Affective disorders and hypertension are among the most common chronic mental and physical conditions that account for a large proportion of premature death and disability worldwide. The two conditions cluster, occurring together more commonly than would be expected from their individual prevalence rates. Reference Davidson, Jonas, Dixon and Markovitz1–Reference Meng, Chen, Yang, Zheng and Hui4 Individuals with hypertension are estimated to be 1.76 times more likely to have moderate to severe symptoms of depression (Patient Health Questionnaire, PHQ-9 > 10) compared with those without hypertension (odds ratio 1.76, 95% CI: 1.14–2.74). Reference Maatouk, Herzog, Böhlen, Quinzler, Löwe and Saum5 Furthermore, both hypertension and depression are risk factors for cardiovascular disease and mortality. Reference Hare, Toukhsati, Johansson and Jaarsma6,Reference Fuchs and Whelton7 However, there is poor understanding of the biological mechanisms underpinning the association between hypertension and depression.
Chronic stress, a factor commonly linked with the development of depression and hypertension, causes tonic activation of the locus coeruleus, which enhances sympathetic tone and increases norepinephrine turnover. This effect is particularly evident in the prefrontal cortex and hippocampus, and is mediated by the amplification of sympathetic tone in the peripheral nervous system. Reference Spruill8–Reference Samuels and Szabadi11 Consequently, pharmacological interventions targeting the sympathetic nervous system could improve depression and hypertension concurrently while reducing the associated risk of cardiovascular disease. However, beyond a small number of studies examining urinary catecholamines, Reference Hughes, Watkins, Blumenthal, Kuhn and Sherwood12,Reference Paine, Watkins, Blumenthal, Kuhn and Sherwood13 the mechanistic link between depression and hypertension has not been explored.
In the present study we examined the association of symptoms of affective disorders with plasma concentrations of normetanephrine, a measure of chronic sympathetic nervous system activation, Reference Peaston and Weinkove14 in patients with primary hypertension. We also examined responses to mental stress and to device-guided breathing (DGB), interventions that act through the same central pathways implicated in chronic stress, to stimulate or inhibit stress-induced vascular responses, and examined the association of these responses with plasma concentrations of normetanephrine. We hypothesised that chronic sympathetic nervous system activation, as indicated by high plasma normetanephrine concentrations, would be associated with worse anxiety and depressive symptoms, and with impaired stress-modulated vasomotor responses.
Method
Study design and population
Subjects with primary hypertension and aged ≥18 years were consecutively recruited from the hypertension clinic at Guy’s and St Thomas’ Hospital NHS Foundation Trust, London, UK. Hypertension was diagnosed based on previous anti-hypertensive treatment and/or daytime ambulatory blood pressure ≥135 mmHg systolic and/or 85 mmHg diastolic, in line with current guidelines from the European Society of Hypertension. Reference Williams, Mancia, Spiering, Rosei, Azizi and Burnier15 Exclusion criteria included pregnancy, subjects with secondary hypertension (including hypertension secondary to drug use), use of antidepressants or beta-blockers, renal impairment, established cardiovascular diseases other than hypertension or any other significant comorbidities apart from hypertension mood disturbance. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation, and with the Helsinki Declaration of 1975 as revised in 2013. All procedures were approved by the local Research Ethics Committee (ref. no. 12/LO/1473). Written informed consent was obtained from all participants.
Demographics and anthropometrics
Detailed social and medical history, including smoking status, alcohol consumption, treatment and comorbidities, was obtained followed by blood sampling for biochemistry, anthropometric measurements and cardiovascular stress responses. Diabetes was diagnosed based on either previous/current diabetes treatment or glycosylated haemoglobin (HbA1c) ≥6.5% (48 mmol/mol) in untreated individuals. Ethnicity was assigned as ‘self-defined ethnicity’, grouped into either White or Black and minority ethnicity.
