Relation of the kynurenine pathway with normal age: A systematic review

Background: The kynurenine pathway (KP) is gaining more attention as a common pathway involved in age-related conditions. However, which changes in the KP occur due to normal ageing is still largely unclear. The aim of this systematic review was to summarize the available evidence for associations of KP metabolites with age. Methods: We used an broad search strategy and included studies up to October 2023. Results: Out of 8795 hits, 55 studies were eligible for the systematic review. These studies suggest that blood levels of tryptophan decrease with age, while blood and cerebrospinal fluid levels of kynurenine and its ratio with tryptophan increase. Studies investigating associations between cerebrospinal fluid and blood levels of kynurenic acid and quinolinic acid with age reported either positive or non-significant findings. However, there is a large heterogeneity across studies. Additionally, most studies were cross-sectional, and only few studies investigated associations with other downstream kynurenines. Conclusions: This systematic review suggests that levels of kynurenines are positively associated with age. Larger and prospective studies are needed that also investigate a more comprehensive panel of KP metabolites and changes during the life-course.


Introduction
Ageing is a process which involves (epi)genetic, environmental, lifestyle, and stochastic factors, and is characterized by alterations in different biological pathways, which have been connected to an increased risk of age-related diseases (Capuron et al., 2011;López-Otín et al., 2013;WHO, 2021).A better understanding of the mechanisms behind age-dependent biological changes is vital for the early diagnosis of age-related disorders and the identification of new treatment targets.
Recently, the kynurenine pathway (KP) has gained attention as a common pathway involved in age-related conditions such as inflammatory diseases, neurodegenerative diseases, neurologic disorders, and cognitive dysregulation, among others (Solvang et al., 2019;Valdiglesias et al., 2017;Xyda et al., 2020;Kepplinger et al., 2019;Theofylaktopoulou et al., 2013).The KP is the main metabolic route of tryptophan (TRP) degradation (Fig. 1).TRP is an essential amino acid and can be degraded into kynurenine (KYN) by the enzymes tryptophan 2,3-dioxygenase (TDO2) and idoleamine 2,3-dioxygenase 1 and 2 (IDO1, IDO2).TDO2 is mainly expressed in the liver and can be induced by stress hormones, while IDO1 and IDO2 are present in most cells and are activated by several pro-inflammatory mediators (Ueland et al., 2017).Metabolites of the KP are called kynurenines, which display neuroactive properties and can modulate immune function (Ueland et al., 2017).As a result, kynurenines have been proposed to mediate diverse age-related processes and diseases (Sorgdrager et al., 2019;Vecsei et al., 2013).
Previous studies suggest that ageing might be associated with changes in activation of the KP.Some have reported lower levels of circulating TRP in blood and cerebrospinal fluid (CSF), and, congruently, levels of kynurenine (KYN) and of more downstream KP metabolites have been found to be higher (Valdiglesias et al., 2017;Theofylaktopoulou et al., 2013;Darst et al., 2019;Sipahi et al., 2013;Bie et al., 2016).However, not all studies found similar associations (Xyda et al., 2020;Wennstrom et al., 2014;Hartai et al., 2007;Frick et al., 2004).As such, the role of the kynurenine pathway in ageing is still largely unclear.Therefore, the aim of this systematic review was to summarize the available evidence for the associations between KP metabolites measured in different tissues and ageing in human populations.

Methods
The current study is part of a larger systematic review in which associations of the KP with a) age and b) age-related neurodegenerative and neurocognitive disorders were investigated.Here, we report on the former part of the study.

Literature search
The systematic review was performed according to the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 2020 guidelines (Page et al., 2021a(Page et al., , 2021b)).A comprehensive systematic literature search was done using PubMed, Embase, PsychINFO, and the Cochrane Database of Systematic Reviews.A search was done to include studies published until October 2023.The final search term was as follows: (tryptophan OR kynuren* OR anthranil* OR xanthurenic OR cinnabar* OR Picolinic OR Quinolinic) AND (Alzheim* OR dementia OR demented OR cogn* OR neurocogn* OR memory OR amnestic OR amnesia OR neuropsychol* OR aging).No filters were used during the search.Additionally, a snowballing systematic literature approach, by screening the references in the selected studies, was carried out.This study was registered in the International Prospective Register of Systematic Reviews (PROSPERO) database on March 11 2020 (ID: CRD42020159274).

