Skip to main content
. Author manuscript; available in PMC: 2022 Feb 1.
Published in final edited form as: Biogerontology. 2020 Oct 16;22(1):63–79. doi: 10.1007/s10522-020-09903-w

Fig. 5.

Fig. 5

Association between physiological dysregulation and frailty risk in full regression models. Estimations (points) together with 95% CIs (segments) are plotted for relationships between dysregulation levels, or the number of dysregulated systems (“# Sys Dysreg”), and frailty risk in WHAS (n = 1194, red) and NuAge (n = 1653, blue). a Frailty risk was assessed with a logistic regression model comparing non-frail/pre-frail to frail (see Fig. S9 for models comparing non-frail to pre-frail/frail), as a function of the number of dysregulated systems categorized either as 1–2, 3–4, and 5+ (upper part) or 1–3 and 4+ (lower part), with no dysregulated system as the reference group. b Frailty risk was assessed with a logistic regression model comparing non-frail and pre-frail to frail, as a function of the number of dysregulated systems. c Frailty risk was also assessed with a proportional odds model on non-frail to pre-frail to frail. All models except “# Sys Dysreg (raw)” controlled for the presence or absence of dysregulation in individual systems (dummy variables). # Sys Dysreg number of dysregulated systems, Electro. electrolytes, KLF kidney/liver function, Leuko. leukopoiesis, Micronut. micronutrients, OT oxygen transport