- Email: [email protected]
Parasympathetic predominance is a risk factor for future depression: A prospective cohort study
Journal of Affective Disorders 260 (2020) 232–237
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research paper
Parasympathetic predominance is a risk factor for future depression: A prospective cohort study
T
Hoyoung Ana,1, Ji Won Hanb,1, Hyun-Ghang Jeongc, Tae Hui Kimd, Jung Jae Leee, Seok Bum Leee, ⁎ Joon Hyuk Parkf, Ki Woong Kimb,g,h, a
Department of Neuropsychiatry, St. Andrew's Hospital, Icheon, Republic of Korea Department of Neuropsychiatry, Seoul National University Bundang Hospital, Seongnam, Republic of Korea c Department of Psychiatry, Korea University Guro Hospital, Korea University College of Medicine, Seoul, Republic of Korea d Department of Psychiatry, Yonsei University Wonju Severance Christian Hospital, Wonju, Republic of Korea e Department of Psychiatry, Dankook University Hospital, Cheonan, Republic of Korea f Department of Neuropsychiatry, Jeju National University Hospital, Jeju, Republic of Korea g Department of Psychiatry, Seoul National University College of Medicine, Seoul, Korea h Department of Brain and Cognitive Science, Seoul National University College of Natural Sciences, Seoul, Korea b
ARTICLE INFO
ABSTRACT
Keywords: Depression Biological markers Geriatric Parasympathetic nervous system Cardiovascular Heart-rate variability
Background: Changes in parasympathetic activity have been associated with depression; however, it is not well understood whether these changes are a result of depression, or represent a compensatory mechanism protecting against it. We examined the association of autonomic nervous system activity with the risk of depression in euthymic individuals and those with subsyndromal depression using heart rate variability (HRV) analysis. Methods: From a community-based longitudinal cohort, 464 subjects from the baseline assessment and 253 who completed the 5-year follow-up visit were included in the cross-sectional and prospective analyses, respectively. Linear regression analysis was used to investigate the association of HRV measures with the current and future GDS scores. Logistic regression analysis examined the effect of HRV on future risk of SSD. Results: Low-frequency power (LFN), high-frequency power (HFN), and the LFN/HFN ratio at the baseline assessment were associated with the GDS score at the 5-year follow-up assessment; however, they were not associated with the GDS score at the baseline assessment. High HFN indicated an increased risk of depression at the 5-year follow-up assessment in euthymic subjects (OR = 3.025, 95% CI = 1.184 – 7.726, p = 0.021). Limitations: HRV was not measured at the follow-up assessment and the interval between the assessments was comparatively long. Five-minute ECG recordings were used, and all participants were 65 years old or older. Conclusions: Parasympathetic predominance may precede the onset of depression in older adults.
1. Introduction Major depressive disorder (MDD) and cardiovascular diseases (CVD) are closely related. More than 30% of patients with a first-episode myocardial infarction develop depression within the first year (Strik et al., 2004), and the presence of depression is associated with poor outcomes (Grippo and Johnson, 2002; Ziegelstein, 2001). Furthermore, MDD increases the risk of CVD four-fold, regardless of the presence of other risk factors (Pratt et al., 1996). Parasympathetic activity has been identified as a mediator that links these two conditions (Grippo and Johnson, 2002). Reduced
parasympathetic activity can increase the likelihood of CVD (Pavlov and Tracey, 2005). It has been associated with depression in large-scale studies (Licht et al., 2008), and also in antidepressant-naïve depressive subjects (Kemp et al., 2012). It can cause disinhibition of the inflammatory response which can lead to both depression and CVD (Halaris, 2017). Reduced parasympathetic activity has also been associated with less effective emotional regulation (Thayer and Lane, 2009), maladaptive self-regulation and social engagement (Geisler et al., 2013), and low mental resilience (Carnevali et al., 2018), all of which may influence the development and progression of depression. These findings suggest that reduced parasympathetic activity may act as a
Corresponding author at: Department of Neuropsychiatry, Seoul National University College of Medicine, Seoul National University Bundang Hospital, 166 Gumiro, Bundang-gu, Seongnam-si, Gyeonggi-do, 13494, South Korea. E-mail address: [email protected] (K.W. Kim). 1 These authors contributed equally to this work. ⁎
https://doi.org/10.1016/j.jad.2019.09.015 Received 1 June 2019; Received in revised form 27 July 2019; Accepted 2 September 2019 Available online 03 September 2019 0165-0327/ © 2019 Elsevier B.V. All rights reserved.
Journal of Affective Disorders 260 (2020) 232–237
H. An, et al.
steppingstone to the development of MDD. However, increases in vagal tone have been reported in the subjects with subsyndromal seasonal affective disorder (Austen and Wilson, 2001), indicating that the parasympathetic activity may increase compensatorily against depression in the subsyndromal stage. Analysis of parasympathetic activity in subjects with subsyndromal depression (SSD) may shed light on this matter. SSD can be defined as a state in which an individual has depressive symptoms but does not meet the diagnostic criteria for MDD. Prevalence rates of SSD among community-dwelling older adults have been reported to be 8–20%, and it has been associated with an increased future risk of MDD (LabordeLahoz et al., 2015; Meeks et al., 2011). These findings indicate that SSD may be a prodromal stage of MDD. Consequently, examination of parasympathetic activity at this stage would provide significant insight into its relationship with depression. Heart-rate variability (HRV) consists of beat-to-beat variabilities between adjacent heartbeats, and is a widely used, noninvasive measure of autonomic function including parasympathetic activity (Shaffer and Ginsberg, 2017). Various metrics have been developed that represent sympathetic and parasympathetic activity. Due to its accuracy and ease-of-use, its clinical significance has been recognized by various institutions (Kadish et al., 2001). These characteristics also make it ideal for use in large-scale community studies, and as a screening measure in primary care settings. In this study, we examined the association of the sympathetic and parasympathetic activities and the severity of depressive symptoms in an elderly cohort using heart rate variability (HRV) analysis. We also examined the association of the parasympathetic activity with the risk of incident depression.
and physical examinations, and laboratory tests, using the Korean version of the Mini International Neuropsychiatric Interview (MINI) (You et al., 2006) and the Korean version of the Consortium to Establish a Registry for Alzheimer's Disease Clinical and Neuropsychological Assessment Battery (CERAD-K) (Lee et al., 2002). Research neuropsychologists or trained nurses administered the Mini Mental Status Examination (MMSE) (Lee et al., 2004) to evaluate global cognition, and the Cumulative Illness Rating Scale (CIRS) (Miller et al., 1992) to evaluate the burden of comorbid chronic medical illnesses. Participants were also asked to complete the Korean version of the Geriatric Depression Scale (GDS)(Cho et al., 1999; Yesavage et al., 1982) to evaluate the severity of depressive symptoms. 2.3. Measurements of heart rate variability All examinations took place between 10 am and 1 pm. Subjects were asked to rest in the supine position, while a 12-lead electrocardiogram (ECG) was administered for at least 5 min. Caffeine intake and smoking was prohibited before the examination (van der Kooy et al., 2006). The following EKG measurements (Shaffer and Ginsberg, 2017) were calculated automatically as indicators of HRV using CARDIOPRO™, Version 2.0 software (Thought Technology, Canada): (a) the root mean square of successive differences between normal heartbeats (RMSSD); (b) relative power of the low-frequency band (0.04–0.15 Hz) in normalized units (LFN); (c) relative power of the high-frequency band (0.15–0.4 Hz) in normalized units (HFN); and (d) the LFN-HFN ratio (LFN/HFN). RMSSD and HFN are widely used to estimate parasympathetic activity, whereas LFN is associated with both sympathetic and parasympathetic activity. LFN/HFN estimates sympathovagal balance; however, its significance has been questioned (Shaffer and Ginsberg, 2017). Normalized units were used to minimize the effect of variations in total power and enable direct comparisons between different individuals.
2. Methods 2.1. Study design and participants
2.4. Statistical analysis
Study participants were selected from the Korean Longitudinal Study on Health and Aging (KLoSHA) (Park et al., 2007). The KLoSHA is a community-based, longitudinal cohort study, in which 1000 community-dwelling elderly Koreans, aged 65 years or older, were randomly sampled from the resident registry of a large satellite city of Seoul, Korea, and assessed at baseline and at 5 years. Baseline assessments were conducted from September 2005 through September 2006, and the 5-year follow-up assessments from October 2010 through September 2011. From the cohort, 733 participants completed the baseline HRV assessments and 464 were included in the current cross-sectional analysis, after excluding participants with the following conditions: major or minor depressive disorders (N = 43), other psychiatric or neurologic disorders (N = 102), and those taking medications that may influence HRV, including antidepressants (N = 124). We did not consider the GDS score in excluding the participants. In the prospective analysis, we included 253 out of 267 participants who were euthymic at the baseline assessment and completed the 5-year follow-up assessment, after excluding those who developed dementia (N = 11) or alcohol dependence (N = 3) during the follow-up period (Fig. 1). At the 5-year follow-up assessment, nine participants were diagnosed with SSD, three were diagnosed with minor depressive disorder, and four were diagnosed with MDD. The study protocol was approved by the Institutional Review Board of the Seoul National University Bundang Hospital. In all cases, a written statement of informed consent was obtained from either the participants themselves or their legal guardians.
