Changes over time in the effect of marital status on cancer survival

BACKGROUND: Rates of all-cause and cause-specific mortality are higher among unmarried than married individuals. Cancer survival is also poorer in the unmarried population. Recently, some studies have found that the excess all-cause mortality of the unmarried has increased over time, and the same pattern has been shown for some specific causes of death. The objective of this study was to investigate whether there has been a similar change over time in marital status differences in cancer survival. METHODS: Discrete-time hazard regression models for cancer deaths among more than 440 000 women and men diagnosed with cancer 1970-2007 at age 30-89 were estimated, using register data encompassing the entire Norwegian population. More than 200 000 cancer deaths during over 2 million person-years of exposure were analyzed. RESULTS: The excess mortality of the never-married compared to the married has increased steadily for men, in particular the elderly. Among elderly women, the excess mortality of the never-married compared to the married has increased, and there are indications of an increasing excess mortality of the widowed. The excess mortality of divorced men and women, however, has been stable. CONCLUSIONS: There is no obvious explanation for the increasing disadvantage among the never-married. It could be due to a relatively poorer general health at time of diagnosis, either because of a more protective effect of partnership in a society that may have become less cohesive or because of more positive selection into marriage. Alternatively, it could be related to increasing differentials with respect to treatment. Today's complex cancer therapy regimens may be more difficult for never-married to follow, and health care interventions directed and adapted more specifically to the broad subgroup of never-married patients might be warranted.