Our main finding is that frailty in older individuals of 65-77 years coincides with lower cellular pSTAT responsiveness. Negative associations between pSTAT responsiveness and frailty were found in both men and women. More and stronger associations were found in men, with signs of lower immune regulatory (IL-10 induced) pSTAT3 responses in frailer men. These lower responses in frailer men were related to higher levels of CRP in the past 20 years. Our results extend previous knowledge that showed lower pSTAT responses at higher age and in cardiovascular dysfunction (16), and underline the importance of the JAK-STAT pathway in frailty and, possibly, immunosenescence.
To our knowledge, no other studies related cellular pSTAT responsiveness to frailty. Lower pSTAT responses were previously found to be related to markers of cardiovascular disease and to chronological age (16). More specifically, IFNα-induced pSTAT3 in CD8+ T cells and pSTAT5 in CD4+ T cells were negatively associated with age. Our data extend these findings by showing that IFNα-induced pSTAT3 responses of CD8+ T cells were also negatively associated with frailty in both men and women and that IFNα-induced pSTAT5 responses in CD4+ T cells were negatively associated with frailty in older men. While previous studies found higher levels of baseline pSTAT levels at higher ages (16, 17) and with higher CRP levels (16), we found an opposite relationship with frailty instead, namely lower baseline pSTAT1 levels in monocytes of frailer men. Given these previous findings, this was unexpected since chronic low-grade inflammation is seen more often in frail individuals (28–30). Thus, it is unclear how this should be interpreted with regard to the theory that immune cells of frail older people show higher baseline JAK-STAT pathway activation. We speculate that baseline STAT phosphorylation is reduced in immune cells of frail elderly due to these cells being continuously triggered by chronic low-grade inflammation, leading to cellular exhaustion; further studies are needed to confirm this hypothesis. On the other hand, impaired cytokine signaling may cause reduced immune responses, leading to accumulation of damage-associated molecular patterns and thereby induction of chronic low-grade inflammation. Thus, while it is unclear whether chronic low-grade inflammation is a cause or a consequence of impaired cytokine signaling, both can enhance each other, are part of a dysfunctional immune system and probably lead to reduced immune responses that are clinically meaningful, due to their associations with frailty.
Of interest is the association we found in men between higher CRP levels in the previous 20 years, a sign of chronic low grade inflammation, and lower IL-10 induced pSTAT3 responses. Since we also observed lower STAT3 responsiveness to IL-10 in lymphocytes of frailer men, this might suggest that their lymphocytes are less responsive to anti-inflammatory signals and therefore less able to control inflammation. These findings are in agreement with results of previous studies showing that chronically elevated levels of CRP are related to frailty (28–30). Recent studies noted that severely ill COVID-19 patients show elevated levels of IL-10 (31) which led some investigators to propose that IL-10 might contribute to the severity of the disease, because IL-10 is known to have both pro- and anti-inflammatory properties (32). Given that the risk of developing severe symptoms in COVID-19 is greater in older men that in older women (33), and since we showed signs of defective IL-10 signaling in frail men, we speculate that defective downstream IL-10 -STAT3 signaling contributes to the severity of infectious diseases such as COVID-19 due to the reduced immune regulatory function of IL-10. Indeed, JAK-STAT pathway signaling has been shown to be pivotal in developing severe symptoms such as a cytokine release syndrome in COVID-19 patients (34), in particular through STAT3 signaling (35).
Many clinical trials are currently ongoing to investigate efficacy of blocking JAK-STAT signaling with JAK inhibitors in the treatment of COVID-19 (34). While phase II studies showed promising results (36, 37), early reports from phase III trials are mixed, with some positive (38) and some negative results (39). These mixed results might be explained by different timing of drug administration, and by improper identification of the patients who benefit most from JAK/STAT inhibition (34). This should be addressed in future studies which should also elucidate whether impaired regulatory IL-10 STAT3 signaling can be a marker of developing severe reactions to infectious diseases. This would strengthen the rationale for targeting this pathway. Thus, cellular pSTAT levels may qualify as a biomarker to help identifying patients that benefit most from JAK inhibitor treatment.
An important factor in the association between frailty and reduced cytokine responsiveness could be overweight, since overweight has been associated with impaired JAK-STAT responses in adipocytes (40). We were unable to adjust the results for BMI directly due to the small sample size. However, we did not find associations of pSTAT responses with BMI, which may imply that BMI is not the main driver in the association of pSTAT1, pSTAT3, and pSTAT5 responses in immune cells and frailty.
While we found that cytokine responsiveness is lower in frail people, we did not find an association of frailty with cellular cytokine production after stimulation with TLR agonists such as LPS, R848 and CpG ODN. This might mean that the initial intracellular reaction to pattern associated molecular patterns is still intact, but that the response to TLR-induced cytokines is reduced in frail individuals due to reduced JAK-STAT pathway signaling. Results should be interpreted with care, as the high dosage of the TLR agonists chosen for in vitro stimulation may overcome or mask possible subtle in-vivo differences in TLR responses.
To the best of our knowledge, this is the first study describing differences in pSTAT responses of leukocytes in men and women. Lower cellular pSTAT responsiveness in frailer participants was seen in both men and women, with the direction of the correlations being the same in men and women for most of the experimental conditions. However, impaired signaling in frailer participants was seen with different pSTATs in men and women, and in general the observed associations were stronger in men. Of interest is that pSTAT1 cytokine signaling was found to be more impaired in frail women than in men, while frail women were also found to have higher monocyte numbers. This is consistent with impaired phagocyte functioning in frail women, with more monocytes being produced due to their impaired cellular signaling capacity as a possible compensatory mechanism. On the other hand, pSTAT5 signaling was found to be impaired in frail men but not in frail women. It is known that pSTAT5 responses can be initiated by growth hormones (41). Growth hormone secretion is higher in women than in men and is influenced by sex hormones such as estrogens (41), which might explain why we found the associations between pSTAT5 responses and frailty in men but not in women.
Strengths of the study are the in-depth pSTAT signaling assay that we used, which allowed us to characterize a wide range of functional cellular cytokine response signals. Another advantage is that we could build on the extensive information and biomaterials gathered in the course of a unique longitudinal cohort study. This gave us the opportunity to compile and score a comprehensive frailty index in older individuals and to measure inflammatory markers longitudinally over a period of 20 years to quantify chronic low-grade inflammation. Our study also comes with limitations. First, our sample size was relatively low, so results cannot be extrapolated to the general population. Furthermore, the cellular pSTAT responsiveness could not accurately predict frailty by a prediction model due to the small sample size and the relatively small changes in effect size (data not shown). Another noteworthy point is that, while we used a large panel of cytokines signaling via JAK/STAT pathways we did not study activation of other known STATs, namely STAT2, STAT4 and STAT6. Larger phospho-flow panels may give a more complete description of cellular responsiveness in frail older people. Other innate signaling cytokines such as IL-4 and IL-8 could also be considered in future studies, especially since the function of myeloid lineage cells is thought to become impaired with age (4, 42) and numbers of myeloid cells are higher in frail older people (27). Also, how the results translate to in-vivo signaling remains to be investigated. In-vivo signaling might differ from in-vitro signaling, as the context is different and in vivo cytokine receptor expression might vary due to e.g. the presence of specific antigen, co-stimulation, or a difference in strength and duration of stimulation.