment' and contacted trial authors for clarification. We presented an adapted PRISMA (Preferred Reporting Items for Systematic Reviews and Meta‐Analyses) flow chart showing the process of trial selection (Liberati 2009).
Data extraction and management
For trials that fulfilled the inclusion criteria, two review authors (of JC, EL, COM, LA, EM, KR) independently abstracted key participant and intervention characteristics and reported data on efficacy outcomes and adverse events using standard data extraction templates as supplied by the CMED, with any disagreements to be resolved by discussion, or, if required, by consultation with a third review author (KR) (for details see Table 1; Appendix 2; Appendix 3; Appendix 4; Appendix 5; Appendix 6; Appendix 7; Appendix 8; Appendix 9) .
| Intervention(s) and comparator(s) | Sample sizea | Screened/eligible | Randomised | ITT | Analysed | Finishing study | Randomised finishing study | Follow‐up timeb | |
| Stark 2014 | I1: LAUNCH home visits | "Because of the small sample size, we were not powered to compare the two LAUNCH groups" | ‐/237 | 15 | 10 | 10 | 7 | 47 | 6 months/12 months |
| I2: LAUNCH clinic visits | 14 | 11 | 11 | 10 | 71 | ||||
| C: enhanced usual care | 13 | 12 | 12 | 11 | 85 | ||||
| total: | 42 | 33 | 33 | 28 | 67 | ||||
| Quattrin 2012 | I: family‐based intervention | In a mixed‐effect model a sample of 108 participants | 171/147 | 52 | 46 | 46 | 30 | 58 | 12 months/24 months |
| C: information control | 53 | 50 | 50 | 40 | 76 | ||||
| total: | 105 | 96 | 96 | 70 | 73 | ||||
| Bocca 2012 | I: multidisciplinary programme | ‐ | 78/75 | 40 | ‐ | 17 | 17 | 43 | 16 weeks/3 years |
| C: usual care | 35 | ‐ | 12 | 12 | 34 | ||||
| total: | 75 | ‐ | 29 | 29 | 39 | ||||
| Taveras 2011 | I: High Five for Kids: behavioural intervention | ‐ | 1486 attempted contact/361 contacted, pre‐eligible, willing | 271 (5 clusters) | ‐ | 253 | 253 | 93 | 12 months (end of intense phase; 2‐year data not reported) |
| C: usual care | 1007 attempted contact/267 contacted, pre‐eligible, willing | 204 (5 clusters) | ‐ | 192 | 192 | 94 | |||
| total: | 475 | 445 | 445 | 94 | |||||
| Stark 2011 | I: LAUNCH | ‐ | 4079/56 | 8 | ‐ | 7 (month 6 and 12) | 7 | 88 | 6 months/12 months |
| C: enhanced standard care | 10 | ‐ | 10 (month 6) 9 (month 12) | 9 | 90 | ||||
| total: | 18 | ‐ | 16 | 16 | 89 | ||||
| Lanigan 2010 | I: Trim Tots multicomponent intervention | ‐ | 105/105 | 49 | 49 | 49 | 21 | 43 | 6 months/2 years (6‐month follow‐up reported only) |
| C: wait‐list control | 39 | 39 | 39 | 21 | 54 | ||||
| total: | 88 | 88 | 88 | 42 | 48 | ||||
| Kelishadi 2009 | I1: dairy‐rich diet | 30 per group | ‐ | 40 | ‐ | ‐ | 36 | 90 | 6 months/3 years |
| I2: energy‐restricted diet | 40 | ‐ | ‐ | 31 | 78 | ||||
| C: control | 40 | ‐ | ‐ | 32 | 80 | ||||
| total: | 120 | 99 | 83 | ||||||
| Grand total | All interventions | 529 | 412 | ||||||
| All comparators | 394 | 317 | |||||||
| All interventions and comparators | 923 | 729 | |||||||
aAccording to power calculation in study publication or report.
bDuration of intervention and/or follow‐up under randomised conditions until end of study.
‐ denotes not reported
C: comparator; I: intervention; ITT: intention‐to‐treat; LAUNCH: Learning about Activity and Understanding Nutrition for Child Health; N/A: not applicable
We provided information including trial identifier about potentially relevant ongoing trials in the Characteristics of ongoing studies table and in Appendix 5 (Matrix of study endpoints (publications and trial documents)). We attempted to find the protocol of each included trial and reported primary, secondary, and other outcomes in comparison with data in publications in Appendix 5.
We e‐mailed all authors of included trials to enquire whether they were willing to answer questions regarding their trials. Appendix 10 shows the results of this survey. Thereafter, we sought relevant missing information on the trial from the primary author(s) of the article, if required.
Dealing with duplicate and companion publications
In the event of duplicate publications, companion documents, or multiple reports of a primary trial, we attempted to maximise yield of information by collating all available data and using the most complete data set aggregated across all known publications. If in doubt, we gave priority to the publication reporting the longest follow‐up associated with our primary or secondary outcomes.
Assessment of risk of bias in included studies
Two review authors (of JC, EL, COM, LA, EM, KR) independently assessed the risk of bias of each included trial. We resolved any disagreements by consensus or by consulting a third review author (KR).
We used the Cochrane 'Risk of bias' assessment tool to evaluate the following criteria (Higgins 2011a; Higgins 2011b).
-
Random sequence generation (selection bias).
-
Allocation concealment (selection bias).
-
Imbalances in baseline characteristics.
-
Blinding of participants and personnel (performance bias).
-
Blinding of outcome assessment (detection bias).
-
Incomplete outcome data (attrition bias).
-
Selective reporting (reporting bias).
-
Other potential sources of bias.
We judged the above '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 presented a 'Risk of bias' graph and a 'Risk of bias' summary figure. We assessed the impact of individual bias domains on trial results at endpoint and trial levels. In case of high risk of selection bias, we marked all endpoints investigated in the associated trial as 'high risk'.
We evaluated whether imbalances in baseline characteristics existed and how these were addressed (Egbewale 2014).
For performance bias (blinding of participants and personnel) and detection bias (blinding of outcome assessors), we evaluated the risk of bias separately for each outcome type (objective and subjective) (Hróbjartsson 2013). We noted whether endpoints were self reported, investigator assessed, or adjudicated outcome measures.
We considered the implications of missing outcome data from individual participants per outcome such as high drop‐out rates (for example above 15%) or disparate attrition rates (for example difference of 10% or more between trial arms).
We assessed outcome reporting bias by integrating the results of 'Examination of outcome reporting bias' (Kirkham 2010) (Appendix 6), in the 'Matrix of study endpoints (publications and trial documents)' (Appendix 5), and section 'Outcomes (outcomes reported in abstract of publication)' of the Characteristics of included studies table. This analysis formed the basis for the judgement of selective reporting (reporting bias).
We distinguished between self reported, investigator assessed, and adjudicated outcome measures.
We defined the following endpoints as potentially self reported ('subjective') outcomes.
-
Adverse events.
-
Health‐related quality of life.
-
Parent‐child relationship or assessment of parenting.
-
Participant views of the intervention.
We defined the following outcomes as potentially investigator assessed ('objective') outcomes.
-
Changes in BMI and body weight.
-
Adverse events.
-
All‐cause mortality.
-
Morbidity.
Measures of treatment effect
We expressed dichotomous data as odds ratios or risk ratios with 95% confidence intervals (CIs). We expressed continuous data as mean differences with 95% CI. We expressed time‐to‐event data as hazard ratios with 95% CIs.
Unit of analysis issues
We took into account the level at which randomisation occurred, such as cross‐over trials, cluster‐randomised trials, and multiple observations for the same outcome.
Dealing with missing data
We obtained relevant missing data from authors, if feasible, and evaluated important numerical data such as screened, eligible, randomised participants as well as intention‐to‐treat, as‐treated, and per‐protocol populations. We investigated attrition rates, for example dropouts, losses to follow‐up, and withdrawals, and critically appraised issues of missing data and imputation methods (for example last observation carried forward).
Where standard deviations for outcomes were not reported, and we did not receive information from trial authors, we calculated these following the methods presented in the Cochrane Handbook for Systematic Reviews of Interventions. Where papers did not report results as change from baseline, we calculated this and for the standard deviation differences followed the methods presented in the Cochrane Handbook for Systematic Reviews of Interventions for imputing these (Section 16.1.3.2 Imputing standard deviations for changes from baseline), and assumed a correlation of 0.5 between baseline and follow‐up measures as suggested by Follmann 1992.
