Autonomic Changes Associated with Clinical Depression as Assessed by Short-term Heart Rate Variability (HRV): A Comparative Study

Abstract

Background:

In the present scenario, mental health issues have become a global burden. Among all the mental disorders, depression is one of the most disabling and has been a known cardiovascular risk. Autonomic attenuation is put forth as the cause. The literature available on the connection between clinical depression and cardiac autonomic attenuation is limited and inconclusive. Hence, to provide more clarity, with the aid of short-term heart rate variability (HRV), autonomic changes in clinically depressed individuals were assessed.

Methods:

Based on the set criteria, we recruited 82 subjects from the hospital’s outpatient department after ethical approval. Among them, 41 were depressed individuals, and the rest were non-depressed healthy controls. Depressed individuals were categorized based on their Hamilton scores. Both the groups were subjected to short-term HRV, the measures obtained were compared, and the HRV measures of the depressed individuals were correlated with their Hamilton scores.

Results:

HRV measures that reflect cardiovagal activity were found to be significantly less (p = .026) in the depressed individuals. No gender-influenced differences were observed among the depressed. Groups with different levels of depression also revealed no significant differences in their autonomic activity. Hamilton scores of the depressed individuals exhibited no significant correlation with their HRV parameters.

Conclusion:

Based on our HRV findings, we conclude that the depressed individuals have reduced cardiovagal activity.

Key Messages:

Since our findings point toward altered cardiac autonomic activity in clinically depressed patients, the potential cardiovascular risk in them may be seriously considered and addressed.

Mental health is essential for the effective functioning of an individual, and mental disorders contribute immensely to the health care burden. ¹ More than 1 billion people globally were affected by mental or addictive disorders in 2016, which is about 16% of the world’s population. ² Among the 25 listed mental health issues, depressive and anxiety disorders are the two most disabling. ³

Depression is a looming global health issue, and in the last 27 years, there has been an astronomical increase of around 50% in its incidence. ⁴ Depression is characterized by a loss of interest in any activities; feeling sad, guilty, or tired; inadequate sleep; and poor appetite and concentration. ⁵ Though it is a mental disorder, depression has been positively associated with physical diseases, especially cardiovascular diseases. ⁶ Autonomic imbalance has been proposed as a possible underlying mechanism for the increased cardiovascular risk in depression. ⁷

Heart rate variability (HRV) is one of the most sensitive and non-invasive tools for assessing autonomic activity, and it is sensitive enough to detect early autonomic changes.^8,9 HRV is measured by two parameters: time domain and frequency domain. Time-domain analysis of HRV is based on the changes in heart rate in the time interval between two successive beats, that is, the normal-to-normal (NN) interval. ¹⁰ Frequency domain analysis decomposes the series of sequential NN intervals into a sum of sinusoidal functions of different amplitudes and frequencies. ¹⁰ Reduced HRV reflects autonomic imbalance and indicates impending cardiac disease. ¹¹

Autonomic changes in depression have already been reported. ¹² Some of the past literature has shown sympathetic overactivity in patients with clinical depression, while a few others have found parasympathetic attenuation.^13-20 However, some other studies found no autonomic changes in the clinically depressed as such.^21-23 Among the depressed, the reported influence of gender was found to be variable.^24-26 Moreover, the association between the severity of the disease and cardiac autonomic changes was also unclear.^27,28

Hence, it is evident that the studies available on the association between cardiac autonomic activity and clinical depression are variable and contradictory and lack a clear pathway. Also, the works available on this association are of Western origin and, therefore, could be more effective in providing adequate regional data/knowledge. Hence, we aimed to fill up the lacuna by assessing the cardiac autonomic activity in clinical depression by short-term HRV.

Materials and Methods

Ours is a cross-sectional, observational, and comparative study. This study was conducted in the outpatient Department of Psychiatry in a tertiary care hospital between April and September 2022, after the necessary approval from the Institutional Human Ethics Committee. We calculated the sample size based on the mean difference. ²⁹ We used data from a previous study to calculate the sample size. ¹⁷ Our total sample size was 82, of which 41 individuals with clinical depression constituted the study group and the remaining 41 acted as controls. We included newly diagnosed clinically depressed patients, both males and females, between the age groups of 18 and 50 in the study group. We recruited age- and sex-matched non-depressed individuals who cleared screening by the in-house psychiatrists for any psychiatric disorders as controls. We selected both groups of participants after carefully excluding hypertensives, patients with artificial pacemakers, diabetics, individuals with a substance abuse history, asthmatics, Parkinson’s disease, hypo- and hyperthyroid patients, cardiac patients, pregnant individuals, and patients with other neurological, endocrinological, and psychiatric issues. Before the commencement of the study, we fully explained the procedure and its implications to the participants and obtained informed written consent.

