Background
Depression is common in late life [
1] and associated with diverse aetiological factors that are poorly understood. Evidence of an association between vascular disease and late-life depression (LLD) led to the 'vascular depression hypothesis’ which proposed that structural damage to the frontostriatal tracts resulting from ischemic cerebrovascular disease creates a vulnerability to depression in late life [
2,
3]. This hypothesis stimulated much research and laid the foundation for examining the mechanisms by which vascular disease influences the development of depression in older adults. As a result, a mechanistic model for vascular depression has recently been proposed that identifies hypoperfusion as a potential mechanistic path to vascular depression [
4].
Blood flow to the brain is influenced by systemic hemodynamics and cerebrovascular autoregulation, whereby cerebral arteries contract or dilate in response to arterial blood pressure (BP) changes. These processes interact to maintain stable perfusion but are impaired in the context of vascular disease leading to perfusion deficits and potentially the development of white matter hyperintensities (WMHs) [
4]. White matter is sensitive to transient ischemia because it is supplied by terminal arterioles with limited collateral flow and so is more susceptible to minor changes in flow due to impaired autoregulation and consequent infarction [
5]. WMHs are believed to contribute to the pathogenesis of LLD since MRI studies have repeatedly found higher densities in the frontal lobes [
6‐
8] and basal ganglia [
8] of older people with major depression (MD) and these regions of the brain are known to regulate mood and cognition. Although the pathology of WMHs varies, autopsy studies have shown that they are associated with ischemic damage and tissue hypoxia in patients with late-life major depression [
9]. In animal models white matter ischemia results from drops in blood pressure [
10] and a recent study [
11] has shown a relationship between the degree of orthostatic drop in BP and WMH volume in LLD.
Hemodynamic homeostasis becomes less effective with aging and is associated with a decreased ability to regulate BP [
12]. Orthostatic hypotension (OH) is an excessive fall in BP after standing which may result from inadequate intravascular volume, autonomic nervous system dysfunction, decreased venous return or an inability to increase cardiac output in response to postural changes [
13]. After essential hypertension, OH is the most common disorder of BP regulation and in adults aged 65 years and older the prevalence ranges between 5 and 30%; depending on the population studied and definition used [
14]. Prospective studies have shown that OH is predictive of ischemic stroke [
15] and all-cause mortality in older adults [
16].
In 1996, a consensus committee defined classical OH as a drop of at least 20 mmHg in systolic blood pressure (SBP) or 10 mmHg in diastolic blood pressure (DBP) within the first 3 minutes of standing [
17]. This may reduce perfusion pressure of organs, especially above heart level, such as the brain. OH can be asymptomatic or manifest as symptoms that range from dizziness and light-headedness to weakness and loss of consciousness. These symptoms arise from periods of hypoperfusion that have the potential to induce ischemic damage and WMHs [
14,
18]. For this reason, it has been suggested that OH may play an important role in the development of depression and accumulating evidence, although limited shows that OH may be more frequent in late-life depression [
19,
20]. Previous studies confirming an association between OH and depression were conducted on small clinical cohorts of patients with a formal diagnosis of MD and using hospital based digital photoplethysmographic measures of phasic BP. It is not known whether the association would persist in the general population; where lower prevalence and severity rates for these conditions are observed, and using traditional oscillometric measures of BP. Consequently, the aim of this study is to investigate if OH is associated with depression in a nationally representative sample of older adults utilising a standard test of OH and a self-report measure of depression.
Discussion
In this study, we found evidence of a relationship between orthostatic hypotension and depression in participants with OH who reported symptoms of hypoperfusion on standing. Emerging evidence suggests that hypoperfusion may be an important factor in the development of late-life vascular depression [
4]. OH can lead to cerebral auto-dysregulation [
28] which can impair cerebral blood flow (CBF) [
29]. Perfusion deficits do not need to cause ischemia in order to influence brain function. Mild CBF reduction impairs protein synthesis [
30] that is crucial for cognitive and affective processing while greater CBF reduction in the context of autoregulatory deficits may cause ischemic injury and WMHs [
4]. Previous results linking OH and depression have been observed among clinical samples [
11,
19,
20] where cases of OH and depression are respectively more prevalent and severe. Our findings suggest that in a representative sample of community dwelling older adults, the relationship is limited to those with more severe OH; since the association is dependent on the presence of symptoms of hypoperfusion.
9% of participants with OH reported symptoms of dizziness/light-headedness on standing. Cerebral autoregulation (CA) explains why OH does not produce symptoms in all adults. Under normal physiologic conditions, changes in mean arterial BP between 60 and 160 mmHg produce little or no change in CBF. Beyond these limits, a sudden decrease in CBF occurs at the lower limit of autoregulation and likely manifests as orthostatic symptoms [
29]. In patients with OH, severe autoregulatory failure is believed to occur in approximately one in four patients and cerebral hypoperfusion occurs in response to relatively small changes in BP [
29]. Participants in our study with symptomatic OH may have impaired CA and more extreme changes in CBF resulting in orthostatic symptoms. In contrast, those with asymptomatic OH may have better preserved autoregulation centrally and therefore no symptoms of hypoperfusion from reduced CBF, despite clinically significant drops in BP during standing. While a causal relationship between WMH and cerebral hypoperfusion induced by OH has yet to be established, both animal models and human studies suggest that this exists [
31,
32] and future waves of TILDA will investigate if OH at baseline predicts depression at follow up.
