Several lines of evidence suggest that clinical depression is a risk factor for cardiac morbidity and mortality in patients with coronary heart disease, especially after an acute myocardial infarction (MI). Major depression is associated with a 4-fold increase in the risk of mortality during the first 6 months after an acute MI, and its prognostic significance is comparable to that of left ventricular dysfunction and a history of MI. Depression is not unusual among individuals with coronary heart disease, with studies indicating that between 15 % and 22 % of patients suffer from depression after a cardiac event.

Not only circadian, but also circannual, circaseptan (about 7-day), and other variations characterize the incidence of adverse vascular events, such as acute myocardial infarctions, strokes, and sudden cardiac deaths, carrying important pathophysiological implications. Most studies showed an increased incidence in the morning, peaking between 06:00 and 12:00. Ambulatory blood pressure monitoring (ABPM) has become an important tool in the diagnosis and management of hypertension. Several studies have indicated that target organ damage and cardiovascular morbidity are more strongly associated with circadian BP endpoints than with office BP values. We reported that circadian and other patterns are synchronized by socio-ecologic factors, such as human lifestyles, as well as economic and environmental conditions. We demonstrated the presence of a weekly BP variation in community-dwelling subjects, including a surge in BP on Mondays, and we showed that it was most prominent in depressed people.

In 2000, we began a community-based study to longitudinally investigate the longevity and aging of a population in Hokkaido County (LILAC) [1, 2], including the evaluation of the younger population aged from 40 to 74 years (health-watch) [3]. Our goal is the prevention of cardiovascular events, notably strokes and myocardial ischemic events in a community. Herein, we examine whether a depressive mood is associated with an increased risk of stroke and cardiovascular mortality, and we estimate the predictive value of a depressive mood on the incidence of adverse vascular events, using a multivariate Cox model.

The 224 subjects monitored between April 2001 and March 2004 were between 24 and 79 years of age (88 men and 136 women; mean age: 56.8 ± 11.2 years), living in Uraus in Hokkaido (latitude: 43.45° N, longitude: 141.85° E). All subjects visited and utilized the free health screening, counseling, and educational services of the Uraus office. Subjects with definite neurological diseases, such as Parkinsonism and stroke, and those who were too severely ill to stand without help were excluded. Lifestyle, including nutritional and sleep quality, was investigated by several types of questionnaires. BP was measured at the beginning of the study in a sitting position, and brachial-ankle pulse wave velocity (baPWV) was measured between the right arm and ankle in a supine position, using an ABI/Form instrument (Nippon Colin Co., Ltd., Komaki, Japan). In all studies, baPWV was measured twice after at least a 5-min rest. For the assessment of the predictive value, only baPWV measures from participants showing normal ankle/brachial pressure index (ABI) values (>0.90) were used.

The 7-day/24-hour BP recordings were obtained with a commercially available ambulatory BP recorder (TM-2431, A&D Company, Japan) [4]. Subjects were fitted with the recorder in the morning, usually between 10:00 and 11:30 a.m. of the first day of monitoring. The recorder was set to take measurements at intervals of 30 min between 7:00 a.m. and 10:00 p.m. and of 60 min between 10:00 p.m. and 7:00 a.m. Subjects were instructed to follow their usual daily routine after they left the office. Subjects were instructed to remain motionless each time a reading was to be taken and then to record their activity on a diary sheet. Stored data were retrieved and analyzed on a personal computer with commercially available software for the device (TM-2430-15, A&D Company, Japan). The data were further analyzed by sphygmochron for the entire 7-day span and separately for each day of monitoring. In addition to the circadian rhythm characteristics, estimates were also obtained for mean values during 24-hour, waking and sleeping spans, and for the day-night ratio (as a gauge of “dipping”) for SBP, DBP, and HR and also for mean blood pressure (MBP = 1/3SBP + 2/3DBP), pulse pressure (PP = SBP−DBP), and the double product (DP = SBP × HR/100) (“dip” is defined as the awake-sleep difference divided by awake (×100), where awake and sleep are average values during the awake and sleep span, respectively).

Numerous depression rating scales are currently available, which represent a mixture of observer-rated and self-rating scales. In this investigation, subjects were asked about 15 items on a depression scale, by means of a self-administered questionnaire [5]. When the depression score was ≥5, subjects were considered to have a depressed mood. A depression score was obtained again 1–2 years later. Subjects with a score during the second screening higher by at least 2 points compared to that of the first screening were assigned to the enhanced depressive mood group (irrespective of the score at the start of study). All other subjects were assigned to the control group.