We classified the subjects who did not have major or minor depressive disorders into two groups (Yesavage et al., 1982): the euthymic (ETH) group whose GDS score was below 10 points and the SSD group whose GDS score was 10 points or higher. We compared baseline demographic and clinical characteristics between the groups using the Chi-square test for categorical variables and Student's t-test for continuous variables. We analyzed the association of the baseline HRV indices with the baseline GDS scores using multiple linear regression models that were adjusted for age, gender, years of education, MMSE score, and CIRS score as covariates. Then, the odds of SSD of high HFN (H-HFN) were compared to those of low HFN (L-HFN) using a logistic regression model that was adjusted for age, gender, years of education, MMSE score, and CIRS score as covariates. We stratified the baseline HFN into two groups: H-HFN group with a baseline HFN of or above the median value of the current study sample, and L-HFN group with a baseline HFN that was below the median value. We also analyzed the association between the baseline HRV indices and the 5-year follow-up GDS scores using multiple linear regression models that were adjusted for the baseline GDS score, age, gender, years of education, MMSE score, and CIRS score as covariates. Then we analyzed the risk of incident depression, including SSD, minor depressive disorder, and MDD, at the 5-year follow-up assessment conferred to the presence of subsyndromal depression at the baseline assessment and the level of baseline HFN using a logistic regression model. In this model, we employed the ETH group with L-HFN as a referent group, and adjusted for age, gender, years of education, MMSE score, and CIRS score as covariates. All statistical analyses were performed using Predictive Analytics SoftWare (PASW) (version 18.0; IBM, USA) and R (version 3.4.3; The R Foundation, USA).
2.2. Assessments In both baseline and 5-year follow-up evaluations, geriatric psychiatrists administered standardized clinical interviews, neurological 233
Journal of Affective Disorders 260 (2020) 232–237
H. An, et al.
Fig. 1. Subject inclusion / exclusion flow-chart.
3. Results
Table 2 Baseline demographic and clinical characteristics of the participants stratified by completion of the follow-up visita.
The participants were 74.98 ± 8.04 years old and educated for 8.32 ± 5.7 years on average at the baseline assessment. About a half were men. The participants in the SSD group were more likely to be women, less educated, and to have lower MMSE scores than those in the ETH group. However, HRV indices were comparable between the two groups (Table 1). Among the 464 participants of the baseline assessment, 253 responded to the 5-year follow-up assessment. The participants who passed away during the follow-up period or refused the follow-up assessment were older, less educated, had lower MMSE scores, and higher GDS scores at the baseline assessment than the responders. However, their baseline HRV indices were comparable to Table 1 Baseline demographic and clinical characteristics of the participantsa. Variables
All (n = 464)
ETH (n = 259)
SSD (n = 205)
Age, years, mean (S.D.) Women, % Education, years, mean (S.D.) BMI, kg/m2, mean (S.D.) MMSE score, mean (S.D.) GDS score, mean (S.D.) CIRS score, mean (S.D.) History of smoking, % HRV Indices RMSSD, ms, mean (S.D.) LFN, nu, mean (S.D.) HFN, nu, mean (S.D.) LFN/HFN, mean (S.D.)
74.98 (8.0) 48.9 8.56 (5.7) 24.00 (3.1) 24.12 (4.2) 9.52 (6.5) 3.74 (2.3) 41.6
74.15 (7.8) 41.7 10.00 (5.6) 24.05 (3.0) 25.09 (4.0) 4.70 (2.7) 3.59 (2.3) 42.1
76.03 (8.3)* 58.0⁎⁎⁎ 6.75 (5.3)⁎⁎⁎ 23.94 (3.2) 22.88 (4.2)⁎⁎⁎ 15.61 (4.5)⁎⁎⁎ 3.93 (2.4) 41.0
55.91 (89.8) 40.34 (21.0) 59.57 (21.1) 1.12 (1.6)
54.57 (88.3) 41.01 (20.9) 58.99 (20.9) 1.16 (1.7)
57.60 (91.8) 39.49 (21.2) 60.32 (21.5) 1.10 (1.4)
Variables
All (n = 464)
Follow-up Yes (n = 268)
No (n = 196)
Age, years, mean (S.D.) Women, % Education, years, mean (S.D.) BMI, kg/m2, mean (S.D.) MMSE score, mean (S.D.) GDS score, mean (S.D.) CIRS score, mean (S.D.) History of smoking, % HRV Indices RMSSD, ms, mean (S.D.) LFN, nu, mean (S.D.) HFN, nu, mean (S.D.) LFN/HFN, mean (S.D.)
74.98 (8.0) 48.9 8.56 (5.7) 24.00 (3.1) 24.12 (4.2) 9.52 (6.5) 3.74 (2.3) 41.6
72.71 (7.0) 46.6 9.43 (5.5) 24.17 (2.9) 25.23 (3.4) 8.55 (6.1) 3.69 (2.4) 40.7
78.10 (8.3)⁎⁎⁎ 52.0 7.38 (5.7)⁎⁎⁎ 23.76 (3.3) 22.59 (4.7)⁎⁎⁎ 10.85 (6.8)⁎⁎⁎ 3.80 (2.3) 42.9
55.91 (89.8) 40.34 (21.0) 59.57 (21.1) 1.12 (1.6)
52.79 (94.2) 41.87 (21.8) 58.02 (21.9) 1.22 (1.7)
60.16 (83.4) 38.25 (19.7) 61.70 (19.9) 1.00 (1.4)
S.D., standard deviation; ETH, euthymic group; SSD, subsyndromal depression group; BMI, body mass index; MMSE, Mini-Mental Status Examination; GDS, Geriatric Depression Scale; CIRS, Cumulative Illness Rating Scale; RMSSD, the root mean square of successive heart rate interval differences; LFN, relative power of the low-frequency band (0.04–0.15 Hz) in normalized units (nu); HFN, relative power of the high-frequency band (0.15–0.4 Hz) in normalized units (nu). a Compared categorical variables using the Chi-square test and continuous variables using Student's t-test. *p < 0.05. ⁎⁎ p < 0.01 ⁎⁎⁎ p < 0.00
those of the responders (Table 2). All baseline HRV indices were not associated with the baseline GDS score (Table 3), and the risk of SSD was not associated with the level of baseline HFN (OR = 1.038, 95% CI = 0.702 – 1.537, p > 0.1). However, LFN, HFN and LFN/HFN ratio were associated with the 5year follow-up GDS score; HFN was positively, and LFN and LFN/HFN ratio were negatively associated with the 5-year follow-up GDS score. RMSSD was not associated with the 5-year follow-up GDS score despite showing a significant correlation with HFN; this was probably due to its wider distribution (Table 3). As shown in Fig. 2, compared to the ETH with L-HFN group, the ETH with H-HFN group showed more than 3 times higher risk of incident SSD at the 5-year follow-up evaluation (OR = 3.025, 95% CI = 1.184 – 7.726, p = 0.021). The SSD with LHFN group (OR: 15.323, 95% CI: 5.603 – 41.903, p < 0.001) and SSD
S.D., standard deviation; ETH, euthymic group; SSD, subsyndromal depression group; BMI, body mass index; MMSE, Mini-Mental Status Examination; GDS, Geriatric Depression Scale; CIRS, Cumulative Illness Rating Scale; RMSSD, the root mean square of successive heart rate interval differences; LFN, relative power of the low-frequency band (0.04–0.15 Hz) in normalized units (nu); HFN, relative power of the high-frequency band (0.15–0.4 Hz) in normalized units (nu). Subjects whose GDS score was below 10 points were categorized into the ETH group, and those whose GDS score was 10 points or higher but were not diagnosed with major or minor depressive disorders were categorized into the SSD group. a Compared categorical variables using the Chi-square test and continuous variables using Student's t-test. ⁎ p < 0.05. ⁎⁎ p < 0.01. ⁎⁎⁎ p < 0.001. 234
Journal of Affective Disorders 260 (2020) 232–237
H. An, et al.
Table 3 Associations of the baseline heart rate variability (HRV) indices with the baseline and 5-year follow-up Geriatric Depression Scale scores. HRV indices
Baseline GDS score (n = 464)a β
95% CI
RMSSD LFN HFN LFN/HFN
−0.008 −0.040 0.038 −0.046
−0.007 −0.039 −0.015 −0.543
5-year follow-up GDS score (n = 253)b β 95% CI to to to to
0.006 0.015 0.038 0.164
0.067 −0.112* 0.114* −0.129⁎⁎
−0.002 to 0.011 −0.061 to −0.006 0.007 to 0.062 −0.878 to −0.146
GDS, Geriatric Depression Scale; CI, confidence interval; RMSSD, the root mean square of successive heart rate interval differences; LFN, relative power of the lowfrequency band (0.04–0.15 Hz) in normalized units; HFN, relative power of the high-frequency band (0.15–0.4 Hz) in normalized units. a Linear regression analyses adjusted for age, gender, years of education, Mini Mental Status Examination (MMSE) score, and Cumulative Illness Rating Scale (CIRS) score at the baseline assessment (n = 464). b Linear regression analyses adjusted for age, gender, years of education, Mini Mental Status Examination (MMSE) score, Cumulative Illness Rating Scale (CIRS) score and baseline GDS score in participants who completed the 5-year follow-up assessment (n = 253). ⁎ p < 0.05. ⁎⁎ p < 0.05. ⁎⁎⁎ p < 0.001
with H-HFN group (OR: 25.660, 95% CI: 8.799 – 74.831, p < 0.001) showed much higher risk of SSD at the 5-year follow-up evaluation than the ETH with L-HFN group. The difference in the risk of SSD at the 5year follow-up evaluation between the SSD with L-HFN group and SSD with HHFN group was not statistically significant (p > 0.1).
