AI 'can spot cognitive decline linked to menopause'

News
Woman on jetty in cold season looking out at lake and surrounding trees
Silvia

An artificial intelligence model has been shown to identify women who are at risk of experiencing a decline in their cognitive function during menopause and who could benefit from additional support and treatment.

Some level of cognitive impairment or brain fog is common in women going through menopause, affecting up to two-thirds according to some studies, and can lead to problems with decision-making and learning new information, concentration problems, forgetfulness, and woolly thinking.

While this subjective cognitive decline (SCD) linked to menopause tends to be transient, it can often occur at an age when neurogenerative disorders like Alzheimer's disease may also emerge, so there is a benefit in having tools that can help determine the likely cause of symptoms.

Moreover, some studies have suggested that SCD may be associated with an increased risk of dementia, so early diagnosis could allow intervention to protect women's cognitive health.

A new study published (PDF) in the journal Menopause tested several machine learning algorithms for their ability to spot subjective cognitive decline (SCD) in 1,264 nurses, a profession that often experiences high occupational stress and is considered to be prone to SCD during the menopause transition.

Using the group, the Chinese researchers were able to develop and validate a machine-learning model, called SVM, that was effective for identifying women experiencing severe SCD and also pointed to new insights into the cognitive decline and its underlying mechanisms.

The model identified more than a dozen factors that seem to have a bearing on risk, including socioeconomic status, age, major life events, chronic diseases, menopausal symptoms, and sleep quality, that could be used by the algorithm to determine risk.

The study also suggested that nurses with poorer economic conditions may be more susceptible to SCD because of "prolonged psychological stress, lack of social support, or increased work burdens."

Existing testing for cognitive performance is largely based on models typically incorporating various laboratory indicators such as blood glucose, blood lipids, and brain imaging. However, the complexity and high cost of these models can make them impractical to implement in regular clinical settings, according to The Menopause Society, which published the Menopause journal.

"Future studies involving objective measures of cognition and longitudinal follow-up are crucial to better understanding these associations," commented Dr Stephanie Faubion, medical director for the Society.

Image by Silvia from Pixabay