Your general point is reasonable enough, but this arithmetic per se is not really supportable. (a) There is no breakdown there of the number of people who ate > 5 vs. < 3 servings of vegetables (either among the general population or among heart disease sufferers), and (b) the risk number is about developing heart disease at all, not deaths (without even mentioning (c) these are cohort studies showing correlations, not proving causal relationships).
It is possible that e.g. serious heart-disease cases likely to cause death result from causes unrelated to vegetable consumption, while many non-vegetable-eaters develop minor heart disease cases which they would not have if they had eaten vegetables, but which still do not result in death.
Likewise, it is likely that of the 610,000 deaths, many (most?) of them were of people eating > 3 servings of vegetables per day.
Your general point is reasonable enough, but this arithmetic per se is not really supportable. (a) There is no breakdown there of the number of people who ate > 5 vs. < 3 servings of vegetables (either among the general population or among heart disease sufferers), and (b) the risk number is about developing heart disease at all, not deaths (without even mentioning (c) these are cohort studies showing correlations, not proving causal relationships).
It is possible that e.g. serious heart-disease cases likely to cause death result from causes unrelated to vegetable consumption, while many non-vegetable-eaters develop minor heart disease cases which they would not have if they had eaten vegetables, but which still do not result in death.
Likewise, it is likely that of the 610,000 deaths, many (most?) of them were of people eating > 3 servings of vegetables per day.