response, and then decrease after the onset of clinical depression. In line with the current study, parasympathetic predominance was reported in patients with subsyndromal seasonal affective disorder (Austen and Wilson, 2001). Second, this study excluded the subjects who took antidepressants at the baseline evaluation. Among the six population-based studies included in the meta-analysis, only one study excluded the subjects who were taking anti-depressants (Chen et al., 2010). In that study, the association of depressive symptoms with sympathetic predominance was found in men, but not in women. Third, we included the MMSE score as a covariate to control for the potential influence of cognitive impairment on HRV. However, none of the studies included in the meta-analysis controlled for the effect of cognitive function in their analyses. As the prefrontal cortex exerts control over the autonomic nervous system, HRV has been associated with various areas of cognition, such as executive function, sustained attention, and working memory (Thayer et al., 2009). Fourth, the studies included in the meta-analysis employed different instruments for evaluating depressive symptoms from those used in this study. Instruments developed and validated in younger subjects tend to have more somatic items than the GDS (Balsamo et al., 2018), and somatic symptoms can influence frequency-domain HRV parameters. Finally, the heterogeneity across the studies included in the meta-analysis was large, and its conclusion should be interpreted with caution. Since parasympathetic predominance was associated with the future risk of SSD, but not with current SSD in the present study and not with current MDD in previous studies, parasympathetic activity may increase before the onset of MDD and start to decrease as symptoms become clinically relevant. This enhancement of parasympathetic activity that precedes MDD may be a compensatory response to counteract depression by its anti-inflammatory effects (Aeschbacher et al., 2017; Pavlov and Tracey, 2005). This has been suggested by several previous studies.
4. Discussion This is the first prospective study investigating the effect of HRV on future subsyndromal depression in older adults. We found that LFN and LFN/HFN ratio were negatively, and HFN was positively associated with the future risk of SSD; however, they were not associated with the current GDS scores in older adults. Compared to participants in the ETH with L-HFN group, those in the ETH with H-HFN group showed more than three times higher risk of incident subsyndromal depression. In participants with SSD, those with a high HFN showed slightly higher risk of SSD at the 5-year follow-up evaluation than those with a low HFN. These results suggest that high parasympathetic activity may precede the onset of depressive symptoms. However, in a meta-analysis of five clinical studies and six observational studies that examined the relationship between HRV and depression in adults with a mean age over 60, the individuals with MDD showed lower LFN than but comparable HFN to the euthymic individuals (Brown et al., 2018). This discrepant result may be attributable to several reasons. First, this study included only euthymic subjects or subjects with SSD after excluding subjects with major and minor depressive disorders at the baseline visit. The meta-analysis, along with all previous reports, was focused on subjects with MDD. To summarize the results of the present and previous studies, high-frequency power of HRV, which reflects parasympathetic activity, may increase during the subsyndromal stages of depression, perhaps as a compensatory
Fig. 2. Risk of incident depression at the 5-year follow-up assessment. Logistic regression analysis adjusted for age, gender, years of education, Mini Mental Status Examination (MMSE) score, and Cumulative Illness Rating Scale (CIRS) score (n = 253). ETH with L-HFN group: euthymic and baseline HFN below the median (n = 89). ETH with H-HFN group: euthymic and baseline HFN of or above the median (n = 72). SSD with L-HFN group: subsyndromal depression and baseline HFN below the median (n = 45). SSD with H-HFN group: subsyndromal depression and baseline HFN of or above the median (n = 47). OR, odds ratio; CI, confidence interval; ETH, euthymic group; SSD, subsyndromal depression group; HFN, relative power of the high-frequency band (0.15–0.4 Hz) in normalized units. 235
Journal of Affective Disorders 260 (2020) 232–237
H. An, et al.
Reduced parasympathetic activity resulted in depression by the disinhibition of the inflammatory response (Halaris, 2017), and anti-inflammatory medications, such as non-steroidal anti-inflammatory drugs and cytokine inhibitors, have showed antidepressant effects (Kohler et al., 2016). Serum anti-inflammatory cytokines, such as interleukin-4, can also decrease as depressive symptoms increase in cancer patients (Li et al., 2017). The serum level of adiponectin, an antiinflammatory adipokine, was elevated in the individuals with SSD, but not in euthymic controls nor individuals with major or minor depressive disorders (Jeong et al., 2012). In contrast to the current study, in a previous prospective study on a population aged 49–62 years (mean = 55.9 years) at the first assessment of HRV for about 10 years, higher HFN was associated with lower risk of incident depressive symptoms in men (Jandackova et al., 2016). This discrepancy may raise the possibility that the parasympathetic responses to depression differ by age. The participants of the current study were about 20 years older than the previous study. Older adults show more somatic symptoms than younger adults when they are depressed (Hegeman et al., 2012; Ismail et al., 2013), and the parasympathetic activity was negatively correlated in younger adults while positively correlated in older adults with somatic symptoms (Tak et al., 2010). There are several limitations in the current study. First, HRV was not measured at the 5-year follow-up assessment. Second, the interval between the baseline and follow-up assessments was too long to directly show the causal relationship between parasympathetic predominance and incident SSD. Third, the duration of ECG recording was only 5 min. However, a recording duration of 5 min is the current gold standard for time domain measurements, and has several benefits compared to 24-h ECG monitoring, especially in population-based studies (Laborde et al., 2017). Fourth, all participants were 65 years old or older. Therefore, the results of the current study may not be generalized to the younger populations. Fifth, about 40% of the participants died or were lost to follow-up. However, none of the baseline HRV indices showed significant differences between subjects who completed the follow-up assessment and those who either died or were lost to follow-up (Table 2). In conclusion, parasympathetic predominance may precede the onset of depression in older adults.
