“ Furthermore, we generated a streamlined aging clock (LinAge), based directly on PCAge, which maintains equivalent predictive power but relies on substantially fewer features.”
This paper has been discussed here in terms of the benefit of a CR diet. But here is a link to the paper that contains a link (see screenshot below) to an excel model and instructions to use the LinAge model for yourself.
Thanks, Joseph. I did download the xls spreadsheet and instructions. Looks a little technical to figure out but I’m working on it.
“Our findings suggest that the additional parameters captured by PCAge enabled the identification of additional healthy aging and at-risk individuals, beyond those identified by PhenoAge.”
“For the construction of LinAge, we selected 61 parameters that are routinely measured clinically or can be extracted from clinical records.”
The blood tests look mostly standard with a few outliers.
“It is impractical to measure all 165 parameters included in PCAge. We, therefore, developed a reduced BA clock derived from PCAge but using a minimal set of parameters (LinAge). Using sensitivity analysis, we selected a subset of clinical parameters for inclusion to retain the predictive power of PCAge. LinAge includes only parameters from the complete blood count, renal function tests, liver function tests, iron panel and lipid panel in addition to vitamin B12, folate, CRP, fibrinogen, LDH, NT-proBNP, uric acid, glucose, HbA1c, urine ACR, blood pressure, pulse rate, BMI, smoking status and medical history.”
“Finally, we compared LinAge’s and PhenoAge’s ability to predict specific causes of death, finding that LinAge overall outperformed PhenoAge in predicting 20-year CVD and non-CVD-related and cancer-related mortality. This advantage was more pronounced for non-CVD deaths.”
" PCAge and LinAge are sensitive to a more complete set of mortality causes, and, unsurprisingly, they generally outperform the ASCVD score in predicting overall future mortality. This advantage is most obvious for individuals who are aging unusually well (whose BA is lower than their CA), because low cardiovascular risk alone does not guarantee healthy aging, but healthy aging is incompatible with substantially elevated cardiovascular risk."
" we recently showed that our clinical aging clock, PCAge, trained against all-cause mortality, could identify metabolic dysfunction and cardiac and renal dysfunction, as well as markers of inflammation as predictors of accelerated biological aging. A reduced clinical clock, LinAge, based directly on PCAge, has equivalent predictive power but uses significantly fewer features. Such clinical clocks are practically optimized for rapid translation into clinical practice after their validation."