Study selection
Two main reviewers (L.B. and K.C.) independently read and assessed the eligibility of the articles based on title and abstract using Endnote x9 software, followed by assessment of the full text, where applicable.Differences in opinion were resolved through structured discussion or by consulting a third reviewer (S.K.).Data was extracted using a prespecified data extraction form (Supplementary Appendix S1).Missing data were handled by contacting the corresponding author.

Quality assessment
Study quality was assessed using the Newcastle-Ottawa Scale (NOS) for cohort studies (Table S4) (Wells et al., 2000).An adapted version was used for cross-sectional data (Appendix S2).The scales consist of three categories (Selection, Exposure, and Comparability), of which a maximum of one star for each numbered item within the category (Selection n = 4, Exposure n = 2, Comparability n = 3) can be awarded, nine in total.

Characteristics of the included studies
After de-duplication, 8774 abstracts were retrieved from the search (Fig. 2).Additionally, 21 studies were found through reference lists or review papers.79 full-text publications were assessed for eligibility, of which 55 were included in the qualitative synthesis.A detailed overview of all the included studies is provided in supplementary table S1.

Evidence from longitudinal studies
Most studies included cross-sectional data only (n = 53).Only two studies contained prospective data with repeated measures of metabolite levels.The first study investigated the influence of age on metabolites over a period of 10 years (Darst et al., 2019).Follow-up took place every two years and blood samples were drawn on the morning of each visit.Each participant had a maximum of three study visits.After adjustment for sex, self-reported race, BMI, storage time, cholesterol lowering medication use, and corrections for multiple testing, TRP decreased with ageing, whereas KYN increased (Darst et al., 2019).The second study investigated associations between kynurenines and age in four different cohorts, of which two (The Melbourne Collaborative Cohort Study [MCCS] and the Western Norway B vitamin Intervention Trial [WENBIT]) contained repeated metabolite measurements in plasma (Solvang et al., 2022).Additionally, one cohort, the Elective Surgery Cohort (COGNORM), contained a smaller subsample (33/109) of participants who donated a second CSF sample after four years (Solvang et al., 2022).In both MCCS (median follow-up of 11 years) and WENBIT (median follow-up of one and three years), levels of KYN, QA, and KTR, and to a lesser extent of 3-HK, KA, and 3-HAA, increased over time, and levels of TRP decreased (Solvang et al., 2022).Levels of AA increased over time in WENBIT, but decreased in MCCS.Additionally, levels of XA increased over time in WENBIT as well.Lastly, in the COGNORM study, CSF levels of KYN, 3-HK, and QA increased over time, while there was no significant change in levels of TRP, KA, AA, or PIC (Solvang et al., 2022).
Levels of KA were also higher in the older group in CSF (Kepplinger et al., 2019(Kepplinger et al., , 2005) ) or blood (Kepplinger et al., 2019;Theofylaktopoulou et al., 2013) or not different among age groups in blood (Kepplinger et al., 2005;Westbrook et al., 2020).CSF levels of QA were higher in the older group in one study (Heyes et al., 1992).One other study investigated differences in serum levels of QA between two age groups, but reported no differences (Westbrook et al., 2020).Other downstream metabolites were again less often studied.Studies investigating levels of 3-HK reported higher blood levels in the older group (Theofylaktopoulou et al., 2013) or no difference in blood levels of 3-HK between two age groups (Westbrook et al., 2020).XA blood levels were either lower in the older group (Theofylaktopoulou et al., 2013) or not significantly different (Westbrook et al., 2020), whereas AA blood levels were either higher in the older group (Theofylaktopoulou et al., 2013) or not significantly different (Westbrook et al., 2020).3-HAA was only reported in one study and was higher in plasma of the older group (Theofylaktopoulou et al., 2013).One study had a relatively small age difference between the groups (74 verus 71 years), but reported higher levels of 3-HK, QA, and KTR in the older group, whereas levels of KA, XA, AA, 3-HAA, and PIC were not significantly different between the age groups (Solvang et al., 2022).

Discussion
This systematic review identified and summarized 55 studies that investigated associations between KP metabolites and age.Results suggest that blood levels of TRP tend to be lower among older individuals, while blood and CSF levels of KYN and KTR are higher.Studies investigating associations between CSF and blood levels of KA and QA and age reported either positive or non-significant findings.Although only a small number of included studies investigated associations of other downstream kynurenines (e.g.3-HK, 3-HAA, AA, XA and PIC) with age, these reported mostly positive or non-significant findings as wel.Almost all studies were cross-sectional with considerable methodological heterogeneity between studies.