Assessment of heterogeneity
In the event of substantial clinical, methodological, or statistical heterogeneity, we did not report trial results as meta‐analytically pooled effect estimates. We identified heterogeneity by visual inspection of the forest plots and by using a standard Chi² test with a significance level of α = 0.1, in view of the low power of this test. We examined heterogeneity using the I² statistic, which quantifies inconsistency across trials to assess the impact of heterogeneity on the meta‐analysis (Higgins 2002; Higgins 2003), where an I² statistic of 75% or more indicates a considerable level of inconsistency (Higgins 2011a).
When we found heterogeneity, we attempted to determine potential reasons for it by examining individual trial and subgroup characteristics.
Assessment of reporting biases
If we had included 10 or more trials for a given outcome, we would have used funnel plots to assess small‐trial effects. Given that there are several explanations for funnel plot asymmetry, we would have interpreted results carefully (Sterne 2011).
Data synthesis
Unless there was good evidence for homogeneous effects across trials, we primarily summarised data that was at low risk of bias by means of a random‐effects model (Wood 2008). We had planned to interpret random‐effects meta‐analyses with due consideration of the whole distribution of effects, ideally by presenting a prediction interval, however relatively few trials were included in each category, of low methodological quality, and so we did not conduct these analyses (Higgins 2009). A prediction interval specifies a predicted range for the true treatment effect in an individual trial (Riley 2011). In addition, we performed statistical analyses according to the statistical guidelines referenced in the latest version of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a).
Quality of evidence
We presented the overall quality of the evidence for each outcome according to the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach, which takes into account issues not only related to internal validity (risk of bias, inconsistency, imprecision, publication bias) but also to external validity such as directness of results. Two review authors (JC, EL) independently rated the quality of the evidence for each outcome. We presented a summary of the evidence in a 'Summary of findings' table, which provides key information about the best estimate of the magnitude of the effect, in relative terms and absolute differences for each relevant comparison of alternative management strategies, numbers of participants and trials addressing each important outcome, and the rating of the overall confidence in effect estimates for each outcome. We created the 'Summary of findings' table based on the methods described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). We presented results for the outcomes as described in Types of outcome measures. In addition, we established an appendix 'Checklist to aid consistency and reproducibility of GRADE assessments' to help with standardisation of 'Summary of findings' tables (Appendix 9) (Meader 2014).
Subgroup analysis and investigation of heterogeneity
We expected the following characteristics to introduce clinical heterogeneity, and planned to carry out subgroup analyses with investigation of interactions where data permitted.
-
Differences in BMI at baseline.
-
Length of follow‐up.
-
The impact of comparator/control: whether concomitant therapy or no treatment (true control).
-
The setting in which the intervention was conducted.
Sensitivity analysis
We planned to perform sensitivity analyses in order to explore the influence of the following factors (when applicable) on effect sizes by restricting analysis to the following.
-
Published trials.
-
Taking into account risk of bias, as specified in the Assessment of risk of bias in included studies section.
-
Very long or large trials to establish the extent to which they dominate the results.
-
Trials using the following filters: diagnostic criteria, language of publication, source of funding (industry versus other), country.
We also tested the robustness of the results by repeating the analysis using different statistical models (fixed‐effect and random‐effects models).
Results
Description of studies
For a detailed description of trials, see the Characteristics of included studies, Characteristics of excluded studies, and Characteristics of ongoing studies sections.
Results of the search
The searches generated 13,784 hits after duplicates were removed. Screening of titles and abstracts identified 105 papers to evaluate for formal inclusion and exclusion. Seven completed randomised controlled trials (RCTs) fulfilled the inclusion criteria and were included in the review. For a detailed description of the included trials, see Characteristics of included studies. We also identified four ongoing trials; see Characteristics of ongoing studies. We have presented the flow of trials through the review in Figure 1.
Study flow diagram.
Included studies
For a detailed description of the characteristics of included trials, see Characteristics of included studies and Appendix 2; Appendix 3; Appendix 4; Appendix 5; Appendix 6; Appendix 7; Appendix 8; Appendix 11). The following is a succinct overview.
Source of data
We obtained the majority of data presented in the review from published literature, including supplementary published data and trials registers where available. For three trials, data were provided via correspondence with the trial authors (Appendix 10).
Comparisons
We identified two main comparisons: six trials compared multicomponent interventions versus control (Bocca 2012; Lanigan 2010; Quattrin 2012; Stark 2011; Stark 2014; Taveras 2011), and one trial compared dietary interventions versus control (Kelishadi 2009).
Overview of trial populations
The seven trials included a total of 923 participants, with over half of these from just one trial (Taveras 2011). In total, 529 participants were randomised to an intervention and 394 to a comparator. The proportion of participants finishing the trial was lowest in Bocca 2012 at three years' follow‐up (39%), however data at this follow‐up are not yet available for inclusion in the review. Similarly, two‐year data (48% follow‐up) are not yet available for Lanigan 2010. For follow‐up reported in this review, the proportion of participants finishing the trial ranged from 47% to 93% in the intervention groups and 71% to 94% in the comparator groups. Individual sample size ranged between 18 and 475. See Table 1 for details.
Trial design
Six trials were parallel comparisons with individual randomisation. One trial was a cluster RCT (Taveras 2011), where the primary care practice was the unit of randomisation. All seven RCTs had a superiority design. Two RCTs had three comparisons (two intervention groups and one usual care or control arm) (Kelishadi 2009; Stark 2014); the remaining trials had two comparison groups. Five trials were single‐centre trials (Bocca 2012; Kelishadi 2009; Lanigan 2010; Stark 2011; Stark 2014), one trial was conducted in four centres (Quattrin 2012), and the cluster RCT was undertaken in 10 centres (Taveras 2011).
Trials were performed from 2003 to 2013. The duration of the interventions was six months in four trials (Kelishadi 2009; Lanigan 2010; Stark 2011; Stark 2014), and ranged from 16 weeks, in Bocca 2012, to two years, in Taveras 2011. The duration of follow‐up ranged from 12 months to three years. One trial terminated before regular end (after 105 families were recruited), as preliminary analysis indicated efficacy (Bocca 2012).
Settings
The interventions were carried out in an outpatient setting in three trials (Bocca 2012; Stark 2011; Stark 2014), primary care in two trials (Quattrin 2012; Taveras 2011), a community setting in one trial (Lanigan 2010), and an obesity research clinic in one trial (Kelishadi 2009). Four trials were conducted in the USA (Quattrin 2012; Stark 2011; Stark 2014; Taveras 2011), and one was conducted in each of the Netherlands (Bocca 2012), UK (Lanigan 2010), and Iran (Kelishadi 2009).
Participants
The diagnostic criteria for overweight and obesity differed between the trials. Two trials included children with BMI on or above the 85th percentile and who had a parent with BMI 27 or more, in Quattrin 2012, or BMI over 25, in Taveras 2011. The latter trial also included children with BMI on or above the 95th percentile and no overweight parent. Three other trials included children with BMI on or above the 95th percentile (Kelishadi 2009; Stark 2011; Stark 2014); two of these trials also specified BMI less than 100% above mean BMI and a parent with BMI 25 or more (Stark 2011; Stark 2014). Lanigan 2010 included children with BMI on or above the 91st percentile or whose weight had crossed centiles upwards, and Bocca 2012 included children who were 'overweight or obese' as defined by the International Obesity Task Force.
The age range for trial eligibility was 1 to 5 years (Lanigan 2010), 2 to 5 years (Quattrin 2012; Stark 2011; Stark 2014), 3 to 5 years (Bocca 2012), and 2 to 6.9 years (Taveras 2011). Kelishadi 2009 did not specify an age range. The mean age of children included in the trials ranged from 4 to under 6 years in six of the seven trials, and was 2.5 years in Lanigan 2010. The proportion of girls varied from 25% to 80%, but was not reported by one trial (Kelishadi 2009). Five trials reported mean BMI (Bocca 2012; Kelishadi 2009; Lanigan 2010; Quattrin 2012; Taveras 2011), which ranged from 18 to 22.7, and five trials reported BMI z score (Bocca 2012; Lanigan 2010; Quattrin 2012; Stark 2014; Taveras 2011), which ranged from 1.0 to 2.7. Stark 2011 reported mean BMI percentile (98 to 99). Mean parental BMI was 36 to 37 in the one trial reporting this (Quattrin 2012). Taveras 2011 reported the proportions of parents with BMI less than 25 (3% to 5%), 25 to 30 (36% to 52%), and 30 or more (44% to 61%).
Five of the seven trials reported ethnicity. The proportion of participants categorised as white was over 70% in four of the trials (Lanigan 2010; Quattrin 2012; Stark 2011; Stark 2014), and 47% to 70% in the fifth trial (Taveras 2011). Five trials reported socioeconomic status using different indicators (Hollingshead score, Hollingshead classification, family income, non‐manual social class, or parental educational attainment) (Lanigan 2010; Quattrin 2012; Stark 2011; Stark 2014; Taveras 2011).
