The Unz Review: An Alternative Media Selection
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
Email This Page to Someone

 Remember My Information

Authors Filter?
Anatoly Karlin Razib Khan
Nothing found
 TeasersGene Expression Blog

Bookmark Toggle AllToCAdd to LibraryRemove from Library • BShow CommentNext New CommentNext New ReplyRead More
ReplyAgree/Disagree/Etc. More... This Commenter This Thread Hide Thread Display All Comments
These buttons register your public Agreement, Disagreement, Thanks, LOL, or Troll with the selected comment. They are ONLY available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also ONLY be used three times during any eight hour period.
Ignore Commenter Follow Commenter
🔊 Listen RSS

Earlier this week I pointed to the controversy which has erupted around the widely reported new paper, Genetic Signatures of Exceptional Longevity in Humans. Newsweek did the most thorough early reporting, but now The New York Times has published a follow up story covering the scientific criticisms to the original paper’s methodology. There’s nothing new in The Times‘ piece as such, but it shows that concerted scientific objection to the reception or interpretation of a particular finding which is widely disseminated in the media can yield results. Too often the mainstream media ends up serving as a glorified press release service, but in this case scientists are making their voices heard, and the media narrative is adjusting to the underlying discussion in the scientific community.

I’ve been told there may be more coming out which may shed light on this controversy next week. Stay tuned…

• Category: Science • Tags: Biology, Genetics, Longevity 
🔊 Listen RSS

When I first heard in the media there was a new study of longevity which had produced a model based on your SNP profile that was “77% accurate” as to whether you’d live to the age of 100 or not, I assumed this was confusion or distortion (perhaps The Daily Mail had broken embargo first and its spin was percolating around the mediasphere). But later I listened to one of the researchers on the radio, and though he seemed to want to tone down the certitude as to that prediction, he did not debunk the claim. Whatever the details, I did not believe that the model was that relevant to most people since very few are going to make it deep into their nineties in any case (I did have a grandfather who made it to 100 [in Bangladesh!], so my chance is presumably greater than the norm). The model would be moving you along the margins. Additionally, over the years it has paid off to be skeptical of the discovery of large effect genes for X, Y and Z. When the X, Y and Z has medical significance I’m even more skeptical, because the non-scientific biases within medical research seem to be really strong. There’s a lot of fame and money to be had. Some of the media were asking the researchers up front whether this might unlock the genetic “Fountain of Youth.” This is entrancing stuff.

So is this post from Dr. Daniel MacArthur, Serious flaws revealed in “longevity genes” study, warrants notice:

If the paper’s claims were true they would be truly remarkable. However, the general feeling from the GWAS community is that the identified associations are likely to be largely or even entirely artefactual, the result of failing to fully control for differences in the genotyping methods used in the cases and controls. The study used a mixture of two different genotyping platforms (albeit both made by Illumina) for their centenarians, while the control data was taken from an online database containing samples examined using multiple platforms. Disturbingly, similar potential genotyping bias also affects their replication cohort.

In the Newsweek piece I mentioned yesterday Kári Stefánsson has this to say about one of the platforms:

Kári Stefánsson, the Icelandic geneticist who founded deCode Genetics, knows something about the 610-Quad—his company has used it too. He says it has a strange and relevant quirk regarding two of the strongest variants linked to aging in the BU study, called rs1036819 and rs1455311. For any given gene, a person will have two “alleles,” or forms of DNA. In the vast majority of people, at the rs1036819 and rs1455311 locations in the genome, these pairs of alleles consist of one “minor” form and one “major” form. But the 610-Quad chip tends to see the wrong thing at those particular locations. It always identifies the “minor” form but not the “major” form, says Stefánsson—even if the latter really is present in the DNA, which it usually is. If you use the error-prone chip in more of your case group than your control group—as the BU researchers did—you’re going to see more errors in those cases. And because what you’re searching for is unusual patterns in your cases, you could very well mistake all those errors (i.e., false positives) for a genetic link that doesn’t actually exist.

