There is growing awareness of the impact of psychosocial stressors, such as urbanization, on the mental and physical health of black South Africans (black Africans). Chronic psychosocial stressors constitute a burden on the ability of the brain to adapt to stress and to initiate the recovery process, especially in susceptible individuals. The neuroendocrine responses are central to how the body copes with stress and recovers from it 1,2. Inability to properly regulate aspects of these responses has been proposed as an essential factor in the pathophysiology of various stress-related disorders, including depression 3. Indeed, evidence from the literature indicates that depression is linked to alterations in neuroendocrine responses to mental stress, more specifically alterations in cortisol and norepinephrine responses, which can be related to increased risk of cardiovascular disease (CVD) 2,4–7. Depression has been associated with increases in norepinephrine responses to stress 8. A recent study by Hamer and Malan 6 supported the link between depressive symptoms, norepinephrine responses and metabolic syndrome in black Africans 6. There remains no clear-cut nor generally accepted model for cortisol responses in depression, as both blunted and increased cortisol activity have previously been noted 5,9–11. To our knowledge, no study has yet attempted to investigate the association between depressive symptoms, neuroendocrine responses and increased cardiovascular risk [left ventricular hypertrophy (LVH)] in black Africans.
LVH is a marker of cardiac structural abnormality and has previously been linked to depressive symptoms in urban black Africans 12. Examining the role of the neuroendocrine responses in the above-mentioned association is of importance because of the high prevalence of depressive symptoms and emerging burden of CVD in this population group 6,13. Therefore, the aim of this study was to examine the association between depressive symptoms, neuroendocrine responses [represented by salivary cortisol and 3-methoxy-phenylglycol (MHPG)] and ECG-LVH in urban black Africans.
Materials and methods
Study design and participants
Participants in this study were recruited as part of the Sympathetic Activity and Ambulatory Blood Pressure in Africans (SABPA) study. The SABPA study was a cross-sectional, multidisciplinary, target population study conducted between February 2008 and May 2009. For the purpose of this study, 200 urban black African teachers working in the Dr Kenneth Kaunda Education district, North West Province, South Africa, were invited to participate. The choice of this selection was based on the need to obtain a homogeneous sample from a similar socioeconomic class. All participants between the ages of 25 and 60 years were invited to participate. The study procedures are described elsewhere 6. Exclusion criteria for this study sample were pregnancy, lactation, use of α and β blockers, psychotropic substance users, ear temperature greater than 37°C, and individuals who had been vaccinated or had donated blood in the 3 months before participating. For the purpose of our study we additionally excluded participants who were HIV positive (N=19) and statin users (N=2). The total sample comprised 179 participants.
The ethical clearance for this study was granted by the Ethics Committee of North-West University (0003607S6), in accordance with the principles outlined by the World Medical Association Declaration of Helsinki 1975 (revised 2008). All participants signed an informed consent form before data collection commenced. Assistance was offered to participants who requested information in their native language.
Clinical assessment procedure
Each morning of the working week, four participants were fitted with a British Hypertension Society-validated ambulatory blood pressure (ABPM) monitoring device and two-lead ECG (Meditech CE120; Cardiotens, Budapest, Hungary). The ABPM device was fitted to the participant’s nondominant arm and programmed to take blood pressure (BP) measurements at 30-min intervals during the day (08:00–22:00 h) and 60-min intervals during the night (22:00–06:00 h) accompanied by a sequential recording of the ECG strips every 5 min for 20 s 14. The successful mean inflation rate in this sample was 82.7% (±3.8%). A 12-lead ECG (PC 1200, software version 5.030; NORAV Medical Ltd, Kiryat Bialik, Israel) device was used to determine LVH, using strip leads RaVL+SV3 in the calculation of a sex-specific formula, the Cornell product: sum of the leads (RaVL+SV3)×QRS>244 mV ms 15,16. Higher Cornell product values imply worse LVH 15,16.
The participants were also fitted with an Actical omnidirectional accelerometer monitor (Mini Mitter, Bend, Oregon, USA; Montréal, Québec, Canada) to measure physical activity during the ABPM period. Participants were requested to continue with normal daily activities, recording any abnormalities such as visual disturbances, headache, nausea, fainting, palpitations and stress on their ambulatory diary charts.
At ∼16:40 h the participants were transported to the Metabolic Unit Research Facility of North-West University (research unit for human studies) and were familiarized with the available facilities and the experimental setup, to minimize the ‘white-coat effect’ 17. This was followed by the completion of a collection of psychosocial questionnaires under the supervision of a registered clinical psychologist. The participants then received a standardized dinner and were advised to retire for the night at ∼22:00 h, fasting overnight.