Interventions
The interventions in six of the seven trials included a combination of nutritional, physical activity, and behavioural components, although approaches differed between the trials (Bocca 2012; Lanigan 2010; Quattrin 2012; Stark 2011; Stark 2014; Taveras 2011). See Characteristics of included studies and Appendix 2 for details of each trial. Two trials reported the same intervention, 'Learning about Activity and Understanding Nutrition for Child Health' (LAUNCH), which involved 18 group‐based clinic sessions and individual home visits over six months, targeting lifestyle behaviour modification and parenting skills (Stark 2011; Stark 2014). Stark 2014 also compared a less intensive mode of delivery, which was identical to the former except that individual home visits were not undertaken. Quattrin 2012 reported a family‐based parenting and behavioural intervention involving 13 group sessions over 12 months, individual meetings to shape goals, and 10 phone calls in between sessions. Bocca 2012 assessed a multidisciplinary intervention involving dietary advice, physical activity sessions, and psychological counselling for parents, with a total of 25 sessions over 16 weeks. Taveras 2011 reported 'High Five for Kids', a behavioural intervention using motivational interviewing face‐to‐face and by telephone, educational modules, and behavioural goal setting. The intensive phase of the intervention lasted for 12 months followed by a 12‐month maintenance phase, although no details were reported of this. Lanigan 2010 assessed the 'Trim Tots' healthy lifestyle programme, which included nutritional education, physical activity, and behavioural change components. Sessions were delivered in the community twice weekly for three months, then weekly for three months.
The comparators in these six trials were enhanced usual care (Stark 2011; Stark 2014), information control (Quattrin 2012), usual care (Bocca 2012; Taveras 2011), or a wait‐list control (Lanigan 2010), with differences between trials in the intensity and amount of contact (Characteristics of included studies; Appendix 2).
One trial compared two dietary approaches (dairy rich and energy restricted) plus healthy lifestyle education versus healthy lifestyle education alone (Kelishadi 2009). The family‐centred education sessions focused on health, nutrition, and physical activity and occurred monthly over six months in all three groups.
Outcomes
Six of seven trials explicitly stated a primary/secondary endpoint in the publication (Appendix 5). The most commonly defined primary outcome in the publications was a BMI variable: BMI z score (Bocca 2012; Lanigan 2010; Stark 2011; Stark 2014), BMI (Taveras 2011), or percent BMI overweight (Quattrin 2012). Where reported, primary outcomes defined in publications reflected those defined in trial registers, although in three trial publications (Bocca 2012; Stark 2011; Stark 2014), a greater number of outcomes were specified as primary. In Lanigan 2010, the primary outcome was specified as BMI in the trial register and BMI z score in the publication. Only one of seven trials reported on adverse events (Lanigan 2010), two reported health‐related quality of life (Bocca 2012; Stark 2011), five reported behaviour change (Bocca 2012; Kelishadi 2009; Stark 2011; Stark 2014; Taveras 2011), two reported participant views of the intervention (Stark 2011; Taveras 2011), and two reported parent‐child relationship or assessment of parenting (Stark 2011; Stark 2014). No trials investigated all‐cause mortality, morbidity, or socioeconomic effects. All seven trials provided a definition of endpoint measurement for BMI (Appendix 7). One trial was published as an abstract only, and a number of secondary outcomes are not yet available (Lanigan 2010).
Excluded studies
After evaluation of the full publication, we excluded 82 of 105 full‐text articles. The main reasons for exclusion were the types of participants included and the trial design not being an RCT with at least six months' duration. Many trials had multiple reasons for exclusion (for further details see Characteristics of excluded studies, which lists the 12 trials with the most uncertainty regarding inclusion).
Risk of bias in included studies
For details on risk of bias of included trials, see Characteristics of included studies. For an overview of review authors' judgements about each 'Risk of bias' item for individual trials and across all trials, see Figure 2 and Figure 3.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included trials (blank cells indicate that the particular outcome was not investigated in some trials).
Risk of bias summary: review authors' judgements about each risk of bias item for each included study (blank cells indicate that the study did not report that particular outcome).
Allocation
All seven trials reported adequate sequence generation, but only three described allocation concealment (Lanigan 2010; Stark 2011; Stark 2014). The risk of selection bias is therefore uncertain in four of the seven included trials.
Blinding
We judged only one trial to have a low risk of performance bias, as participants and personnel were blinded to treatment allocation (Quattrin 2012). Three trials did not undertake blinding of participants and personnel (Lanigan 2010; Stark 2011; Stark 2014); we judged these trials to have a high risk of performance bias for both objective and subjective outcomes (where reported). The risk of performance bias was unclear in three trials: one was described as single blind (Bocca 2012), and another as double‐blind (in the trial record only), but it was unclear who was blinded and how this was achieved (Taveras 2011), and in the third blinding was not reported (Kelishadi 2009).
We judged the risk of detection bias to be low for objective outcomes in all trials, regardless of whether or not outcome assessors were blinded. Two trials reporting subjective outcomes did not undertake blinding of outcome assessors (Stark 2011; Stark 2014); we judged these trials to have a high risk of detection bias for subjective outcomes. The risk of detection bias for subjective outcomes was unclear in three trials, either because blinding was not reported or it was unclear who was blinded and how this was achieved (Bocca 2012; Kelishadi 2009; Taveras 2011).
Incomplete outcome data
We judged the risk of attrition bias to be low in three trials (Bocca 2012; Kelishadi 2009; Stark 2011). Three trials had an imbalance in attrition between groups and were judged to have a high risk of attrition bias (Lanigan 2010; Quattrin 2012; Stark 2014). The risk of attrition bias was unclear for one trial that did not report reasons for attrition and only reported baseline and results data for participants completing the trial (Taveras 2011).
Selective reporting
We judged only two trials to have a low risk of selective reporting bias (Stark 2011; Taveras 2011). We judged the remaining five trials to have a high risk of selection bias, due to some outcomes being incompletely reported or not reported at all.
Other potential sources of bias
There was a low risk of bias from other sources in four trials (Bocca 2012; Lanigan 2010; Stark 2011; Taveras 2011). We judged the risk of bias from other sources to be unclear in the remaining three trials.
Effects of interventions
See: Summary of findings for the main comparison ; Summary of findings 2
Multicomponent interventions versus control
Six trials compared multicomponent interventions with usual care, enhanced usual care, information control, or wait‐list control. One trial assessed two interventions, LAUNCH with home visits and LAUNCH without home visits (described in the analyses as LAUNCH clinic only) (Stark 2014). Stark 2011 also assessed LAUNCH (with home visits). The other four trials each assessed different multicomponent interventions (Bocca 2012; Lanigan 2010; Quattrin 2012; Taveras 2011).
We considered outcomes here at the end of the intervention (if six months or longer) and at any follow‐up period. The intervention period was six months in three trials (Lanigan 2010; Stark 2011; Stark 2014), and 12 months in two trials (Quattrin 2012; Taveras 2011). The intervention was 16 weeks in one trial (Bocca 2012); we did not include outcomes at this time point in this review. Four trials reported outcomes at 12 months, in Bocca 2012, Stark 2011 and Stark 2014, or 18 months, in Quattrin 2012, from baseline (six to eight months from end of intervention), and one trial, Quattrin 2012, reported outcomes at 24 months (12 months from end of intervention). Lanigan 2010 followed participants for 24 months, but data are not yet available. Bocca 2012 followed participants for 36 months, but data at this time point were not reported in a useable format.
The proportion of participants finishing the trial (or the longest reported follow‐up, if different) was less than 80% in four trials, Bocca 2012, Lanigan 2010, Quattrin 2012, Stark 2014, out of the six (Table 1), moreover there was a differential rate of losses to follow‐up between groups in three of these trials (Lanigan 2010; Quattrin 2012; Stark 2014). Attrition was highest in the multicomponent intervention arms of Lanigan 2010 and Quattrin 2012, and the LAUNCH home visits arm of Stark 2014 (although both intervention arms had higher attrition than the enhanced usual care arm in this trial).
