Tau affects cortex, while amyloid affects the hippocampus?

That’s what this article by Wang et al suggests. Quantitatively, they found that:

  • In CSF tau-positive individuals (i.e. those with >= 500 pg/mL; ~ 1/4 of the total cohort), CSF tau correlates positively with AD-like cortical thinning, explaining 15% of the variance after covariate correction, under their model, whereas amyloid beta has no significant association.
  • In CSF amyloid-positive individuals (i.e. those with >= 500 pg/mL; ~ 1/2 of the total cohort), CSF amyloid levels correlates positively with hippocampal volume [1], explaining 7% of the total variance after covariate correction, under their model, whereas now tau has no significant association.

[1]: This direction makes sense: low amyloid CSF levels correlate with higher brain amyloid lesions.

Four thoughts on how amyloid beta mediates cognitive decline

A new paper addresses the question “how does the presence of amyloid beta (Ab) on neuroimaging correlate with memory deficits?” in a variety of interesting ways. Here are some assorted thoughts:

1) Their strongest direct association (in Tbl 2) is the -0.46 correlation between logical memory scores and Ab positivity on PET imaging in MCI patients. “Positive” status is based on a previously defined threshold basically meaning “lots of Ab.” This is much stronger than the -0.1 correlation between logical memory and Ab positivity on PET imaging in cognitively normal controls. Per their discussion, that Ab correlates well with memory deficits in MCI and not cognitively normal people has been found before, and is rightly part of the justification for why people say that Alzheimers pathology is not normal.

2) This is actually quite nitpicky, since it is not the central point of their article, but to me it’s worth pointing out that the line in their methods, “we examined whether the p value for the association between Ab and cognition changed from significant to nonsignificant when adjusting for GM volume or FDG-PET” is troubling, since differences in significance are not necessarily significant differences. But they address the same question in three ways, and this is just one of them.

3) I like their use of ridge regression in the case of correlated predictors. I became especially interested in how it is a special case of the more general Bayesian ridge regression method, and tried out an example of it, the code for which you can find here, if you’re interested in wasting 1-3 minutes of your life.

4) Stepping back, and although this point has been made before, the general proposition that Ab positivity on PET imaging has a moderate-to-strong correlation with worse memory in MCI populations is a non-trivial finding and is in clear support of the amyloid hypothesis. While it seems that tau correlates more strongly with memory deficits, this is still an important point for those who doubt the AH [1] to keep in mind.


Mattsson et al, 2015. Brain structure and function as mediators of the effects of amyloid on memory. Pubmed.


[1]: Then again, people may be right to doubt the AH. I’m not trying to use “doubter” in the pejorative sense here. But, I don’t immediately see an easier way to phrase this.

Tau imaging levels correlate with worse memory in living AD patients

Gabrielle Strobel at the inimitable AlzForum has a nice article about how tau imaging levels correlate more strongly with memory deficits than amyloid imaging levels in patients with AD, based on recent conference data.

In med school (read: First Aid for the USMLE S1, 2014), we also learned that tau levels correlate more strongly with memory deficits, but that was based on post-mortem neuropath, which is slightly more likely to be confounded.

I await the actual publications to get a sense of what the correlation numbers are. In postmortem neuropath exam, the correlations (r^2) are 0.74 for tau tangles and 0.50 for amyloid plaques with final MMSE, respectively.


How does occupation affect Alzheimers risk?

Here are three tidbits that I picked up about this on a very preliminary read-through:

  • Most studies find no large effect of occupation per se on AD risk, and most of the effect that is there appears to be mediated by education (see here)
  • Low job autonomy (which has been previously associated with higher levels of job stress) is associated with an increased risk of AD (n = 900 in Sweden; HR = 2.2; see here)
  • At least one case report suggests that increased exposure to aluminum in one’s occupation can predispose to an increased risk of AD. Larger studies would be needed to endow this finding with more confidence (see here)

Can non-human primates have Alzheimers?

Attention conservation notice: This is an important but hotly debated question and I’m not nearly well-versed enough in the field to provide an answer. Instead, I’ll summarize a 2012 review and a little bit of external data in table format.

Why does this matter? Well, if some non-human primates get Alzheimers dementia and others don’t, or, even better, if there were a spectrum of Alzheimers pathology and clinical symptoms among NHPs, then this could be correlated with genomic or transcriptomic data to get a better sense of the mechanism of the disease.

According to this 2012 review by Heuer et al, there are two major areas where research has been done across NHP species: histological and behavioral.

Histologically, the major areas are:

  • The presence of amyloid plaques (+/- soluble oligomers) that show up at ~ 80% (+/-, say, 20%) of the species’s maximum lifespan
  • The presence of abnormally phosphorylated tau
  • The presence of cerebral amyloid angiopathy
  • The presence of substantial neuron loss

Behaviorally, the major areas are:

  • Executive deficits
  • Memory deficits

This google spreadsheet summarizes the type of pathology present for each of the above, for each type of NHP species, where available.

A summary of the data is that great apes (i.e., the NHPs closest to humans) have executive but not memory deficits in aging, the latter of which is consistent with them having amyloid plaques but no major neuron loss.

If you have any suggestions, please make comments on the spreadsheet and I’ll incorporate them.


Heuer E, Rosen RF, Cintron A, Walker LC. Nonhuman primate models of Alzheimer-like cerebral proteopathy. Curr Pharm Des. 2012;18(8):1159-69.

Erwin J, Hof PR. Aging in Nonhuman Primates. Karger Medical and Scientific Publishers; 2002.