At 06:00 h the following morning, the ABPM and the Actical devices were removed. The participants’ anthropometric measurements were then obtained followed by the resting 12-lead ECG and psychophysiological stress testing while the participants were in a semirecumbent position. Fasting venous blood and saliva samples were obtained. Saliva samples were collected using the Sarstedt salivette device (Sarstedt, Nümbrecht, Germany), which involves placing a small cotton roll in the mouth of the participant for several minutes. Participants stayed overnight at the Metabolic Unit, and refrained from consuming alcohol and caffeine, smoking and exercising 8 h before data collection. In addition, they were advised to take caution with their dental hygiene the previous night to prevent bleeding gums during sampling the following morning. Participants were requested not to brush their teeth before data sampling 18.
Mental stress testing
The Stroop mental stress task has been used extensively in psychophysiological stress testing. The task was administered over a 1-min period followed by a 30-min recovery, using cards. The participants were required to identify the colours of the colour word cards in contrasting colours of ink under time pressure. As a motivation to complete the task, each participant received a monetary incentive of ∼1 euro with the completion of every column. Saliva samples were collected at baseline and 30 min post-task for detection of cortisol and MHPG. Salivary MHPG has been shown to be a reliable indicator of sympathetic activity during acute mental stress, with optimum responses occurring 30 min poststress task 4.
Depressive symptoms were assessed with the use of the nine-item self-administered Patient Health Questionnaire (PHQ-9). The PHQ-9 is a measure of depressive symptom severity and has been validated in various ethnic groups including sub-Saharan Africans 19–21. The questionnaire is designed for use in primary healthcare settings adapting diagnostic criteria from the Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV). Each item of the PHQ-9 evaluates the presence of one of the nine DSM-IV criteria of major depression 19. In the current study, the Cronbach α-reliability index for the total PHQ-9 score was 0.81. Items on the questionnaire are scored to reflect the frequency of symptom occurrence during the prior 2 weeks on a scale of 0–3, with 0 reflecting ‘not at all’ and 3 ‘nearly every day’, thus providing a continuous score between 0 and 27 19. We used the recommended and established PHQ-9 cutoff point of greater than 10 indicating the presence of depressive symptoms, stratifying the study sample into participants with and without depressive symptoms 20.
Anthropometric data were standardized and taken in triplicate to the nearest 0.1 cm. Body weight and height measurements were taken and used to calculate BMI (kg/m2). Waist circumference was measured with a metal tape at the midpoint between the lower costal border and the iliac crest perpendicular to the long axis of the trunk 22. Intraobserver and interobserver variability were less than 10%.
Biochemical sampling and analyses
Fasting blood samples were handled and prepared according to standardized methods, and frozen at −80°C until analysis. Fasting sodium fluoride (glucose) and serum samples for triglycerides, HDL-cholesterol, gamma glutamyl transferase (γ-GT) and cotinine were analysed using the sequential multiple analyser computer (Konelab 20i; Thermo Scientific, Vantaa, Finland) and by immunoassay (Integra 400; Roche, Basel, Switzerland). As a marker of alcohol abuse, γ-GT was measured 23, whereas serum cotinine levels were measured as a marker of smoking status, with values above 14.99 μg/l considered an indication of exposure to first- or second-hand smoking 24. The intracoefficient and intercoefficients of variation for all assays were below 10%.
Cortisol levels were determined from salivary samples, which were obtained before 09:00 h. To avoid the cortisol awakening response, resting saliva samples were collected 45 min after awakening 25. Cortisol levels were determined using a high-sensitivity enzyme-linked immunosorbant assay. Cortisol showed intracoefficient and intercoefficients of variation of 7.7 and 9.8%, respectively. The major metabolite of norepinephrine, MHPG, closely reflects plasma metabolite levels and was analysed using high-performance liquid chromatography coupled to an electrochemical detector 26,27. The interday and intraday coefficients of variation for all assays were less than 10%.
