Chocolate and risk of chronic disease: a systematic review and dose-response meta-analysis.

The non-linear dose-response analyses showed a borderline significance for CHD ( p non-linearity p = 0.07, n = 8 studies) and stroke ( p non-linearity p = 0.06, n = 7 studies) and significance for HF ( p non-linearity p < 0.001, n = 5 studies) ( Fig. 1 ) .

also in 7,017 other papers

In the non-linear dose-response analysis, we found a significant non-linear pattern with a peak risk reduction at 12 g/d intake and a slight trend of increasing risk when amount consumed exceeds 35 g/d, which is confirmed by another meta-analysis [ 9 ].

also in 4,078 other papers