Depression, anxiety and sleep disturbance
Self-report questionnaires in written format were used to assess depression, anxiety and sleep symptom severity. The Patient Health Questionnaire (PHQ-9) and the 16-item Quick Inventory of Depressive Symptomatology- Self Report (QIDSSR-16) were used to determine depressive symptoms; Generalized Anxiety Disorder-7 (GAD-7) for anxiety; and Epworth Sleepiness Score (ESS) and STOP-BANG for evaluation of sleep disturbance and presence of obstructive sleep apnoea (OSA). For these questionnaires, higher scores indicate worse symptom severity. These questionnaires are psychometrically robust and commonly used in adult populations, exhibiting high test–retest reliability and high concurrent validity. Reference Trivedi, Rush, Ibrahim, Carmody, Biggs and Suppes16–Reference Johns20
Biochemistry analysis
Blood biochemistry analysis was performed at ViaPath Laboratories, Guy’s and St Thomas’ Trust, London, UK. Plasma biochemistry was determined from samples obtained following 15 min of rest in supine position. Blood samples were collected from the antecubital fossa into ethylenediaminetetra-acetic acid-containing tubes, mixed, placed on ice then centrifuged within 30 min of collection. Plasma was aliquoted into cryovials and stored at −80°C. Assays were performed in duplicate from these aliquots with internal quality controls, maintaining inter- and intra-assay coefficients of variation below 10%. Measurements included creatinine, electrolytes, fasting glucose, lipid profile, plasma renin, aldosterone, normetanephrine and metanephrine. Normetanephrine is a metabolite of norepinephrine that, in the absence of a catecholamine-secreting tumour, derives predominantly from peripheral sympathetic nerves. From the perspective of the present application, it has the advantages over norepinephrine in that it is less influenced by acute stress and that plasma concentrations of normetanephrine taken under basal conditions hence represent chronic activity of the sympathetic nervous system. Reference Peaston and Weinkove14
Stress-induced cardiovascular responses
Participants were asked to abstain from caffeine, strenuous exercise and alcohol for 24 h before the visit. All measurements were performed in a quiet, temperature-controlled vascular laboratory (23–25°C). Cardiovascular measurements were performed at baseline following ≥15 min of rest in supine position, then during mental stress and DGB with these interventions performed consecutively in random order and with the second intervention following a period of rest, after which further baseline measurements were recorded. Mental stress was elicited using the Stroop colour-word test for a period of 5 min. This is a standardised test utilised as a cognitive stressor capable of inducing emotional responses and heightened levels of autonomic reactivity. Reference MacLeod21,Reference Parris, Wadsley, Hasshim, Benattayallah, Augustinova and Ferrand22 DGB was performed using the Resperate device (InterCure Ltd., Lod, Israel). This is a biofeedback device consisting of a respiration sensor and headphones with feedback sounds generated to prolong the expiratory phase of each breath, thereby reducing the breathing rate to <10/min.
Cardiovascular measurements comprised blood pressure, heart rate and forearm blood flow (FBF). Brachial blood pressure was obtained using an Omron HEM 705-CP semiautomatic oscillometric recorder (Omron Health Care, Tokyo, Japan). An average of three consecutive readings of systolic blood pressure (SBP), diastolic blood pressure (DBP) and heart rate was recorded. FBF was measured using venous occlusion strain gauge plethysmography (Hokanson, Inc., Washington, USA) with strain gauges that were electrically calibrated. Cuffs placed on the wrist were inflated to a suprasystolic blood pressure level to prevent any contribution of hand blood flow, whereas venous collecting cuffs on the biceps were inflated intermittently to 40 mmHg for 10 s. FBF was measured from the initial increase in arm circumference for an average of five or more upper arm cuff inflations. Forearm vascular resistance (FVR) was calculated from mean arterial pressure (MAP) divided by FBF.
Statistical analysis
Descriptive summary statistics were stratified according to tertiles of plasma normetanephrine determined following the recruitment of the cohort. Mean ± s.d. was used for continuous variables (unless otherwise stated), and counts and percentages for categorical variables. Vasodilation/vasoconstriction and FVR were analysed as absolute numerical change, as well as by percentage change from immediately preceding baseline measurements. Differences between groups were analysed using one-way analysis of variance (ANOVA) and analysis of covariance (ANCOVA) for normally distributed values. Post hoc Bonferroni analysis was used to assess the significance of differences between groups. The independent-samples Kruskal–Wallis test was utilised for non-normally distributed variables, with ANCOVA used to adjust for confounding factors following log-transformation of non-normally distributed variables. A χ 2 test was used for categorical values.