The health-watch study was started on April 12, 2001. One or two doctors of our team visited and provided repeated consultations every 3 months. They offered advice in relation to the rehabilitation of disordered functions and healthy lifestyle modifications (promoting complete cessation of smoking, weight reduction, reduction of salt intake, moderation in the consumption of fruits and vegetables, and alcohol intake). They also advised in terms of medical prescriptions for the local general practitioner.

The follow-up herein ended on November 30, 2004. During this time, 9 subjects died of myocardial infarction or stroke. The follow-up time was defined as the time elapsed between the date of the first (reference) examination and the date of first cardiovascular event or death.

We used Cox’s regression analysis to calculate the unadjusted and adjusted relative risks (RRs) and 95 % confidence intervals (CIs) for cardiovascular death. To identify independent predictors of stroke and cardiovascular disease in relation to a depressive mood, we used a multivariate Cox regression analysis with stepwise selection, including as variables age, gender, BMI, lifestyle, sleep quality, QOL, baPWV, and ABP endpoints. The independent correlation of the enhanced depressive mood was determined by means of a logistic regression analysis. Kaplan-Meier event probability curves were computed with two groups, and the cumulative probability of events of two groups was compared by means of the log-rank test. Significance was considered at a value of p < 0.05.

The 224 subjects comprised 88 men and 136 women. The mean age of participants at entry was 56.8 years. The mean follow-up time was 1064 days, during which 4 subjects suffered from adverse vascular outcomes (myocardial infarction, 1 man and 1 woman; stroke, 2 men).

As a comprehensive assessment in this chronoecological health-watch, we checked out the following items: indices of environmental conditions, lifestyle, quality of sleep, QOL, and 7-day ABP endpoints, such as the 7-day (from Thursday to Wednesday) and daily averages of SBP, DBP, and HR; the incidence of SBP and DBP CHAT (circadian hyper-amplitude-tension) and of SBP, DBP, and HR “dipping” (SBP-dip, DBP-dip, and HR-dip); estimates of the circadian amplitude and acrophase of SBP, DBP, and HR; as well as estimates of the percentage time elevation (PTE) of SBP, DBP, and HR. Averages over 7 days were also computed (not shown) as SBP 1-7, DBP 1-7, pulse pressure (PP) 1-7, double product (DP) 1-7, 24-hour SD of HR (HRSD1-7), SBP-dip 1-7, DBP-dip 1-7, HR-dip 1-7, MAP-dip 1-7, PP-dip 1-7, and DP-dip 1-7.

Depression scores were obtained for 218 of the 224 citizens at the start of study and from 179 subjects after 1–2 years (177 subjects provided a score during both sessions). Scores obtained during the first and second sessions averaged 4.11 ± 3.11 and 4.42 ± 3.14, respectively. Of the 177 citizens who answered both questionnaires, 39 were assigned to the enhanced depressive mood group. A comparison of characteristics between the two groups is shown in Table 11.1. It can be seen that depressive subjects had a higher body mass index (BMI), they consumed smaller meals (smaller portions of rice), and their total serum cholesterol was higher as compared to the control subjects. The time to fall asleep was also slightly longer in the depressive group (p = 0.051). Kaplan-Meier curves for event-free survival revealed a significant difference between the two groups (p = 0.0124, log-rank test) (Fig. 11.1). Subjects in the control group had a better event-free survival than depressed subjects.

Among the variables used in Cox proportional hazard models, depressive mood, assessed by GDS, as well as DBP-MESOR and SBP-Amplitude showed a statistically significant association with the occurrence of vascular outcomes. It should be noted that GDS during the second but not during the first session was statistically significantly associated with adverse vascular outcomes. In univariate analyses, an increase by 2 or 3 points of the GDS score during the second session was associated with a RR of 2.121 (95 % CI: 1.237–3.637) or 3.088 (95 % CI: 1.375–6.935), respectively (P = 0.0063). A 5 mmHg increase in DBP-MESOR and a 3 mmHg increase in SBP-Amplitude were associated with RRs of 2.143 (95 % CI: 1.232–3.727, P = 0.0070) and 0.700 (95 % CI: 0.495–0.989, p = 0.0430), respectively. In multivariate analyses, when both the second GDS score and DBP-MESOR were used as continuous variables in the same model, GDS remained statistically significantly associated with the occurrence of cardiovascular death. After adjustment for DBP-MESOR, a 3-point increase in GDS score was associated with a RR of 2.172 (95 % CI: 1.123–4.200).