Primary outcomes
Changes in body mass index (BMI) and body weight
All six trials reported BMI z score (Bocca 2012; Lanigan 2010; Quattrin 2012; Stark 2011; Stark 2014; Taveras 2011). Pooling the studies in a random‐effects meta‐analyses (Analysis 1.1) demonstrated a reduction in BMI z score in the intervention groups compared with controls at the end of intervention: mean difference (MD) ‐0.26 units (95% confidence interval (CI) ‐0.37 to ‐0.16); P < 0.00001; 210 participants; 4 trials; low‐quality evidence). At 12 to 18 months' follow‐up, the MD was ‐0.38 units (95% CI ‐0.58 to ‐0.19); P = 0.0001; 202 participants; 4 trials; low‐quality evidence). One trial, Quattrin 2012, reported outcomes at 24 months' follow‐up (12 months' postintervention) and found the benefit was maintained (MD ‐0.25 units (95% CI ‐0.40 to ‐0.10)). One cluster RCT comparing intervention with control reported a change in BMI z score from baseline to one year of 0.05 units (95% CI ‐0.14 to 0.04); P = 0.28 (Taveras 2011).
Three trials reported BMI at the end of intervention or postintervention: Lanigan 2010 showed a MD of ‐0.40 kg/m² (95% CI ‐0.85 to 0.05) in 64 participants, and the cluster RCT by Taveras 2011 an adjusted (child age, sex, ethnicity, parent education, overweight/obesity status at baseline, household income, time elapsed from baseline to follow‐up) MD of ‐0.21 kg/m² (95% CI ‐0.50 to 0.07) in 445 participants. Eight months' postintervention, Bocca 2012 reported a MD of ‐1.00 kg/m² (95% CI ‐1.79 to ‐0.21) in 57 participants (Analysis 1.2).
Three of the trials reporting BMI z score also reported additional BMI variables, including changes in percent over BMI (Analysis 1.3) and BMI percentile (Analysis 1.4), which reflected the results in BMI z score.
Five trials reported change in body weight (Bocca 2012; Lanigan 2010; Quattrin 2012; Stark 2011; Stark 2014) (Analysis 1.5). Pooling the trials in a random‐effects meta‐analysis showed less weight gain with the interventions compared with control at the end of the intervention: MD ‐1.18 kg (95% CI ‐1.91 to ‐0.45); P = 0.001; 210 participants; 4 trials; low‐quality evidence) and at 12 to 18 months' follow‐up: MD ‐2.81 kg (95% CI ‐4.39 to ‐1.22); P = 0.0005; 202 participants; 4 trials; low‐quality evidence).
One trial, Quattrin 2012, reported change in body weight at 24 months' follow‐up (12 months' postintervention) and found the benefit was maintained: MD ‐1.60 kg (95% CI ‐2.42 to ‐0.78); 96 participants). Follow‐up was low (58%) in the intervention arm of this trial.
Three trials reported change in parental BMI or weight (Quattrin 2012; Stark 2011; Stark 2014); these trials required the parent to have a BMI of least 25 or 27 for trial inclusion. Pooling the trials in a meta‐analysis demonstrated a reduction in parental BMI in the intervention group compared with controls at the end of intervention (6 or 12 months) (MD ‐2.00 kg/m² (95% CI ‐2.52 to ‐1.48); P < 0.00001; 113 participants; 2 trials; low‐quality evidence) and at 12 to 18 months' follow‐up (MD ‐2.08 kg/m² (95% CI ‐2.65 to ‐1.51); P < 0.00001; 112 participants; 2 trials; low‐quality evidence). Quattrin 2012 also reported outcomes at 24 months' follow‐up, where the effect remained stable (MD ‐2.00 kg/m² (95% CI ‐2.57 to ‐1.43); 96 participants) (Analysis 1.6).
The effect on parental body weight was similar (Analysis 1.7), with parents in the intervention group reducing weight by around 5 kg more than the control group at the end of the intervention (MD ‐4.69 kg (95% CI ‐7.27 to ‐2.11); P = 0.0004; 146 participants; 3 trials; low‐quality evidence) and at 12 to 18 months' follow‐up (MD ‐5.14 kg (95% CI ‐8.96 to ‐1.33); P = 0.008; 49 participants; 2 trials; low‐quality evidence). Only one trial reported outcomes at 24 months' follow‐up (Quattrin 2012), where the effects remained (MD ‐6.70 kg (95% CI ‐8.42 to ‐4.98)).
Adverse events
Only one trial reported, as an abstract only, on adverse events (Lanigan 2010), stating that no adverse events were observed (Appendix 8). Further details, such as how adverse events were monitored, were not provided.
Secondary outcomes
Health‐related quality of life and self esteem
Three trials reported health‐related quality of life (Bocca 2012; Stark 2011; Stark 2014), but measures and scores used varied between trials (Appendix 11). Bocca 2012 reported two tools, the Dutch Child AZL TNO Quality‐of‐Life (DUX‐25), which measures daily activities, and the Dutch edition of the Child Health Questionnaire Parent Form (CHQ‐PF50), which measures health perception. At 12 months' follow‐up (eight months postintervention), a statistically significant higher increase was found in 40 participants in the multidisciplinary intervention group compared with the usual care group in the DUX‐25 total score (change in median of the total score: +5 in the intervention group versus ‐5 in the control group; 0 to 100 scale with higher scores indicating better health‐related quality of life) and in one of four domains (physical score; change in median: +8 in the intervention group versus ‐4 in the control group; with higher scores indicating better health‐related quality of life), but no substantial differences were found in any of the 15 items on the CHQ‐PF50 (Analysis 1.8; Analysis 1.9). Parents of 50% to 57% of randomised children completed questionnaires.
Stark 2011 reported the Pediatric Quality of Life Inventory (PedsQL) Generic Core scales, using the total score and the physical functioning, emotional functioning, and social functioning subscales, however only data for physical functioning were reported. A statistically significant improvement (higher scores indicate better quality of life) in the change in health‐related quality of life in physical functioning was found at the end of intervention (six months, mean +9.5 units in the intervention group versus ‐1.7 units in the control group) and at 12 months' follow‐up (mean +13.8 units in the intervention group versus ‐2.7 units in the control group) (Analysis 1.10).
Stark 2014 used the parent version of the PedsQL (total score) and found no substantial differences between multicomponent interventions and control in the total score (Analysis 1.11).
No trials measured self esteem.
All‐cause mortality
Not reported.
Morbidity
Not reported.
Anthropometric measures other than BMI
One trial reported waist circumference at the end of the six‐month intervention (Lanigan 2010) (Analysis 1.12), and one trial reported a number of anthropometric measures at 12 months' follow‐up (eight months after intervention end) (Bocca 2012), including waist circumference (Analysis 1.12), waist circumference z score (Analysis 1.13), hip circumference (Analysis 1.14), hip circumference z score (Analysis 1.15), upper arm circumference (Analysis 1.16), per cent body fat (Analysis 1.17), fat‐free mass (Analysis 1.18), visceral fat (Analysis 1.19), and subcutaneous fat (Analysis 1.20). Although these outcomes tended to favour the intervention, differences were only statistically significant for upper arm circumference and visceral fat. Caution is required in the interpretation of these results due to the number of outcomes measured.
Behaviour change
Four trials reported some form of assessment of behaviour change (Bocca 2012; Stark 2011; Stark 2014; Taveras 2011).
Four trials measured physical activity, however different methods and outcomes were used (Appendix 7). Taveras 2011 found no substantial difference in change in number of hours per day of outdoor active playtime at the end of intervention (Analysis 1.21), and Bocca 2012 found no substantial difference in change in number of steps per day at 12 months' follow‐up (Analysis 1.22) Of the two trials evaluating LAUNCH (Stark 2011; Stark 2014), both reported change in average daily minutes of moderate and vigorous physical activity (Analysis 1.23; Analysis 1.24). Stark 2011 found no substantial difference in physical activity, and although data from Stark 2014 suggest a statistically significant effect in the LAUNCH clinic only group compared with control at 12 months' follow‐up, this result should be viewed with caution as the treatment effect was not statistically significant when computed using maximum likelihood estimation to account for missing data by Stark 2014.
One cluster RCT, Taveras 2011, reported a greater reduction in the number of servings per day of sugar‐sweetened drinks (MD ‐0.26 (95% CI ‐0.49 to ‐0.03); Analysis 1.25), but this was not statistically significant in an adjusted analysis (MD ‐0.22 (95% CI ‐0.52 to 0.08)). Similarly, there was no substantial difference in the increase in servings per day of fruits and vegetables (MD 0.06 (95% CI ‐0.21 to 0.33); Analysis 1.26), adjusted analysis: MD 0.12 (95% CI ‐0.17 to 0.42). However, we found a statistically significant difference in the reduction in hours per day of television and video viewing: MD ‐0.46 hrs (95% CI ‐0.70 to ‐0.22); Analysis 1.27), adjusted analysis: MD ‐0.36 hrs (95% CI ‐0.64 to ‐0.09). Analyses were adjusted for child age, sex, and race/ethnicity; parent education and overweight/obesity status at baseline; household income; and exact time elapsed from baseline to follow‐up visit.
