This is often said to older men, even slightly older men (Like a 23 y/o with a 19y/o or a 23 y/o with a 32 y/o) to shame them for being with young women. This sentiment is misleading.

While it is true, human brain development is not “finished at age 25”, but rather somewhere between the 30s and never. There are three arguments which are strongly needed to contextualise how age effects the brain.

  1. The PopSci cut-off point around age 25 refers specifically to measurable structural development (white/grey matter ratio etc.) in the frontal areas which, according to some longitudinal studies, plateaus around that age (IBID). Yet in other studies and depending on what parameter you look at, the leveling off happens earlier or later than that.
  2. There isn’t a 100% clear connection between specific measurable changes (in chemistry, connectivity, volume etc) and real-life behaviour. This means that lack of maturity in a part of a brain does not necessarily correlate to an actual difference in cognition or psychosocial maturity.
  3. Culture and environment can heavily effect maturity of the brain and levels of psychosocial maturity independently of age as a predictive factor.

TLDR: Humans hit cold cognition 'peak' (or adult levels in late/mid teens on average). Psychosocial maturity generality reaches adult levels a little later or a lot later; however when this is, is heavily dependent on culture and environment.

Is the age of 25 specifically significant for the brain? “The mid twenties number doesn’t come entirely out of the blue as it is an age where many different brain regions will have reached their maximum volume for example. However, this absolutely does not imply that the brain then stops being malleable to change nor does it mean that up until that point the brain would not be capable of functioning at a developed level,” Bethlehem and Seidlitz explained.

Longer discussion

In terms of functionality rather than chemistry, psychosocial maturity which is associated with the frontal and prefrontal areas is assumed to be a theoretical linear curve that continues far beyond age 16 - the age where a person’s logical development (again, along a theoretical curve) is assumed to plateau.

The problem with taking something abstract like theoretical neuroscience and applying it to real people is that the direct connection between the two isn’t 1-1. Human beings are not theoretical curves and empirically, there is a huge range of Darwinistic individual variance within any population.

According to Robert Sapolsky, a Stanford neurobiologist, the frontal cortex has a larger than average inter-individual variation than the rest of the brain (Behave, p. 45).

For example, this 238-participant MRI study of functional maturity based on an multi-factor brain connectivity index has observations of kids aged 8-12 which show the same functional maturity level as the asymptote level of age 22, the age where development speed in the brain connectivity in the study dropped off.

This is obviously a few extreme outliers, but if you look how dispersed the points on their graph are for any given age, it goes to show that the trend is not at all a predictor for the individual. You can quite generally have lots of people at any age that have a brain development or maturity correspondent to another age - both upwards and downwards.

The speed of convergence is also very different between individuals and according to these authors, a function of cumulative experience.

Secondly, we need to look at how age specifically is affecting a given person’s decision making. An often quoted author is the behavioral scientist Laurence Steinberg who made an experiment where 18-21 year olds had an increased risk appetite (in economical terms) about halfway between teenagers and a 22+ age cohort defined in the study as adults.

This was when the logical decision they had to make had a background of a negative or positive emotional arousal state, possibly because the limbic system would take over from the prefrontal areas of the brain by some evolutionary mechanism that increased risk tolerance.

The specific relevance of Steinberg’s work is that it includes (either in this one, or in other studies) many of the characteristics that are mentioned with regard to immaturity: responses to peer pressure, impulse inhibition and a skewed analysis of risk/reward.

So, was everyone in the 2 younger groups scoring what was expected from their age-predicted maturity? Absolutely not. The data (Figure 3) shows plenty of test subjects that score within 1-3 years higher or lower on predicted maturity compared to their real age.

This would actually be enough to move a 19-20 year old into the adult group. The Rho (correlation coefficient) for both emotional contexts is close to 0 which is obvious because the individual variation between data points on the graph is really significant.

Another group of authors actually found that only the negative emotional context changes the logical decision for young adults, whereas a positive emotional context does not.

Generally speaking, saying that a brain or decision making apparatus is “developing” says nothing about at when the development threshold for making sound decisions about a given situation is reached.

Again, this would in practice vary immensely between individuals as the above empirical data clearly shows. It’s also not clear according to science (see first link in the post) WHERE that threshold is - because different measures would yield different answers, and brain chemistry is not 100% translatable to real life decisions.

At least according to Steinberg’s study, the risk appetite increase that presents itself in young adults who score younger than their age on maturity are present, but more limited after age 22. (“By adulthood (ages 22+), many of these findings were reduced or limited, but present to a degree“).

This is in line with the general commentary by Diego Luna referred to in this popular article as well as the multifactor MRI study above (plateau at age 22) which measures maturity in terms of functional connectivity. The above suggests that, (without looking at individual variation), at least on some measures, the leveling off in growth rates happens earlier than 25.

Finally, any individual’s psychological maturity will at least to some extent be a reflection of culture and how soon society allows him/her independence. In some Western European countries like Sweden, people on average move out from their parents’ homes very early after the age of majority, which is 18 which forces growth earlier.

In the US, said people would not be able to drink legally for 3 more years. According to FT.com, in 2020, over 50% of American 18-29 year olds lived with their parents, the highest share ever - whereas some Europeans were living on their own and paying rent at age 16-17.

The empirical data from this 11-country, 5000-participant study (also involving Steinberg) based on a number of behavioral experiments has some interesting results about country variations in mental maturity. If you look at the psychological maturity index data for most countries, it’s best estimated by a linear curve that does not flatten out (supporting the theory about continuous development for a long time).

Yet data from Sweden and Italy(1) suggest that psychosocial maturity index levels of 15-20 are reached around ages 22-23 on the curve, whereas the US psychosocial maturity curve slows down after age 20 (at which age the index level is lower than both for Sweden and Italy) and actually never reaches a trend index level of 20 for any age (the curve is not linear and flattens out with a maximum index value at age 26-27). Ironically this same maximum level is reached at ages 22-23 in Sweden.

There may be some weird data in play for the US, and it is important to distinguish individuals from trends (the US standard error in Table 5 is multitudes higher than the Swedish and Italian one, suggesting a large individual spread).

Still, it’s interesting that another study using the same study participants shows that Swedish and Italian risk taking patterns, which could be a proxy for gaining experience, peak at earlier ages and have higher peaks than America does.

Robert Sapolsky specifically says (Behave, p. 51 and p. 173) that it is experiences rather than genes than create differences in brain maturity - for example, social network size influences the size of certain PFC regions and adolescent studies show that peer pressure resistance is higher for people who spend more time with friends. Which could explain the above findings.