All tests were two-tailed, and differences were considered significant with at P < 0.05. Using G*Power 3.1, we calculated that a sample size of 100 patients would be sufficient to detect moderate to large differences (Cohen’s f = 0.32) across groups at 80% power and P < 0.05. Reference Cohen23
SPSS Statistics (version 25 for Windows) was used for all statistical analyses, and GraphPad Prism 9 (for Windows; GraphPad Software Inc., La Jolla, California, USA; https://graphpad.com) for all graphical representation of data.
Results
The characteristics of hypertensive subjects (n = 100, from a total of 106 approached and 6 declining due to time constraints), stratified according to tertiles of normetanephrine, are shown in Table 1. The study population had a mean age of 43.4 ± 10.9 years, with 54% of participants male; 60% of participants were from Black and minority ethnic groups, in line with the demographic of South-east London where the hospital is located. The majority of subjects (88%) were being treated with anti-hypertensive medications, and 4% had type 2 diabetes. Age, gender and ethnicity were similar across the tertiles of normetanephrine, as was the proportion of smokers and those regularly consuming alcohol. However, body mass index (BMI) was greater in those in the highest compared with the lowest tertile of normetanephrine distribution (30.8 ± 5.0 v. 27.1 ± 4.0 kg/m2, P = 0.012).
Table 1 Demographics, questionnaire scores and biochemistry analysis stratified by tertiles of plasma normetanephrine (NM)
BMI, body mass index; ACE, angiotensin-converting enzyme; Ca, calcium; PHQ-9, Patient Health Questionnaire-9; QIDSSR-16, Quick Inventory of Depressive Symptomatology Self Report-16; GAD-7, Generalized Anxiety Disorder-7; HbA1c, glycosylated haemoglobin; CRP, C-reactive protein.
*P < 0.05, **P < 0.001, tertile 3 versus tertile 1.
Depression, anxiety and sleep questionnaires
A high proportion of patients had scores for PHQ-9 and QIDSSR-16 indicating significant mood disturbance (31 and 34% of the study population scored moderate to severe depression for PHQ-9 and QIDSSR-16, respectively). Scores representing low mood and high anxiety were two- to threefold higher for hypertensive subjects in the higher compared with the lower tertiles of normetanephrine (Fig. 1). The mean scores for PHQ-9 were 9.16 ± 5.80 in the highest tertile compared with 3.76 ± 2.67 in the lowest (difference in score (95% CI) 5.40 (2.92–7.88), P < 0.001), and those for QIDSSR-16 were 12.26 ± 4.91 v. 4.48 ± 3.48 (difference 7.78 (95% CI: 4.71–10.84), P = 0.003). Similarly, GAD-7 scores were 8.96 ± 5.45 v. 4.58 ± 3.84 (difference 4.38 (95% CI: 1.27–7.49), P = 0.003) in the highest versus lowest normetanephrine tertiles, respectively. ESS and STOP-BANG scores were also higher in subjects with higher normetanephrine levels (Table 1).
Fig. 1 Depression and anxiety questionnaire scores stratified according to tertiles of plasma normetanephrine (NM). PHQ-9, Patient Health Questionnaire-9; QIDSSR-16, Quick Inventory of Depressive Symptomatology Self Report-16; GAD-7, Generalized Anxiety Disorder-7. *P < 0.05, **P < 0.001, tertile 3 (high NM) versus tertile 1 (low NM).
Blood pressure and stress-induced cardiovascular responses
Baseline blood pressure and FBF were similar across the tertiles of normetanephrine, but heart rate was higher in subjects in the highest compared with the lowest tertile (Table 2). Mental stress resulted in an increase in SBP and DBP that was similar across the tertiles of normetanephrine, but an increase in FBF was attenuated in subjects in the highest compared with the lowest tertile: FBF increased by 47.1 ± 29.7% in the lowest compared with 28.3 ± 21.2% in the highest tertile and FVR decreased by 22.8 ± 12.6% in the low normetanephrine tertile versus 10.5 ± 11.8% in the high tertile (both P < 0.01). The increase in heart rate was also lower in subjects in the highest compared with the lowest tertiles (7.8 ± 6.7% v. 14.9 ± 10.5%, P = 0.01).
