Genome-wide analysis identifies molecular systems and 149 genetic loci associated with income, 2019, Hill et al

Might well have been him I was thinking of.

I have, however, found this,
"Personality and perfectionism in chronic fatigue syndrome: A closer look", https://www.tandfonline.com/doi/abs/10.1080/08870440802403863, which says "We thank Ian Deary at Edinburgh University for advice on personality and on statistical analysis.".

 

I think the political correctness that shuts out any talk of genes is the same mindset that says that PWME need to get off their butt...

 

From the paper:

This kind of reminds me of what someone once said about the correlation between intelligence and longevity. "Einstein was a very smart guy, but he didn't live orders of magnitude longer than the rest of us."

 

I came upon this offering, I don't remember how or why.

General practitioners' attitudes to patients with a self diagnosis of myalgic encephalomyelitis
BMJ 1995; 310 doi: https://doi.org/10.1136/bmj.310.6978.508 (Published 25 February 1995) Cite this as: BMJ 1995;310:508

• Article
• Related content
• Metrics
• Responses

1. Shonagh Scott, psychology studenta,
2. Ian Deary, readera,
3. Anthony J Pelosi, consultant psychiatristb

1. 
   Author affiliations

1. Correspondence to: Dr Pelosi.

• Accepted 9 December 1994

Interest in the symptom of tiredness has increased with the suggestion of a syndrome of prolonged fatigue caused by infection. The syndrome is referred to as myalgic encephalomyelitis, even though no evidence exists that sufferers have encephalitis or myelitis. Active support organisations encourage self diagnosis1 and advise how to approach a general practitioner who “doesn't believe in ME.”2 Problems in doctor-patient relationships may be a factor in persistent disability in fatigue states.3 We therefore used a case vignette method to examine how self diagnosis of myalgic encephalomyelitis could influence general practitioners.4

 

At least that makes it easy to know to dismiss anything he says. Words and their meaning, do they matter? Clearly not to everyone.

 

Studies like this are only ever suggestive. Even with these sorts of p values, there is still a lack of specificity and sensitivity, hence some well characterised mechanism needs to be found before we claim there is a cause/effect relationship.

 

I completely agree; however, a number of replies seem to take a different view and dismiss the inquiry out of hand.

I don't think that's right as genes are essentially causal in a way other sorts of factors aren't.

Of course there are all sorts of conceivable family advantages such as inheritance of wealth, connections, sense of status, early exposure to necessary knowledge in some field, etc.. But in modern society there are so many effectively independent networks of people interacting reasonably spontaneously over time that I would accept that it would all wash out. That point is probably worth deeper examination, though!

I agree and I think it urgently needs to be dispelled. And not just for the sake of ill or otherwise unusually unlucky people. I worry that the structure of the economy is shifting such that even perfectly capable people with 'average' circumstances will not be favored to achieve a reasonable standard of living in the coming decades.

 

Haven't read the full study so just reflecting in general on the link between genes and wealth. What bothers me is that this skips a step and muddles two concepts.

It's one thing to look at the link between genes and natural general intelligence (as in intelligence independent of education), just like you can look at genes that make you a better sprinter or a better endurance runner, or at genes that give you perfect pitch or tone deafness.

But it's a different thing to look at what factors make you rich. It's entirely possible that natural general intelligence contributes to your chances of becoming wealthy in our current society - but that's because we have a culture that values jobs that require certain 'talents', which could be intelligence or ruthlessness or self-promotion or whatever. If empathy and caring - which may or may not have some genetic base, too, has anybody looked? - were highly valued in our culture, resthome workers would be the rich ones (ok, not every single resthome worker is caring, but you get the point).

 

I agree that talk of genes should not be shut out by political correctness. But it should be taken with a very cautious pinch of salt when being applied to a concept like "intelligence", which we are fairly rubbish at defining and measuring (fill in this questionnaire, we'll turn it into a number and call it science - it's only one step away from personality traits) or "income", for which there are so many other factors to control for, many of which we haven't got much of a grasp on either, that it makes any correlation fairly meaningless.

 

Researchers Link Genes to Income. Other Scientists Beg to Differ.

https://undark.org/2019/12/20/abstracts-genetics-income-science-backlash/

 

But not all lines of enquiry are equally worthy of study. Some are rubbish. Some are even harmful because their results can be misused. Do any of us here want to see another PACE trial? Of course not, and that's not because we want to "suppress" research into CBT and GET, but because we think there's no justification for asking that question yet again, and the problems inherent in answering it mean that the results may be misleading and harmful.

The problem with this GWAS study is very similar: it is that with GWAS, its almost impossible to get a "no" answer to your question.

So, if you wanted to identify the genome associated with preferring the colour blue, I'm pretty sure you'd find it. Because even though liking b blue is pretty meaningless in real terms, it - like most other things - is probably a non-random phenomenon. It could be linked to ethnicity, to upbringing, to socioeconomic status, to certain kinds of mental health variables, and so on. But in this instance, you can see how silly it would be to conclude that some people are genetically predisposed for liking blue. Because that's vacuous and meaningless.