References Aeschbacher, S., Schoen, T., Dorig, L., Kreuzmann, R., Neuhauser, C., Schmidt-Trucksass, A., Probst-Hensch, N.M., Risch, M., Risch, L., Conen, D., 2017. Heart rate, heart rate variability and inflammatory biomarkers among young and healthy adults. Ann. Med. 49, 32–41. Austen, M.L., Wilson, G.V., 2001. Increased vagal tone during winter in subsyndromal seasonal affective disorder. Biol. Psychiat. 50, 28–34. Balsamo, M., Cataldi, F., Carlucci, L., Padulo, C., Fairfield, B., 2018. Assessment of latelife depression via self-report measures: a review. Clin. Interv. Aging 13, 2021–2044. Brown, L., Karmakar, C., Gray, R., Jindal, R., Lim, T., Bryant, C., 2018. Heart rate variability alterations in late life depression: a meta-analysis. J. Affect. Disorders 235, 456–466. Carnevali, L., Koenig, J., Sgoifo, A., Ottaviani, C., 2018. Autonomic and brain morphological predictors of stress resilience. Front. Neurosci. 12, 228. Chen, H.C., Yang, C.C., Kuo, T.B., Su, T.P., Chou, P., 2010. Gender differences in the relationship between depression and cardiac autonomic function among community elderly. Int. J. Geriatr. Psych. 25, 314–322. Cho, M.J., Bae, J.N., Suh, G.H., Hahm, B.J., Kim, J.K., Lee, D.W., Kang, M.H., 1999. Validation of geriatric depression scale, Korean version (GDS) in the assessment of DSM-III-R major depression. J. Korean Neuropsychiatr. Assoc 38, 48–63. Geisler, F.C., Kubiak, T., Siewert, K., Weber, H., 2013. Cardiac vagal tone is associated with social engagement and self-regulation. Biol. Psychol. 93, 279–286. Grippo, A.J., Johnson, A.K., 2002. Biological mechanisms in the relationship between depression and heart disease. Neurosci. Biobehav. Rev. 26, 941–962. Halaris, A., 2017. Inflammation-Associated co-morbidity between depression and cardiovascular disease. In: Dantzer, R., Capuron, L. (Eds.), Current Topics in Behavioral Neurosciences 31. Springer International Publishing, Switzerland. Hegeman, J.M., Kok, R.M., van der Mast, R.C., Giltay, E.J., 2012. Phenomenology of depression in older compared with younger adults: meta-analysis. Brit. J. Psychiat. 200, 275–281. Ismail, Z., Fischer, C., McCall, W.V., 2013. What characterizes late-life depression? Psychiat. Clin. N. Am. 36, 483–496. Jandackova, V.K., Britton, A., Malik, M., Steptoe, A., 2016. Heart rate variability and depressive symptoms: a cross-lagged analysis over a 10-year period in the Whitehall II study. Psychol. Med. 46, 2121–2131. Jeong, H.G., Min, B.J., Lim, S., Kim, T.H., Lee, J.J., Park, J.H., Lee, S.B., Han, J.W., Choi, S.H., Park, Y.J., Jang, H.C., Kim, K.W., 2012. Plasma adiponectin elevation in elderly individuals with subsyndromal depression. Psychoneuroendocrinology 37, 948–955. Kadish, A.H., Buxton, A.E., Kennedy, H.L., Knight, B.P., Mason, J.W., Schuger, C.D., Tracy, C.M., Boone, A.W., Elnicki, M., Hirshfeld Jr., J.W., Lorell, B.H., Rodgers, G.P., Tracy, C.M., Weitz, H.H., 2001. ACC/AHA clinical competence statement on electrocardiography and ambulatory electrocardiography. A report of the ACC/AHA/ ACP-ASIM Task Force on Clinical Competence (ACC/AHA committee to develop a clinical competence statement on electrocardiography and ambulatory electrocardiography). J. Am. Coll. Cardiol. 38, 2091–2100. Kemp, A.H., Quintana, D.S., Felmingham, K.L., Matthews, S., Jelinek, H.F., 2012. Depression, comorbid anxiety disorders, and heart rate variability in physically healthy, unmedicated patients: implications for cardiovascular risk. PLoS ONE 7, e30777. Kohler, O., Krogh, J., Mors, O., Benros, M.E., 2016. Inflammation in depression and the potential for anti-inflammatory treatment. Curr. Neuropharmacol. 14, 732–742. Laborde-Lahoz, P., El-Gabalawy, R., Kinley, J., Kirwin, P.D., Sareen, J., Pietrzak, R.H., 2015. Subsyndromal depression among older adults in the USA: prevalence, comorbidity, and risk for new-onset psychiatric disorders in late life. Int. J. Geriatr. Psych. 30, 677–685. Laborde, S., Mosley, E., Thayer, J.F., 2017. Heart rate variability and cardiac vagal tone in psychophysiological research - Recommendations for experiment planning, data analysis, and data reporting. Front. Psychol. 8, 213. Lee, D.Y., Lee, K.U., Lee, J.H., Kim, K.W., Jhoo, J.H., Kim, S.Y., Yoon, J.C., Woo, S.I., Ha, J., Woo, J.I., 2004. A normative study of the CERAD neuropsychological assessment battery in the Korean elderly. J. Int. Neuropsychol. Soc. 10, 72–81. Lee, J.H., Lee, K.U., Lee, D.Y., Kim, K.W., Jhoo, J.H., Kim, J.H., Lee, K.H., Kim, S.Y., Han, S.H., Woo, J.I., 2002. Development of the Korean version of the Consortium to Establish a Registry for Alzheimer's Disease Assessment Packet (CERAD-K): clinical and neuropsychological assessment batteries. J. Gerontol. B Psychol. Sci. Soc. Sci. 57, P47–P53. Li, M., Kouzmina, E., McCusker, M., Rodin, D., Boutros, P.C., Paige, C.J., Rodin, G., 2017. Pro- and anti-inflammatory cytokine associations with major depression in cancer patients. Psychooncology 26, 2149–2156. Licht, C.M., de Geus, E.J., Zitman, F.G., Hoogendijk, W.J., van Dyck, R., Penninx, B.W., 2008. Association between major depressive disorder and heart rate variability in the Netherlands Study of Depression and Anxiety (NESDA). JAMA Psychiat 65, 1358–1367. Meeks, T.W., Vahia, I.V., Lavretsky, H., Kulkarni, G., Jeste, D.V., 2011. A tune in "a minor" can "b major": a review of epidemiology, illness course, and public health implications of subthreshold depression in older adults. J. Affect. Disorders 129, 126–142. Miller, M.D., Paradis, C.F., Houck, P.R., Mazumdar, S., Stack, J.A., Rifai, A.H., Mulsant, B., Reynolds 3rd, C.F., 1992. Rating chronic medical illness burden in geropsychiatric practice and research: application of the Cumulative Illness Rating Scale. Psychiat. Res. 41, 237–248. Park, J.H., Lim, S., Lim, J., Kim, K., Han, M., Yoon, I., Kim, J., Chang, Y., Chang, C.B., Chin, H.J., Choi, E.A., Lee, S.B., Park, Y.J., Paik, N., Kim, T.K., Jang, H.C., Kim, K.W., 2007. An overview of the Korean longitudinal study on health and aging. Psychiatr. Invest. 4, 84–95.
Institutional Board Review The study protocol was approved by the Institutional Review Board of the Seoul National University Bundang Hospital. Role of the funding source This study was supported by a grant from the Korean Health Technology R&D Project, Ministry of Health and Welfare, Republic of Korea [grant no. HI09C1379 (A092077)]. The funding source was not involved in study design, the collection, analysis and interpretation of data, or the writing of the report. CRediT authorship contribution statement Hoyoung An: Formal analysis, Writing - original draft, Writing review & editing. Ji Won Han: Formal analysis, Writing - original draft, Writing - review & editing, Data curation, Writing - review & editing. Hyun-Ghang Jeong: Data curation, Writing - review & editing. Tae Hui Kim: Data curation, Writing - review & editing. Jung Jae Lee: Data curation, Writing - review & editing. Seok Bum Lee: Data curation, Writing - review & editing. Joon Hyuk Park: Data curation, Writing review & editing. Ki Woong Kim: Writing - original draft, Writing review & editing. Declaration of Competing Interest None. 236
Journal of Affective Disorders 260 (2020) 232–237
H. An, et al. Pavlov, V.A., Tracey, K.J., 2005. The cholinergic anti-inflammatory pathway. Brain Behav. Immun 19, 493–499. Pratt, L.A., Ford, D.E., Crum, R.M., Armenian, H.K., Gallo, J.J., Eaton, W.W., 1996. Depression, psychotropic medication, and risk of myocardial infarction. Prospective data from the Baltimore ECA follow-up. Circulation 94, 3123–3129. Shaffer, F., Ginsberg, J.P., 2017. An overview of heart rate variability metrics and norms. Front. Public Health 5, 258. Strik, J.J., Lousberg, R., Cheriex, E.C., Honig, A., 2004. One year cumulative incidence of depression following myocardial infarction and impact on cardiac outcome. J. Psychosom. Res. 56, 59–66. Tak, L.M., Janssens, K.A., Dietrich, A., Slaets, J.P., Rosmalen, J.G., 2010. Age-specific associations between cardiac vagal activity and functional somatic symptoms: a population-based study. Psychother. Psychosom. 79, 179–187. Thayer, J.F., Hansen, A.L., Saus-Rose, E., Johnsen, B.H., 2009. Heart rate variability, prefrontal neural function, and cognitive performance: the neurovisceral integration perspective on self-regulation, adaptation, and health. Ann. Behav. Med. 37,
141–153. Thayer, J.F., Lane, R.D., 2009. Claude Bernard and the heart-brain connection: further elaboration of a model of neurovisceral integration. Neurosci. Biobehav. R. 33, 81–88. van der Kooy, K.G., van Hout, H.P., van Marwijk, H.W., de Haan, M., Stehouwer, C.D., Beekman, A.T., 2006. Differences in heart rate variability between depressed and non-depressed elderly. Int. J. Geriatr. Psych. 21, 147–150. Yesavage, J.A., Brink, T.L., Rose, T.L., Lum, O., Huang, V., Adey, M., Leirer, V.O., 1982. Development and validation of a geriatric depression screening scale: a preliminary report. J. Psychiat. Res 17, 37–49. You, S.W., Koong, K.N., Kim, S.J., Kim, C.H., Chae, J.H., Oh, K.S., Min, K.J., Choi, Y.H., Kim, Y.S., Noh, J.S., Lee, K.C., Jeon, S.I., Oh, D.J., Joo, E.J., Park, H.J., 2006. Validity of Korean version of the miniinternational neuropsychiatric interview. Anxiety Mood 2, 50–55. Ziegelstein, R.C., 2001. Depression in patients recovering from a myocardial infarction. JAMA 286, 1621–1627.
