Types of studies

Eligible studies were RCTs. We included studies reported as full text, those published as abstract only, and unpublished data.

Types of participants

Adults (age 18 and over) who are at high risk of CVD and adults from the general population in primary prevention trials were eligible. We excluded participants with the following co‐morbidities/characteristics:

• those who have experienced a previous myocardial infarction (MI) or stroke, or both;

• those who have undergone a revascularisation procedure (coronary artery bypass grafting (CABG) or percutaneous transluminal coronary angioplasty (PTCA));

• those with angina or angiographically‐defined coronary heart disease (CHD);

• those with type 2 diabetes, although this is a major risk factor for CVD, as interventions for the treatment and management of type 2 diabetes are covered by reviews registered with the Cochrane Metabolic and Endocrine Disorders Group.

Types of interventions

We included trials comparing dietary fibre with no intervention or minimal intervention (for example leaflets with no person‐to‐person intervention or reinforcement). The intervention was in the form of advice to increase consumption or the provision of fibre supplements or high‐fibre foods. Where we found a sufficient number of trials, we stratified results by the type of fibre (soluble or insoluble), dose of supplementation, duration of intervention, and type of intervention (advice, diet, or supplementation).

We excluded multi‐factorial lifestyle intervention trials and trials focused on weight loss in order to avoid confounding. We also focused on follow‐up periods of 12 weeks (or three months) or more, as longer follow‐up periods are more relevant for public health interventions.

Types of outcome measures

Electronic searches

We identified trials through systematic searches of the following bibliographic databases:

• The Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library Issue 12, 2014)

• The Database of Abstracts of Reviews of Effects (DARE) (The Cochrane Library Issue 4, 2014)

• The NHS Economic Evaluation Database (NEED) (The Cochrane Library Issue 4, 2014)

• The Health Technology Assessment (HTA) Database (The Cochrane Library Issue 4, 2014)

• Ovid MEDLINE (1946 to January Week 1 2015)

• Ovid EMBASE and EMBASE Classic (1947 to 12 January 2015)

• Science Citation Index Expanded, Social Sciences Citation Index, and Conference Proceedings Citation Index ‐ Science on Web of Science Core Collection (Thomson Reuters) (1970 to 12 January 2015)

We adapted the preliminary search strategy for MEDLINE (Ovid) (Appendix 1) for use in the other databases. We applied the Cochrane sensitivity‐maximising RCT filter (Lefebvre 2011) to MEDLINE (Ovid) and adaptations of it to the other databases, except CENTRAL, DARE, NEED and HTA.

We also conducted a search of ClinicalTrials.gov (www.ClinicalTrials.gov) and the World Health Organization International Clinical Trials Registry Platform Search Portal (http://apps.who.int/trialsearch/).

We searched all databases from their inception to the present, and we imposed no restriction on language of publication.

Searching other resources

We checked reference lists of all primary studies and review articles for additional references. We also, where necessary, contacted authors for additional information.

Selection of studies

Two authors (LH and MM) independently screened titles and abstracts for inclusion and coded them as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve'. We retrieved the full‐text study reports/publications, and two authors (LH and MM) independently screened the full text to identify studies for inclusion, and identify and record reasons for exclusion of the ineligible studies. We resolved any disagreement through discussion or, where required, we consulted a third author (KR). We identified and excluded duplicates and collated multiple reports of the same study so that each study, rather than each report, was the unit of interest in the review. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram (Figure 1) and Characteristics of excluded studies table (Moher 2009).

Figure 1

---

---

Study flow diagram.

Data extraction and management

We used a piloted data collection form for study characteristics and outcome data. Two authors (LH, MM or JC) extracted study characteristics from included studies. We extracted the following study characteristics.

• Methods: study design, total duration of study, details of any 'run in' period, number of study centres and location, study setting, withdrawals, and date of study.

• Participants: number (N), mean age, age range, gender, severity of condition, diagnostic criteria, inclusion criteria, and exclusion criteria.

