are about 25% of people genetically early chronotypes and 16% late ones?
“that is about, by the way, 25% of the human population.”
a guest, a cognitive neuroscientist said on The Diary of a CEO, 2026-10-15, at 1:41:57. hear it in context → the speaker cited: research from her lab.
first, the part that holds
the numbers are in the right range. in the largest chronotype dataset we read, 23.9% of about 450,000 uk adults called themselves "definitely a morning person". a us time-use study put 16% of people in its "late" chronotype group and found that a quarter of people sleep noticeably earlier than the rest. chronotype is also partly inherited, and hundreds of gene variants are linked to it.
what we found when we went and looked
we looked for the speaker's lab data and did not find it, and "am shifted" and "pm shifted" are not standard labels in the papers we read. the closest published numbers depend on where the line is drawn. in the uk biobank, 23.9% say they are "definitely" morning people and 8.0% "definitely" evening. a study of 53,689 american time use survey diaries found 40% early, 44% intermediate and 16% late chronotypes, using fixed sleep-midpoint cutoffs. the same study's "25% earlier than 2:24" is a quartile, so it equals 25% by construction. the weaker part is "genetically". the uk biobank team estimated that common gene variants explain 13.7% of the variation in chronotype. people carrying the most morning-linked variants (top 5%) sleep about 25 minutes earlier than those carrying the fewest (bottom 5%). that is a real genetic effect, but it does not divide people into genetic groups of 25% and 16%.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| about 25% of people are 'am shifted' (early) chronotypes | unverifiable | the speaker's lab data was not found. published figures depend on the cutoff: 23.9% 'definitely morning' in the uk biobank (self-report, n=449,732); 25% by definition in the us time-use study's earliest quartile (midsleep before 2:24); 40% early by that study's fixed cutoff (midsleep before 3:00). 25% is plausible under one definition, but nothing ties it to the speaker's category [1][2] |
| about 16% of people are 'pm shifted' (late) chronotypes | unverifiable | the us time-use study found 16% late chronotypes (midsleep after 5:00 on weekend diary days), which matches the figure. but the uk biobank found 8.0% 'definitely evening' and 25.6% 'more evening than morning', so the share depends on where the line is drawn. the speaker's own data was not found [1][2] |
| these groups are set by genetics | overstated | common gene variants explain 13.7% (95% ci 13.3-14.0%) of the variation in self-reported chronotype. the top and bottom 5% by morning-linked variants differ by about 25 minutes in measured sleep timing (03:06 vs 03:32). genes push chronotype in a direction; they do not sort a set share of people into genetic types. the us figures come from diaries and say nothing about genes [1][2] |
both directions
for the speaker: the uk biobank figure of 23.9% "definitely morning" is very close to 25%. the 16% "late" figure from the us time-use study matches the speaker's number exactly. chronotype is partly heritable: the same genome-wide study found 351 gene regions linked to being a morning person, and those variants shift measured sleep timing.
against: the two numbers seem to come from different studies with different definitions. the 25% is either a self-report category or a quartile, and a quartile is 25% by construction. the 16% is a diary-based cutoff, and in that same study the matching early cutoff gives 40%, not 25%. measured genetic effects are modest: common variants explain 13.7% of the variation, and the extreme genetic groups differ by about 25 minutes. age and sex also shift chronotype a lot, according to the us study. so calling these shares "genetic" chronotypes claims more certainty than the evidence supports.
what this cannot say
we did not find any publication from the speaker's lab that uses the "am shifted" and "pm shifted" labels. the lab may have unpublished data with different cutoffs. both studies we read rely on self-report: a questionnaire answer in one and weekend sleep diaries in the other. both cover mostly one country each (uk and us). snp heritability counts only common variants, so it understates total heritability; twin studies, which we did not read here, usually give higher estimates. our check is limited to two primary papers.
sources, each one fetched and read
- Jones SE et al., Genome-wide association analyses of chronotype in 697,828 individuals provides insights into circadian rhythms, Nature Communications, 2019, PMC6351539 What we read there: uk biobank (n=449,732): definitely morning 23.9%, more morning than evening 32.2%, don't know 10.3%, more evening than morning 25.6%, definitely evening 8.0%. snp heritability 13.7% (95% ci 13.3-14.0%). top vs bottom 5% of morningness alleles: activity-measured sleep timing 25.1 min earlier (03:06 vs 03:32). Fetched 2026-10-27 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC6351539/
- Fischer D et al., Chronotypes in the US - influence of age and sex, PLOS ONE, 2017, PMC5479630 What we read there: american time use survey 2003-2014, n=53,689: 40% early (midsleep before 3:00), 44% intermediate, 16% late (after 5:00). '25% show a chronotype earlier than 2:24, 50% fall between 2:24 and 4:15, and another 25% show a chronotype later than 4:15.' does not look at genetics. Fetched 2026-10-27 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC5479630/
speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam