The inverse association between increased fruit and vegetable consumption and mortality was particularly noticeable for cardiovascular mortality; higher consumption was not significantly linked with reduced risk of cancer mortality
All of the pitfalls of the component studies are present in the meta-analysis, the authors warn, including the fact that fruit and vegetable intake is typically self-reported. Not all studies adjusted for other types of foods consumed, including saturated fats, processed meats, etc. The definition of “portions” may also have been different across studies, the authors caution.
Vegans and radical vegetarians take note. Bold in quote above is my emphasis.