Vascular ageing and the complex relationship between cognitive decline and cardiovascular disease

This editorial refers to ‘Health status and cognitive function for risk stratification in chronic coronary and peripheral artery disease’, by K.G. Smolderen et al. https://doi.org/10.1093/eurjpc/zwac282 Several observational studies indicate that cardiovascular risk factors such as smoking and hypertension support the development of cognitive deficits in later life.1–3 In addition, cardiovascular disease (CVD) and cognitive disorders share pathophysiological pathways such as inflammation and increased oxidative stress.3–6 Further, associations of both clinical and subclinical CVD with impaired cognitive function1,2,5 were found. However, longitudinal studies regarding the association of cardiovascular risk factors such as diabetes and cholesterol levels with cognitive function and decline were inconclusive.2 A possible explanation for these ambiguous results is the complex pathways and biological processes that are involved in cognitive performance. To what extent cognitive function may augment cardiovascular risk prediction beyond the traditional cardiovascular risk factors, in CVD, is hence a matter of debate.5 In this issue, Smolderen et al.7 provide an intriguing answer to the question whether the assessment of cognitive function may be of practical use to (i) augment risk stratification in coronary artery disease patients and (ii) serve as a target for preventive intervention. The authors derive their strongly affirmative conclusions from a secondary analysis of the ‘Cardiovascular Outcomes for People Using Anticoagulation Strategies’ COMPASS trial data. The COMPASS trial was an international, double-blind, randomized superiority trial designed to compare, in patients with stable CVD, the preventive effects of rivaroxaban plus aspirin, rivaroxaban monotherapy, or aspirin monotherapy. A total of 27 395 patients [mean (standard deviation) age, 68.2 (7.9) years; 78.0% men] were assessed, with roughly equal numbers of patients randomized to each of the three intervention arms. The primary efficacy outcome was defined as the composite of cardiovascular death, stroke, or myocardial infarction [hereinafter summarily referred to as major adverse cardiovascular event (MACE)]. The assessment of cognitive function at baseline was performed using the Digit Symbol Substitution Test (DSST). The data of the COMPASS trial are an attractive source to probe for a suspected inverse association of cognitive function with the risk for MACEs for two reasons: its large number of participants and the DSST’s acknowledged sensitivity to both the presence and the change of cognitive dysfunction across a wide range of clinical populations.8 Smolderen et al. found the association between DSST score and MACE to be significant and strong: being in the lower two quartiles of the DSST carried a 33% (hazard ratio 1.33; 95% confidence interval 1.17–1.51) increased risk for suffering a MACE compared with the upper two quartiles. As expected, the associations between the number of risk factors and quartiles of health status (EQ VAS scores) and cognitive functioning (DSST, MoCA) were inverse (more risk factors, lower quartile). Scoring in the upper quartiles of health status correlated with higher scoring on cognitive function and younger age. The association between DSST-assessed cognitive dysfunction and CVD is well established. In the Atherosclerosis Risk in Communities (ARIC) study population, an inverse association between baseline cognitive function and subclinical CVD, expressed as elevated hs-CTnT (≥14 ng/L) and NT-proBNP (≥300 pg/mL), was observed, which was independent of all other baseline biomarkers of risk.1 Interestingly, study participants with the lowest DSST performance scored the highest on modifiable CVD risk factors, particularly unfavourable lifestyles. This observation aligns with another investigation of the ARIC study population demonstrating a strong correlation between cardiovascular health, assessed with the American Heart Association (AHA)’s Life’s Simple 7 (LS7) concept and score, and cognitive function.2 Each unit increase in LS7 scores confers a 2-year protection against global cognitive decline over the course of 20 years. Individuals scoring the highest on the LS7 (or its successor, the Life’s Essential 8 score, LE8) also have more than 60% reduced risk for CVD events.3 Taken together, these observations support the notion that health behaviour is an upstream cause shared by both CVD and cognitive dysfunction. While the pathways from lifestyle behaviours to cognitive impairment and CVD events are complex and incompletely understood, the nearly complete overlap of risk factors and causative health behaviours between cognitive and CVD has been well established.4 Equally well established is the centrality of systemic vascular function as a determinant and benchmark for cognitive and cardiovascular health.5 However, vascular function also progressively declines with increasing age. As the aforementioned risk factors and health behaviours accelerate or decelerate the rate of ageing-associated decline, the corresponding concepts of an early vascular ageing syndrome and its benevolent antagonist, supernormal vascular ageing (SUPERNOVA), have emerged.9 Early vascular ageing has been identified as a cause and determinant of cognitive dysfunction6 as well as of CVD.10 In conclusion, these observations suggest that health behaviour may be considered a confounder of the supposedly causal correlation between cognitive function and CVD risk, and vascular function may serve as the objectively measurable proxy for this confounder. Hence, the authors’ suggestion that bringing patients to a higher performance level of cognitive function will offer the potential to reduce MACE by 22–29% must be taken with an important caveat: unless the improvement in cognitive performance is paralleled by a clinically relevant improvement of vascular function, any improvement in DSST score is unlikely to yield a commensurate risk reduction for MACE. An acknowledged proxy of vascular function is arterial stiffness, of which pulse wave velocity (PWV) is an easily obtainable surrogate marker that independently predicts cardiovascular and cognitive decline, and which is modifiable by lifestyle interventions.11 Pulse wave velocity, aside from being considered a marker for hypertension-mediated organ damage (PWV >10 m/s), has emerged as an independent risk factor for CVD. Enabled by new oscillometric devices, its automated and therefore simplified measurement should facilitate a wider adoption of PWV assessment in daily clinical practice. A hoped-for consequence is to get a step closer to individualized preventive care. Since PWV responds sensitively to lifestyle changes and other treatments, it could serve as an early biomarker for the effectiveness of interventions that are aimed at bringing the right treatment to the right patient at the right dose and time. What has prevented us from achieving this goal is the inextricable complexity between an individual’s physiology, genome, microbiome, lifestyle, and environment.12 Smolderen et al.’s investigation draws our attention to the fact that cognition adds one more dimension in this space of complex interactions. But a sensitive marker such as PWV would facilitate a trial-and-error approach to finding the optimally individualized preventive treatment, while the complexity of interactions still remains a black box. We might be closer than we think to realizing this goal. A methodically and statistically sound trial-and-error approach necessitates repeated, ideally daily, measurement of the outcome variable (PWV). The advent of wearables and layperson-capable mobile health tech makes daily remote assessment of relevant biomarkers (including PWV) accessible to almost everyone. Also, current computer technology enables processing of these data within the context of N-of-1 trial methods, the 1-subject equivalent of clinical trials, using validated statistical and experimental techniques to investigate the effects of treatments on a single individual.13 We thank the authors for their very valuable contribution to what may well turn out to be the starting point of a lively discourse about how to more efficiently prevent the two most disabling health scourges afflicting our societies: cognitive impairment and CVD.