Monday and/or Tuesday endpoints of the 7-day ABP showed a statistically significant association with adverse vascular outcomes. A 5-mmHg increase in DBP-5 or DBP-6, namely, in the arithmetic mean of DBP from data collected on Mondays and Tuesdays, was associated with RRs of 1.576 (95 % CI: 1.011–2.457, P = 0.0446) and 1.666 (95 % CI: 1.052–2.639, p = 0.0297), respectively. A 20-unit increase in DP-5 (double product on Mondays) and a 3 % increase in MAP-dip-5 were associated with RRs of 3.067 (95 % CI: 1.075–8.753, p = 0.0362) and 0.824 (95 % CI: 0.683–0.995, P = 0.0439), respectively. Saturday endpoints also related to adverse vascular outcomes. A 3 % increase in PP-dip-3 (dipping ratio of pulse pressure on Saturday nights) and a 5-unit increase in DP-dip-3 (dipping ratio of Saturday’s double product) were inversely associated with adverse vascular outcome, with RRs of 0.865 (95 % CI: 0.756–0.989, P = 0.0337) and 0.752 (95 % CI: 0.596–0.948, P = 0.0158), respectively.

Depression is a risk factor for cardiac morbidity and mortality in patients with coronary heart disease [6–8]. Anda et al. [9] followed a cohort of 2832 persons for an average of 12.4 years. During the reference stage, 11.1 % of the study cohort had a depressed affect. The adjusted relative risk of fatal coronary heart disease (CHD) was 1.4 for patients with depressed affect, compared with patients who were not depressed. For nonfatal CHD, the adjusted relative risk was 1.6. For both fatal and nonfatal CHD, the increased risk associated with depression was independent of established risk factors for CHD (e.g., smoking, cholesterol concentration, family history). Barefoot and Schroll [8] reported that high scores on a measure of symptoms of depressed mood were associated with an increased risk for acute myocardial infarction (AMI) and early mortality during a 27-year follow-up and that the impact of a depressed affect on health did not differ between men and women. We also observed in middle-aged community-dwelling people that a depressive mood, assessed by GDS, predicted the occurrence of vascular diseases beyond the prediction provided by age, gender, ABP, lifestyle, and environmental conditions, as assessed by means of a multivariate Cox model. A depressive mood, especially when enhanced for 1–2 years, was associated with adverse vascular outcomes. Results herein suggest the clinical importance of repeated assessment of a depressive mood and of taking sufficient care of depressed subjects.

Another result of the present study is that in community-dwelling people, circadian and circaseptan characteristics of BP variability derived from 7-day/24-hour ABP predicted the occurrence of vascular diseases beyond the prediction provided by age, gender, depressive mood, and lifestyle, as assessed by means of a multivariate Cox model. Ambulatory 24-hour blood pressure monitoring has become quite popular in recent years, offering a number of advantages from the viewpoint of both diagnosis and treatment [10]. One advantage rests on the fact that clinic blood pressure is not representative of blood pressure values outside the clinic. Results obtained herein show that blood pressure is characterized not only by a circadian but also by a weekly variation. The 24-hour averages of SBP and DBP are seen to differ from one day to another in the same citizen in the same 7-day record. In middle-aged community-dwelling people, DBP-MESOR and SBP-Amplitude predicted the occurrence of vascular diseases, assessed by means of a multivariate Cox model. We found that several endpoints from data collected on Mondays predicted adverse outcomes, suggesting that circaseptan as well as circadian characteristics are associated with the occurrence of vascular disease. We reported on weekly BP characteristics earlier [11]. Most citizens in a community showed a novelty effect (the first day effect), a holiday dip, and a Monday morning surge of SBP. A Monday peak has been reported for the incidence of acute myocardial infarction, sudden cardiac death, and stroke [12–16]. Our previous studies showed that the morning BP surge on Mondays was statistically significantly higher compared with other weekdays [17]. Although a direct association between the Monday surge in BP and cardiovascular events could not be demonstrated directly, it is possible that a morning surge of BP on Mondays may trigger cardiovascular events. This possibility is supported by the fact that we found that depressive citizens had a more prominent circaseptan component for SBP, DBP, and DP, compared with nondepressive subjects [18].