The prevalence of hypertension was higher in participants with OH which could have confounded results since hypertension is a known risk factor for OH [
33] and WMHs [
34]. However, hypertension was not associated with increased CES-D score in univariate analysis and the addition of systolic and diastolic BP levels to the model only slightly attenuated the association between SOH and depression. The use of antihypertensive medication was also somewhat higher in the OH group; however, the relationship between SOH and depression also persisted even after adjusting for these medications in the final model. In healthy community-dwelling adults such as our participants these drugs are associated with only low rates of orthostatic hypotension [
11,
35] making them unlikely to have influenced our findings.
Participants currently taking antidepressants were excluded from our analysis. There is no simple way to 'adjust’ for anti-depressant use in our baseline data since the model we are trying to estimate is essentially cyclical (i.e. our hypothesis is that OH causes depression, but we know depression results in anti-depressant use and anti-depressant use can cause OH).
7% of our target population were currently taking antidepressants (see Figure
1). Sensitivity analysis including participants on antidepressants (and controlling for antidepressants in the model) attenuated but did not eliminate the significant association between SOH and depression and overlapping confidences intervals between estimates derived from the sample with antidepressant users, and without them, suggests no significant difference between estimates .
The STST is commonly used in everyday clinical practice, however it is believed to have inferior sensitivity for detecting OH compared to other orthostatic stress tests (Active Stand Test or the Head-Up Tilt-Table Test) [
36]. Automated BP machines can take 25–40 seconds to record a measurement. We therefore expect that our findings underestimate the prevalence of OH (compared with prevalence rates estimated from phasic BP measurement) and potentially any associations uncovered; given that in some participants BP may have recovered before the recording was completed. Interestingly, orthostatic symptoms were only reported by participants with OH which would suggest that the STST has good sensitivity for detecting those with severe OH. Previous studies linking OH and depression have utilised continuous, beat-to-beat BP monitoring to measure OH in patients with MD at centre based assessments. Home respondents are known to differ markedly from the centre based attendees and in a TILDA pilot study we showed that participants who selected a home assessment had poorer physical health with higher levels depressive symptomology [
37]. The STST is easily administered in an individual’s home therefore it maximised the participation of older participants in our study and was effective for identifying the association between symptomatic OH and depression.
The main strengths of this study are the large population representative sample and comprehensive health assessment. Limitations are that the data was analysed cross-sectionally so we cannot make any direct inferences about temporality. Although many relevant risk factors were assessed and statistically controlled for in the analysis, the possibility of residual confounding exists. In particular psychotrophic medications are captured by a single variable which could pool agents that cause OH with those that do not. Depression was not formally diagnosed in our study although the CES-D instrument is a well validated measure of depressive symptomatology [
22,
23]. The prevalence of MD in community samples of older adults ranges from 1-5%, with the majority of studies reporting prevalence in the lower end of the range [
38‐
41]. It is therefore likely that the CES-D overestimated the prevalence of MD and our measure of depression represents both major and subthreshold depression. Another limitation is that our primary conclusions were drawn from the SOH group which only included 20 individuals; the potential for spurious findings therefore exists. It must also be acknowledged that in some participant’s vestibular dizziness rather than hypotension could have caused the symptoms of dizziness although in our study all participants who reported symptoms had a clinically significant drop in blood pressure. Moreover, dizziness may also be a feature of anxiety which is highly co-morbid with depression in this age group [
42].
Conclusion
In summary, we have identified an association between symptomatic OH and depression in a large community-based sample of older adults. Symptomatic orthostatic hypotension may be an important, but not widely recognised, risk factor for late-life depression which needs to be considered in studies examining the relationship between vascular disease and depression. Clinically, our results are important as they show that the relationship between SOH and depression is not limited to clinical samples but extends to the general population and potentially individuals with subthreshold depression. Clearly longitudinal studies are required to clarify the direction of this association.
Acknowledgements
The authors would like to acknowledge the contribution of the participants in the study, members of the TILDA research team, study nurses and administrators.
Funding for the TILDA study is provided by The Atlantic Philanthropies, Irish Life plc and the Irish Government.
Competing interests
None of the authors have any financial or non-financial competing interests to disclose.
Authors’ contributions
All authors were involved in the conception and design, analysis and interpretation of data. COR drafted the article and PK HC BL and RAK revised and edited the manuscript. GS provided statistical input. All authors gave final approval of the version to be published.