• Interventions: intervention, comparison, concomitant medications, and excluded medications.

• Outcomes: primary and secondary outcomes specified and collected, and time points reported.

Two authors (LH, MM or JC) independently extracted outcome data from included studies. A third author (EL) resolved disagreements between the two reviewers. One author (EL) transferred data into a Review Manager (RevMan 2014) file. We double‐checked that data was entered correctly by comparing the data presented in the systematic review with the study reports. A second author (KR) spot‐checked study characteristics for accuracy against the trial report.

Assessment of risk of bias in included studies

Two authors (LH, MM or JC) independently assessed risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). We resolved any disagreements by discussion or by involving another author (EL or KR). We assessed the risk of bias according to the following domains.

• Random sequence generation

• Allocation concealment

• Blinding of participants and personnel

• Blinding of outcome assessment

• Incomplete outcome data

• Selective outcome reporting

• Other bias (e.g. industry funding)

We graded each potential source of bias as high, low, or unclear and provide a quote from the study report together with a justification for our judgement in the Risk of bias section as part of the Characteristics of included studies table. We summarised the risk‐of‐bias judgements across different studies for each of the domains listed. Where information on risk of bias relates to unpublished data or correspondence with a trialist, we noted this in the Risk of bias section.

When considering treatment effects, we took into account the risk of bias for the studies that contribute to that outcome.

Measures of treatment effect

We analysed dichotomous data as odds ratios (OR) or risk ratios (RR) with 95% confidence intervals (CI) and continuous data as mean difference (MD) or standardised mean difference (SMD) with 95% CI. For continuous variables we presented data for the change from baseline rather than end‐point data. We entered data presented as a scale with a consistent direction of effect, with the exception of HDL cholesterol where an increase in this outcome is a positive finding.

We narratively described skewed data reported as medians and interquartile ranges.

Unit of analysis issues

Dealing with missing data

We contacted investigators or study sponsors to verify key study characteristics and obtain missing numerical outcome data where possible (for example when a study is identified as abstract only). 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 (16.1.3.2 Imputing standard deviations for changes from baseline Higgins 2011b), and assumed a correlation of 0.5 between baseline and follow‐up measures as suggested by Follman 1992.

Assessment of heterogeneity

We used the I² statistic to measure heterogeneity among the trials in each analysis (Higgins 2003). Where we identified substantial heterogeneity (greater than 50%), we reported it and explored possible causes by pre‐specified subgroup analysis. 

Assessment of reporting biases

Where we were able to pool more than 10 trials, we created and examined a funnel plot to explore possible small‐study biases for the primary outcomes (Sterne 2011).

Data synthesis

We undertook meta‐analyses only where this was meaningful, that is if the treatments, participants, and the underlying clinical question were similar enough for pooling to make sense. Where there was no heterogeneity between included studies, we performed a fixed‐effect meta‐analysis. Where we detected substantial heterogeneity (I² greater than 50%) and could not explain it, we considered the following options: providing a narrative overview and not aggregating studies, or using a random‐effects model with appropriately cautious interpretation.

Subgroup analysis and investigation of heterogeneity

We planned to carry out the following subgroup analyses.

• Type of fibre (soluble and insoluble)

• Dose of supplement

• Duration of intervention

• Type of intervention (fibre supplementation, provision of high‐fibre foods, and advice to increase fibre consumption)

However, data were only available to undertake subgroup analyses on the type of fibre and the type of intervention.

We used the following outcomes in subgroup analyses.

• Changes in blood pressure (systolic and diastolic)

• Blood lipids (total cholesterol, HDL cholesterol, LDL cholesterol, triglycerides).

We used the formal test for subgroup interactions in Review Manager (RevMan 2014).

Sensitivity analysis

We planned to carry out sensitivity analyses looking at studies with a low risk of bias. However, no studies met this criteria (see Risk of bias in included studies). One study reported a high loss to follow up (> 20%) but this study did not report any data that could be included in the meta‐analysis and therefore no sensitivity analysis was required.