Participant views of the intervention
Two trials reported the participants' views of the interventions. Taveras 2011 reported that 97% were “somewhat” or “very satisfied” with the High Five for Kids program, and that 91% would recommend the program to their family and friends. Using the Barrier to Treatment Participation Scale questionnaire, Stark 2011 reported no statistically significant difference between LAUNCH and comparator (paediatric counselling) on parent perceptions of treatment demands (11 for both groups) or relevance of treatment (11.3 versus 10.6, respectively). Parents in LAUNCH reported significantly greater stressors and obstacles compared to parents in the comparator group (33 ± 8.2 versus 25.6 ± 4.7, P = 0.038). Parents in LAUNCH and comparator were highly satisfied with treatment and did not differ substantially in their satisfaction ratings for information on nutrition (4.86 ± .38 versus 4.30 ± 1.25, P > 0.05) or physical activity (4.71 ± .49 versus 4.00 ± 1.25, P > 0.05), or in their satisfaction with ability to make recommended changes (4.26 ± .49 versus 4.20 ± 1.23, P > 0.05).
Parent‐child relationship or assessment of parenting
Two trials reported the Child Feeding Questionnaire (CFQ), a self report measure to assess parental beliefs, attitudes, and practices regarding child feeding (Stark 2011; Stark 2014) (Appendix 7). However, the trial authors report limited data. The CFQ contains 31 items, loading on seven factors. Four items relate to parental perception of child and parent weight, and concern about weight, which may elicit parental control in feeding. Pressure to eat scores were reduced by 0.9 and 0.6 at six months and by 0.6 and 0.3 at 12 months in the intervention and control groups, respectively (Stark 2011). Restriction to eat was described as "stable at approximately 4" with no statistically significant differences. Stark 2014 state "from baseline to months 6 and 12, CFQ restriction and pressure to eat remained relatively low (< 2.3) across all time points with no significant changes between groups".
Socioeconomic effects
No trials reported socioeconomic effects.
Diet interventions versus control
One three‐arm trial compared two diets (dairy rich and energy restricted) plus healthy lifestyle education with health lifestyle education alone (Kelishadi 2009). The intervention was of six months' duration, and follow‐up was three years from baseline, with 90.1%, 77.5%, and 90% follow‐up in the dairy‐rich, energy‐restricted, and control groups, respectively. We judged the trial to have a high risk of selective outcome bias and an unclear risk of selection bias, performance bias, detection bias, and other bias. The reported standard deviations appear very small and could possibly be standard errors.
Primary outcomes
Changes in body mass index (BMI) and body weight
For change in BMI z score, see Analysis 2.1. At the end of the intervention, BMI z score was reduced in all groups, however the reduction was greater with both the dairy‐rich diet (MD ‐0.10 units (95% CI ‐0.11 to ‐0.09)) and energy‐restricted diet (MD units ‐0.10 (95% CI ‐0.11 to ‐0.09)) than with control. At 12, 24, and 36 months' follow‐up, the difference between the dairy‐rich diet and control increased (MD units ‐0.20 (95% CI ‐0.21 to ‐0.19); MD ‐0.60 units (95% CI ‐0.61 to ‐0.59); MD ‐0.70 units (95% CI ‐0.71 to ‐0.69), respectively). In contrast, at 12 months' follow‐up, the MD between the energy‐restricted diet and control was 0.00 units (95% CI ‐0.01 to 0.01), and at 24 and 36 months' follow‐up, the outcome favoured control (MD 0.10 units (95% CI 0.09 to 0.11); MD units 0.10 (95% CI 0.09 to 0.11), respectively).
Adverse events
Adverse events were not reported.
Secondary outcomes
Health‐related quality of life and self esteem
Not reported.
All‐cause mortality
Not reported.
Morbidity
Not reported.
Anthropometric measures other than BMI
Reductions in waist circumference were greater with both the dairy‐rich (MD ‐0.30 cm (95% CI ‐0.39 to ‐0.21)) and energy‐restricted diet (MD ‐0.80 cm (95% CI ‐0.91 to ‐0.69)) at the end of the intervention (Analysis 2.2). At three years' follow‐up, only the dairy‐rich diet remained better on waist circumference than control (MD ‐0.70 cm (95% CI ‐0.84 to ‐0.56)), while the energy‐restricted diet had a worse outcome than the control group (MD 0.40 cm (95% CI 0.23 to 0.57)).
The dairy‐rich diet group had a greater reduction in percentage body fat than control at the end of the intervention (MD ‐2.00% (95% CI ‐2.18 to ‐1.82)), but this was not maintained at 24 months' follow‐up (Analysis 2.3). There were no substantial differences in percentage body fat between the energy‐restricted diet and control at either time point. Data at three years' follow‐up were not reported for this outcome.
Behaviour change
Kelishadi 2009 stated that "mean daily energy expenditure did not differ significantly by group at each time period throughout the study”, however data were presented in a figure only and could not be accurately estimated by review authors.
Participant views of the intervention
Participant views were not reported.
Parent‐child relationship or assessment of parenting
Not reported.
Socioeconomic effects
Socioeconomic effects were not reported.
Subgroup analyses
We did not perform subgroups analyses because there were not enough trials to estimate effects in various subgroups.
Sensitivity analyses
We did not perform any sensitivity analyses because there were not enough trials included in the analyses.
Assessment of reporting bias
We did not draw funnel plots due to limited number of trials per outcome (n = 7).
Ongoing studies
We found four ongoing RCTs. All are parallel RCTs, with estimated sample sizes of 28 to 240. For descriptions of the interventions, see Characteristics of ongoing studies; one trial focuses on parenting and parent lifestyle, and the other three trials are multicomponent interventions. The primary outcome includes BMI z score or BMI in all trials. The trial completion date ranges from November 2015 to August 2016, where reported.
Discussion
Summary of main results
This systematic review summarised seven RCTs examining the effect of diet, physical activity, and behavioural interventions for treating overweight and obesity in preschool children up to the age of 6 years. We only included trials with at least a six‐month outcome assessment, with the aim of assessing the longer‐term effects of these interventions. Interventions and comparators varied between the included trials, and we divided the trials into two main groups to ease interpretation: multicomponent interventions and diet‐only interventions. Outcomes assessed also varied between trials; the most commonly reported measure was BMI z score, but one trial reported BMI only. To allow comparison across trials, we analysed outcome data at the end of the intervention (whether this was 6 or 12 months) and at each reported follow‐up period. Four trials had an unclear risk of selection bias, and most had an unclear or high risk of performance bias. All but two trials had a high risk of selective reporting bias.
Overall, multicomponent interventions were more successful than the comparators in reducing BMI and body weight in preschool children and their parents, and the effects were maintained two years after the start of the intervention. Improvements were found in some, but not all, aspects of health‐related quality of life. The trials measured behaviour change inconsistently, and the effects of the interventions were more equivocal. There was limited assessment of participant views, parent‐child relationship, or assessment of parenting; where these were reported, there was no difference between groups. Only one trial commented on adverse events, stating that none were reported.
One three‐arm trial found that both the dairy‐rich and energy‐restricted dietary interventions resulted in greater reduction in BMI than the comparator at the end of the intervention period, but only the dairy‐rich diet maintained this at 12 to 36 months' follow‐up. Limited assessment was made of behaviour change; mean daily energy expenditure did not differ substantially between groups. Health‐related quality of life, adverse effects, participant views, and parenting were not measured.
A number of outcomes such as all‐cause mortality, morbidity, self esteem, and socioeconomic effects were not measured by any of the included trials.
Overall completeness and applicability of evidence
This review identified just seven trials assessing the effects of diet, physical activity, and behavioural interventions at six months or longer.
Six trials included multicomponent interventions. One of these trials had not yet fully reported at the time of writing, but will be included in the next review update. The duration of the interventions and length of follow‐up varied slightly between these trials, with most reporting outcomes at 12 to 18 months' follow‐up. Only one trial reported outcomes at two years' follow‐up. Long‐term effects of the interventions therefore remain uncertain.
One trial included two different diet interventions. The dairy‐rich diet was high in curd and dough, which are traditional dairy products prepared and consumed in Iran, and may not be relevant to other countries. The benefits of the dairy‐rich diet found in this trial would need to be confirmed in other trials.
Few of the trials reported secondary outcomes of interest to the review, and where these were reported there was little agreement in the types of measures used, making comparison difficult.