In-house psychiatrists made the diagnoses of clinical depression based on the guidelines of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5). ⁵ Our in-house psychiatrists scored and categorized the study participants using the Hamilton Depression Rating Scale. ³⁰ Then, we subjected both the study and control groups to short-term HRV.

We instructed the participants to avoid smoking and coffee for at least 2 hours and alcohol for at least 36 hours before HRV recording. We advised them to have adequate rest and at least 8 hours of good sleep the night before they underwent HRV recording. On the day of the assessment, a normal breakfast was recommended. During the recording, subjects were made to lie quietly for 5 minutes on a couch to alleviate any anxiety. The study room was a sound-attenuated, dimly lit one with a temperature of about 20–25°C. After explaining the procedure to the subject, we performed a short-term HRV analysis using an ambulatory ECG system (INCO digital NIVIQURE, Bangalore, India) in standard lead II configuration for five minutes. The HRV device is fitted with a multichannel digital data acquisition system that acquires and analyzes ECG data obtained at a sampling rate of 1,024 Hz. An interface RS232C-compatible module transfers data to the computer. The transferred data was analyzed using an in-built software system.

We collected both the time domain and frequency domain measures from the participants. The time domain measures collected here are indices that are the standard deviation of all NN intervals (SDNN), the square root of the mean of the sum of the squares of the differences between adjacent NN intervals (RMSSD), and the number of pairs of adjacent NN intervals differing by more than 50 milliseconds in the entire recording divided by the total number of NN intervals (pNN50). SDNN measures total variability, while RMSSD and pNN50 reflect high-frequency variations in heart rate. ¹⁰

The frequency domain measures considered here are high-frequency (HF) and low-frequency (LF) components that calculate the absolute or relative amount of signal energy within the component bands, with each frequency band denoting its significance. HF reflects cardio-vagal response, LF assesses sympathetic drive, and the LF/HF ratio measures sympathovagal balance. ¹⁰

We compared the collected HRV measures between depressed and non-depressed individuals and subgroups of depressed individuals and correlated HRV measures with the social science software (SPSS) version 24. Appropriate statistical methods such as t-tests for independent samples, Mann–Whitney U, Kruskal–Wallis, and Pearson correlation were employed for comparison and correlation analysis. Values were expressed as mean ± SD; p ≤ .05 was taken as statistically significant.

Results

1. On comparison of the HRV measures between depressed individuals and healthy controls, pNN50 (p = .026*), HF (p = .026*), and LF/HF ratio (p = .048*) were found to be significantly reduced (Table 1). When the HRV parameters between depressed and non-depressed males were compared, there were no statistically significant differences (Table 2). Similarly, females also showed no statistically significant differences between the depressed and non-depressed (Table 3).

2. Among the clinically depressed group, the HRV parameters between males and females on comparison showed no statistically significant differences (Table 4). Similarly, when depressed individuals were grouped according to their levels of depression (mild, moderate, and severe) and compared among themselves (Kruskal–Wallis test), their distribution across all three groups was the same and was found to be not statistically significant.

3. When the HRV parameters of the depressed individuals were correlated with Hamilton score after correcting for age and body mass index (BMI), there was no statistical significance (Table 5).

Table 1.

Comparison of the HRV Measures Between Depressed Individuals and Healthy Controls.

Variable	Group (N = 41 for Both Groups)	Mean ± Standard Deviation	Significance (p)
Age (years)	Control	43.02 ± 5.78	.483
	Cases	45.73 ± 6.21
BMI (kg/m2)	Control	25.02 ± 3.38	.862
	Cases	24.68 ± 3.49
RR	Control	652.63 ± 221.95	.861
	Cases	571.83 ± 212.69
SDNN (ms)	Control	124.91 ± 90.11	.083
	Cases	124.18 ± 69.01
PNN50 (%)	Control	7.44 ± 7.63	.026*
	Cases	6.58 ± 5.25
RMSSD (ms)	Control	93.58 ± 94.52	.183
	Cases	95.98 ± 70.30
LF (ms2)	Control	897.31 ± 777.44	.891
	Cases	961.09 ± 805.37
HF (ms2)	Control	6215.54 ± 5943.87	.026*
	Cases	5129.81 ± 4712.75
LF/HF	Control	0.29 ± 0.46	.048*
	Cases	0.24 ± 0.15

% = percentage; ms = milliseconds; kg/m² = kilograms/meter²; *p < .05.

Table 2.

