25-year study critique
Commentary surfaced criticizing a large cohort study described as having up to a 25‑year span because its mean follow‑up was only 5.5 years despite that 25‑year maximum. (Defenders said the study’s scale still makes it one of the longest cohorts, while critics argued the short average follow‑up weakens long‑term inferences.) (x.com) (x.com)
A cohort study can run for decades on paper while most participants are observed for far less time. That gap is driving a new fight over a study described as spanning up to 25 years. (pmc.ncbi.nlm.nih.gov) In cohort research, “follow-up” means how long each participant is actually tracked before an event happens or the record stops. Methodology papers say investigators should report that time clearly because a maximum span and an average or median follow-up can tell very different stories. (pmc.ncbi.nlm.nih.gov) The current dispute centers on a study promoted as having follow-up of up to 25 years, while critics pointed to a mean follow-up of 5.5 years. In epidemiology, “up to” names the longest observation for any participant, not the typical observation across the whole cohort. (ebn.bmj.com) (pmc.ncbi.nlm.nih.gov) That distinction changes what a reader can infer. A cohort with some 20-year records can still produce estimates driven mostly by participants observed for five or six years if most enrollees joined late, had events early, or were censored before the study’s outer limit. (pmc.ncbi.nlm.nih.gov 1) (pmc.ncbi.nlm.nih.gov 2) Researchers have long treated follow-up completeness as a validity issue, not a formatting choice. A 2015 PLOS One paper said study validity depends on how representative measured outcomes remain as follow-up thins out over time. (pmc.ncbi.nlm.nih.gov) Defenders of the study’s framing have leaned on scale and rarity. Large prospective resources such as UK Biobank are widely used because hundreds of thousands of participants can generate substantial numbers of outcomes even before every participant reaches long durations of follow-up. (science.org) Critics are making a narrower point: a long calendar window does not automatically support strong claims about very long-term risk. Statistical guidance says follow-up length should be reported in ways that show where the data are most stable, because estimates become less secure when fewer participants remain under observation. (pmc.ncbi.nlm.nih.gov) This is not unusual in population studies. Early Framingham findings shaped cardiovascular medicine after about six years of median follow-up, and later papers extended those conclusions with much longer observation as more time accumulated. (mdpi.com) (nature.com) The practical question for readers is simple: was the claim about risk over five years, over 25 years, or over a mix of both? If a paper emphasizes the 25-year ceiling without equal emphasis on the 5.5-year average, readers can overestimate how much of the cohort actually informs the longest-range conclusion. (pmc.ncbi.nlm.nih.gov) The argument is likely to keep surfacing whenever large biobanks publish early long-horizon analyses. “Up to 25 years” and “mean follow-up 5.5 years” can both be true at the same time, but they do not describe the same evidence. (science.org) (pmc.ncbi.nlm.nih.gov)