Results to date indicate that the magnitude of change in BMI z score with multicomponent interventions is likely to be clinically significant at all follow‐up periods. Available data in adolescents show that a change of 0.25 or more is associated with improvements in adiposity and metabolic health, with larger changes eliciting greater benefits (Ford 2010). Though the populations are not directly comparable, this nevertheless gives some indication of clinical relevance.
Quality of the evidence
Overall, the quality of the evidence was low, with six of seven included trials judged to have a high risk of bias on individual 'Risk of bias' criteria, and only three trials having a low risk of selection bias. GRADE assessments of the outcomes in this review led to trials being downgraded for risk of bias and also imprecision owing to the small number of trials and small sample sizes. This makes overall interpretation of the data difficult.
Potential biases in the review process
We carried out a comprehensive search across major databases for interventions for childhood overweight and obesity for this review. In addition, we screened the reference lists of systematic reviews. Each included trial in the review was comprehensively selected, assessed, data extracted, and quality assessed by two review authors to minimise potential biases in the review process. We made no decisions about the analysis or investigation of heterogeneity after seeing the data. There were differences between the trials in both the interventions delivered and the comparators. We divided the comparisons in this review into two groups, multicomponent interventions and dietary interventions, in order to improve comparability of trials within groups. We made this decision after examination of trial characteristics of included trials, but before we viewed the data. Where relevant data were missing, either to allow assessment of eligibility or at the data extraction stage, the review authors contacted the trial authors for further information.
Agreements and disagreements with other studies or reviews
Previous systematic reviews identified an absence of RCTs assessing interventions for treating overweight or obesity in preschool‐aged children (Bluford 2007; Bond 2009; Bond 2011; Oude Luttikhuis 2009). Since these reviews were undertaken, seven trials have been published and four trials are still ongoing. The current review is the first to synthesise the most up‐to‐date and highest‐quality research available on the effectiveness of interventions for treating overweight or obesity in preschool‐aged children.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included trials (blank cells indicate that the particular outcome was not investigated in some trials).
Risk of bias summary: review authors' judgements about each risk of bias item for each included study (blank cells indicate that the study did not report that particular outcome).
Comparison 1 Multicomponent intervention versus control, Outcome 1 Changes in BMI z score.
Comparison 1 Multicomponent intervention versus control, Outcome 2 Changes in BMI.
Comparison 1 Multicomponent intervention versus control, Outcome 3 Changes in % over BMI.
Comparison 1 Multicomponent intervention versus control, Outcome 4 Changes in BMI percentile.
Comparison 1 Multicomponent intervention versus control, Outcome 5 Changes in body weight.
Comparison 1 Multicomponent intervention versus control, Outcome 6 Changes in parental BMI.
Comparison 1 Multicomponent intervention versus control, Outcome 7 Changes in parental weight.
| Study | Domain | Intervention median (25th‐75th percentile), N | Control median (25th‐75th percentile), N | P value |
| 12 months follow‐up (8 months post intervention) | ||||
| Bocca 2012 | Total | 5.0 (‐1.8; 6.5), 20 | ‐4.5 (‐15.3; 4.0), 20 | P = 0.04 |
| Bocca 2012 | Physical | 8.3 (‐6.3; 16.7), 20 | ‐4.2 (‐12.5; 4.7), 20 | P = 0.03 |
| Bocca 2012 | Home | ‐4.2 (‐12.5; 4.7), 20 | 0.0 (‐15.0; 10.0), 20 | P = ns |
| Bocca 2012 | Emotional | 0.0 (‐3.6; 3.6), 20 | ‐3.6 (‐17.9; 10.7), 20 | P = ns |
| Bocca 2012 | Social | 1.8 (‐9.8; 7.1), 20 | ‐8.9 (‐14.3; 1.8), 20 | P = ns |
Comparison 1 Multicomponent intervention versus control, Outcome 8 Changes in health‐related quality of life: DUX 25.
| Study | Domain | Intervention median (25th;75th percentile), N | Control median (25th;75th percentile), N | P value |
| 12 months follow‐up (8 months post intervention) | ||||
| Bocca 2012 | Global health Physical functioning | 0.0 (‐12.5; 20.0), 20 0.0 (‐1.4; 12.5), 20 | 0.0 (‐25.0; 0.0), 20 0.0 (0.0; 9.7), 20 | P = ns P = ns |
| Bocca 2012 | Role functioning – Emotional/ behaviour Role functioning ‐ physical | 0.0 (0.0; 0.0), 20 0.0 (0.0; 0.0), 20 | 0.0 (0.0; 0.0), 20 0.0 (0.0; 0.0), 20 | P = ns P = ns |
| Bocca 2012 | Bodily pain General behaviour | 0.0 (‐5.0; 20.0), 20 0.0 (0.0; 25.0), 20 | 0.0 (‐17.5; 27.5), 20 0.0 (‐22.5; 25.0), 20 | P = ns P = ns |
| Bocca 2012 | Behaviour Mental health | 4.2 (‐9.4; 18.3), 20 0.0 (‐5.0; 10.0), 20 | ‐9.6 (‐16.7; 6.3), 20 ‐5.0 (‐15.0; 10.0), 20 | P = ns P = ns |
| Bocca 2012 | Self‐esteem General health | 12.5 (‐5.2; 29.2), 20 0.0 (‐3.3; 15.4), 20 | 4.2 (‐8.3; 12.5), 20 4.2 (‐16.0; 17.9), 20 | P = ns P = ns |
| Bocca 2012 | Change in health Parental impact ‐ emotional | 25.0 (0.0; 31.3), 20 0.0 (‐8.3; 8.3), 20 | 12.5 (0.0; 25.0), 20 ‐4.2 (‐14.6; 8.3), 20 | P = ns P = ns |
| Bocca 2012 | Parental impact ‐ time Family activities | 0.0 (‐11.1; 11.1), 20 ‐4.2 (‐13.5; 8.3), 20 | 0.0 (‐11.1; 0.0), 20 0.0 (‐8.3; 12.5), 20 | P = ns P = ns |
| Bocca 2012 | Family cohesion | 0.0 (‐15.0; 25.0), 20 | 0.0 (‐25.0; 22.5), 20 | P = ns P = ns |
Comparison 1 Multicomponent intervention versus control, Outcome 9 Changes in health‐related quality of life: CHQ‐PF50.
Comparison 1 Multicomponent intervention versus control, Outcome 10 Changes in health‐related quality of life: PEDsQL physical functioning.
Comparison 1 Multicomponent intervention versus control, Outcome 11 Changes in health‐related quality of life: PEDsQL total score.
Comparison 1 Multicomponent intervention versus control, Outcome 12 Changes in waist circumference.
Comparison 1 Multicomponent intervention versus control, Outcome 13 Changes in waist circumference z‐score.
Comparison 1 Multicomponent intervention versus control, Outcome 14 Changes in hip circumference.
Comparison 1 Multicomponent intervention versus control, Outcome 15 Changes in hip circumference z‐score.
Comparison 1 Multicomponent intervention versus control, Outcome 16 Changes in upper arm circumference.
Comparison 1 Multicomponent intervention versus control, Outcome 17 Changes in per cent body fat.
Comparison 1 Multicomponent intervention versus control, Outcome 18 Changes in fat‐free mass.
Comparison 1 Multicomponent intervention versus control, Outcome 19 Changes in visceral fat.
Comparison 1 Multicomponent intervention versus control, Outcome 20 Changes in subcutaneous fat.
Comparison 1 Multicomponent intervention versus control, Outcome 21 Changes in outdoor active play.
Comparison 1 Multicomponent intervention versus control, Outcome 22 Changes in steps.
Comparison 1 Multicomponent intervention versus control, Outcome 23 Changes in physical activity, moderate.
Comparison 1 Multicomponent intervention versus control, Outcome 24 Changes in physical activity, vigorous.
Comparison 1 Multicomponent intervention versus control, Outcome 25 Changes in sugar‐sweetened drinks.
Comparison 1 Multicomponent intervention versus control, Outcome 26 Changes in fruit and vegetable intake.
Comparison 1 Multicomponent intervention versus control, Outcome 27 Changes in TV and video viewing.
Comparison 2 Diet intervention versus control, Outcome 1 Changes in BMI z score.
Comparison 2 Diet intervention versus control, Outcome 2 Changes in waist circumference.
Comparison 2 Diet intervention versus control, Outcome 3 Changes in per cent body fat.
| Diet, physical activity, and behavioural interventions for the treatment of overweight or obesity in preschool children aged 0 to 6 years | ||||||
| Population: preschool children (aged 0 to 6 years) with overweight or obesity Settings: various Intervention: multicomponent interventions Comparison: usual care/enhanced usual care/information control/wait‐list control | ||||||
| Outcomes | Control | Multicomponent intervention | Relative effect | No of participants | Quality of the evidence | Comments |
| Changes in BMI and body weight a. BMI z scorea [units] Follow‐up: 12 to 18 months b. Weight [kg] Follow‐up: 12 to 18 months | a. The mean change in BMI z score ranged across control groups from ‐0.3 units to +0.4 units b. The mean change in weight ranged across control groups from +3.1 kg to +5.2 kg | b. The mean change in BMI z score in the intervention groups was 0.4 units lower (0.6 to 0.2 lower) b. The mean change in weight in the intervention group was 2.8 kg lower (4.4 to 1.2 lower) | ‐ | a. 202 (4) b. 202 (4) | a. ⊕⊕⊝⊝ b. ⊕⊕⊝⊝ | Lower units indicate more weight loss |
| Adverse events Follow‐up: 24 months | See comment | See comment | See comment | 88 (1) | ⊕⊝⊝⊝ | Only 1 trial (abstract only) reported on adverse events, stating no adverse events were observed |
| HrQoL and self esteem a. DUX 25 (Dutch Child AZL TNO Quality‐of‐Life tool: total score and 4 domains; scale 0 to 100; higher score indicates better HrQoL) Follow‐up: 12 months b. CHQ‐PF50 (Dutch edition of the Child Health Questionnaire Parent Form: 15 items; score 0 to 100; higher score indicates better HrQoL) Follow‐up: 12 months c. PedsQL (Pediatric Quality of Life Inventory, physical functioning subscale; higher score indicates better HrQoL) Follow‐up: 6 months/12 months d. PedsQL (total score) Follow‐up: 12 months | See comment | See comment | See comment | a. 40 (1) b. 40 (1) c. 17 (1) d. 16 (1) | a/b/c/d | No trials reported self esteem a. Change in median of the total score: +5 in the intervention group versus ‐5 in the control group; change in median of 1 of 4 domains (physical functioning): +8 in the intervention group versus ‐4 in the control group b. No statistically significant differences in any of the 15 items c. 6 months' change in mean: +9.5 units in the intervention group versus ‐1.7 units in the control group, data not reported for total score and 3 other subscales; 12 months' change in mean +13.8 units in the intervention group versus ‐2.7 units in the control group, data not reported for total score and 3 other subscales d. No substantial differences between multicomponent intervention and control group |
| All‐cause mortality | See comment | See comment | See comment | See comment | See comment | No trials reported all‐cause mortality |
| Morbidity | See comment | See comment | See comment | See comment | See comment | No trials reported morbidity |
| Parent‐child relationship or assessment of parenting | See comment | See comment | See comment | 44 (2) | ⊕⊝⊝⊝ | Limited data were reported, no substantial differences between intervention and control groups |
| Socioeconomic effects | See comment | See comment | See comment | See comment | See comment | No trials reported socioeconomic effects |
| BMI: body mass index; CI: confidence interval; HrQoL : health‐related quality of life | ||||||
| GRADE Working Group grades of evidence | ||||||
| a"A BMI z score or standard deviation score indicates how many units (of the standard deviation) a child's BMI is above or below the average BMI value for their age group and sex. For instance, a z score of 1.5 indicates that a child is 1.5 standard deviations above the average value, and a z score of ‐1.5 indicates a child is 1.5 standard deviations below the average value" (NOO NHS 2011). | ||||||
| Diet, physical activity, and behavioural interventions for the treatment of overweight or obesity in preschool children aged 0 to 6 years | ||||||
| Patient or population: preschool children (aged 0 to 6 years) with overweight or obesity Settings: obesity research clinic Intervention: dietary interventions + healthy lifestyle education Comparison: healthy lifestyle education | ||||||
| Outcomes | Healthy lifestyle education | Dietary intervention + healthy lifestyle education | Relative effect | No of participants | Quality of the evidence | Comments |
| Changes in BMI and body weight 1. Dairy‐rich diet a. BMI z score [units]a Follow‐up: 6 months b. BMI z score [units] Follow‐up: 36 months 2. Energy‐restricted diet a. BMI z score [units] Follow‐up: 6 months b. BMI z score [kg/m²] Follow‐up: 36 months | 1. Dairy‐rich diet a. The mean change in BMI z score was ‐0.5 units in the control group b. The mean change in BMI z score was +0.6 units in the control group 2. Energy‐restricted diet a. The mean change in BMI z score was ‐0.5 units in the control group b. The mean change in BMI z score was +0.6 units in the control group | 1. Dairy‐rich diet a. The mean change in BMI z score in the intervention group was b. The mean change in BMI z score in the intervention group was 2. Energy‐restricted diet a. The mean change in BMI z score in the intervention group was b. The mean change in BMI z score in the intervention group was | ‐ | 1. Dairy‐rich diet a. 59 (1) b. 52 (1) 2. Energy‐restricted diet a. 57 (1) b. 47 (1) | 1. Dairy‐rich diet a/b ⊕⊝⊝⊝ 2. Energy‐restricted diet a/b ⊕⊝⊝⊝ | Lower units indicate more weight loss 2 dietary interventions and 1 control compared in one 3‐arm randomised controlled trial (the number of participants in the control group was halved for the analysis and is shown here) |
| Adverse events | See comment | See comment | See comment | See comment | See comment | Not reported |
| Health‐related quality of life and self esteem | See comment | See comment | See comment | See comment | See comment | Not reported |
| All‐cause mortality | See comment | See comment | See comment | See comment | See comment | Not reported |
| Morbidity | See comment | See comment | See comment | See comment | See comment | Not reported |
| Parent‐child relationship or assessment of parenting | See comment | See comment | See comment | See comment | See comment | No trials reported parent‐child relationship or assessment of parenting |
| Socioeconomic effects | See comment | See comment | See comment | See comment | See comment | Not reported |
| *The basis for the assumed risk (e.g. the median control group risk across trials) 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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| a"A BMI z score or standard deviation score indicates how many units (of the standard deviation) a child's BMI is above or below the average BMI value for their age group and sex. For instance, a z score of 1.5 indicates that a child is 1.5 standard deviations above the average value, and a z score of ‐1.5 indicates a child is 1.5 standard deviations below the average value" (NOO NHS 2011). bDowngraded by three levels because of reporting bias, indirectness, and imprecision (one trial only with small number of participants); see Appendix 9. | ||||||
| Intervention(s) and comparator(s) | Sample sizea | Screened/eligible | Randomised | ITT | Analysed | Finishing study | Randomised finishing study | Follow‐up timeb | |
| Stark 2014 | I1: LAUNCH home visits | "Because of the small sample size, we were not powered to compare the two LAUNCH groups" | ‐/237 | 15 | 10 | 10 | 7 | 47 | 6 months/12 months |
| I2: LAUNCH clinic visits | 14 | 11 | 11 | 10 | 71 | ||||
| C: enhanced usual care | 13 | 12 | 12 | 11 | 85 | ||||
| total: | 42 | 33 | 33 | 28 | 67 | ||||
| Quattrin 2012 | I: family‐based intervention | In a mixed‐effect model a sample of 108 participants | 171/147 | 52 | 46 | 46 | 30 | 58 | 12 months/24 months |
| C: information control | 53 | 50 | 50 | 40 | 76 | ||||
| total: | 105 | 96 | 96 | 70 | 73 | ||||
| Bocca 2012 | I: multidisciplinary programme | ‐ | 78/75 | 40 | ‐ | 17 | 17 | 43 | 16 weeks/3 years |
| C: usual care | 35 | ‐ | 12 | 12 | 34 | ||||
| total: | 75 | ‐ | 29 | 29 | 39 | ||||
| Taveras 2011 | I: High Five for Kids: behavioural intervention | ‐ | 1486 attempted contact/361 contacted, pre‐eligible, willing | 271 (5 clusters) | ‐ | 253 | 253 | 93 | 12 months (end of intense phase; 2‐year data not reported) |
| C: usual care | 1007 attempted contact/267 contacted, pre‐eligible, willing | 204 (5 clusters) | ‐ | 192 | 192 | 94 | |||
| total: | 475 | 445 | 445 | 94 | |||||
| Stark 2011 | I: LAUNCH | ‐ | 4079/56 | 8 | ‐ | 7 (month 6 and 12) | 7 | 88 | 6 months/12 months |
| C: enhanced standard care | 10 | ‐ | 10 (month 6) 9 (month 12) | 9 | 90 | ||||
| total: | 18 | ‐ | 16 | 16 | 89 | ||||
| Lanigan 2010 | I: Trim Tots multicomponent intervention | ‐ | 105/105 | 49 | 49 | 49 | 21 | 43 | 6 months/2 years (6‐month follow‐up reported only) |
| C: wait‐list control | 39 | 39 | 39 | 21 | 54 | ||||
| total: | 88 | 88 | 88 | 42 | 48 | ||||
| Kelishadi 2009 | I1: dairy‐rich diet | 30 per group | ‐ | 40 | ‐ | ‐ | 36 | 90 | 6 months/3 years |
| I2: energy‐restricted diet | 40 | ‐ | ‐ | 31 | 78 | ||||
| C: control | 40 | ‐ | ‐ | 32 | 80 | ||||
| total: | 120 | 99 | 83 | ||||||
| Grand total | All interventions | 529 | 412 | ||||||
| All comparators | 394 | 317 | |||||||
| All interventions and comparators | 923 | 729 | |||||||