Stefánsson says he is “convinced that the reported association between exceptional longevity and most of the 33” variants found in the Science study, including all the variants that other scientists hadn’t already found, “is due to genotyping problems.” He has one more piece of evidence. Given what he knows about the 610-Quad, he says he can reverse-engineer the math in the BU study and estimate what fraction of the centenarians were analyzed with that chip. His estimate is about 8 percent. The actual fraction, which wasn’t initially provided in the Science paper, is 10 percent, the BU researchers tell NEWSWEEK. That’s close, given that Stefánsson’s calculations look at just two of the variants found in the study and there may be similar problems with others.

Stefánsson recognizing one of the 150 SNPs as a problematic one is another red flag. The effect sizes of the SNPs in the study seem really large, so that should make you curious as to what’s going on. Here’s a post from 23andMe suggesting we should be cautious of the results for that reason:

-A large study combining results of four genome-wide association studies of longevity was published in May in the Journals of Gerontology. That study found no associations meeting their pre-specified criteria for genome-wide significance. While they used a more inclusive phenotype (age 90 or older), it is surprising that there could be so many loci associated with survival to age 100 in the new study, some with very large effect sizes, yet none were found in the larger study from earlier this year.

23andMe applied the model (the SNPs) outlined in the paper and attempted to see if it had any utility in to their admittedly small sample within their own database. They found nothing of note:

We took a preliminary look in our customer data to see if the proposed SNP-based model described in Sebastiani et al. is predictive of exceptional longevity. A commonly used measure of test discrimination is to calculate how often, for a randomly selected case and control, a test correctly assigns a higher score to the case. This is known as the “c statistic” or “area under the curve”. The authors of the new study say their model scored a 0.93 for this statistic. But when we compared 134 23andMe customers with age ≥ 95 to more than 50,000 controls, we obtained a test statistic of 0.532, with a 95% confidence interval from 0.485 to 0.579. Using 27 customers with age ≥ 100, we get a value of 0.540, with a 95% confidence interval from 0.434 to 0.645. A random predictor of longevity would give a 0.5 on this scale, so based on our data, performance of this model is not significantly better than random. Even with our small sample size, we can also clearly exclude values as high as the published result of 0.93.

If you go back to Dr. MacArthur’s post he has a chart which indicates that even by eyeballing their are indications that the results in the Science paper were artifacts of the methodological limitations. Newsweek ends with this caution:

Still, one has to wonder how the paper wound up in Science, which, along with Nature, is the top basic-science journal in the world. Most laypeople would never catch a possible technical glitch like this—who reads the methods sections of papers this complicated, much less the supplemental material, where a lot of the clues to this mystery were?—but Science’s reviewers should have. It’s clear that the journal—which hasn’t yet responded to the concerns raised here—was excited to publish the paper, because it held a press conference last week and sent a representative to say as much.

This isn’t about the media. They didn’t have to sensationalize too much; the findings themselves if correct are moderately sensational. But if Dr. Daniel MacArthur could spot something indicative of serious problems by scanning the supplements presumably it shouldn’t have made it through the review process without the issue being mooted and addressed. But then again, it’s medical genetics, and there’s a lot of pressure to find the roots of human morbidity and mortality. It’s a field where results like ALH 84001 abound. The heart wants what is wants. That’s why it’s nice to focus on less practical evolutionary genetic questions; no one really cares that much whether we’re descended from Neandertals. Right?

Note: And earlier post from Nature with more quotes from scientists who are skeptical of the findings. Also, after reading the posts I did read the original paper. Obviously I was cued to fixate on the particular issues highlighted above, but it is often rather illuminating to contrast the clear and spare summary presented to the public of findings to the numerous moving parts in the guts of the original paper.

• Category: Science • Tags: Genetics, Longevity, Medicine 
🔊 Listen RSS

My post below alluded to the fact that there seems to be a non-trivial between region difference in male life expectancy, even controlling for race, in the United States. From what I can tell Americans seem to have a somewhat schizophrenic attitude toward the reality of regionalism. On the one hand we are a relatively mobile people, and the original social-political aspect of states has been superseded by states as simply arbitrary sub-national units. And yet regional identities are still alive, most notably in the case of Southerners (with Texas as perhaps a special particular case even in the South, along with other areas such as Cajun Country). The differences are obvious in the case of accent and dialect, but one might think of these as simply indicators of a host of implicit underlying variables which are often imperceptible until one takes oneself “out of region.” In Albion’s Seed David Hackett Fisher explored the possible cultural roots of American regionalism as a work of history, while in The Nine Nations of North America Joel Garreau treated the subject in the manner of contemporary human geography.