Darusman HS, Pandelaki J, Mulyadi R, et al. Poor memory performance in aged cynomolgus monkeys with hippocampal atrophy, depletion of amyloid beta 1-42 and accumulation of tau proteins in cerebrospinal fluid. In Vivo. 2014;28(2):173-84.

Is Alzheimers just normal aging? Part 1

A paper by Robinson et al provides a data point against, evaluating the post-mortem brains of around 150 people 90+ years old. They split their cohort three groups: those with dementia, those with cognitive impairment but no dementia, and those without cognitive impairment. Here is their main result table.

robinson et al

Thus, in their study, we can see that

a) there are pathological differences between dementia and other cognitive impairment, and

b) of the pathology types that they measured, there are no clear pathological differences between cognitive impairment and non-cognitive impairment.

To me this seems to be pretty good data emphasizing the standard theory that Alzheimers is distinct from normal aging. And that normal aging can also present with cognitive impairment, although these individuals can still convert to Alzheimers later in life.

Case studies of dementia reversal through combination therapy

Attention conservation notice: Discussion of mostly non-randomized studies based on small sample sizes. Also, I am not a doctor.

Over the past few years a number of case reports for lifestyle or nutritional interventions for decreasing the rate of decline or even reversing dementia, usually with a presumed clinical diagnosis of AD, have gotten some attention.

As a slight digression, the most famous of these is Mary Newport’s book propounding the idea that coconut oil reversed her husbands dementia.

As I noted earlier, coconut oil is the most commonly discussed treatment for Alzheimers in YouTube comments. This may be because it is actionable for YouTube commenters, whereas secretase inhibitors, outside of a clinical trial, are not. Also probably because the Mary Newport videos on YT are pretty popular.

The user neglect on the web forum sdn has expressed skepticism about coconut oil, noting that trials for fatty acids in AD have been negative outside of APOE subgroup analysis. Also, a 2008 study found that giving rats hydrogenated coconut oil led to worse memory performance, although that is probably not a particularly fair diet comparison to the one proposed in humans.

But that is not the point of this blog post.

The point of this blog post is to point out this interesting series of case reports by Dale Bredesen, which found that optimizing a wide breadth of patient characteristics through lifestyle, nutritional, and pharmacological means led to at least some improvement in 9/10 of the patients treated. See below for a summary of the initial characteristics of the patients and the results of the treatments.

case study results

Notably, for some of the patients, increasing medium chain triglycerides via coconut oil or a food containing coconut oil was one of the treatment modalities.

As Bredesen notes, the majority of success stories in diseases like HIV and CVD have been due to combination therapy, and such an approach may also lead to better outcomes in AD. This might make it harder for researchers to tease out exactly which treatment led to the improvement, but that’s why we have statistics.

There’s no reason, theoretically, that a clinical trial like this couldn’t be personalized, as long as the algorithm by which patient characteristics are converted into treatment decisions is codified prior to beginning the trial.

Thoughts on APP Expression in AD

It’s well known that Down syndrome patients have a greatly increased incidence of early onset Alzheimers (around 75% by year 65), and this is usually attributed to the presence of the gene APP on chromosome 21, leading to its triplication and increased levels.*

If an increase in APP expression really is the cause of the early onset Alzheimers in Down syndrome patients, then we would also expect the following:

1) Triplications of APP should also lead to Alzheimers. Importantly, we do see this. Eg, see this paper reporting 19 such cases. This is strong evidence that APP is indeed the causative gene of the early dementia in Down syndrome patients. Caveat: based on this paper’s data, it seems possible, though pretty unlikely, that the causative gene is instead one of the others in the region that is also duplicated, such as GABPA. From their Figure 1C, a picture is worth a few hundo words:

Also, not a caveat, but to me a worthy question: why does Down syndrome associated early-onset AD start at ~ year 40 and the early-onset AD in these patients start at ~ year 52?

2) We’d expect to see cis-eQTLs near APP that likely regulate levels of APP expression and increase it in Alzheimers association studies. We do not see this; eg, see this paper, which does not mention APP in the main text nor mention it in the supplementary tables I checked. However, who knows: maybe there just aren’t any SNPs in the regions of the DNA that happen to control APP expression? This is only very weak, circumstantial evidence against APP being the causative gene.

3) Now, would we expect to see increased APP mRNA expression in Alzheimers patients vs controls? I think basically yes. And indeed, APP expression is increased in Down syndrome patients, by about 1.5-fold, which is around exactly what you’d expect based on the triplication. What makes this confusing is that its levels in Alzheimers are increased in some studies, such as this one, but not mentioned as upregulated in others. Confusing.

* = Interestingly, according to this article, nearly all Down syndrome patients have amyloid plaques in their brains by age 40, but by age 65, 25% still do not have Alzheimers dementia.

What words do commenters on YouTube use to discuss Alzheimers?

Using a hacky YouTube API call and nltk I made a quick chart of the top 30 non-stopwords found in comments for videos about “Alzheimers.”

Screen shot 2014-07-27 at 8.20.08 PM

Here are two aspects that jump out to me:

– “Years” is the third most common word used. That makes sense for two reasons: a) according to a Mayo clinic summary, the life expectancy of a patient diagnosed with Alzheimers is 8-10 years, so people often live with the disorder for many years, and b) many of the users were likely discussing age at diagnosis, since Alzheimers is associated with aging.

– The word “oil” likely refers to coconut oil, which is often discussed as a treatment for Alzheimers. This is the only specific treatment mentioned in the top 30 words. According to alz.org, coconut oil has never been tested in clinical trials.