237
Contents lists available at ScienceDirect
Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad
Research paper
Parasympathetic predominance is a risk factor for future depression: A prospective cohort study
T
Hoyoung Ana,1, Ji Won Hanb,1, Hyun-Ghang Jeongc, Tae Hui Kimd, Jung Jae Leee, Seok Bum Leee, ⁎ Joon Hyuk Parkf, Ki Woong Kimb,g,h, a
Department of Neuropsychiatry, St. Andrew's Hospital, Icheon, Republic of Korea Department of Neuropsychiatry, Seoul National University Bundang Hospital, Seongnam, Republic of Korea c Department of Psychiatry, Korea University Guro Hospital, Korea University College of Medicine, Seoul, Republic of Korea d Department of Psychiatry, Yonsei University Wonju Severance Christian Hospital, Wonju, Republic of Korea e Department of Psychiatry, Dankook University Hospital, Cheonan, Republic of Korea f Department of Neuropsychiatry, Jeju National University Hospital, Jeju, Republic of Korea g Department of Psychiatry, Seoul National University College of Medicine, Seoul, Korea h Department of Brain and Cognitive Science, Seoul National University College of Natural Sciences, Seoul, Korea b
ARTICLE INFO
ABSTRACT
Keywords: Depression Biological markers Geriatric Parasympathetic nervous system Cardiovascular Heart-rate variability
Background: Changes in parasympathetic activity have been associated with depression; however, it is not well understood whether these changes are a result of depression, or represent a compensatory mechanism protecting against it. We examined the association of autonomic nervous system activity with the risk of depression in euthymic individuals and those with subsyndromal depression using heart rate variability (HRV) analysis. Methods: From a community-based longitudinal cohort, 464 subjects from the baseline assessment and 253 who completed the 5-year follow-up visit were included in the cross-sectional and prospective analyses, respectively. Linear regression analysis was used to investigate the association of HRV measures with the current and future GDS scores. Logistic regression analysis examined the effect of HRV on future risk of SSD. Results: Low-frequency power (LFN), high-frequency power (HFN), and the LFN/HFN ratio at the baseline assessment were associated with the GDS score at the 5-year follow-up assessment; however, they were not associated with the GDS score at the baseline assessment. High HFN indicated an increased risk of depression at the 5-year follow-up assessment in euthymic subjects (OR = 3.025, 95% CI = 1.184 – 7.726, p = 0.021). Limitations: HRV was not measured at the follow-up assessment and the interval between the assessments was comparatively long. Five-minute ECG recordings were used, and all participants were 65 years old or older. Conclusions: Parasympathetic predominance may precede the onset of depression in older adults.
1. Introduction Major depressive disorder (MDD) and cardiovascular diseases (CVD) are closely related. More than 30% of patients with a first-episode myocardial infarction develop depression within the first year (Strik et al., 2004), and the presence of depression is associated with poor outcomes (Grippo and Johnson, 2002; Ziegelstein, 2001). Furthermore, MDD increases the risk of CVD four-fold, regardless of the presence of other risk factors (Pratt et al., 1996). Parasympathetic activity has been identified as a mediator that links these two conditions (Grippo and Johnson, 2002). Reduced
parasympathetic activity can increase the likelihood of CVD (Pavlov and Tracey, 2005). It has been associated with depression in large-scale studies (Licht et al., 2008), and also in antidepressant-naïve depressive subjects (Kemp et al., 2012). It can cause disinhibition of the inflammatory response which can lead to both depression and CVD (Halaris, 2017). Reduced parasympathetic activity has also been associated with less effective emotional regulation (Thayer and Lane, 2009), maladaptive self-regulation and social engagement (Geisler et al., 2013), and low mental resilience (Carnevali et al., 2018), all of which may influence the development and progression of depression. These findings suggest that reduced parasympathetic activity may act as a
Corresponding author at: Department of Neuropsychiatry, Seoul National University College of Medicine, Seoul National University Bundang Hospital, 166 Gumiro, Bundang-gu, Seongnam-si, Gyeonggi-do, 13494, South Korea. E-mail address: [email protected] (K.W. Kim). 1 These authors contributed equally to this work. ⁎
https://doi.org/10.1016/j.jad.2019.09.015 Received 1 June 2019; Received in revised form 27 July 2019; Accepted 2 September 2019 Available online 03 September 2019 0165-0327/ © 2019 Elsevier B.V. All rights reserved.
Journal of Affective Disorders 260 (2020) 232–237
H. An, et al.
steppingstone to the development of MDD. However, increases in vagal tone have been reported in the subjects with subsyndromal seasonal affective disorder (Austen and Wilson, 2001), indicating that the parasympathetic activity may increase compensatorily against depression in the subsyndromal stage. Analysis of parasympathetic activity in subjects with subsyndromal depression (SSD) may shed light on this matter. SSD can be defined as a state in which an individual has depressive symptoms but does not meet the diagnostic criteria for MDD. Prevalence rates of SSD among community-dwelling older adults have been reported to be 8–20%, and it has been associated with an increased future risk of MDD (LabordeLahoz et al., 2015; Meeks et al., 2011). These findings indicate that SSD may be a prodromal stage of MDD. Consequently, examination of parasympathetic activity at this stage would provide significant insight into its relationship with depression. Heart-rate variability (HRV) consists of beat-to-beat variabilities between adjacent heartbeats, and is a widely used, noninvasive measure of autonomic function including parasympathetic activity (Shaffer and Ginsberg, 2017). Various metrics have been developed that represent sympathetic and parasympathetic activity. Due to its accuracy and ease-of-use, its clinical significance has been recognized by various institutions (Kadish et al., 2001). These characteristics also make it ideal for use in large-scale community studies, and as a screening measure in primary care settings. In this study, we examined the association of the sympathetic and parasympathetic activities and the severity of depressive symptoms in an elderly cohort using heart rate variability (HRV) analysis. We also examined the association of the parasympathetic activity with the risk of incident depression.
and physical examinations, and laboratory tests, using the Korean version of the Mini International Neuropsychiatric Interview (MINI) (You et al., 2006) and the Korean version of the Consortium to Establish a Registry for Alzheimer's Disease Clinical and Neuropsychological Assessment Battery (CERAD-K) (Lee et al., 2002). Research neuropsychologists or trained nurses administered the Mini Mental Status Examination (MMSE) (Lee et al., 2004) to evaluate global cognition, and the Cumulative Illness Rating Scale (CIRS) (Miller et al., 1992) to evaluate the burden of comorbid chronic medical illnesses. Participants were also asked to complete the Korean version of the Geriatric Depression Scale (GDS)(Cho et al., 1999; Yesavage et al., 1982) to evaluate the severity of depressive symptoms. 2.3. Measurements of heart rate variability All examinations took place between 10 am and 1 pm. Subjects were asked to rest in the supine position, while a 12-lead electrocardiogram (ECG) was administered for at least 5 min. Caffeine intake and smoking was prohibited before the examination (van der Kooy et al., 2006). The following EKG measurements (Shaffer and Ginsberg, 2017) were calculated automatically as indicators of HRV using CARDIOPRO™, Version 2.0 software (Thought Technology, Canada): (a) the root mean square of successive differences between normal heartbeats (RMSSD); (b) relative power of the low-frequency band (0.04–0.15 Hz) in normalized units (LFN); (c) relative power of the high-frequency band (0.15–0.4 Hz) in normalized units (HFN); and (d) the LFN-HFN ratio (LFN/HFN). RMSSD and HFN are widely used to estimate parasympathetic activity, whereas LFN is associated with both sympathetic and parasympathetic activity. LFN/HFN estimates sympathovagal balance; however, its significance has been questioned (Shaffer and Ginsberg, 2017). Normalized units were used to minimize the effect of variations in total power and enable direct comparisons between different individuals.
2. Methods 2.1. Study design and participants
2.4. Statistical analysis
Study participants were selected from the Korean Longitudinal Study on Health and Aging (KLoSHA) (Park et al., 2007). The KLoSHA is a community-based, longitudinal cohort study, in which 1000 community-dwelling elderly Koreans, aged 65 years or older, were randomly sampled from the resident registry of a large satellite city of Seoul, Korea, and assessed at baseline and at 5 years. Baseline assessments were conducted from September 2005 through September 2006, and the 5-year follow-up assessments from October 2010 through September 2011. From the cohort, 733 participants completed the baseline HRV assessments and 464 were included in the current cross-sectional analysis, after excluding participants with the following conditions: major or minor depressive disorders (N = 43), other psychiatric or neurologic disorders (N = 102), and those taking medications that may influence HRV, including antidepressants (N = 124). We did not consider the GDS score in excluding the participants. In the prospective analysis, we included 253 out of 267 participants who were euthymic at the baseline assessment and completed the 5-year follow-up assessment, after excluding those who developed dementia (N = 11) or alcohol dependence (N = 3) during the follow-up period (Fig. 1). At the 5-year follow-up assessment, nine participants were diagnosed with SSD, three were diagnosed with minor depressive disorder, and four were diagnosed with MDD. The study protocol was approved by the Institutional Review Board of the Seoul National University Bundang Hospital. In all cases, a written statement of informed consent was obtained from either the participants themselves or their legal guardians.