Comparison of the HRV Parameters Between Depressed and Healthy Males.

Variable	Group	N	Mean ± Standard Deviation	Significance (p)
Age (years)	Control	23	43.78 ± 4.92	.077
	Cases	18	46.06 ± 6.01
BMI (kg/m2)	Control	23	25.31 ± 3.27	.205
	Cases	18	24.28 ± 3.83
RR	Control	23	651.57 ± 187.84	.062
	Cases	18	54.06 ± 222.88
SDNN (ms)	Control	23	106.85 ±78.70	.093
	Cases	18	167.48 ±77.18
PNN50 (%)	Control	23	5.46 ± 5.93	.098
	Cases	18	7.77 ± 4.99
RMSSD (ms)	Control	23	72.72 ± 76.71	.078
	Cases	18	106.14 ±71.87
LF (ms2)	Control	23	883.78 ± 830.28	.854
	Cases	18	828.35 ± 740.11
HF (ms2)	Control	23	5689.23 ± 6121.19	.875
	Cases	18	4658.60 ± 4972.40
LF/HF	Control	23	0.29 ± 0.35	.618
	Cases	18	0.24 ± 0.15

% = percentage; ms = milliseconds; kg/m² = kilograms/meter²; *p < .05.

Table 3.

Comparisons of the HRV Parameters Between Depressed and Healthy Females.

Variable	Group	N	Mean ± Standard Deviation	Significance (p)
Age (years)	Control	18	42.06 ± 6.75	.075
	Cases	23	45.48 ± 6.49
BMI (kg/m2)	Control	18	24.67 ± 3.58	.712
	Cases	23	25 ± 3.25
RR	Control	18	654 ± 265.03	.393
	Cases	23	585.74 ± 208.33
SDNN (ms)	Control	18	147.98 ± 100.42	.386
	Cases	23	113.77 ± 61.63
PNN50 (%)	Control	18	9.97 ± 8.91	.176
	Cases	23	5.65 ± 5.36
RMSSD (ms)	Control	18	120.24 ± 109.84	.401
	Cases	23	88.04 ± 69.61
LF (ms2)	Control	18	914.61 ± 727.63	.694
	Cases	23	1064.97 ± 854.61
HF (ms2)	Control	18	6888.04 ± 5812.53	.462
	Cases	23	5498.58 ± 4577.39
LF/HF	Control	18	0.30 ± 0.58	.115
	Cases	23	0.24 ± 0.16

% = percentage; ms = milliseconds; kg/m² = kilograms/meter²; *p < 0.05.

Table 4.

Comparison of the HRV Parameters Between Males and Females with Clinical Depression.

Variable	Group	N	Mean ± Standard Deviation	Significance (p)
Age (years)	Females	23	45.48 ± 6.49	.833
	Males	18	46.06 ± 6.01
BMI (kg/m2)	Females	23	25 ± 3.24	.303
	Males	18	25.28 ± 3.83
Hamilton Score	Females	23	21.74 ± 6.86	.33
	Males	18	23.61 ± 5.23
RR	Females	23	585.74 ± 208.33	.462
	Males	18	554.06 ± 222.88
SDNN (ms)	Females	23	113.77 ± 61.63	.331
	Males	18	137.48 ± 77.18
PNN50 (%)	Females	23	5.65 ± 5.36	.203
	Males	18	7.77 ± 4.99
RMSSD (ms)	Females	23	88.04 ± 69.61	.338
	Males	18	106.14 ± 71.87
LF (ms2)	Females	23	1064.97 ± 854.61	.306
	Males	18	828.35 ± 740.11
HF (ms2)	Females	23	5498.58 ± 4577.39	.636
	Males	18	4658.61 ± 4972.40
LF/HF	Females	23	0.24 ± 0.16	.895
	Males	18	0.24 ± 0.15

% = percentage; ms = milliseconds; kg/m² = kilograms/meter²; *p < .05.

Table 5.

Correlation Between HRV Parameters and Hamilton Scores in Depressed Individuals.

HRV Parameters	Hamilton Score
RR	r 0.121
	p .450
SDNN (ms)	r 0.239
	p .132
PNN50 (%)	r 0.169
	p .289
RMSSD (ms)	r 0.229
	p .159
LF (ms2)	r 0.058
	p .718
HF (ms2)	r 0.071
	p .661
LF/HF	r 0.028
	p .864

% = percentage; ms = milliseconds; *p < .05.

Discussion

Depression is becoming a major public health concern, ⁵ and it is known to have a significant association with cardiovascular diseases ⁶ due to its possible negative influence on cardiac autonomic activity. ¹² Since the kinds of literature available on the association between depression and cardiac autonomic activity are variable and inconclusive,^{13,15,17,18,22} this study was undertaken.

