10.10.2015 Views

Carbohydrates and Health

1OqQDo3

1OqQDo3

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

• Limited – three to four cohort studies were identified <strong>and</strong> there was some<br />

indication that the results were in the same direction but no meta-analysis was<br />

performed. However, the evidence may also have been considered limited if<br />

there were a number of studies, but the biological plausibility was unclear or<br />

the methodology was not precise or the exposures were heterogeneous.<br />

R<strong>and</strong>omised controlled trials<br />

A2.19 Generally, there needed to be three or more r<strong>and</strong>omised controlled trials to<br />

determine if there was evidence of an effect or not. Expert judgement was used<br />

to determine if the quality of these data permitted a conclusion to be drawn.<br />

Where an effect appeared to be present, the evidence was graded as adequate,<br />

moderate or limited. The number of studies was used as an approximate guide<br />

only, since issues surrounding study design, study size, methodology <strong>and</strong> definition<br />

of exposures affected the final grading of the evidence.<br />

A2.20 Evidence was graded as follows:<br />

• Adequate – a meta-analysis was performed on three or more r<strong>and</strong>omised<br />

controlled trials. Alternatively, a total of four or more studies were identified in<br />

the systematic review <strong>and</strong>/or the update search; these consistently showed the<br />

same outcome although a meta-analysis was not performed.<br />

• Moderate – a meta-analysis was performed on three or more r<strong>and</strong>omised<br />

controlled trials. Alternatively, three or more r<strong>and</strong>omised controlled trials were<br />

identified in the systematic review <strong>and</strong>/or the update search, which consistently<br />

showed the same outcome although a meta-analysis was not performed.<br />

• Limited – three to four r<strong>and</strong>omised controlled trials were identified <strong>and</strong> there<br />

was some indication that the results were in the same direction, but no metaanalysis<br />

was performed. However, the evidence may also have been considered<br />

limited if there were a number of studies, but the biological plausibility was<br />

unclear, the methodology was not precise or the exposures were heterogeneous.<br />

A2.21 The evidence was normally considered as insufficient where there were fewer than<br />

three studies of the same type. Alternatively, if there were more studies but they<br />

were of poor quality, the evidence may have also been considered as insufficient.<br />

A2.22 Evidence was deemed inconsistent according to statistical considerations i.e.<br />

in a meta-analysis, when I 2 >75%, the confidence intervals did not overlap or if<br />

the results of individual studies were not in the same direction. When the I 2 was<br />

greater than 75%, but the forest plot suggested there was evidence of a direction<br />

for an outcome, expert judgement was used to upgrade the conclusion, where<br />

appropriate.<br />

A2.23 In addition to the evidence grading, if an effect or association was shown,<br />

statements have been provided on the following:<br />

• Whether the effect or association indicated beneficial or adverse effects with<br />

higher intakes of carbohydrate or carbohydrate components.<br />

206

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!