Associations between blood levels of TRP, KYN, and KTR and age
The findings of lower blood levels of TRP and higher levels of KYN and KTR with older age suggest that ageing is associated with increased degradation of TRP via the KP.This is in line with an in vivo study in rats showing that TRP levels decreased in the liver and kidney with ageing, while KYN levels showed a tendency to increase (Braidy et al., 2011).Studies have shown an age-associated increase in chronic low-grade inflammation, also known as inflammaging (Furman et al., 2017;Ferrucci and Fabbri, 2018), reflected by an increase in levels of pro-inflammatory biomarkers including IFN-γ and TNF-α (Michaud et al., 2013).It is well-established that these inflammatory stimuli activate IDO, which is one of the rate-limiting enzymes responsible for TRP degradation (Harden et al., 2016a).As a result, KTR is often included in studies as a marker for IDO activation and inflammation (Baumgartner et al., 2019).In a study in mammals, KTR in the liver was negatively correlated with longevity (Ma et al., 2015).Similarly, according to our systematic summary, studies most consistently report positive associations of KTR with age.In line with this, those studies that incidentally reported on neopterin (it was not included in our search term) also generally found a positive association with age (Capuron et al., 2011;Valdiglesias et al., 2017;Theofylaktopoulou et al., 2013;Sipahi et al., 2013;Bie et al., 2016;Frick et al., 2004;Dugué et al., 2022) (supplementary tables S2 and S3).Neopterin is an inflammatory marker and it is suggested to reflect activation of pro-inflammatory cytokines and immune cells such as macrophages and microglia (Schroecksnadel et al., 2004) and is associated with activation of the KP (de Bie et al., 2016).
antioxidant responses is disrupted with an increase in oxidative damage (Sastre et al., 2000).Though TRP is best known for its role as a biochemical precursor for the kynurenine and serotonin pathways, studies have shown that TRP has antioxidant properties such as scavenging free radicals, as well as reactive oxygen species (ROS) and reactive chlorine species, and has the highest antiradical activity compared to other amino acids (Bitzer-Quintero et al., 2010;Weiss et al., 2002;Pazos et al., 2006).KYN is a central substrate to produce several downstream kynurenines (e.g.KA, 3-HK, and AA), and studies have reported both pro-and anti-oxidant properties through the generation of-or as a scavenger of ROS (Mor et al., 2021).As such, differences in levels of TRP and KYN with age, as suggested in our systematic review, might play a role in mediating changes in antioxidant capacity that are typically observed in ageing.Clearly, more research in this area is needed.
QA is another well-known metabolite, known for its agonistic properties of NMDA receptors, with the potential to induce glutamateinduced excitoxicity (Badawy, 2017).Additionally, QA induces oxidative stress, (neuro)inflammation, and mitochondrial dysfunction (Chen and Guillemin, 2009;Stone and Perkins, 1981;Bo et al., 2018).Both results from longitudinal and cross-sectional studies included in our review showed predominantly positive associations of QA with age, which were consistent across CSF (Bie et al., 2016;Solvang et al., 2022;Heyes et al., 1992;Dugué et al., 2022) and blood (Darst et al., 2019;Solvang et al., 2022;Dugué et al., 2022).This is in line with an older study in rodents, which reported that cortical levels of QA increased with ageing (Moroni et al., 1984).Additionally, studies in humans have shown that QA is positively correlated with C-reactive protein and neopterin, which are markers indicative of a pro-inflammatory immune status (de Bie et al., 2016;Zheng et al., 2022).Thus, through activation of the KP, and of QA in particular, inflammaging may contribute to an increased risk for age-related diseases.
At the same time, QA is a precursor of nicotinamide adenine dinucleotide (NAD+), a key metabolite involved in a large array of cellular metabolic pathways.NAD+ decreases with aging and is declined in various age-related diseases (Wang et al., 2020;Covarrubias et al., 2021).Interestingly, while activation of the KP increases upon inflammation, de novo synthesis of NAD+ in macrophages is suppressed (Minhas et al., 2019).QA is metabolized into NAD+ by quinolate phosphoribosyl transferase (QPRT), an enzyme that decreased in aging rats (Braidy et al., 2011).As such, one reason for increased QA levels with aging might be due to a decrease of QA degradation by QPRT.

