Sho Communica ion
Rebu al o De enses Ci ing “In e nally Ma ched Coho s, Time- o-E en Models, and
Pe son-Time Denomina o s” - A Response o De enses Agains ou C i ique on Coho
Composi ion in he Kim e al. COVID-19 Vaccine S udy (Bioma k Res, 13:114, 2025)
Au ho : Ma co Rocce i
A ilia ion: Uni e si y o Bologna, Depa men o Compu e Science and Enginee ing,
Bologna, I aly
Co espondence: [email p o ec ed]
In oduc ion and Disclaime
The de ails o ou quan i a i e c i ique on he me hodology o he Kim e al. s udy [1] can be
ound in ou p ep in s [2-6]. Ou analysis is no ocused on whe he COVID-19 accines a e
biologically associa ed wi h Se ious Ad e se E en s (SAE) o cance isk. Ins ead, ou co e
inding is ha he coho used in he sc u inized s udy is undamen ally lawed and un eliable
due o a se e e me hodological e o in pa icipan selec ion. I will now explain, using
quan i a i e me ics, why he s udy's epo ed associa ion is a s a is ical a i ac and no a
genuine biological signal.
The$Co e Thesis: Ex e nal Validi y Vi ia es In e nal Resul s
The cen al a gumen is ha he s udy by Kim e al. ailed he undamen al equi emen o
ex e nal alidi y (Gene alizabili y – STROBE I em 21) [7]. A ailu e o ex e nal alidi y,
demons ably caused by se e e selec ion bias, ende s any subsequen s a is ical a emp s a
in e nal alidi y (such as P opensi y Sco e Ma ching) i ele an .
The Pa adox
The coho 's o e all C ude Incidence Ra e (CR) o cance was 26% lowe han he
es ablished na ional a e age a e o Sou h Ko ea (40.78 s. 55.02 pe 10,000 in 2022) [8].
This d as ic supp ession means he coho does no ep esen he a ge popula ion's ue
baseline isk. Simila sup ession holds i one conside p e ious yea s [9,10].
The Quan i a i e E idence o Asymme ic Bias
The sou ce o his supp ession is a p onounced, asymme ic unde ep esen a ion in he high-
isk, non- accina ed subg oup:
• O e all Demog aphic De ici : The o al coho shows a ela i e de ici la ge han
32% in he high- isk demog aphic g oup (indi iduals aged >= 65) compa ed o he
Sou h Ko ean na ional s anda d o 18.0% [11].
• The C i ical Asymme y: The mos c ucial inding is he incidence de ici wi hin he
non- accina ed elde ly pa icipan s (>= 65 yea s old): This subg oup's obse ed
cance incidence a e was 85.2 pe 10,000, compa ed o he o icial na ional a e o
155.2 pe 10,000 [12]. This cons i u es an unp eceden ed la ge han 45% unde coun
o expec ed cance cases in he e e ence g oup o un accina ed indi iduals.
The Mechanism o he A i ac
The 45% de ici in he non- accina ed elde ly g oup is he speci ic mechanism ha
ma hema ically gene a ed he spu ious associa ion: The selec ion p ocess (likely a lawed
P opensi y Sco e Ma ching p ocedu e) led o he non- accina ed e e ence g oup being
disp opo iona ely composed o younge , heal hie , lowe - isk indi iduals ( he Heal hy Use
Bias) [13]. By a i icially supp essing he baseline cance isk in he e e ence g oup, he less-
supp essed a e obse ed in he accina ed g oup was made o alsely appea as an excess
isk.
Add essing he Rebu als: P ecision Canno Co ec Bias
C i ics a gue ha he s udy's use o in e nally ma ched coho s, ime- o-e en models, and
pe son- ime denomina o s alida es hei indings. This a gumen is unsound because i
con uses p ecision (In e nal Validi y) wi h u h ulness (Ex e nal Validi y).
• In e nal s. Ex e nal Validi y: These sophis ica ed echniques add ess In e nal Validi y
(p ecision and balancing wi hin he coho ). Howe e , as demons a ed, he coho
was undamen ally comp omised by selec ion bias, a ailu e o Ex e nal Validi y.
• The Failu e o Denomina o s: A undamen al law in he denomina o ( he popula ion
a isk) canno be co ec ed by ad anced s a is ical modeling. I he popula ion
denomina o is missing 45% o he expec ed cance cases in he highes - isk s a um,
he esul ing incidence a e will be inhe en ly lawed. P ecision canno co ec o
bias.