Results herein indicate the clinical importance of taking care of depressed subjects. In view of the strong relation between depression and adverse cardiac events, studies should be done to ascertain whether treatment of depression, especially from the standpoint of chronodiagnosis and chronotherapy, can reduce elevated blood pressure values and improve the altered variability in BP and HR, so as to reduce the incidence of adverse cardiac events. Fewer than 7 days of monitoring means a greater chance of a false diagnosis, on which a treatment decision may depend for the long term, including perhaps a potentially unnecessary treatment for decades. Less than 7 days of monitoring can be compared with taking the pulse for less than a cardiac cycle, if the circaseptan aspect of time structures (chronomes) is viewed as one of many new spectral components, with even longer periods, including transyears characterizing series of blood pressure and heart rate covering at least 5 and up to 38 years [19]. When the ECG and EEG, as well as the blood pressure of small rodents, can be monitored continuously for most of their lifetime, this study is but a small step toward universal 7-day and eventually continuous monitoring of BP and HR, or the ECG.

Depression is a risk factor for cardiac morbidity and mortality in patients with coronary heart disease, especially after an acute myocardial infarction [8, 9]. The link between depression and mortality in older persons is now hotly debated. Fried et al. [20] observed an association between depression and mortality, and Schulz et al. [21] followed to show in a large population-based study of older persons, the Cardiovascular Health Study, that high scores of symptoms of depressed mood are an independent risk factor for mortality. Motivational depletion may be a key underlying mechanism for the depression-mortality effect.

Much research has shown that depression predicts total mortality and the prognosis of patients after MI, but evidence concerning the role of depression in the pathogenesis of coronary disease has been less clear. Ambulatory blood pressure monitoring (ABPM) has become an important tool in the diagnosis and management of hypertension. Several previous studies have indicated that target organ damage and cardiovascular morbidity are more strongly associated with circadian variations and the average BP over a 24-hour span than with office BPs.

In recent years, not only circadian, but also circannual, circaseptan (about 7-day) [12–14, 16, 15], and other variations in the incidence of adverse, notably vascular events, such as acute myocardial infarctions, strokes, and sudden cardiac deaths, have been noted in many studies and may carry important pathophysiological implications. Most studies showed an increased onset in the morning with a peak incidence between 06:00 and 12:00, which has been aligned with the concomitant rapid increase in BP, HR, cardiac output and platelet aggregability, and the corresponding decrease in cardiac vagal activity and in fibrinolytic function. These patterns are synchronized by socio-ecologic factors such as human lifestyles and economic and environmental conditions. We have already reported the presence of a weekly variation in BP in community-dwelling subjects. In this investigation, we focus on how depressed mood is associated with abnormal about-weekly and half-weekly variations in BP and HR, including the BP surge on Mondays.

The 217 subjects monitored between April 2001 and April 2003 were between 24 and 79 years of age (85 men and 132 women; mean age, 56.8 ± 11.3 years), living in Uraus in Hokkaido (latitude 43.45° N and longitude 141.85° E). All subjects visited and utilized the free health screening, counseling, and educational services of the Uraus office. Subjects with definite neurological diseases, such as Parkinsonism and stroke, and those who were too severely ill to stand without help were excluded. The 7-day/24-hour BP recordings were obtained from all participants as outlined above.

Numerous depression rating scales are currently available, which represent a mixture of observer-rated and self-rating scales. In this investigation, subjects were asked about 15 items on a depression scale, through the use of a self-administered questionnaire [5]. When the depression score was ≥5, subjects were considered to have a depressed mood.

Results are reported as mean ± SD. Student’s t-tests were used to compare subjects with depression scores <5 or ≥5. Analyses of variance (ANOVA) with repeated measures were used to assess any day-of-week and group (depression score) effects. Cosinor methods with parameter tests [22] compared the circadian, circasemiseptan, and circaseptan characteristics of SBP, DBP, and HR of subjects in both groups. A P-value below 0.05 was considered to indicate statistical significance (and below 0.10, borderline statistical significance).

Table 11.2 lists the 24-hour averages and standard deviations (SD) of the 7 consecutive days of monitoring for SBP, DBP, PP, DP, HR-SD, SBP-dip, DBP-dip, PP-dip, DP-dip, and HR-dip. The day-to-day variation is statistically significant for the 24-hour average of SBP, DBP, DP, HR-SD, and SBP-dip, which shows a novelty effect. Namely, the 24-hour averages of SBP and DBP and the dipping ratio of SBP were highest and HR-SD and HR-dip were lowest on the first day of the 7-day monitoring. On the other hand, a novelty effect was not detected for DP.