"In 2006, a study of nearly 3,000 nurses with breast cancer found that women without close friends were four times as likely to die from the disease as women with 10 or more friends."
The results cited in the article seem to only imply a correlation. It seems like the author of the article implied causation to add a bit of journalistic flair. After all, it would be a much less interesting article if it simply stated correlations.