To the editors:
The notion that things could never be as good as they were back in the day is timeless. Early in The Iliad, Agamemnon and Achilles start a hot-headed quarrel prompting an aged and honey-tongued Nestor to deliver a memorable lecture:
I am older than either of you; therefore be guided by me. Moreover I have been the familiar friend of men even greater than you are, and they did not disregard my counsels. Never again can I behold such men… These were the mightiest men ever born upon this earth: mightiest were they… Not a man now living could withstand them.1
His plea for cooler heads to prevail is ignored, putting the gods offside and setting off an inexorable march to the disastrous Trojan War.
This vignette contains many common perspectives on aging: generational decline, the wisdom of our elders, the impetuosity of youth, and the rosy-hue of positive recall bias. Calamity. Have things changed?
Von Hippel touches on many of these issues in his thought-provoking essay. We are in agreement on many insights and conclusions, particularly in recognizing the accrual of wisdom by many, but certainly not all, elders. Some were new to me, and gratefully acknowledged; here, in the spirit of combat (and good humor) I focus on those where I think the old man is plain wrong.2
Universality of Decline
But even the most resilient decline with age. Just as there are no eighty-year-olds who could be mistaken for a twenty-year-old, there are no eighty-year-old brains that could be confused with a twenty-year-old brain.
As direct contrary evidence is the case of the 115-year old who died with a sprightly 60-year-old-like brain.3 A rare case perhaps, but a counterfactual en plein soleil.
More generally, von Hippel makes a popular error when considering the impact of aging, which is to ignore the principle of non-ergodicity. This error assumes that averages appropriately represent the sample, or even worse, individuals in the population.
We can all imagine a graph that charts some biological or psychological characteristic—take memory performance, for example—against age on the x-axis: a more or less continuous fall from grace starting in the twenties, right?That is certainly what one sees when charting age-group averages. More revealing are those epic longitudinal studies that have followed individuals for ten, twenty and even thirty years. What this tends to show are three clusters: stabilizers, slow decliners, and fast decliners.4 If the dependent variable was memory, or general cognition, then the fast decline group probably includes those at risk for or already succumbing to dementia. For present purposes, it is the stable cluster that is most interesting and germane; clearly there is a non-trivial subset of individuals where decline is not inevitable as they age.
Hence, the reviewer who noted the value in understanding why some elders do not exhibit inhibitory loss was not dreaming. Indeed, we understand a lot more about the cluster that declines quickly—the changes are dramatic, statistically powered, and often linked to in vivo brain changes—than how and why the stabilizers stay stable. Both phenomena exist in nature.
Causal Attitudes to Ageing
One of the hairiest issues in determining causality in complex areas like medicine or psychology is temporal primacy.5
In the example used by von Hippel, it is indeed impossible to parse the chicken from the egg when solely armed with cross-sectional or even longitudinal data. So, for those associations between positive thinking in youth and better health outcomes in later life, it is certainly possible that there might be a causal link, or more simply, healthier young people tend to have better long-range expectations, which, unsurprisingly, are borne out.
That is why randomized controlled trials (RCTs) are so important. We manipulate a key putative causal factor—e.g. mental exercise—let the power of randomization spread evenly all the other relevant known and unknown factors, and test the idea.
It is therefore misplaced to suggest that, “the effects of complex thought on long-term functioning may, in fact, be negligible.” Firstly, just what is the editorial position since an immediate prior conclusion was: “These activities [complex mental ones] play a demonstrable role in the delayed onset of cognitive decline”?
This is my home ground so I will resist the temptation to elaborate. Perhaps von Hippel considers delaying cognitive decline to be of negligible value? Unlikely, since the same association extends to a strong protection from dementia, the object of almost universal terror.6 More likely is the skeptic’s mistrust of epidemiology, with all its dirty correlations and associations observed “in the absence of controls”?
