U-shaped curve
More than half of the sample (52.96%) showed no changes in Internet use from baseline during the study period, while one-fifth (20.54%) did show changes in use.
Investigators found a robust link between Internet usage and lower dementia risk (cause-specific hazard ratio, 0.57 [95% CI, 0.46-0.71]) – a finding that remained even after adjusting for self-selection into baseline usage (csHR, 0.54 [0.41-0.72]) and signs of cognitive decline at baseline (csHR, 0.62 [0.46-0.85]).
Each additional wave of regular Internet usage was associated with a 21% decrease in the risk of dementia (95% CI, 13%-29%), wherein additional regular periods were associated with reduced dementia risk (csHR, 0.80 [95% CI, 0.68-0.95]).
“The difference in risk between regular and nonregular users did not vary by educational attainment, race-ethnicity, sex, and generation,” the investigators note.
A U-shaped association was found between daily hours of online engagement, wherein the lowest risk was observed in those with 0.1-2 hours of usage (compared with 0 hours of usage). The risk increased in a “monotonic fashion” after 2 hours, with 6.1-8 hours of usage showing the highest risk.
This finding was not considered statistically significant, but the “consistent U-shaped trend offers a preliminary suggestion that excessive online engagement may have adverse cognitive effects on older adults,” the investigators note.
“Among older adults, regular Internet users may experience a lower risk of dementia compared to nonregular users, and longer periods of regular Internet usage in late adulthood may help reduce the risks of subsequent dementia incidence,” said Ms. Cho. “Nonetheless, using the Internet excessively daily may negatively affect the risk of dementia in older adults.”
Bidirectional relationship?
Commenting for this article, Claire Sexton, DPhil, Alzheimer’s Association senior director of scientific programs and outreach, noted that some risk factors for Alzheimer’s or other dementias can’t be changed, while others are modifiable, “either at a personal or a population level.”
She called the current research “important” because it “identifies a potentially modifiable factor that may influence dementia risk.”
However, cautioned Dr. Sexton, who was not involved with the study, the findings cannot establish cause and effect. In fact, the relationship may be bidirectional.
“It may be that regular Internet usage is associated with increased cognitive stimulation, and in turn reduced risk of dementia; or it may be that individuals with lower risk of dementia are more likely to engage in regular Internet usage,” she said. Thus, “interventional studies are able to shed more light on causation.”
The Health and Retirement Study is sponsored by the National Institute on Aging and is conducted by the University of Michigan, Ann Arbor. Ms. Cho, her coauthors, and Dr. Sexton have disclosed no relevant financial relationships.
A version of this article originally appeared on Medscape.com.