Comparison of the HRV Measures Between Depressed Individuals and Healthy Controls

Among the HRV measures, pNN50 and HF power were significantly reduced in clinically depressed individuals (p = .026, p = .026) (Table 1). Since both of these measures reflect high-frequency variations of heart rate, ¹⁰ our study shows reduced cardiovagal activity in depressed individuals. Our finding concurs with the finding of Rechlin et al., who also showed a reduction of HF power in their set of depressed individuals, but instead of pNN50, they found a significant reduction in RMSSD. ¹⁸ Our hypothesis of reduced cardiovagal activity in depression is further corroborated by Udupa et al., who found a significantly lesser Valsalva ratio among the depressed. ¹⁷ Schiweck et al. also agreed with the negative effect of depression on the vagal tone, but they theorized it was evident only during exposure to stress. ¹⁹ Furthermore, studies by Imaoka et al., who proposed an autonomic attenuation, ¹² and Robinson et al., who found a possible parasympathetic inhibition, ¹⁶ support our findings. Thayer et al. also concurred with this neural modulation in depression and suggested it can act as an index of affective and cognitive function. ²⁰

However, our finding is contradicted by Yeragani et al., who found no HRV changes due to depression, and they were limited by their short observation period. ²² Also, Licht et al. and Lehofer et al. contradict our hypothesis by postulating that the cardiovagal effects seen in depression might not be due to depression at all, but by the influence of tricyclic antidepressants.^21,23 Here, our study provides evidence that this may not hold since we included newly diagnosed individuals with depression who were not on any antidepressants.

We also found a marginally significant reduction in the LF/HF ratio in depressed individuals (p = .048), which can be a statistical effect influenced by extreme values. We also found no statistically significant differences between depressed and healthy males (Table 2). Similarly, depressed and healthy females also exhibited no differences (Table 3). These findings prompt us to propose a larger sample in the future to study gender influences.

To sum it up, our finding of altered HRV in depression points to reduced cardiovagal activity. Though this contradicts findings of past literature that implicate sympathetic predominance as a cause,^13-15 it can be partly explained by the polyvagal theory, which states that the vagus, which promotes social engagement and adjustments through its regulatory role in emotions, is affected by depression. ²⁸ This is further supported by Thayer et al., who proposed a circuit involving the prefrontal cortex that might have a tonic inhibitory effect on the heart rate. ²⁰ Won et al. provide us with the link between depression and an increase in inflammatory cytokines due to a reduction in acetylcholine mediated by reduced cardiovagal activity. ³¹

Gender Differences in HRV Parameters in Clinical Depression

Our study shows no HRV differences between depressed males and depressed females (Table 2). Our finding concurs with the findings of Kuang et al., who also found no significant differences between males and females only when they were at rest. ²⁶ Our findings are contradicted by Verkuil et al., who found a significantly higher cardiovagal activity in depressed women. ²⁴ Chen et al. also contradicted this in their study when they found reduced vagal activity in elderly males. ²⁵ Hence, a study with a larger sample size might reveal more information regarding the influence of sex on depressed individuals.

Levels of Depression and HRV

Based on our findings, there are no significant differences among the groups of depressed individuals based on their Hamilton Depression Rating (HAM–D) scores. Our study found no significant correlation between HRV parameters and Hamilton scores (Table 5) after correcting for age and BMI. Our study shows that the severity of clinical depression does not impact cardiac autonomic activity, contradicting Agelink et al. ²⁷ We feel a much larger sample consisting exclusively of patients with MDD will reveal more about this association.

Recommendation

Since our study points toward an autonomic impairment in clinically depressed individuals, this finding should be cautiously interpreted, and the potential cardiovascular risk may be seriously considered. Addressing this risk is left to the discretion of clinicians.

Limitations

Apart from not using a standardized structured screening tool in selecting controls, the main limitation of our study was the sample size of the depressed individuals. These numbers might not be enough to reveal the influence of sex and the severity of the disease on HRV changes associated with depression. Moreover, the effect of possible confounders such as age and BMI can be more effectively studied in a larger sample. Rather than a cross-sectional study, a cohort study involving a larger population might throw more light on these shortcomings.

Conclusion

Based on our study, we conclude that depressed individuals have an altered HRV and autonomic impairment due to a reduction in their cardiovagal activity. Hence, we propose that clinical depression might be directly involved in parasympathetic attenuation. There is no evidence in this study to indicate any gender-influenced differences among depressed individuals.