Meta-analysis or Systematic Review

M - Click here for explanation of classification scheme.

POSITIVE: See Quality Criteria Checklist below.

To quantitatively assess the relation between fruit and vegetable intake and the risk of stroke by a meta-analysis of prospective cohort studies.

• Published as a full article in English
• Prospective cohort design
• Report relative risks of hazard ratios (HR) and 95% confidence intervals (CI) of stroke in relation to each category of fruit and vegetable intake
• Provide frequency of amount of fruit and vegetable consumption (in order to standardize classification of fruit and vegetable intake).

 

• Case-control study design
• Mixed healthy diet was reported, where the effect of fruits and vegetables could not be separated
• Only surrogate nutrients of fruits or vegetables were reported, whereas fruits or vegetables themselves were not reported
• Only two categories of fruit and vegetable intake were reported, which could not allow for adequate characterization of fruit and vegetable intake.

Recruitment (Search Strategy)

• Studies that reported the association between fruit and vegetable intake and coronary heart disease (CHD) were searched for in MEDLINE (1966-November 2005) and EMBASE (1980-November 2005) using MeSH terms and text words
• The Cochrane Library was also searched with the terms, 'fruit' and 'vegetables'
• The reference lists of review and original articles were searched for more studies.

Design

Meta-analysis.

Dietary Intake/Dietary Assessment Methodology

• Fruit and vegetable intake was standardized among studies and grouped into three categories (less than three servings per day, three to five servings per day, and more than five servings per day)
• The average serving size was 80g for fruits, 77g for vegetables, and 391g for the average serving per day of fruits and vegetables combined.

Blinding Used

Not applicable.

Intervention 

Not applicable.

Statistical Analysis

• Relative risks or hazard ratios were used as a measure of the relation between fruit and vegetable intake and stroke
• Relative risks and hazard ratios in each study were transformed by taking their natural logarithms
• Relative risks were combined if individual studies reported risks based on multiple outcomes (e.g., subtypes of stroke) or multiple exposures (e.g., fruit alone)
• By comparison with the lowest category of fruit and vegetable consumption, the pooled relative risks of stroke for the middle and highest categories was calculated using a random effects model due to the presence of significant heterogeneity among studies. 

Timing of Measurements

Not applicable.

Dependent Variables

Stroke.

Independent Variables

Fruit and vegetable intake in servings per day.

Control Variables

Potential confounders controlled for in the original studies.

• Initial N: Nine cohort studies
• Attrition (final N): Nine studies with a total of 257,551 men and women, 4,917 events, and a median of 13 years of follow-up
• Age range: 34-103 years, mean not reported
• Ethnicity: Not reported
• Other relevant demographics: Not reported
• Anthropometrics: Not reported
• Location: Studies conducted in the US, Japan and Europe.

 

 

  Key findings

• Compared with individuals who had lass than three servings per day of fruit and vegetables, the pooled relative risk of stroke was 0.89 (95% CI: 0.83-0.97; P=0.005) for those with three to five servings per day and 0.74 (0.69-0.79, P<0.0001) for those with more than five servings per day
• Compared with individuals who had less than three servings per day of fruits and vegetables, those with more than five servings per day had a significantly reduced risk of stroke, irrespective of sex, duration of follow-up, method of dietary assessment, dietary instrument administration, or stroke subtype.
   

This meta-analysis of prospective cohort studies demonstrates that increased consumption of fruit and vegetables from less than three to more than five serving per day is related to a 26% reduction in the risk of stroke, whereas increased intake to three to five servings per day is associated with an 11% reduction in the risk of stroke.

Study strengths:

• Heterogeneity among studies and publication bias was assessed
• Average serving size was standardized for exposure measurement across studies
• Included only prospective cohort designs
• Large sample size and long duration of follow-up
• Data were extracted by two independent people and resolved by discussion with a third
• Studies adjusted for major confounding factors.

Study limitations:

• Heterogeneity among studies was present
• Residual confounding may have been present (e.g., due to lifestyle factors)
• Measurement error in studies' dietary assessment may have been present (only three of the eight took account of changes in dietary intake over time)
• Biases may have existed due to misclassification of fruit and vegetable intake because there was study variation in dietary assessment method, the number of groups of fruit and vegetable consumption, and the reference category.