Stress responses were calculated with the following formula as changes from baseline (delta, Δ): X stressor−X resting. All data were analysed using the software package STATISTICA 10 (StatSoft Inc., Tulsa, Oklahoma, USA; 2010). γ-GT levels were logarithmically transformed to normalize distribution. Differences in characteristics of participants in relation to depressive symptoms were analysed using t tests (mean±SD) to examine continuous variables and χ2 to calculate prevalence (%). A single two-way analysis of covariance (ANCOVA) was applied to assess interaction between the main effects (depressive symptoms×cortisol median split response), independent of a-priori covariates (age, smoking prevalence, resting cortisol and MHPG levels). Subsequent one-way ANCOVA analyses followed where cardiovascular and biochemical data of low and high cortisol responses were compared using least squared means.
Multiple unadjusted and adjusted regression analyses were computed to identify independent predictors of ECG-LVH in the participants with and without depressive symptoms in the low- and high-cortisol response groups. In forward stepwise regression analyses, ECG-LVH was the dependent predictor and the following independent predictors were included: age, sex, smoking prevalence, body surface area, resting cortisol and MHPG levels, MPHG responses and 24 h systolic BP. Significance was noted as a P-value of up to 0.05 and trend for P as less than 0.1.
We computed sensitivity analyses on the relationship between dipping status, depressive symptoms and cortisol levels in separate sex groups.
The participant sample is described in Table 1. Africans with depressive symptoms were more likely to smoke (P=0.05), although no other differences in baseline characteristics were observed. Both groups displayed similar resting MHPG and cortisol levels.
Mental stress responses
Two-way ANCOVAs revealed significant interaction on main effects (depressive symptoms×cortisol median split responses) for ECG-LVH [F(1,179)=5.86; P=0.02] with a trend for systolic BP [F(1,179)=2.96; P=0.09] and diastolic BP [F(1,179)=3.18; P=0.08]. Therefore, our group was stratified into with depressive symptoms and without, as well as median split cortisol responses using a cutoff point of greater than 1.5 ng/l. As shown in Table 2, the Africans with depressive symptoms and low cortisol responses demonstrated a significantly blunted MHPG response (P=0.004) in comparison with the high-cortisol response group with lower triglyceride levels (P=0.04). Interestingly, Africans without depressive symptoms and high cortisol responses displayed significantly higher ECG-LVH (P=0.01) and a trend for elevated systolic (P=0.08) and diastolic BP (P=0.06).
As shown in Fig. 1, cortisol responses at rest and responses during acute mental stress Stroop test were compared in Africans with depressive and without, independent of covariates. A blunted cortisol response was revealed in Africans displaying depressive symptoms (P=0.02).
Predictors of ECG-left ventricular hypertrophy
In forward stepwise regression analysis models (Table 3), blunted MHPG acute mental stress responses were positively associated with ECG-LVH in Africans with depressive symptoms and with low cortisol stress responses [adjusted R2=0.20; β=0.92 (95% confidence interval 0.74, 1.10); P≤0.02].
No associations existed between MHPG responses and ECG-LVH in participants without depressive symptoms.
No relationship existed between dipping status, depressive symptoms and cortisol levels in separate sex groups.
Possible associations were assessed between neuroendocrine acute mental stress responses (represented by salivary MHPG and cortisol), depressive symptoms and ECG-LVH (a marker of structural left ventricular wall abnormalities) in urban black Africans. To our knowledge, this is the first well-controlled study focusing on the contribution of selected stress responses and depression to cardiovascular risk in black Africans. Main findings demonstrated no direct association between depressive symptoms and resting cortisol levels. However, blunted neuroendocrine responses linked depressive symptoms and ECG-LVH in black Africans. When coupled to their hypertensive status, these vasoconstrictive agent responses may underpin the increased long-term depression and vascular disease risk in urban Africans.
Earlier work demonstrated elevated salivary MHPG levels in individuals with greater depressive symptoms 4,29. We could not establish similar significant findings. Conversely, salivary MHPG is a major metabolite of norepinephrine that closely resembles plasma levels and was investigated in this African target population 27. The positive association revealed between salivary MHPG stress responses and depressive symptoms may be related to psychosocial stressors 30 such as urbanization, demonstrating a challenged nervous system and heightened sympathetic activity. Findings are supported by other studies where urbanization has been associated with increased hypertension prevalence, vascular responsiveness and myocardial ischaemic risk in this ethnic group 31–33.
A secondary aim of the study was to link depressive symptoms to neuroendocrine responses and potential target organ damage. Depressive symptoms may act via attenuation of norepinephrine responses to stress, potentially contributing to allostatic load, which can predispose Africans to structural LVH changes 34,35. Indeed, depressed and nondepressed individuals did not show a difference with respect to basal cortisol and MHPG levels, however, blunted salivary stress MHPG and cortisol responses were apparent in individuals with depressive symptoms after exposure to the Stroop test. This could imply that the presence of depressive symptoms sensitizes the individual to stress and the subsequent development of vascular disease and/or other lifestyle illnesses.
