Well, with a 35-45 lb. annual weight gain, after 40 years you would literally weigh as much as a car. Obviously that doesn't happen with any regularity, so clearly an equilibrium is reached and the curve flattens off at some relatively modest point. Therefore, the heavier you are, the more food you need to maintain your weight. Asymptotically, calorie intake predicts your final weight, not weight gain.

Note that I meant that BMR can vary between 1100 kcal/day to 2500 kcal/day (so a variance of 1400 kcal/day between individuals), not that it can vary by 2500 kcal/day.

Here are two studies that support slightly smaller numbers than I claimed, but still significantly larger than 300-400.

[0] studied the factors affecting BMR in 150 Scottish adults. It found a variance between ~1100 kcal/day to ~2100 kcal/day between its subjects (for the record, 60% of the variance was attributable to lean body mass).

[1] is a meta-study that looked at RMR (expressed as kcal/kg*h) for adults at various ages. The largest variance from Figure 1 is, assuming all subjects weighed 100kg, the highest RMR would be ~2500 kcal/day, while the lowest would be 1600 kcal/day. Given the varying weights, the actual variance recorded could be bigger or smaller.

In addition to this, even people with the same BMR/RMR eating the same food will not gain the same weight - they may differ in the amount of exercise, or in their calorie calorie consumption while exercising.

[0] https://academic.oup.com/ajcn/article/82/5/941/4607670 [1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4535334/