Associations between kynurenines and age-related diseases
Through their modulatory role in immune response and neuroactive properties, metabolites of the KP have been associated with a variety of age-related diseases, including neurodegenerative-and inflammatory diseases (Chen and Guillemin, 2009;Guillemin et al., 2001;Belladonna et al., 2006;Harden et al., 2016b).For instance, a recent meta-analysis reported lower blood levels of TRP in Alzheimer's disease (AD), Parkinson's disease and Huntington's disease compared to healthy controls (Fathi et al., 2022).In individuals with AD, no differences were reported in levels of KYN in either blood or CSF.However, CSF levels of KA were higher (Fathi et al., 2022).With respect to post-mortem brain tissue, studies generally report no differences in levels of TRP or QA levels between individuals with AD and healthy controls (Mourdian et al., 1989;Paglia et al., 2016;Storga et al., 1996).Additionally, associations have been found of kynurenines with cancer (Chen and Guillemin, 2009;Litzenburger et al., 2014;Liebau et al., 2002;Yu et al., 2014), respiratory disorders (Fox et al., 2015), arthritis (Lim et al., 2016), and obesity (Favennec et al., 2015).
Next to age-related diseases, the KP has been associated with psychiatric disorders as well.For instance, two independent meta-analyses reported lower blood levels of TRP (Marx et al., 2020), KYN (Marx et al., 2020;Ogyu et al., 2018), and KA (Marx et al., 2020;Ogyu et al., 2018) in individuals with depression in comparison to healthy controls, while blood levels of QA were higher (Ogyu et al., 2018).One hypothesis is that serotonergic dysregulation in patients with a major depressive disorder is strongly associated with cytokine-induced IDO activation, switching degradation of TRP via the serotonin pathway to TRP degradation via the KP (Oxenkrug, 2013;Miura et al., 2008).While there is no direct association between ageing and depression, the association between inflammation and depressive symptoms might increase with age (Straka et al., 2021).Whether changes in the KP might be a modulating pathway in this respect needs further research.

Methodological considerations and future directions
The main strength of this systematic review includes the extensive search, incorporating a structured search strategy and snowballing approach by which, in the end, 55 studies could be included.The main limitations of our systematic review are the relatively small sample sizes of the studies included and almost complete lack of longitudinal studies, which indicates the need for more and larger studies with repeated analyses to investigate the role of the KP in ageing.Most studies also investigated kynurenines in blood, whereas studies investigating levels in the CSF were limited.Additionally, the heterogeneity of the included studies likely led to considerable inconsistencies between findings, and also made it impossible to do a meta-analysis.For instance, age ranges were substantially different between studies, limiting the comparability of results.Finally, we limited publications to articles written in English and as such might have excluded other articles that would otherwise have met inclusion criteria.
In order to assess the changes over time of levels of kynurenines, the field would highly benefit from more longitudinal studies.Additionally, future studies should investigate various samples (e.g.plasma, serum, CSF, and urine) simultaneously.Furthermore, studies should investigate more downstream kynurenines and relevant ratios to allow a full picture of the KP and its association in the pathophysiology of ageing and disease.Lastly, clinical studies should adjust for important factors that are known to be associated with kynurenine concentrations such as sex, renal function, and lifestyle factors (e.g.smoking habits, alcohol use, and body mass index) (Theofylaktopoulou et al., 2013;Darst et al., 2019;Sarwar et al., 1991;Deac et al., 2015;Tomioka et al., 2020;Ngandu et al., 2007;Stirland et al., 2018;Anstey and Peters, 2018;Pawlak et al., 2002;Seliger et al., 2004;Mangge et al., 2014;Cournot et al., 2006;Saito et al., 2000;Schefold et al., 2009;Lischka et al., 2022).Previous studies for instance reported higher blood levels of TRP, KYN, KA, XA, AA, and 3-HAA in men than in women (Theofylaktopoulou et al., 2013;Darst et al., 2019;Sarwar et al., 1991;Deac et al., 2015;Tomioka et al., 2020) and results from our systematic review suggest that associations between kynurenines and age might be different according to sex as well, emphasizing the need for more research investigating these sex-specific differences.Additionally, kynurenines were higher in individuals with reduced renal function (Theofylaktopoulou et al., 2013) and renal disease (Saito et al., 2000;Schefold et al., 2009).

Conclusion
This systematic review indicates large heterogeneity across studies and suggests that levels of some kynurenines are either positively or not significantly associated with age.Larger and prospective studies are needed that investigate a more comprehensive panel of KP metabolites.

Table 1
Associations between kynurenines and age in cerebrospinal fluid, serum and plasma.

Table 2
Differences in kynurenines between older and younger age groups.