Table 2 Blood pressure, mental stress and device-guided breathing responses stratified by tertiles of plasma normetanephrine (NM)
SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; HR, heart rate; FBF, forearm blood flow; FVR, forearm vascular resistance; bpm, beats per minute; Δ, delta; DGB, device-guided breathing.
*P < 0.05, **P < 0.001, tertile 3 versus tertile 1.
DGB significantly reduced blood pressure, with an effect that was similar across the tertiles of normetanephrine. DGB also reduced FBF but with an effect that was attenuated in the high compared with the low normetanephrine group: FBF reduction of −3.7 ± 20.9% in the highest compared with −18.6 ± 15.2% in the lowest tertiles of normetanephrine (P = 0.01). The reduction in heart rate by DGB was greater in the highest compared with the lowest tertile of normetanephrine.
Similar results were seen for both mental stress and DGB when analysing changes in forearm vascular conductance to take into account concurrent changes in blood pressure (as expected, because these changes were similar across normetanephrine tertiles). Variation in FBF responses to mental stress and DGB across the tertiles of normetanephrine was similar whether adjusted or unadjusted for BMI (Fig. 2).
Fig. 2 Forearm blood flow (FBF) responses to mental stress and device-guided breathing (DGB) stratified according to tertiles of plasma normetanephrine (NM). The values are unadjusted and adjusted for body mass index (BMI). *P < 0.05, **P < 0.001, tertile 3 (high NM) versus tertile 1 (low NM).
Discussion
These data suggest a high prevalence of symptoms of depression and anxiety in individuals with hypertension, consistent with findings from other studies involving hypertensive populations. Reference Li, Li, Chen, Chen and Hu3 Specifically, in our sample of hypertensive individuals, 31–34% exhibited symptoms of moderate to severe depression based on PHQ-9 scores of ≥10, which is markedly higher than the 7–10% prevalence reported for the general adult population using the same assessment tool. Reference Thibodeau and Asmundson24,Reference Hinz, Mehnert, Kocalevent, Brähler, Forkmann and Singer25 A major finding of the present study is that symptoms of low mood and high anxiety are strongly associated with circulating plasma normetanephrine; thus, scores for PHQ-9 and QIDSSR-16 were approximately threefold higher for those in the highest compared with the lowest tertile of normetanephrine. These differences persisted following adjustment for BMI or sleep disturbance, factors that are associated with high sympathetic activity and potential mediators of the association. Reference Esler, Straznicky, Eikelis, Masuo, Lambert and Lambert26,Reference Narkiewicz and Somers27
Higher plasma catecholamine concentrations and higher urinary catecholamine excretion have previously been reported in subjects with depression and anxiety, including in normotensive and untreated hypertensive populations. Reference Hughes, Watkins, Blumenthal, Kuhn and Sherwood12,Reference Paine, Watkins, Blumenthal, Kuhn and Sherwood13 Hypertension is associated with increased levels of circulating catecholamines compared with those seen in normotensive subjects. Reference Goldstein28 However, in both depression and hypertension there is large range of concentrations of circulating catecholamines, with many individuals having concentrations that are within the range seen in the general population. Reference Goldstein28–Reference Wyatt, Portnoy, Kupfer, Snyder and Engelman30 The finding that depression and anxiety symptoms in those with hypertension are largely restricted to subjects with high concentrations of normetanephrine raises the possibility that these may reflect a common neurogenic mechanism linking these conditions. It is also possible that one condition may contribute directly to the onset or exacerbation of the other, underscoring the complex interplay between these comorbidities. However, one possibility is that chronic stress leads to persistent or more frequent/intense bursts of activation of the sympathetic nervous system via the ‘fight or flight’ reaction, which contributes to hypertension and can also result in anxiety and secondary depression. Reference Holwerda, Luehrs, Gremaud, Wooldridge, Stroud and Fiedorowicz31 