That’s not how life expectancy (or any other statistic) works. If life expectancy reduces for the overall population by a few weeks, it doesn’t neatly follow that everyone within that population will simply die a few weeks earlier.
Some will die much earlier than expected, whilst others won’t be affected at all. The overall average might only be reduced by a few weeks, but for the individuals concerned the impact will be much more dramatic.
If you think about it, overall life expectancy at the moment is something like 79 for men and 83 for women, but we all know of people who died much, much younger than that. If you reduce overall life expectancy (even by a few weeks weeks), you shift the distribution of those deaths to the left and increase the number of deaths at ALL ages below the previous average.