When it comes to one operationalized form of complex mental activity, computerized cognitive training, evidence from RCTs is overwhelming. In a series of meta-analyses across >70RCTs, we have shown positive effects on: general cognition in healthy elderly, memory and psychosocial wellbeing in Mild Cognitive Impairment (a precursor to dementia), and executive function in Parkinson’s disease.7 In these studies, both active and passive controls were used and so there can be no reverse causality error. When viewed in conjunction with extensive epidemiological and behavioral neuroscience work, there should be no further ambiguity about a causal link between complex mental activity and enhanced cognition in the aging brain.8
Evolutionary Fables
Von Hippel teaches us that monkeys and humans both restrict their social circles as they age. Interesting. Equally interesting is the association between positive memory bias and better immunological state years later. What is the link? If I have read von Hippel correctly, marshalling eighteenth-century Finnish graveyard records, he seems to suggest that rather than social condensation being an emotional or psychological response, it is perhaps simply an evolutionary by-product. More positively-inclined persons are more likely to live longer, outlive their less-positive peers and so have children who have children.
This may well be the case but how could we know? Evolutionary psychology ideas like this are not falsifiable, and so not scientific theories, but rather stories that sink or swim on aesthetic and political grounds.9 Indeed, it is somewhat ironic that the proponent is at the same time so ready to reject epidemiological data because of lack of scientific rigor.
Obviously, any factor that tends to extend longevity makes it more likely to have children, and grandchildren, and great-grandchildren etc. One does not need to go on an expedition. And why the selective focus on CD4+ or immune state? Why not telomere length, lipid peroxidation, epigenetic methylation, DNA mutations, or any other biological process that in general deteriorates with age, perhaps even tooth length?
I am willing to wager older persons with a good attitude have better dentition. And those with flashy teeth live longer and end up with more grandchildren. Evolution.
Michael Valenzuela
William von Hippel replies:
Michael Valenzuela is a rising star in aging research, so my heart quickened when I saw he had thrown down the gauntlet in response to my recent essay. His letter is spirited, it provides a useful corrective to my over-statement regarding the universality of decline, and I learned something from it. So why bother responding? After all, I chose not to respond to Thomas Suddendorf and Susan Charles, whose letters engaged me similarly. Unfortunately, despite the adroitness of Valenzuela’s not-yet atrophying brain, he makes three errors that would benefit from correction.
First, Valenzuela argues that “the reviewer who noted the interest value in understanding why some elders do not exhibit inhibitory loss was not dreaming.” If that is what the reviewer had actually noted, I would agree. But if you look back at my essay, you’ll see that the reviewer had a different agenda altogether. (S)he wrote, “although there is something intriguing about the idea that declines in executive function could negatively impact social functioning, the more interesting question may be why they do not.” I labelled this claim wishful thinking because the reviewer is asserting that declines in executive function have no impact on social outcomes. In other words, the reviewer is essentially saying that older adults are doing just fine despite age-related losses in the brain’s control center. The fact that some older adults show no decline in executive function is irrelevant here, and indeed we predicted and found that such individuals also show no losses in social functioning.
Second, Valenzuela cites three of his recent papers in an effort to refute my claim that “the effects of complex thought on long-term functioning may, in fact, be negligible” (emphasis added). When I went back to re-read these papers, it was evident that our disagreement is more apparent than real. In the first of these papers, Valenzuela and colleagues write, “These results … do not address the durability of training effects.”10 In the second paper, they note that “results indicated a substantial waning of training benefits after training cessation.”11 Finally, in the third paper, Valenzuela and colleagues do not raise the issue of long-term change, noting that their goal was to detect “change from baseline to immediately post-training.”12 These are great papers, but I don’t see how they would lead Valenzuela or anyone else to disagree with what I wrote.
Finally, and most egregiously, Valenzuela attacks my own beloved field of study in his claim that evolutionary psychology is unfalsifiable—and hence unscientific. This oft-repeated foolishness has been refuted so many times that I won’t bother addressing it here, other than to state that like every other scientific enterprise, evolutionary psychology proposes testable hypotheses.13 I will note, however, that in a recent survey of social psychologists, David Buss and I found that approximately one percent of our colleagues thought Darwinian evolution was impossible.14 Perhaps Valenzuela is a member of that one percent? If not—if he is indeed a proud member of the ninety-nine percent—I’d encourage him to read the literature before passing judgment on it. He may be surprised by what he finds there.