These works paint with a broad brush, and explore the variation on a relatively coarse scale. Most Americans are aware of local religionalisms to a far greater level of detail, something which they are often not explicitly cognizant of. As a personal example I spent my adolescence in an area of the Intermontane West where both Mormons and “cowboys” were well represented. Though both groups were politically conservative, culturally there were stark differences which everyone was implicitly aware of. It was only later on that I learned that this region had experienced an influx of people from the Upper South in the 19th century, and later “Okies”, which was evident in the speech patterns of some individuals. On the other hand many of the Mormons had roots in Utah and eastern Idaho, and were cultural descendants of New England Yankees or later Northwest European converts who emigrated to Utah (the Mormon fixation on genealogy meant that if you had Mormon friends you would usually find out where their family was from through casual conversation since they knew). Last fall Steve Sailer pointed out that the counties where Barack Obama underperformed John Kerry, against the national trend, were those settled by and dominated by the Scots-Irish in the 18th century. Greg Cochran has told me that he was aware as a child the differences between Midwesterners whose origins were in the Upper South, and those who were Yankees.

Why does this matter? Because American public policy is often predicated on ceteris paribus assumptions once race and income are accounted for. Public policy prescriptions generated on the federal level will make the nod to race and class as interaction effects, but rarely allude to the possibility that white Americans even controlling for class may behave differently because of distinct cultural traditions. American regionalism is often conceived of as how you speak and what you eat, but I believe that these are simply the most obvious aspects of whole folkways, which are often assumptions and behaviors we take for granted.

But I come here not to talk, but to explore. The paper Eight Americas: Investigating Mortality Disparities across Races, Counties, and Race-Counties in the United States has the data for the white male longevity for each county in the United States. The Census has data on median household income, as well as the proportion of non-Hispanic whites in each county, or at least a subset. Unfortunately the tables I found had many counties missing for income and the proportion non-Hispanic white, so when I merged them with the one from the supplemental data from the paper above I was left with far fewer counties. I invite readers to point to better data sets in the comments than what I found poking through the Census website. There are certainly many likely variables which might explain longevity differences between regions, from climate to military service to participation in risky behaviors (the Mormon ban on alcohol probably means fewer men die of stupid acts at younger ages). But income is the primary predictor people think of, so it is what I focused on. Below are a set of charts and maps where I try and tease out regional variation. The x-axis is always median household income, while the y-axis is male life expectancy. Keep in mind that I filtered and constrained the data set in various ways when viewing the results, as my choices naturally have an effect. My point in presenting these results is to leverage reader knowledge about local variation. I am not interested in offering general explanations of why variation exists within the United States, rather, I am interested in outliers, and sharp local gradients. As the data was limited to counties which are at least 80% or more non-Hispanic white, there is a strong skew toward some regions, rural areas and less populous counties. This is not optimal, but I think it does the trick for this cursory examination.

All counties where non-Hispanic whites are 80% or more, male life expectancy vs. median household

All counties where non-Hispanic whites are 80% or more, male life expectancy vs. median household, labeled only with states

What I’m really interested in is the middle of the distribution, not the really rich or really poor counties. So I limited to incomes between $35,000 and $65,000 dollars. So the same as above, but now constrained as noted.

Focus on the outliers. What is going on in Baker County, Florida? Raw data is below, but I want to map these results above. Again these are the counties from the chart above (income between $35 and $65 K) shaded in proportion to the value of of the residual. In other words, a “dark” blue county is far deviated from the trendline by being above it, while a “dark” red county is deviated by being below it. Being above the trendline means that the county has a high life expectancy for its income, while below means it has one below what one would expect for income.

As I said above, there are constraints with these data. Some counties are missing from the source tables which I used, and only those counties present in all of the source datasets remain. Additionally, the map excludes very wealthy areas (parts of New England) and very poor ones (much of Appalachia), as well as those areas where less than 80% of the population is non-Hispanic white. The income data here surely exaggerations differences in real consumption; it isn’t taking into account cost of living. But, I think the general insight from the earlier map remains: being close to Canada is good for a county’s average life expectancy.