Conclusion
The de ense o he Kim e al. s udy based on in e nal alidi y me ics is insu icien . Ou
quan i a i e analysis demons a es ha he signal o inc eased cance isk is he likely
consequence o se e e, asymme ic selec ion bias ha supp essed he ue baseline isk. A
me hodologically igo ous and demog aphically balanced coho would p edic ably show no
s a is ically impo an di e ence in cance incidence be ween g oups.
Re e ences
[1] Kim HJ, e al. (2025) 1-yea isks o cance s associa ed wi h COVID-19 accina ion: a
la ge popula ion-based coho s udy in Sou h Ko ea. Bioma k Res. 13(114). DOI:
10.1186/s40364-025-00831-w
[2] Rocce i M. (2025) A C i ical No e on Con adic ions in Sou h Ko ean Cance Incidence
Ra es: The Pa adox o C ude Ra es De i ed om he Kim HJ e al. Coho (Bioma k Res,
13:114, 2025) Showing Concu en Inc eases in he Vaccina ed and O e all Dec ease.
P ep in n. 202510.0883, P ep in s.o g. DOI:10.20944/p ep in s202510.0883. 1.
[3] Rocce i M. (2025) Addendum o “A C i ical No e on Con adic ions in Sou h Ko ean
Cance Incidence Ra es: The Pa adox o C ude Ra es De i ed om he Kim HJ e al. Coho
(Bioma k Res, 13:114, 2025)” - On Possible Sampling Bias and In e ed P opensi y Sco e
Ma ching. P ep in n. 202510.1664, P ep in s.o g. DOI:10.20944/p ep in s202510.1664. 1.
[4] Rocce i M. (2025) Me hodological Conside a ions on he Ex e nal Validi y o he Kim HJ
e al. COVID-19 Vaccina ion S udy (Bioma k Res 13:114, 2025): A Quan i a i e Analysis.
P ep in n. 17658026. Zenodo.o g. DOI: 10.5281/zenodo.17434738.
[5] Rocce i M. (2025) A Bios a is ical Reapp aisal Un eiling he Mechanism Behind
Appa en Cance Risk Signals in a COVID-19 Vaccina ed Coho . P ep in n. 17651590.
Zenodo.o g. DOI: 10.5281/zenodo.17508346.
[6] Rocce i M. (2025) In e en ial Z-Tes Valida ion o Dual S uc u al Bias in Cance Risk
Assessmen wi hin La ge COVID-19 Vaccine Coho s. P ep in n. 17688859. Zenodo.o g.
DOI: 10.5281/zenodo.17434738.
[7] Elm E, Al man DG, Egge M, e al. (2007) The S eng hening he Repo ing o
Obse a ional S udies in Epidemiology (STROBE) S a emen : guidelines o epo ing
obse a ional s udies. PLoS Med.4(10):e296. DOI: 10.1016/j.jclinepi.2007.11.008.
[8] Pa k EH, Jung K-W, Pa k NJ, e al. (2025) Cance S a is ics in Ko ea: Incidence,
Mo ali y, Su i al, and P e alence in 2022. Cance Res T ea . 57(2):312-330. DOI:
10.4143/c .2025.264.
[9] Kang MJ, Jung K-W, Bang SH, e al. (2023). Cance S a is ics in Ko ea: Incidence,
Mo ali y, Su i al, and P e alence in 2020. Cance Res T ea ., 55(2):385-399. DOI:
10.4143/c .2023.447.
[10] Pa k EH, Jung K-W, Pa k NJ, e al. (2024). Cance S a is ics in Ko ea: Incidence,
Mo ali y, Su i al, and P e alence in 2021. Cance Res T ea ., 56(2):357-371. DOI:
10.4143/c .2024.253.
[11] Wo ld Bank. Popula ion ages 65 and abo e (% o o al popula ion) - Ko ea, Rep. Wo ld
Popula ion P ospec s, Uni ed Na ions (UN). [Accessed 2025 No 2]. A ailable om:
h ps://da a.wo ldbank.o g/indica o /SP.POP.65UP.TO.ZS?loca ions=KR.
[12] S a is a. Sou h Ko ea: Cance c ude incidence a e by age, 2022. S a is a; 2024.
[Accessed 2025 No 2]. A ailable om: h ps://www.s a is a.com/s a is ics/1440818/sou h-
ko ea-cance -c ude-incidence- a e-by-age/.
[13] Chemai elly H, Ayoub H, Coyle P, e al. (2025) Assessing heal hy accinee e ec in
COVID-19 accine e ec i eness s udies: a na ional coho s udy in Qa a . eLi e 2025;
14:e103690. DOI: 10.7554/eLi e.103690.