| aAccording to power calculation in study publication or report. ‐ denotes not reported C: comparator; I: intervention; ITT: intention‐to‐treat; LAUNCH: Learning about Activity and Understanding Nutrition for Child Health; N/A: not applicable | |||||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Changes in BMI z score Show forest plot | 5 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 1.1 End of intervention (6‐12 months) | 4 | 210 | Mean Difference (IV, Random, 95% CI) | ‐0.26 [‐0.37, ‐0.16] |
| 1.2 12‐18 months follow‐up (6‐8 months post intervention) | 4 | 202 | Mean Difference (IV, Random, 95% CI) | ‐0.38 [‐0.58, ‐0.19] |
| 1.3 24 months follow‐up (12 months post intervention) | 1 | 96 | Mean Difference (IV, Random, 95% CI) | ‐0.25 [‐0.40, ‐0.10] |
| 2 Changes in BMI Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 2.1 End of intervention (6‐12 months) | 1 | 64 | Mean Difference (IV, Random, 95% CI) | ‐0.40 [‐0.85, 0.05] |
| 2.2 12 months follow‐up (8 months post intervention) | 1 | 57 | Mean Difference (IV, Random, 95% CI) | ‐1.0 [‐1.79, ‐0.21] |
| 3 Changes in % over BMI Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 3.1 End of intervention (12 months) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 3.2 18 months follow‐up (6 months post intervention) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 3.3 24 months follow‐up (12 months post intervention) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 4 Changes in BMI percentile Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 4.1 End of intervention (6 months) | 2 | 50 | Mean Difference (IV, Random, 95% CI) | ‐1.54 [‐2.82, ‐0.26] |
| 4.2 12 months follow‐up (6 months post intervention) | 2 | 49 | Mean Difference (IV, Random, 95% CI) | ‐3.47 [‐5.11, ‐1.82] |
| 5 Changes in body weight Show forest plot | 5 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 5.1 End of intervention (6‐12 months) | 4 | 210 | Mean Difference (IV, Random, 95% CI) | ‐1.18 [‐1.91, ‐0.45] |
| 5.2 12‐18 months follow‐up (6‐8 months post intervention) | 4 | 202 | Mean Difference (IV, Random, 95% CI) | ‐2.81 [‐4.39, ‐1.22] |
| 5.3 24 months follow‐up (12 months intervention) | 1 | 96 | Mean Difference (IV, Random, 95% CI) | ‐1.60 [‐2.42, ‐0.78] |
| 6 Changes in parental BMI Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 6.1 End of intervention (6‐12 months) | 2 | 113 | Mean Difference (IV, Random, 95% CI) | 0.00 [‐2.52, ‐1.48] |
| 6.2 12‐18 months follow‐up (6 months post intervention) | 2 | 112 | Mean Difference (IV, Random, 95% CI) | ‐2.08 [‐2.65, ‐1.51] |
| 6.3 24 months follow‐up (12 months post intervention) | 1 | 96 | Mean Difference (IV, Random, 95% CI) | ‐2.0 [‐2.57, ‐1.43] |
| 7 Changes in parental weight Show forest plot | 3 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 7.1 End of intervention (6‐12 months) | 3 | 146 | Mean Difference (IV, Random, 95% CI) | ‐4.69 [‐7.27, ‐2.11] |
| 7.2 12‐18 months follow‐up (6 months post intervention) | 2 | 49 | Mean Difference (IV, Random, 95% CI) | ‐5.14 [‐8.96, ‐1.33] |
| 7.3 24 months follow‐up (12 months post intervention) | 1 | 96 | Mean Difference (IV, Random, 95% CI) | ‐6.7 [‐8.42, ‐4.98] |
| 8 Changes in health‐related quality of life: DUX 25 Show forest plot | Other data | No numeric data | ||
| 8.1 12 months follow‐up (8 months post intervention) | Other data | No numeric data | ||
| 9 Changes in health‐related quality of life: CHQ‐PF50 Show forest plot | Other data | No numeric data | ||
| 9.1 12 months follow‐up (8 months post intervention) | Other data | No numeric data | ||
| 10 Changes in health‐related quality of life: PEDsQL physical functioning Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 10.1 End of intervention (6 months) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 10.2 12 months follow‐up (6 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 11 Changes in health‐related quality of life: PEDsQL total score Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 11.1 End of intervention (6 months) | 1 | 28 | Mean Difference (IV, Random, 95% CI) | 4.35 [‐2.35, 11.06] |
| 11.2 12 months follow‐up (6 months post intervention) | 1 | 28 | Mean Difference (IV, Random, 95% CI) | 0.74 [‐5.80, 7.27] |
| 12 Changes in waist circumference Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 12.1 End of intervention (6 months) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 12.2 12 months follow‐up (8 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 13 Changes in waist circumference z‐score Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 13.1 12 months follow‐up (8 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 14 Changes in hip circumference Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 14.1 12 months follow‐up (8 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 15 Changes in hip circumference z‐score Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 15.1 12 months follow‐up (8 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 16 Changes in upper arm circumference Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 16.1 12 months follow‐up (8 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 17 Changes in per cent body fat Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 17.1 12 months follow‐up (8 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 18 Changes in fat‐free mass Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 18.1 12 months follow‐up (8 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 19 Changes in visceral fat Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 19.1 12 months follow‐up (8 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 20 Changes in subcutaneous fat Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 20.1 12 months follow‐up (8 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 21 Changes in outdoor active play Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 21.1 End of intervention (12 months) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 22 Changes in steps Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 22.1 12 months follow‐up (8 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 23 Changes in physical activity, moderate Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 23.1 At end of intervention (6 months) | 2 | 48 | Mean Difference (IV, Random, 95% CI) | 6.57 [‐0.47, 13.61] |
| 23.2 12 months follow‐up (6 months post intervention) | 2 | 46 | Mean Difference (IV, Random, 95% CI) | 10.14 [‐3.80, 24.08] |
| 24 Changes in physical activity, vigorous Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 24.1 End of intervention (6 months) | 2 | 48 | Mean Difference (IV, Random, 95% CI) | 2.78 [‐1.30, 6.85] |
| 24.2 12 months follow‐up (6 months post intervention) | 2 | 47 | Mean Difference (IV, Random, 95% CI) | 7.40 [2.81, 12.00] |
| 25 Changes in sugar‐sweetened drinks Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 25.1 End of intervention (12 months) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 26 Changes in fruit and vegetable intake Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 26.1 End of intervention (12 months) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| 27 Changes in TV and video viewing Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 27.1 End of intervention (12 months) | 1 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [0.0, 0.0] | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Changes in BMI z score Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.1 Dairy rich: end of intervention (6 months) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.2 Energy restricted: end of intervention (6 months) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.3 Dairy rich: 12 months follow‐up (6 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.4 Energy restricted: 12 months follow‐up (6 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.5 Dairy rich: 24 months follow‐up (18 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.6 Energy restricted: 24 months follow‐up (18 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.7 Dairy rich: 36 months follow‐up (30 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.8 Energy restricted: 36 months follow‐up (30 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2 Changes in waist circumference Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 2.1 Dairy rich: end of intervention (6 months) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.2 Energy restricted: end of intervention (6 months) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.3 Dairy rich: 12 months follow‐up (6 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.4 Energy restricted: 12 months follow‐up (6 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.5 Dairy rich: 24 months follow‐up (18 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.6 Energy restricted: 24 months follow‐up (18 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.7 Dairy rich: 36 months follow‐up (30 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2.8 Energy restricted: 36 months follow‐up (30 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 3 Changes in per cent body fat Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 3.1 Dairy rich: 12 months follow‐up (6 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 3.2 Energy restricted: 12 months follow‐up (6 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 3.3 Dairy rich: 24 months follow‐up (18 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 3.4 Energy restricted: 24 months follow‐up (18 months post intervention) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
