My mistake, "causation" was what I was trying to get at, not "correlation".

It appears that your closing examples of life expectancy during the great depression don't appear to show a correlation to food supply issues. 
 
Diet is mentioned as one of multiple factors that may have contributed to an increase in life span, but the University of Michigan researchers appear to concede that the actual explanations go beyond the scope of their study.