Papers on protein score

Who was studied44,917 participants with plasma protein measurements, 13.64 years

What changedA blood-protein score accounted for 31.69% of the association between fine-particle pollution and type 2 diabetes

The catchDiabetes prevention through targeting these proteins was not measured

Who was studied6,468 CARDIA, MESA, and FHS participants

What changedHigher blood protein scores were linked to more calcium buildup in heart arteries

The catchSelecting proteins and evaluating their calcium association in the same dataset risks overstating the association

Who was studied52,483 people without previous heart failure, against clinical risk factors alone

What changedThe protein score distinguished future heart failure cases better than clinical risk factors alone: AUC 0.80 versus 0.76

The catchThe analysis did not measure whether changing these proteins would prevent heart failure

Who was studied12,308 people with metabolic syndrome, against the clinical PREVENT model

What changedCardiovascular prediction accuracy was higher with proteins: C-statistic 0.730 against 0.673 with the clinical PREVENT model

The catchParticipants were healthier than the broader population, potentially biasing the model’s performance

Who was studied462,516 UK Biobank participants, against people without chronic kidney disease

What changedLater metabolic fatty liver disease was higher among people with chronic kidney disease (hazard ratio 1.36)

The catchAcute kidney dysfunction could inflate chronic kidney disease prevalence in the hospital cohort

Who was studied44 male C57BL/6J mice

What changedAge-linked protein patterns differed across mouse organs: 93% of core clock proteins were selected in only one organ

The catchCorrelated proteins can produce different clock features despite shared underlying aging processes

Who was studied35,269 UK Biobank participants, median 13.7 years

What changedTriglyceride-rich particles and lipoprotein(a) shared 36 proteins; higher scores for both patterns were linked to higher coronary disease risk

The catchThe analyses did not measure whether targeting these proteins prevents coronary disease

Who was studied45,788 UK Biobank participants without valvular heart disease, median 13.8 years

What changedInflammatory protein scores predicted incident aortic valve stenosis with C-indices of 0.934 in females and 0.890 in males

Who was studied115,971 UK Biobank participants, more than 15 years

What changedThe four-protein score predicted death from any cause better than established noninvasive scores

The catchThe abstract does not quantify improvement in mortality prediction over existing scores