Chronic stress is known to increase activation of the sympathetic nervous system and is a risk factor for development of hypertension; Reference Spruill8 it forms the basis of the most well-known animal model of depression and is frequently associated with depression in humans. Reference Yang, Zhao, Wang, Liu, Zhang and Li9 Numerous mechanisms have been proposed to explain the link between chronic stress and depression, including effects secondary to alterations in immune function and inflammation induced by chronic activation of the hypothalamic–pituitary–adrenal axis. Reference Yang, Zhao, Wang, Liu, Zhang and Li9 A direct effect is thought to involve activation of noradrenergic neurons in the locus coeruleus. Reference Morris, McCall, Charney and Murrough10 The locus coeruleus is a centre implicated in attention and arousal, with projections throughout the brain, and modulates outflow from the midbrain to the spinal cord, regulating peripheral sympathetic nervous system activity. Reference Samuels and Szabadi11 Norepinephrine released by these neurons engages with low-affinity α-adrenergic receptors in the prefrontal cortex (PFC), leading to a generalised reduction in neuronal activity and dendritic atrophy primarily in the dendritic spines of pyramidal neurons within the PFC, which is commonly observed in depression. Reference Arnsten, Raskind, Taylor and Connor32,Reference Qiao, Li, Xu, Chen, An and Ma33 It is also thought to decrease the level and function of the inhibitory neurotransmitter GABA, which is critical for maintaining the balance between excitatory and inhibitory signalling in cortical circuits. A deficit in GABAergic signalling is associated with increased cortical excitability and dysregulated neural networks involved in emotional regulation and stress response, both of which are central to the pathophysiology of depression. Reference Luscher, Shen and Sahir34
Seeking support for such a direct role of stress on the locus coeruleus and its projections, we examined vasomotor responses to mental stress and DGB. These are acute interventions that act on the same pathways as does chronic stress: PFC and noradrenergic neurons in the locus coeruleus. Reference Arnsten, Raskind, Taylor and Connor32,Reference Zaccaro, Piarulli, Laurino, Garbella, Menicucci and Neri35,Reference Yennu, Tian, Smith-Osborne, Gatchel, Woon and Liu36 Mental stress activates a number of centres in the PFC, Reference Milham, Banich and Barad37,Reference Van Veen and Carter38 which then leads to peripheral vascular responses, the most notable of which is an increase in FBF. Reference Dietz, Rivera, Eggener, Fix, Warner and Joyner39 At a local level, forearm vasodilation is mediated by release of the neurotransmitter and vasodilator nitric oxide (NO) synthesised from neuronal NO synthase (nNOS), probably in skeletal muscle. Reference Dietz, Rivera, Eggener, Fix, Warner and Joyner39,Reference Seddon, Chowienczyk, Brett, Casadei and Shah40 Interconnections between the PFC and peripheral nNOS activation have not been fully characterised, but are likely to involve projections from the PFC via the locus coeruleus to the autonomic nervous system (ANS). Reference Morris, McCall, Charney and Murrough10 Slowing the expiratory phase of breathing via DGB has long been known to reduce stress and alter the balance of sympathetic and parasympathetic activity in the ANS in favour of a reduction in sympathetic activity and increase in parasympathetic activity. Reference Zaccaro, Piarulli, Laurino, Garbella, Menicucci and Neri35 It is now thought that alterations in breathing pattern are detected by specific neurons in the preBötzinger complex, the primary breathing rhythm generator. Reference Yackle, Schwarz, Kam, Sorokin, Huguenard and Feldman41 These neurons project onto and regulate noradrenergic neurons in the locus coeruleus, with slow breathing reducing activation of the latter, Reference Yackle, Schwarz, Kam, Sorokin, Huguenard and Feldman41 an effect opposite to mental stress. That subjects with hypertension and raised plasma normetanephrine concentrations had both mood disturbance and impaired vasomotor responses to mental stress and DGB is, therefore, consistent with involvement of the PFC and locus coeruleus in this hypertension/mood phenotype.
