COPENHAGEN – Losing a spouse has been associated with the broken heart syndrome and other adverse health outcomes, but the loss seems to have a paradoxical effect on the development of dementia.
An analysis of a large national database showed that individuals who had a mild cognitive impairment (MCI) and lost a spouse developed dementia significantly later than did those who hadn’t lose a spouse.
"Initially, we said this can’t be right," Dr. Bryan K. Woodruff of the Mayo Clinic in Scottsdale, Ariz., said in an interview at the annual meeting of the Alzheimer’s Association International Conference.
Dr. Bryan K. Woodruff
"But what we think is happening is when somebody loses their spouse and they’ve got MCI, that individual isn’t going to continue to live independently, so we think extended family is coming in and recognizing that this person needs more help now that the spouse is gone, getting them into assisted living or other arrangements. And the married couple continues to plug along without getting extended support," said Dr. Woodruff.
The finding hints at the need for mobilizing more support for married couples, too, Dr. Woodruff said. "This needs to start at the MCI stage, to increase support for the MCI person and also the spouse."
Dr. Woodruff and his colleagues selected 3,783 married individuals with MCI using the National Alzheimer’s Coordinating Center (NACC) database from September 2005 to September 2013. The subjects were enrolled in a National Institute of Aging-Alzheimer’s Disease Center, and stayed in the study for up to 7.5 years. Those who were divorced or separated were excluded, and so were those without follow-up.
Of the remaining 2,457 individuals, 134 were widowed and the rest stayed married. A total of 1,078 developed dementia.
At first visit, the widowed individuals were significantly more often older, female, a carrier of the apolipoprotein E epsilon-4 (APOE epsilon-4) allele, and less educated.
Results showed that the median age of dementia onset in the widowed group was 92 years, compared with 83 years in the married group (hazard ratio, 0.36; 95% confidence interval, 0.26-0.48; P less than .001). Adjustment for sex, presence of an APOE epsilon-4 allele, and age did not substantially affect the hazard ratio calculated for the risk of dementia after losing a spouse, the authors wrote.
Aside from the possible effect of additional support after loss of a spouse, the authors also wondered if the findings could be the result of selection bias "if the excluded widowed subject with MCI developed dementia at a younger age."
Dr. Woodruff and his associates ran another analysis on the same database, focusing on individuals who had normal cognitive function. This time, both groups, married or widowed, developed dementia around the same age, showing that "loss of spouse does not predict earlier development of dementia," among those with normal cognition, the authors wrote in their poster.
In the second analysis, there were 6,088 individuals who were married and cognitively normal at entry. Those who were divorced or separated were excluded, and so were those with no follow-up.
Of the remaining 4,446 subjects, 398 were widowed, and the rest remained married. Overall, 218 individuals developed dementia.
Both groups developed dementia at 96 years of age. Adjustments for sex, age, and presence of APOE epsilon-4 allele did not increase the hazard ratio for loss of a spouse, according to the authors.
"The absence of a "widowhood effect" may reflect increased resilience in individuals who elect to participate in such research studies, or could be result of selection bias if excluded subjects developed dementia at a younger age," the authors wrote. "Alternatively, the impact of widowhood may lessen with age, noting the advanced age of this sample," they added.
A recent study also found that loss of a spouse didn’t increase the risk of dementia, "yet women demonstrate a seemingly temporary decline in executive function following the death of a partner" (Am. J. Epidemiol. 2014;179:674-83)
Dr. Woodruff and his colleagues are planning to conduct the studies in community settings to find out if the results can be replicated.
Dr. Woodruff had no financial disclosures.
On Twitter @naseemmiller