We were able to establish MHPG responses as a predictor of ECG-LVH in these participants. As cortisol has a permissive effect on norepinephrine functioning, the vasoconstrictive effect of norepinephrine secretion will enhance higher α-adrenergic vascular responsiveness and a hyperkinetic state. Indeed, a hypertension prevalence of 71.67% may support the notion of a shift from central cardiac (β-adrenergic) to vascular (α-adrenergic) BP responses 31, increasing CVD risk.
In a previous study we demonstrated an association between systolic BP and ECG-LVH in African men with depression 12. We support these findings as we demonstrated that blunted cortisol and MHPG responses, acting as vasoconstrictive agents, may induce structural wall abnormalities by increasing preload and afterload to the heart 31. Depression may enforce these changes and augment cardiovascular morbidity. Blunted cortisol responses to laboratory and naturalistic psychosocial stressors have been demonstrated in both clinical and subclinical depression 36,37. For instance, Burke et al.5 demonstrated that women with high depressive symptoms exhibited blunted cortisol stress responses to a naturalistic stressor 5. This response is possibly not unexpected as depression is typically associated with elevated cortisol, while at the same time demonstrating a blunted response to dexamethasone challenge 1,38,39. However, it could be speculated that as depression is a constant state of perceived stress, further exposure to a challenging urban environment or psychosocial stress may result in habituation of the neuroendocrine pathways 40. Alternatively, although the hypothalamic–pituitary–adrenal-axis in depression is moderately activated, possibly due to the initial (primary) hippocampal degeneration in this condition, the ensuing structural changes are likely to illicit a maladaptive cortisol response as described here 38. Chronic mental or psychosocial stress triggers sympathetic hyperactivity resulting in increased release of catecholamine and ultimately depletion, a process that is abrogated by cortisol 9. Unmitigated increases in cortisol, as a result of maladaptive responses to stress, lead to the dysregulation of hypothalamic–pituitary–adrenal-axis activity resulting in the above-described patterns of cortisol release, eventually leading to the development of an anxiety and/or mood disorder 9,40. The failure to appropriately regulate aspects of this stress axis represents a crucial factor in the pathophysiology of various stress-related disorders 3.
Therefore, our findings suggest that Africans with depressive symptoms and blunted MHPG-cortisol responses are at a greater risk for vascular disease and stress-related disorders. Subsequently, various maladaptive neurobiological and behavioural changes ensue that lay the foundation for altered metabolic and redox function, as has recently been described in this population, which in turn drive the increased risk for cardiovascular illness 41.
Certain strengths and limitations of this study should be acknowledged. The nature of the study design and the uniqueness of the study sample are noteworthy strengths, and this is the first well-controlled study in sub-Saharan Africa examining the role of neuroendocrine responses and depressive symptoms and their association with cardiovascular health. The limitations of our study include the inability to make inferences about causality, because of the cross-sectional nature of the study. Prospective data are needed to address this issue. Furthermore, we used a self-report questionnaire to assess depressive symptoms, which may introduce reporting biases. The study sample consists only of black Africans and it is suggested that the collectivistic nature of this population group, where most actions are reflected by group/social interaction, coupled with culture-specific manifestations of depressive symptoms may influence the interpretation of the items in the PHQ-9 42,43. However, the questionnaire is a well-established diagnostic tool that has been validated in various ethnic groups, including Africans 21.
Blunted neuroendocrine responses were linked to depressive symptoms and ECG-LVH in black Africans. When coupled to their hypertensive status, these vasoconstrictive agent responses may underpin the increased long-term depression and vascular disease risk in urban Africans. These findings may have important health implications given the emerging burden of CVD among urban black Africans and the fact that a three-fold risk of depressive symptoms has been observed in this population group in comparison with white Africans 6,13. Identifying and treating depressive symptoms in Africans may play a role in reducing stress-related cardiovascular risk in this ethnic group.
The authors thank the participants of the SABPA study for their invaluable and voluntary participation.
The Sympathetic Activity and Ambulatory Blood Pressure in Africans study is funded by North-West University, the National Research Foundation, Medical Research Council, Roche Diagnostics, South Africa and the Metabolic Syndrome Institute, France. The funders played no role in the design and conduct of the study; collection, management, analysis and interpretation of the data; and preparation, review or approval of the manuscript.