We classified the subjects who did not have major or minor depressive disorders into two groups (Yesavage et al., 1982): the euthymic (ETH) group whose GDS score was below 10 points and the SSD group whose GDS score was 10 points or higher. We compared baseline demographic and clinical characteristics between the groups using the Chi-square test for categorical variables and Student's t-test for continuous variables. We analyzed the association of the baseline HRV indices with the baseline GDS scores using multiple linear regression models that were adjusted for age, gender, years of education, MMSE score, and CIRS score as covariates. Then, the odds of SSD of high HFN (H-HFN) were compared to those of low HFN (L-HFN) using a logistic regression model that was adjusted for age, gender, years of education, MMSE score, and CIRS score as covariates. We stratified the baseline HFN into two groups: H-HFN group with a baseline HFN of or above the median value of the current study sample, and L-HFN group with a baseline HFN that was below the median value. We also analyzed the association between the baseline HRV indices and the 5-year follow-up GDS scores using multiple linear regression models that were adjusted for the baseline GDS score, age, gender, years of education, MMSE score, and CIRS score as covariates. Then we analyzed the risk of incident depression, including SSD, minor depressive disorder, and MDD, at the 5-year follow-up assessment conferred to the presence of subsyndromal depression at the baseline assessment and the level of baseline HFN using a logistic regression model. In this model, we employed the ETH group with L-HFN as a referent group, and adjusted for age, gender, years of education, MMSE score, and CIRS score as covariates. All statistical analyses were performed using Predictive Analytics SoftWare (PASW) (version 18.0; IBM, USA) and R (version 3.4.3; The R Foundation, USA).
2.2. Assessments In both baseline and 5-year follow-up evaluations, geriatric psychiatrists administered standardized clinical interviews, neurological 233
Journal of Affective Disorders 260 (2020) 232–237
H. An, et al.
Fig. 1. Subject inclusion / exclusion flow-chart.
3. Results
Table 2 Baseline demographic and clinical characteristics of the participants stratified by completion of the follow-up visita.
The participants were 74.98 ± 8.04 years old and educated for 8.32 ± 5.7 years on average at the baseline assessment. About a half were men. The participants in the SSD group were more likely to be women, less educated, and to have lower MMSE scores than those in the ETH group. However, HRV indices were comparable between the two groups (Table 1). Among the 464 participants of the baseline assessment, 253 responded to the 5-year follow-up assessment. The participants who passed away during the follow-up period or refused the follow-up assessment were older, less educated, had lower MMSE scores, and higher GDS scores at the baseline assessment than the responders. However, their baseline HRV indices were comparable to Table 1 Baseline demographic and clinical characteristics of the participantsa. Variables
All (n = 464)
ETH (n = 259)
SSD (n = 205)
Age, years, mean (S.D.) Women, % Education, years, mean (S.D.) BMI, kg/m2, mean (S.D.) MMSE score, mean (S.D.) GDS score, mean (S.D.) CIRS score, mean (S.D.) History of smoking, % HRV Indices RMSSD, ms, mean (S.D.) LFN, nu, mean (S.D.) HFN, nu, mean (S.D.) LFN/HFN, mean (S.D.)
74.98 (8.0) 48.9 8.56 (5.7) 24.00 (3.1) 24.12 (4.2) 9.52 (6.5) 3.74 (2.3) 41.6
74.15 (7.8) 41.7 10.00 (5.6) 24.05 (3.0) 25.09 (4.0) 4.70 (2.7) 3.59 (2.3) 42.1
76.03 (8.3)* 58.0⁎⁎⁎ 6.75 (5.3)⁎⁎⁎ 23.94 (3.2) 22.88 (4.2)⁎⁎⁎ 15.61 (4.5)⁎⁎⁎ 3.93 (2.4) 41.0
55.91 (89.8) 40.34 (21.0) 59.57 (21.1) 1.12 (1.6)
54.57 (88.3) 41.01 (20.9) 58.99 (20.9) 1.16 (1.7)
57.60 (91.8) 39.49 (21.2) 60.32 (21.5) 1.10 (1.4)
Variables
All (n = 464)
Follow-up Yes (n = 268)
No (n = 196)
Age, years, mean (S.D.) Women, % Education, years, mean (S.D.) BMI, kg/m2, mean (S.D.) MMSE score, mean (S.D.) GDS score, mean (S.D.) CIRS score, mean (S.D.) History of smoking, % HRV Indices RMSSD, ms, mean (S.D.) LFN, nu, mean (S.D.) HFN, nu, mean (S.D.) LFN/HFN, mean (S.D.)
74.98 (8.0) 48.9 8.56 (5.7) 24.00 (3.1) 24.12 (4.2) 9.52 (6.5) 3.74 (2.3) 41.6
72.71 (7.0) 46.6 9.43 (5.5) 24.17 (2.9) 25.23 (3.4) 8.55 (6.1) 3.69 (2.4) 40.7
78.10 (8.3)⁎⁎⁎ 52.0 7.38 (5.7)⁎⁎⁎ 23.76 (3.3) 22.59 (4.7)⁎⁎⁎ 10.85 (6.8)⁎⁎⁎ 3.80 (2.3) 42.9
55.91 (89.8) 40.34 (21.0) 59.57 (21.1) 1.12 (1.6)
52.79 (94.2) 41.87 (21.8) 58.02 (21.9) 1.22 (1.7)
60.16 (83.4) 38.25 (19.7) 61.70 (19.9) 1.00 (1.4)
S.D., standard deviation; ETH, euthymic group; SSD, subsyndromal depression group; BMI, body mass index; MMSE, Mini-Mental Status Examination; GDS, Geriatric Depression Scale; CIRS, Cumulative Illness Rating Scale; RMSSD, the root mean square of successive heart rate interval differences; LFN, relative power of the low-frequency band (0.04–0.15 Hz) in normalized units (nu); HFN, relative power of the high-frequency band (0.15–0.4 Hz) in normalized units (nu). a Compared categorical variables using the Chi-square test and continuous variables using Student's t-test. *p < 0.05. ⁎⁎ p < 0.01 ⁎⁎⁎ p < 0.00
those of the responders (Table 2). All baseline HRV indices were not associated with the baseline GDS score (Table 3), and the risk of SSD was not associated with the level of baseline HFN (OR = 1.038, 95% CI = 0.702 – 1.537, p > 0.1). However, LFN, HFN and LFN/HFN ratio were associated with the 5year follow-up GDS score; HFN was positively, and LFN and LFN/HFN ratio were negatively associated with the 5-year follow-up GDS score. RMSSD was not associated with the 5-year follow-up GDS score despite showing a significant correlation with HFN; this was probably due to its wider distribution (Table 3). As shown in Fig. 2, compared to the ETH with L-HFN group, the ETH with H-HFN group showed more than 3 times higher risk of incident SSD at the 5-year follow-up evaluation (OR = 3.025, 95% CI = 1.184 – 7.726, p = 0.021). The SSD with LHFN group (OR: 15.323, 95% CI: 5.603 – 41.903, p < 0.001) and SSD
S.D., standard deviation; ETH, euthymic group; SSD, subsyndromal depression group; BMI, body mass index; MMSE, Mini-Mental Status Examination; GDS, Geriatric Depression Scale; CIRS, Cumulative Illness Rating Scale; RMSSD, the root mean square of successive heart rate interval differences; LFN, relative power of the low-frequency band (0.04–0.15 Hz) in normalized units (nu); HFN, relative power of the high-frequency band (0.15–0.4 Hz) in normalized units (nu). Subjects whose GDS score was below 10 points were categorized into the ETH group, and those whose GDS score was 10 points or higher but were not diagnosed with major or minor depressive disorders were categorized into the SSD group. a Compared categorical variables using the Chi-square test and continuous variables using Student's t-test. ⁎ p < 0.05. ⁎⁎ p < 0.01. ⁎⁎⁎ p < 0.001. 234
Journal of Affective Disorders 260 (2020) 232–237
H. An, et al.
Table 3 Associations of the baseline heart rate variability (HRV) indices with the baseline and 5-year follow-up Geriatric Depression Scale scores. HRV indices
Baseline GDS score (n = 464)a β
95% CI
RMSSD LFN HFN LFN/HFN
−0.008 −0.040 0.038 −0.046
−0.007 −0.039 −0.015 −0.543
5-year follow-up GDS score (n = 253)b β 95% CI to to to to
0.006 0.015 0.038 0.164
0.067 −0.112* 0.114* −0.129⁎⁎
−0.002 to 0.011 −0.061 to −0.006 0.007 to 0.062 −0.878 to −0.146
GDS, Geriatric Depression Scale; CI, confidence interval; RMSSD, the root mean square of successive heart rate interval differences; LFN, relative power of the lowfrequency band (0.04–0.15 Hz) in normalized units; HFN, relative power of the high-frequency band (0.15–0.4 Hz) in normalized units. a Linear regression analyses adjusted for age, gender, years of education, Mini Mental Status Examination (MMSE) score, and Cumulative Illness Rating Scale (CIRS) score at the baseline assessment (n = 464). b Linear regression analyses adjusted for age, gender, years of education, Mini Mental Status Examination (MMSE) score, Cumulative Illness Rating Scale (CIRS) score and baseline GDS score in participants who completed the 5-year follow-up assessment (n = 253). ⁎ p < 0.05. ⁎⁎ p < 0.05. ⁎⁎⁎ p < 0.001
with H-HFN group (OR: 25.660, 95% CI: 8.799 – 74.831, p < 0.001) showed much higher risk of SSD at the 5-year follow-up evaluation than the ETH with L-HFN group. The difference in the risk of SSD at the 5year follow-up evaluation between the SSD with L-HFN group and SSD with HHFN group was not statistically significant (p > 0.1).
