BACKGROUND: Evidence that long-term cancer survivors (LTCS) experience greater cognitive dysfunction than individuals without cancer is mixed. This analysis of a population-based sample of United States residents compares self-reported cognitive difficulties in LTCS and cancer naive controls (CNC), controlling for psychiatric disorders.

METHODS: National Comorbidity Survey-Replication (NCS-R) interviews were conducted in a nationally-representative sample of 9282 people, of whom 5692 were assessed for cancer history. Long-term survivors of adult cancers were defined as individuals who were at least 18 years old at time of diagnosis; greater than 5 years following diagnosis; and with cancer reportedly in remission or cured. Cognitive dysfunction queries included two screening questions and five items drawn than the World Health Organization (WHO) Disability Assessment Schedule 2.0. Psychiatric disorders were identified using the WHO's Composite International Diagnostic Interview. Odds ratios and 95% confidence intervals were obtained from multivariable logistic regression models fit to evaluate the relationship between cancer status (LTCS versus CNC) and cognitive symptoms.

RESULTS: Of NCS-R participants, 225 met criteria for LTCS and 3953 for CNC. Long-term cancer survivors and CNC screened positive for cognitive symptoms at a rate of 19.5% and 20.8%, respectively, for the first screen and 16.0% and 16.4%, respectively, for the second. Adjusting for demographic and psychiatric variables, LTCS did not carry increased odds of experiencing cognitive symptoms [(OR) 1.00 (95% CI, .59 to 1.68); (OR) .95 (95 %CI, .57 to 1.59)].

CONCLUSIONS: Self-reported cognitive symptoms are common among LTCS and CNC. LTCS do not report cognitive symptoms with greater frequency.