Conflicts of interest
There are no conflicts of interest.
1. Harvey BH, McEwen BS, Stein DJ.Neurobiology of antidepressant withdrawal: implications for the longitudinal outcome of depression.Biol Psychiatry2003;54:1105–1117.
2. Kavelaars A, Heijnen CJ.Stress, genetics, and immunity.Brain Behav Immun2006;20:313–316.
3. Gold PW, Chrousos GP.The endocrinology of melancholic and atypical depression: relation to neurocircuitry and somatic consequences.Proc Assoc Am Physicians1999;111:22–34.
4. Hamer M, Tanaka G, Okamura H, Tsuda A, Steptoe A.The effects of depressive symptoms on cardiovascular and catecholamine responses to the induction of depressive mood.Biol Psychol2007;74:20–25.
5. Burke HM, Fernald LC, Gertler PJ, Adler NE.Depressive symptoms are associated with blunted cortisol stress responses in very low-income women.Psychosom Med2005;67:211–216.
6. Hamer M, Malan L.Sympathetic nervous activity, depressive symptoms, and metabolic syndrome in black Africans: the sympathetic activity and ambulatory blood pressure in Africans study.Stress2012;15:562–568.
7. Light KC, Kothandapani RV, Allen MT.Enhanced cardiovascular and catecholamine responses in women with depressive symptoms.Int J Psychophysiol1998;28:157–166.
8. Grossman F, Potter WZ.Catecholamines in depression: a cumulative study of urinary norepinephrine and its mayor metabolites in unipolar and biopolar depression patients versus healthy volunteers at the NIMH.Psychiatry Res1999;87:21–27.
9. Hellhhammer J, Schlotz W, Stone AA, Pirke KM, Hellhammer DH.Allostatic load, perceived stress, and health: a prospective study in two age groups.Ann N Y Acad Sci2004;1032:8–13.
10. Hellhammer DH, Wüst S, Kudielka BM.Salivary cortsiol as a marker in stress research.Psychoneuroendocrinology2009;34:163–171.
11. Holsboer F, Von bardeleben U, Gerken A, Stalla GK, Muller OA.Blunted corticotrophin and normal cortisol response to human corticotrophin-releasing factor in depression.N Engl J Med1984;311:1127.
12. Mashele N, Malan L, Van Rooyen JM, Harvey BH, Potgieter JC, Hamer M.Depression, cardiometabolic function and left ventricular hypertrophy in African men and women: the SABPA study.Clin Exp Hypertens2013;35:213–219.
13. Gaillard T.Insulin resistance and cardiovascular risk in Black people of the African diaspora.Curr Cardiovasc Risk Rep2010;4:186–194.
14. Mancia G, Fagard R, Narkiewicz K, Redon J, Zanchetti A, Böhm M, et al..2013 ESH/ESC Guidelines for the management of arterial hypertension: the task force for the management of arterial hypertension of the European Society of Hypertension (ESH) and of the European Society of Cardiology.J Hypertens2013;31:1281–1357.
15. Ishikawa J, Ishikawa S, Kabutoya T, Gotoh T, Kayaba K, Schwartz JE, et al..Cornell product left ventricular hypertrophy in electrocardiogram and the risk of stroke in a general population.Hypertension2009;53:28–34.
16. Verdecchia M.Prognostic value of a new electrocardiographic method for diagnosis of left ventricular hypertrophy in essential hypertension.J Am Coll Cardiol1998;31:383–390.
17. Obrist PA.Cardiovascular psychophysiology: a perspective1981.New York:Plenum Press.
18. Rydstedt LW, Cropley M.The relationship between long-term job strain and morning and evening saliva cortisol secretion among white-collar workers.J Occup Health Psychol2008;13:105–113.
19. Kroenke K, Spitzer RL.The PHQ-9: a new depression diagnostic and severity measure.Psychiatr Ann2002;32:1–7.
20. Löwe B, Kroenke K, Herzog W, Gräfe K.Measuring depression outcome with a brief self-report instrument: sensitivity to change of the Patient Health Questionnaire (PHQ-9).J Affect Disord2004;81:61–66.
21. Monahan PO, Shacham E, Reece M, Kroenke K, Ong’or WO, Omollo O, et al..Validity/reliability of PHQ-9 and PHQ-2 depression scales among adults living with HIV/AIDS in Western Kenya.J Gen Intern Med2009;24:189–197.