response, and then decrease after the onset of clinical depression. In line with the current study, parasympathetic predominance was reported in patients with subsyndromal seasonal affective disorder (Austen and Wilson, 2001). Second, this study excluded the subjects who took antidepressants at the baseline evaluation. Among the six population-based studies included in the meta-analysis, only one study excluded the subjects who were taking anti-depressants (Chen et al., 2010). In that study, the association of depressive symptoms with sympathetic predominance was found in men, but not in women. Third, we included the MMSE score as a covariate to control for the potential influence of cognitive impairment on HRV. However, none of the studies included in the meta-analysis controlled for the effect of cognitive function in their analyses. As the prefrontal cortex exerts control over the autonomic nervous system, HRV has been associated with various areas of cognition, such as executive function, sustained attention, and working memory (Thayer et al., 2009). Fourth, the studies included in the meta-analysis employed different instruments for evaluating depressive symptoms from those used in this study. Instruments developed and validated in younger subjects tend to have more somatic items than the GDS (Balsamo et al., 2018), and somatic symptoms can influence frequency-domain HRV parameters. Finally, the heterogeneity across the studies included in the meta-analysis was large, and its conclusion should be interpreted with caution. Since parasympathetic predominance was associated with the future risk of SSD, but not with current SSD in the present study and not with current MDD in previous studies, parasympathetic activity may increase before the onset of MDD and start to decrease as symptoms become clinically relevant. This enhancement of parasympathetic activity that precedes MDD may be a compensatory response to counteract depression by its anti-inflammatory effects (Aeschbacher et al., 2017; Pavlov and Tracey, 2005). This has been suggested by several previous studies.
4. Discussion This is the first prospective study investigating the effect of HRV on future subsyndromal depression in older adults. We found that LFN and LFN/HFN ratio were negatively, and HFN was positively associated with the future risk of SSD; however, they were not associated with the current GDS scores in older adults. Compared to participants in the ETH with L-HFN group, those in the ETH with H-HFN group showed more than three times higher risk of incident subsyndromal depression. In participants with SSD, those with a high HFN showed slightly higher risk of SSD at the 5-year follow-up evaluation than those with a low HFN. These results suggest that high parasympathetic activity may precede the onset of depressive symptoms. However, in a meta-analysis of five clinical studies and six observational studies that examined the relationship between HRV and depression in adults with a mean age over 60, the individuals with MDD showed lower LFN than but comparable HFN to the euthymic individuals (Brown et al., 2018). This discrepant result may be attributable to several reasons. First, this study included only euthymic subjects or subjects with SSD after excluding subjects with major and minor depressive disorders at the baseline visit. The meta-analysis, along with all previous reports, was focused on subjects with MDD. To summarize the results of the present and previous studies, high-frequency power of HRV, which reflects parasympathetic activity, may increase during the subsyndromal stages of depression, perhaps as a compensatory
Fig. 2. Risk of incident depression at the 5-year follow-up assessment. Logistic regression analysis adjusted for age, gender, years of education, Mini Mental Status Examination (MMSE) score, and Cumulative Illness Rating Scale (CIRS) score (n = 253). ETH with L-HFN group: euthymic and baseline HFN below the median (n = 89). ETH with H-HFN group: euthymic and baseline HFN of or above the median (n = 72). SSD with L-HFN group: subsyndromal depression and baseline HFN below the median (n = 45). SSD with H-HFN group: subsyndromal depression and baseline HFN of or above the median (n = 47). OR, odds ratio; CI, confidence interval; ETH, euthymic group; SSD, subsyndromal depression group; HFN, relative power of the high-frequency band (0.15–0.4 Hz) in normalized units. 235
Journal of Affective Disorders 260 (2020) 232–237
H. An, et al.
Reduced parasympathetic activity resulted in depression by the disinhibition of the inflammatory response (Halaris, 2017), and anti-inflammatory medications, such as non-steroidal anti-inflammatory drugs and cytokine inhibitors, have showed antidepressant effects (Kohler et al., 2016). Serum anti-inflammatory cytokines, such as interleukin-4, can also decrease as depressive symptoms increase in cancer patients (Li et al., 2017). The serum level of adiponectin, an antiinflammatory adipokine, was elevated in the individuals with SSD, but not in euthymic controls nor individuals with major or minor depressive disorders (Jeong et al., 2012). In contrast to the current study, in a previous prospective study on a population aged 49–62 years (mean = 55.9 years) at the first assessment of HRV for about 10 years, higher HFN was associated with lower risk of incident depressive symptoms in men (Jandackova et al., 2016). This discrepancy may raise the possibility that the parasympathetic responses to depression differ by age. The participants of the current study were about 20 years older than the previous study. Older adults show more somatic symptoms than younger adults when they are depressed (Hegeman et al., 2012; Ismail et al., 2013), and the parasympathetic activity was negatively correlated in younger adults while positively correlated in older adults with somatic symptoms (Tak et al., 2010). There are several limitations in the current study. First, HRV was not measured at the 5-year follow-up assessment. Second, the interval between the baseline and follow-up assessments was too long to directly show the causal relationship between parasympathetic predominance and incident SSD. Third, the duration of ECG recording was only 5 min. However, a recording duration of 5 min is the current gold standard for time domain measurements, and has several benefits compared to 24-h ECG monitoring, especially in population-based studies (Laborde et al., 2017). Fourth, all participants were 65 years old or older. Therefore, the results of the current study may not be generalized to the younger populations. Fifth, about 40% of the participants died or were lost to follow-up. However, none of the baseline HRV indices showed significant differences between subjects who completed the follow-up assessment and those who either died or were lost to follow-up (Table 2). In conclusion, parasympathetic predominance may precede the onset of depression in older adults.
