The apolipoprotein E4 (ApoE4) allele, the strongest known genetic risk factor for late-onset sporadic Alzheimer’s disease, puts women at higher risk for the disease than men, but how remains unclear. Now, a large meta-analysis of genetic, biomarker, and autopsy data from 10 longitudinal cohorts indicates that ApoE4 raises CSF tau in women more than in men, particularly in people with amyloid. Other markers of AD pathology—CSF Aβ, or plaques and tangles—showed no such effect of sex in E4 carriers. The study, led by Timothy Hohman, Vanderbilt University in Nashville, Tennessee, appeared in the May 7 JAMA Neurology.
While the results suggest that ApoE4 promotes plaques and tangles equally in women and men, it hints that women carriers are at higher risk for neurodegeneration downstream since elevated CSF tau signals ongoing neuronal loss.
“[This study] is not the first to document a link among sex, APOE4, and tau, but its combined analyses of several data sets that include broad ranges of age and diagnoses and both biomarker and autopsy measures is the most complete investigation of the issue to date,” wrote Christian Pike, University of Southern California, Los Angeles, in an email to Alzforum.
For reasons that are not clear, more women than men with E4 get clinical Alzheimer’s disease (Farrer et al., 1997). This dichotomy varies with age, reaching its peak between the ages of 55 and 75, and then declining (Sep 2017 news).
What could explain the sex difference? It’s known that ApoE4 drives earlier and higher amyloid deposition in carriers, possibly by blocking amyloid clearance from the brain (Apr 2013 news). ApoE4 also affects tau: Carriers have more tau in the CSF and more neurofibrillary tangles in the brain, and animal work suggests E4 accelerates tau deposition and toxicity (Sep 2017 news). However, not much work has addressed how sex factors in, and results have been mixed.
To remedy that, Hohman and colleagues analyzed data from the 10 studies. To determine sex effects on CSF biomarkers, the investigators pooled data from four longitudinal aging studies that measured CSF Aβ42, total tau, and phospho-tau in a total of 1,792 participants in the Alzheimer’s Disease Neuroimaging Initiative (ADNI), and three smaller studies—the Biomarkers of Cognitive Decline among Normal Individuals (BIOCARD), the Vanderbilt Memory and Aging Project (VMAP), and the Wisconsin Registry of Alzheimer’s Prevention (WRAP). The average age of participants was 70, half were women, and between 34 and 46 percent of each cohort carried at least one ApoE4 allele. Of the nearly 1,800 subjects, 226 had a clinical diagnosis of AD, and the rest had mild cognitive impairment or were cognitively normal. To evaluate sex effects on brain pathology, the researchers also aggregated postmortem neuropathologic data on 5,109 subjects from six studies into an autopsy cohort. The average age was 84, with an equal split between men and women, and approximately 40 percent carried at least one ApoE4 allele.
As expected, Apoe4 status correlated with plaque and tangle pathology in the group as a whole, whether measured by biomarkers or postmortem analysis. ApoE4 carriers had higher CSF total tau (t-tau) and phosphorylated tau (p-tau) and lower CSF Aβ42 than noncarriers. Carrier status was associated with plaque and tangle positivity, based on the Consortium to Establish a Registry for Alzheimer's Disease (CERAD) neurotic plaque score, and Braak staging, respectively.
However, when the investigators broke out the results by sex, they found stronger effects of ApoE4 on the elevation of CSF tau in women compared to men. The association depended on amyloidosis being established, because when the investigators separately analyzed amyloid-positive and -negative women, they only saw it in the former. No interaction of sex and ApoE was seen for CSF Aβ42.
The results support previous results from the ADNI cohort on CSF t-tau (Jun 2012 news, Altmann et al., 2014), and extend the sex difference to p-tau. That they saw similar results in three additional cohorts, with widely varying recruitment criteria, length of follow-up, age and number of people with AD, lend credence to the findings, Hohman told Alzforum. “The four data sets are different, yet the results consistently show the interaction of ApoE4 and sex on tau. That makes for a more convincing association,” he said.
The findings are consistent with previous work showing that female ApoE4 carriers have more hippocampal atrophy and cognitive decline at lower amyloid loads compared to men (Koran et al., 2017). Why so remains to be seen, but estrogen boosts ApoE gene expression and release from microglia, and its loss in post-menopausal E4 carriers could leave them with a greater deficit of “good” ApoE than women with ApoE2/3 alleles.
In contrast to the CSF results, the researchers found no sex difference in either neurofibrillary tangle or amyloid load in ApoE4 carriers in the autopsy samples. This might be because these subjects were much older. However, the investigators saw no sex effect even among those younger than 75. The results further support the idea that the sex-specific increase in CSF tau levels signals an excess of neurodegeneration due to ApoE4, rather than higher cortical tau or amyloid deposition. Alternatively, the Braak staging scheme might be too crude to pick up subtle, sex-specific ApoE effects, Hohman said. He suggested more comprehensive, continuous measures of tauopathy will be needed to explore that idea.
Comorbidities might explain the results as well. In epidemiological studies, researchers can't account for men who have already died from cardiovascular disease, but who would also have been a high risk for dementia had they lived long enough. Because ApoE4 affects cardiovascular risk, this survival effect could weigh on men and women differently, creating an apparent sex difference in dementia risk. Hohman is now using mortality data to better estimate and account for survivor effects in his statistical models.
Finally, in a subtle twist, Hohman assumed going into the study he would see sex differences in the ApoE4 effect on CSF Aβ42, but he did not. Nonetheless, the study does not rule out sex-specific effects on amyloid, noted Michael Greicius, Stanford University, Palo Alto, California. “This is a nice study and very well done, but a limitation of healthy aging studies of people in their 60s and 70s is that some of the ApoE-related amyloid effects might already be in play ahead of that,” he told Alzforum. “It’s not to say there aren’t amyloid-independent effects of ApoE—that’s an exciting and testable hypothesis—but we can’t exclude the possibility that the effect they see is the result of amyloid overproduction and misprocessing that starts five years earlier in women E4 carriers compared to men,” Greicius said. To sort that out will take longitudinal studies beginning earlier than even ADNI, which enrolls people 55 and over. And while a few groups are starting to study younger people, as yet nobody has enough subjects for a statistical analysis of people in the critical window between 45 and 65, Greicius said. —Pat McCaffrey