Ask a stupid question, get a stupid answer.

The problem I see is too much high-tech methodology, and not enough attention to formulating a good research question.

And don't get me started on "Intelligence", a theoretically vacuous construct that we should have left behind several decades ago - once we had enough understanding of cognitive function to generate something with a bit more theoretical coherence. How could any psychology researcher believe "intelligence" is a worthy target of research in this day and age? And it is an even bigger crime to ignore the problem of reverse causation: measures of intelligence are associated with socioeconomic status, but that doesn't mean the first causes the second. Its very likely that a child raised in a high SES environment is better equipped to do well on IQ tests than a child from a low SES background.

 

I don't think this is rubbish (not necessarily the study itself - I don't have the expertise to assess it - but the line of inquiry). Understanding how fundamental social outcomes come about with hard data is the most prime information both for arguing against retrograde 'pull yourself up by the bootstraps' fundamentalism and for designing public policy that actually has a chance to be effective.

In any case, they came up with a pretty low number for how much genetic variation plays a part at this time. If the study is to be trusted, I think it is telling us not to worry too much about genetics for broad public policy applications (- even though the authors don't really communicate this).

Intelligence is certainly a fraught topic. However, I think we are in an in-between sort of place with it. Obviously almost nobody is really happy with the measures we have, but there seems to me to be enough substance to them that we wouldn't really want to throw it all out. For instance, most colleges in the US use the SAT and ACT, intelligence tests of a sort, because they are strong standardized predictors of who will succeed at an institution (and it is certainly not in a student's interest to go where they will not succeed when we are all paying out the nose here!). Furthermore, if schools drop these as a factor in admissions, that would very likely further bias admissions towards the wealthy and well-connected because personal statements and extracurricular activities would become much more important; wealthy families can support more extracurricular activities and expensive tutors to help craft personal statements. On the other hand, SAT prep programs and coaching apparently don't improve scores by much, so the standardized test turns out to be surprisingly 'meritocratic' - even if we all know that they don't fully capture any person's intellectual capabilities, and may grievously misjudge that of many individuals.

 

It's very easy to get a no answer, but most researchers prefer to stack the cards in their favor by weakening the methodology and statistical thresholds. The problem isn't so much the study, as the inappropriate conclusions.

As @Forbin suggested, the results are weak to say the least.

Wow!

You can't have it both ways, study of intelligence (if done properly) will increase our understanding of cognitive function. Secondly, while it is true that global SES has a substantial influence on IQ test scores for a majority of the population (within 1-1.5 SD of the mean), and can certainly alter the rate of cognitive development of children (eg lack of nutrition, educational opportunities etc), high intelligence (admittedly, the measures of this are flawed) as defined as greater than 2 SD from the mean has a much greater proportion of the variance explained by genetics, rather than environment.

 

I could be wrong here, but my understanding of the GWAS approach is that it all comes down to how much you correct for false positives, and even stringent methods generally let some false positive results through. So the studies I've seen have all "found something", but given the sheer number of comparisons I'm not at all surprised by that. There have been some attempts at replication in GWAS studies in psychiatry, and they've been disappointing - so even the associations that appear to be strongest in one cohort are not necessarily the ones that appear strongest in a different cohort. That makes me a bit suspicious, and worried that this kind of broad hypothesis-free data mining might be particularly vulnerable to false positives. Especially when measure of interest is ill-defined. But I admit that I'm not an expert on the methodology.

I don't agree here. I think intelligence as a construct has held us back a lot in understanding the how people vary in their capacities for learning and reasoning. The problem is that, instead of examining and refining our theoretical models of those capacities, we've kind of taken the idea of intelligence as a given, and focussed all our efforts on measurement. The result is endless refinement in testing and scales that all piggy back on the preceding scales. The latest standardised measures of intelligence do pay lip service to some more recent cognitive science ideas, like working memory and verbal versus nonverbal capacities (and they do try not to measure too much semantic memory - aka crystallised knowledge - which is good). But they still fundamentally make the same assumptions.

Something we're picking up in IQ tests is probably related to genetics. But we need better ways of conceptualising what it is those tests are measuring. We've historically made a strict division between intellectual ability on the one hand and temperament on the other, and I think they might be closely tied. I wonder if anxiety/levels of arousal (whatever you want to call that dimension of temperament) might be important. To do well on certain IQ tasks, you need just the right level of arousal - not too much and not too little - to give you the kind of optimal attentional control you need.

Anyway, I'm veering way off topic...

 

I agree that focusing on the narrow/superficial standardised scales of intelligence is part of the problem. It's the usual problem of studying questionnaires/tests separately from the underlying biological phenomena and assuming the two are inherently linked without ever demonstrating the mechanism.

 

Thinking on this, I was reflecting on the myriad historical factors that determine socioeconomic status in white UK residents. Groups of multiple ethic origins - norman, saxons, celtic - have blended together for centuries, but I would be prepared to bet that even now, those groups are not equally represented in the rich vs the poor.

So you might observe genetic variation with income that is related to those sorts of non-random factors too.

 

Why do you think that?