References Aeschbacher, S., Schoen, T., Dorig, L., Kreuzmann, R., Neuhauser, C., Schmidt-Trucksass, A., Probst-Hensch, N.M., Risch, M., Risch, L., Conen, D., 2017. Heart rate, heart rate variability and inflammatory biomarkers among young and healthy adults. Ann. Med. 49, 32–41. Austen, M.L., Wilson, G.V., 2001. Increased vagal tone during winter in subsyndromal seasonal affective disorder. Biol. Psychiat. 50, 28–34. Balsamo, M., Cataldi, F., Carlucci, L., Padulo, C., Fairfield, B., 2018. Assessment of latelife depression via self-report measures: a review. Clin. Interv. Aging 13, 2021–2044. Brown, L., Karmakar, C., Gray, R., Jindal, R., Lim, T., Bryant, C., 2018. Heart rate variability alterations in late life depression: a meta-analysis. J. Affect. Disorders 235, 456–466. Carnevali, L., Koenig, J., Sgoifo, A., Ottaviani, C., 2018. Autonomic and brain morphological predictors of stress resilience. Front. Neurosci. 12, 228. Chen, H.C., Yang, C.C., Kuo, T.B., Su, T.P., Chou, P., 2010. Gender differences in the relationship between depression and cardiac autonomic function among community elderly. Int. J. Geriatr. Psych. 25, 314–322. Cho, M.J., Bae, J.N., Suh, G.H., Hahm, B.J., Kim, J.K., Lee, D.W., Kang, M.H., 1999. Validation of geriatric depression scale, Korean version (GDS) in the assessment of DSM-III-R major depression. J. Korean Neuropsychiatr. Assoc 38, 48–63. Geisler, F.C., Kubiak, T., Siewert, K., Weber, H., 2013. Cardiac vagal tone is associated with social engagement and self-regulation. Biol. Psychol. 93, 279–286. Grippo, A.J., Johnson, A.K., 2002. Biological mechanisms in the relationship between depression and heart disease. Neurosci. Biobehav. Rev. 26, 941–962. Halaris, A., 2017. Inflammation-Associated co-morbidity between depression and cardiovascular disease. In: Dantzer, R., Capuron, L. (Eds.), Current Topics in Behavioral Neurosciences 31. Springer International Publishing, Switzerland. Hegeman, J.M., Kok, R.M., van der Mast, R.C., Giltay, E.J., 2012. Phenomenology of depression in older compared with younger adults: meta-analysis. Brit. J. Psychiat. 200, 275–281. Ismail, Z., Fischer, C., McCall, W.V., 2013. What characterizes late-life depression? Psychiat. Clin. N. Am. 36, 483–496. Jandackova, V.K., Britton, A., Malik, M., Steptoe, A., 2016. Heart rate variability and depressive symptoms: a cross-lagged analysis over a 10-year period in the Whitehall II study. Psychol. Med. 46, 2121–2131. Jeong, H.G., Min, B.J., Lim, S., Kim, T.H., Lee, J.J., Park, J.H., Lee, S.B., Han, J.W., Choi, S.H., Park, Y.J., Jang, H.C., Kim, K.W., 2012. Plasma adiponectin elevation in elderly individuals with subsyndromal depression. Psychoneuroendocrinology 37, 948–955. Kadish, A.H., Buxton, A.E., Kennedy, H.L., Knight, B.P., Mason, J.W., Schuger, C.D., Tracy, C.M., Boone, A.W., Elnicki, M., Hirshfeld Jr., J.W., Lorell, B.H., Rodgers, G.P., Tracy, C.M., Weitz, H.H., 2001. ACC/AHA clinical competence statement on electrocardiography and ambulatory electrocardiography. A report of the ACC/AHA/ ACP-ASIM Task Force on Clinical Competence (ACC/AHA committee to develop a clinical competence statement on electrocardiography and ambulatory electrocardiography). J. Am. Coll. Cardiol. 38, 2091–2100. Kemp, A.H., Quintana, D.S., Felmingham, K.L., Matthews, S., Jelinek, H.F., 2012. Depression, comorbid anxiety disorders, and heart rate variability in physically healthy, unmedicated patients: implications for cardiovascular risk. PLoS ONE 7, e30777. Kohler, O., Krogh, J., Mors, O., Benros, M.E., 2016. Inflammation in depression and the potential for anti-inflammatory treatment. Curr. Neuropharmacol. 14, 732–742. Laborde-Lahoz, P., El-Gabalawy, R., Kinley, J., Kirwin, P.D., Sareen, J., Pietrzak, R.H., 2015. Subsyndromal depression among older adults in the USA: prevalence, comorbidity, and risk for new-onset psychiatric disorders in late life. Int. J. Geriatr. Psych. 30, 677–685. Laborde, S., Mosley, E., Thayer, J.F., 2017. Heart rate variability and cardiac vagal tone in psychophysiological research - Recommendations for experiment planning, data analysis, and data reporting. Front. Psychol. 8, 213. Lee, D.Y., Lee, K.U., Lee, J.H., Kim, K.W., Jhoo, J.H., Kim, S.Y., Yoon, J.C., Woo, S.I., Ha, J., Woo, J.I., 2004. A normative study of the CERAD neuropsychological assessment battery in the Korean elderly. J. Int. Neuropsychol. Soc. 10, 72–81. Lee, J.H., Lee, K.U., Lee, D.Y., Kim, K.W., Jhoo, J.H., Kim, J.H., Lee, K.H., Kim, S.Y., Han, S.H., Woo, J.I., 2002. Development of the Korean version of the Consortium to Establish a Registry for Alzheimer's Disease Assessment Packet (CERAD-K): clinical and neuropsychological assessment batteries. J. Gerontol. B Psychol. Sci. Soc. Sci. 57, P47–P53. Li, M., Kouzmina, E., McCusker, M., Rodin, D., Boutros, P.C., Paige, C.J., Rodin, G., 2017. Pro- and anti-inflammatory cytokine associations with major depression in cancer patients. Psychooncology 26, 2149–2156. Licht, C.M., de Geus, E.J., Zitman, F.G., Hoogendijk, W.J., van Dyck, R., Penninx, B.W., 2008. Association between major depressive disorder and heart rate variability in the Netherlands Study of Depression and Anxiety (NESDA). JAMA Psychiat 65, 1358–1367. Meeks, T.W., Vahia, I.V., Lavretsky, H., Kulkarni, G., Jeste, D.V., 2011. A tune in "a minor" can "b major": a review of epidemiology, illness course, and public health implications of subthreshold depression in older adults. J. Affect. Disorders 129, 126–142. Miller, M.D., Paradis, C.F., Houck, P.R., Mazumdar, S., Stack, J.A., Rifai, A.H., Mulsant, B., Reynolds 3rd, C.F., 1992. Rating chronic medical illness burden in geropsychiatric practice and research: application of the Cumulative Illness Rating Scale. Psychiat. Res. 41, 237–248. Park, J.H., Lim, S., Lim, J., Kim, K., Han, M., Yoon, I., Kim, J., Chang, Y., Chang, C.B., Chin, H.J., Choi, E.A., Lee, S.B., Park, Y.J., Paik, N., Kim, T.K., Jang, H.C., Kim, K.W., 2007. An overview of the Korean longitudinal study on health and aging. Psychiatr. Invest. 4, 84–95.
Institutional Board Review The study protocol was approved by the Institutional Review Board of the Seoul National University Bundang Hospital. Role of the funding source This study was supported by a grant from the Korean Health Technology R&D Project, Ministry of Health and Welfare, Republic of Korea [grant no. HI09C1379 (A092077)]. The funding source was not involved in study design, the collection, analysis and interpretation of data, or the writing of the report. CRediT authorship contribution statement Hoyoung An: Formal analysis, Writing - original draft, Writing review & editing. Ji Won Han: Formal analysis, Writing - original draft, Writing - review & editing, Data curation, Writing - review & editing. Hyun-Ghang Jeong: Data curation, Writing - review & editing. Tae Hui Kim: Data curation, Writing - review & editing. Jung Jae Lee: Data curation, Writing - review & editing. Seok Bum Lee: Data curation, Writing - review & editing. Joon Hyuk Park: Data curation, Writing review & editing. Ki Woong Kim: Writing - original draft, Writing review & editing. Declaration of Competing Interest None. 236
Journal of Affective Disorders 260 (2020) 232–237
H. An, et al. Pavlov, V.A., Tracey, K.J., 2005. The cholinergic anti-inflammatory pathway. Brain Behav. Immun 19, 493–499. Pratt, L.A., Ford, D.E., Crum, R.M., Armenian, H.K., Gallo, J.J., Eaton, W.W., 1996. Depression, psychotropic medication, and risk of myocardial infarction. Prospective data from the Baltimore ECA follow-up. Circulation 94, 3123–3129. Shaffer, F., Ginsberg, J.P., 2017. An overview of heart rate variability metrics and norms. Front. Public Health 5, 258. Strik, J.J., Lousberg, R., Cheriex, E.C., Honig, A., 2004. One year cumulative incidence of depression following myocardial infarction and impact on cardiac outcome. J. Psychosom. Res. 56, 59–66. Tak, L.M., Janssens, K.A., Dietrich, A., Slaets, J.P., Rosmalen, J.G., 2010. Age-specific associations between cardiac vagal activity and functional somatic symptoms: a population-based study. Psychother. Psychosom. 79, 179–187. Thayer, J.F., Hansen, A.L., Saus-Rose, E., Johnsen, B.H., 2009. Heart rate variability, prefrontal neural function, and cognitive performance: the neurovisceral integration perspective on self-regulation, adaptation, and health. Ann. Behav. Med. 37,
141–153. Thayer, J.F., Lane, R.D., 2009. Claude Bernard and the heart-brain connection: further elaboration of a model of neurovisceral integration. Neurosci. Biobehav. R. 33, 81–88. van der Kooy, K.G., van Hout, H.P., van Marwijk, H.W., de Haan, M., Stehouwer, C.D., Beekman, A.T., 2006. Differences in heart rate variability between depressed and non-depressed elderly. Int. J. Geriatr. Psych. 21, 147–150. Yesavage, J.A., Brink, T.L., Rose, T.L., Lum, O., Huang, V., Adey, M., Leirer, V.O., 1982. Development and validation of a geriatric depression screening scale: a preliminary report. J. Psychiat. Res 17, 37–49. You, S.W., Koong, K.N., Kim, S.J., Kim, C.H., Chae, J.H., Oh, K.S., Min, K.J., Choi, Y.H., Kim, Y.S., Noh, J.S., Lee, K.C., Jeon, S.I., Oh, D.J., Joo, E.J., Park, H.J., 2006. Validity of Korean version of the miniinternational neuropsychiatric interview. Anxiety Mood 2, 50–55. Ziegelstein, R.C., 2001. Depression in patients recovering from a myocardial infarction. JAMA 286, 1621–1627.
237