What it is
How many servings of nuts you eat in a typical week — a small handful counts as one serving. This is nuts, not peanut butter; one study on this page found peanut butter did not carry the same association even though peanuts themselves did.
How it is modeled (not scored)
Technical notes
The curve is Aune 2016: dose-response meta-analysis, 20 studies. Per 28 g/day (about a small handful) increase in nut intake, all-cause mortality RR 0.78 (0.72–0.84). Main source: Aune. Supporting: Chen 2017 (18 studies, 81,034 deaths, highest-vs-lowest RR 0.81 (0.78–0.84), risk levels off around 3 servings/week) / Grosso 2015 (a smaller, steeper pool of only 7 studies, one-serving/day RR 0.73 (0.60–0.88), and authors flag that nut eaters also tend to smoke less and eat more fruit and vegetables — confounding cannot be excluded) / Bao 2013 (Nurses' Health Study + Health Professionals Follow-up Study, 27,429 deaths; a monotonic frequency-response with no plateau in this cohort, HR 0.80 (0.73–0.86) at ≥7 servings/week, funded in part by a nut-industry group, disclosed by the authors). van den Brandt 2015 corroborates the plateau shape and is the source for the peanut-butter gap.
How to improve it
- Add a small daily handful of nuts (about 1 oz / 28 g)
- Choose whole nuts over peanut butter
- Swap nuts in for a less healthy daily snack
- Either tree nuts or peanuts were associated with lower risk — no single type is required