Study of Frontal Alpha Asymmetry in Mild Depression: A Potential Biomarker or Not?

INTRODUCTION

Depression is one of the most taxing mental illness regarding global burden and days affected life years.[1] Regarding remission, it is interesting to observe that we have some space left for “normalcy of illness”, and rating scales for depression do not have cutoff of zero for remission. The clinical endpoint of depression is often chosen by the patient by stating how he/she feels or back to functioning.[2] Even score on rating scale are subjective account of patient measured objectively. In other branches of medicine, biomarkers are invariably used to measure outcome of treatment or define remission. We have seen many such biomarkers especially in the field of cancer. These biomarkers are either reflection of normalcy or underlying pathological changes in person.[2] Use of biomarker brings more of objectivity in understanding and management. However, many response or remission indicated by biomarker does not coincide with clinical response or remission.[3] In mental health, this holds true due to heterogeneity of disorders, differential response to treatments and varied subjective interpretation for what remission or response means. Hence, search of biomarker in psychiatry has been difficult. There is ongoing research for the quest of biomarkers of depression.

Peripheral biomarkers hold some promise, but with limited utility due to lack of specificity and sensitivity.[45] Similarly, functional imaging and electroencephalography (EEG) also has shown variable results.[6] Considering all the research ongoing in the field of biomarkers of depression, EEG is one of the most cost-effective, noninvasive and largely available tool for clinicians.

There is evidence to suggest predictive the value of alpha and theta band as well as alpha asymmetry and theta asymmetry in depression.[78] There are certain measures calculated from quantitative assessment of EEG which are proposed as biomarkers.[89] Frontal lobe of brain especially midline region, is highlighted in the quest for physiological correlates of depression.[10] Lower alpha activity in the right prefrontal cortex and higher alpha activity in the left prefrontal cortex is assumed to be associated with depression where higher alpha activity reflects resting, nonactive state.[11]

For what sounds so promising in a theoretical sense may not always translate into actual findings. There is no clear evidence yet to put frontal alpha asymmetry (FAA) as biomarker of depression. Meta-analysis[12] suggests moderate level of correlation (r = 0.19) between depression score and FAA measure at F3-F4; however, it was convenient to include only F3-F4 and exclude other scalp site. There is further need of research and across all frontal electrode. We additionally also explore correlation of resting and activated alpha activity as well. With all this in mind, we devised current study and aimed to explore these lacunae.

METHODOLOGY

RESULTS

We evaluated 24 cases (15 females, 09 males) and 17 controls (11 females. 06 males). Mean age of the cases was 34.82 (±11.05) years and that of controls was 29.52 (±09.80) years. Frontal alpha power was measured across FP1, FP2, F3, F4, F7, and F8. Mean frontal alpha power was compared while resting (before), activated (during), and post activated resting stage (after). There was no statistically significant difference between the corresponding frontal electrode when compared before, during and after in both cases as well as controls [Table 1]. Further, we calculated alpha power asymmetry in cases between respective paired electrode (FP1–FP2, F3–F4, F7–F8). Alpha power calculated were compared on respective electrodes and difference of power was statistically significant across F7-F8 and not across FP1-FP2; F3-F4. In pair F7-F8, significance was observed on all three phases of recording (Before activation, during activation and after activation) [Table 2]. Comparison of difference of alpha power across pair of electrodes in controls did not reveal any statistical difference across any pair or any phase of recording [Table 3]. Subsequently, we compared respective electrode's mean alpha power difference between cases and control, revealing no statistical difference across any electrode except F7. Electrode F7 had statistically significant difference in all before activation, during activation, and after activation phase [Table 4].

DISCUSSION

We examine the relationship between frontal alpha power asymmetry and depression during a resting state and a musically valance task using Indian classical music with aim to assess the feasibility of FAA (resting and activated) as biomarker in depression.