22. Marfell-Jones M, Olds T, Stewart A, Carter L.International standards for anthropometric assessment2006.Potchefstroom:International Society for the Advancement of Kinanthropometry.
23. Sharpe PC.Biochemical detection and monitoring of alcohol abuse and abstinence.Ann Clin Biochem2001;38:652–664.
24. Jarvis MJ, Feyerabend C, Bryant A, Hedges B, Primatesta P.Passive smoking in the home: plasma cotinine concentrations in non-smokers with smoking partners.Tob Control2001;10:368–374.
25. O’ Donnell K, Baddrick E, Kumari M, Steptoe A.Psychological coping styles and cortisol over the day in healthy older adults.Psychoneuroendocrinology2008;33:601–611.
26. Yamada S, Yamauchi K, Yajima J, Hisadomi S, Maeda H, Toyomasu K, Tanaka M.Saliva level of free 3-methoxy-4-hydroxyphenylglycol (MHPG) as a biological index of anxiety disorders.Psychiatry Res2000;93:217–223.
27. Yang RK, Yehuda R, Holland DD, Knott PJ.Relationship between 3-methoxy-4-hydroxyphenylglycol and homovanillic acid in the saliva and plasma of healthy volunteers.Biol Psychiatry1997;42:821–826.
28. Alberti KGMM, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI, Donato KA, et al..Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity.Circulation2009;120:1640–1645.
29. Mausbach BT, Dimsdale JE, Ziegler MG, Mills PJ, Ancoli-Isreal S, Patterson TL, Grant I.Depressive symptoms predict norepinehrine responses to psychological stressor task in Alzheimer’s caregivers.Psychosom Med2005;67:638–642.
30. Van Lill L, Van Rooyen JM, Steyn F, Reimann M, Zeimssen T.Baroreceptor sensitivity, cardiovascular responses and ECG left ventricular hypertrophy in men: the SABPA study.Blood Press2001;20:355–361.
31. Malan L, Hamer M, Schlaich MP, Lambert GW, Harvey BH, Reimann M, et al..Facilitated defensive coping, silent ischemia and ECG left ventricular hypertrophy: the SABPA study.J Hypertens2012;30:543–550.
32. Seedat YK.Perspectives on research in hypertension.Cardiovasc J Afr2009;20:39–42.
33. Van Rooyen JM, Huisman HW, Eloff FC, Laubscher PJ, Malan L, Steyn HS, Malan NT.Cardiovascular reactivity in black South African males of different age groups: the influence of urbanization.Ethn Dis2002;12:69–75.
34. Schlaich MP, Kaye DM, Lambert E, Sommervilee M, Socratous F, Esler MD.Relation between cardiac sympathetic activity and hypertensive left ventricular hypertrophy.Circulation2003;108:108560–108565.
35. McEwen BS.Physiology and neurobiology of stress and adaptation: central role of the brain.Physiol Rev2007;81:873–904.
36. Burke HM, Davis MC, Otte C, Mohr DC.Depression and cortisol stress response: a meta-analysis.Psychoneuroendocrinology2005;30:846–856.
37. Peeters F, Nicholson NA, Berkhof J.Cortisol responses to daily events in major depressive disorder.Psychosom Med2003;65:836–841.
38. Harvey BH.Adaptive plasticity during stress and depression and the role of glutamate-nitric oxide pathways.S Afr Psychiatry Rev2006;9:132–139.
39. Meador Woodruff JH, Greden JF, Grunhaus LJ, Haskett RF.Severity of depression and hypothalamic-pituitary-adrenal axis dysregulation: identification of contributing factors.Acta Psychiatr Scand1990;81:364–371.
40. Gerra G, Zaimovic A, Mascetti GG, Gardini S, Zambelli U, Timpano M, et al..Neuroendocrine responses to experimentally induced psychological stress in healthy humans.Psychoneuroendocrinology2001;26:97–107.
41. Harvey BH, Hamer M, Louw R, van der Westhuizen FH, Malan L.Metabolic and glutathione redox markers associated with brain-derived neurotropic factor in depressed African men and women: evidence of counter-regulation?Neuropsychobiology2013;67:33–40.
42. Happner P, Heppner MJ, Wang L.Development and validation of a collectivist coping styles inventory.J Couns Psychol2006;53:107–125.
43. Tomlinson M, Swartz L, Kruger M, Gureje O.Manifestations of affective disturbance in sub-Saharan Africa: key themes.J Affect Disord2007;1021–3191–198.