The grandmother hypothesis suggests that humans have "given up" their reproductive potential in later years in order to invest in the children they already have as well as their grandchildren. Naturally, this is an unconscious, biological adaptation that emerges over many generations and is not the result of individual decision-making. For such a hypothesis to be confirmed it would have to be demonstrated that children are significantly more likely to survive when a grandmother is present than when she isn't.
Dr. Lummaa has done just that in her study published in the journal Nature, demonstrating that children are 12% more likely to survive to adulthood when they have a grandmother's support than when they don't.
Let's hear it for grandmas! Link
The rest is just ice cream and dreams of Darth Vader costumes for Halloween.
Biologically speaking, just about any event that happens after procreation can't be explained by evolution. The genes have already been passed on. There's no genetic way to "filter out" that event if it results in, say, an increased mortality rate. Because of the limited lifespan, women tended not to be menopausal grandmothers. I find it doubtful that they contributed significantly to the success of their offspring to procreate.
I think this account is a better explaination as to why women live for so long after their reprodictive cycles have ceased, as most animals don't outlive their ability to produce offspring for long.