We made three hypotheses for assessment of FAA as biomarker in depression

1. FAA (left <right) would be specific for patients of mild depression when compared with controls

2. Frontal alpha activity would differ in resting and activated state in patients with mild depression

3. Frontal alpha power difference would be significant on respective electrodes of cases and controls.

Many researchers studied FAA in perspective of neuroanatomical correlation in search to pinpoint anatomical basis of depression. Zotev et al.[16] compared electroencephalographic FAA with functional magnetic resonance imaging using blood oxygen level–dependent concurrently on same patient to understand neuroanatomical correlation. Study found spatial correlation in the left lateral orbitofrontal cortex, the left middle temporal gyrus, the left amygdala, the right medial frontotempolar cortex, the left insula, the right hippocampus, the right parahippocampal gyrus, the left pregenual anterior cingulate cortex, and the right amygdala.[16] Above findings displays widespread anatomical basis of depression. Clinically, depression in itself is heterogeneous entity, differing in its presentation patient to patient; both could explain inconsistency of FAA findings obtained in our study.

Our finding on FAA is inconsistent across electrode not only in case but also in control. Similar inconclusive results were also found in many previous studies, where the comparison between depressed and nondepressed individual was made by measuring frontal alpha activity at resting stage;[1718192021] in addition, inconsistent pattern of findings was also reported in frontal alpha activity when calculated from average, Cz, and lateral mastoid references.[22] A recent metanalysis[23] discusses the FAA research in depression to be very heterogeneous with small effect size. Although we used auricular reference only, we failed to obtain consistent finding with respect to FAA. In an interesting study by Sun et al.,[24] it is concluded that, the neuromodulation which impacts affective circuits, modulates FAA, which in turn modulates emotions and behavior. We investigated only cases with mild depression and our finding to certain extent suggest that FAA association with depression may be relevant to subsyndromal cases of depression.[25]

Statistically significant finding at F7–F8 level is partially in line with two meta-analyses[1012] where analysis of 26 and 19 (respectively) studies shown mild-to-moderate correlation (r = 0.26 and r = 0.19, respectively). However, same time both meta-analyses could not explain heterogeneity of FAA across various pair of electrodes. Most of the study chosen for meta-analysis had statistically significant difference across F3–F4 in contrast with our result. However, few studies also have shown significant difference in F7–F8 and not in other. In addition, Thibodeau et al.[10] point out that such inconsistent finding may be due to publication bias, journal or researcher favoring positive studies for publication. A recent EEG research in depression[26] finds inconclusive evidence, and further hypothesized of different reference electrodes and possible gender differences which held true at static recording in recent study.[27] The emotional challenge task assumes statistically significant consistency across reference modes, concurring hypothesis that emotional challenge produces frontal alpha activity asymmetry in patients of depression.[27] While our study gives evidence that musical valance does not produce asymmetry in patients of depression. Meaning FAA in activated state is not consistent across various stimuli but largely dependent on nature of stimulus, and in turn processing area of brain.

Our study found statistical difference in FAA across F7-F8 pair, but failed to find on PF1-PF2 and F3-F4 pairs, partially accepting first hypothesis. Testing second hypothesis, there was no statistical difference in any electrode (FP1, FP2, F3, F4, F7, and F8) in resting as well as stimulated stage could be found. This finding is replicated in cases as well as controls. Final hypothesis was also partially acceptable, as mean alpha activity on F7 electrode had statistically significant difference in cases and controls. Summarizing, on basis on our hypothesis, we could not be certain about specificity of FAA as diagnostic indicator of depression. Resent meta-analysis suggests limited value of FAA in diagnosis and as biomarker for depression.[23] Yet, consistent F7-F8 asymmetry in resting and activated state should not be prematurely invalidated.

CONCLUSION

Quest for biomarker for depression churned out FAA as frontrunner. Despite of vast amount of research on it, practical utility eludes us. We need to revisit our approach from conventional search of diagnostic biomarker; as FAA might reflect component of depression but not the syndrome which is termed depression. In our opinion we are not yet ready for it and have a road ahead to travel.