Both hypertension and affective disorders are highly heterogenous conditions with respect to aetiology, clinical characteristics and response to treatment. That the overlap of these two conditions may identify a specific phenotype is likely to have implications for the response to treatment, and raises the possibility that treatment of one condition could benefit the other. Early experiences with drugs such as hexamethonium, a ganglion-blocking agent for hypertension, did not suggest any benefit on depressive symptoms. Reference Paton42 However, it is possible that drugs acting in the central rather than peripheral nervous system to block sympathetic activity could improve depression. Identification of noradrenergic neurons in the projections of the locus coeruleus raises the possibility that α1 antagonists or α2 agonists that reduce central noradrenergic activity could benefit both conditions. There is some indirect evidence for this in that the α1 antagonist prazosin and α2 agonist clonidine are used in the treatment of post-traumatic stress disorder (PTSD) and may also improve depressive symptoms, although a recent large trial of prazosin for PTSD was negative. Reference Wendell and Maxwell43,Reference Raskind, Peskind, Chow, Harris, Davis-Karim and Holmes44 The atypical antipsychotic quetiapine is being increasingly used for depression, Reference Daly and Trivedi45 has α1 antagonist properties and reduces blood pressure, a feature shared by many other antipsychotics. Reference Tan, Hoppe and Heard46 In targeting an underlying disturbance of the hypertension/mood phenotype, such drugs may be particularly effective for their primary indication as well as treating the associated comorbidity. Conversely, noradrenergic uptake inhibitors that increase noradrenergic activity, such as duloxetine and venlafaxine, could be less effective in depression associated with hypertension and exacerbate the latter. However, we stress that this is purely speculative and needs to be tested in clinical trials.
Limitations
Our study is subject to several limitations. We studied a relatively small sample of individuals referred to a secondary hypertension service, the majority of whom had well-controlled blood pressure and were on treatment with anti-hypertensive medications that could potentially have influenced normetanephrine levels. Further studies with a larger sample size, including in primary care populations, will be required to verify the present findings, in order to examine to what extent they may be influenced by treatment and common comorbidities such as obesity, obstructive sleep apnoea and diabetes. Additionally, participants did not undergo diagnostic interviews to confirm depression or anxiety, and symptom prevalence was based on questionnaire assessments which, although psychometrically robust with high test–retest reliability and strong validity, may not fully capture diagnostic criteria. The study also did not include a control group of a normotensive population, limiting direct comparisons. Although the association of affective disturbance and raised normetanephrine with impaired stress-mediated vasomotor responses is most readily explained by chronic stress impinging on pathways regulating central arousal and peripheral sympathetic nerve activity, we cannot make firm conclusions regarding causality from the results of the present study. Further studies examining the results of interventions that reduce chronic stress will be required to test this hypothesis.
In conclusion, a hyperadrenergic state in hypertension is associated with affective disturbance and impaired stress-modulated vasomotor responses. This association may be mediated by chronic stress impinging on pathways regulating central arousal and peripheral sympathetic nerve activity. The possibility that affective disturbance associated with hypertension may identify a phenotype with distinct neurophysiology should be explored.
Data availability
All data requests should be submitted to the corresponding author for consideration. Access to anonymised data may be granted following review.
Author contributions
All authors developed the study concept and designed the research. L.F., B.F. and R.J.M. were involved in study implementation and data collection. R.J.M., L.F. and B.F. performed data analysis, with C.D.M. and A.H.Y. helping to interpret the results. B.F. and P.J.C. wrote the majority of the manuscript. All authors read and approved the final version of the manuscript.
Funding
The work was supported by a Stratified Medicines Award from the British Heart Foundation and the Medical Research Council – the Ancestry and Biological Informative Markers for the Stratification of Hypertension (AIM HY) study (no. MR/M016560/1).
Declaration of interest
A.H.Y. is a member of the BJPsych Open editorial board; we confirm that the authors did not take part in the review or decision-making process of this paper.
Open access
eLetters
No eLetters have been published for this article.