Here are the counties 2 or more years above the trendline:
FL – Charlotte 2.008085
ND – Ward 2.015647
SD – Lawrence 2.050383
MT – Gallatin 2.058849
ND – Cass 2.079347
WI – Marathon 2.112865
WA – Kittitas 2.146117
WI – Dunn 2.153582
MN – Steele 2.156426
IA – Bremer 2.200
TX – Bandera 2.202348
MN – Stearns 2.262364
WA – Whatcom 2.280879
MN – Winona 2.289539
MN – Crow Wing 2.296179
ID – Kootenai 2.319326
WI – Wood 2.365409
NE – Madison 2.386554
MN – Martin 2.407487
MI – Emmet 2.418643
NY – Tompkins 2.437007
NY – Seneca 2.546057
PA – Union 2.568985
CO – Larimer 2.582040
NE – Buffalo 2.583082
IA – Henry 2.662992
MN – Freeborn 2.683949
MN – Mower 2.770022
KS – Douglas 2.811094
CO – La Plata 2.815263
WI – Eau Claire 2.821042
WI – Clark 2.920767
MN – Brown 2.980544
MN – Kandiyohi 3.064475
WA – Island 3.071250
IA – Mahaska 3.080397
UT – Iron 3.114267
WA – Jefferson 3.229158
PA – Centre 3.274080
IA – Winneshiek 3.305467
MI – Leelanau 3.378293
ID – Latah 3.605875
IA – Johnson 3.618503
OR – Polk 3.661479
MO – Nodaway 3.750706
IA – Story 3.761283
KS – Riley 3.812826
UT – Washington 3.857329
MN – Douglas 3.871383
SD – Brookings 3.893517
ID – Madison 4.116757
UT – Cache 4.261088
IA – Sioux 4.312095
OR – Benton 4.544464

And 2 or more years below:
FL – Baker -7.775926
AL – Walker -4.976348
AR – Greene -4.273662
MD – Cecil -3.862680
TX – Hardin -3.856310
TN – Carroll -3.577800
GA – Bartow -3.459386
IN – Starke -3.429478
WV – Berkeley -3.344611
GA – Jackson -3.282126
MS – George -3.254395
TN – Wilson -3.244293
AL – Chilton -3.208314
TX – Orange -3.199165
AL – Marshall -3.176659
OK – Garvin -3.107875
TN – Henry -3.006837
NC – Currituck -2.953442
WV – Jefferson -2.951280
GA – Walker -2.876994
VA – Warren -2.823080
AL – St. Clair -2.801636
TX – Fannin -2.779233
AR – Lonoke -2.673197
MS – Hancock -2.639797
FL – Nassau -2.636036
KY – Scott -2.605132
TN – Robertson -2.579745
GA – Murray -2.565466
TN – Lawrence -2.544601
TN – Maury -2.534866
MO – Jefferson -2.503979
TN – Dickson -2.490682
GA – Walton -2.475931
GA – Gordon -2.433042
MI – Osceola -2.378020
FL – Clay -2.370529
GA – Paulding -2.369467
TX – Wise -2.366306
IA – Marshall -2.331662
MS – Pearl River -2.283195
OK – Grady -2.256928
340 MO – St. Francois -2.224602
WY – Sweetwater -2.212283
IL – Lee -2.204632
AZ – Mohave -2.203554
TX – Van Zandt -2.147798
MI – Calhoun -2.143441
TN – Obion -2.138999
KY – Kenton -2.124380
WV – Kanawha -2.121422
OH – Madison -2.115574
IN – Dearborn -2.089985
GA – Oconee -2.077321
KY – Nelson -2.059997
TN – Rhea -2.056843
TN – Cheatham -2.053162
WV – Raleigh -2.006031

(all these are the counties between $35 and $65 K in median household income. The trendline was generated from this constrained sample as well)

• Category: Science • Tags: Longevity, Regions 
Razib Khan
About Razib Khan

"I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. If you want to know more, see the links at"