Frontal alpha asymmetry (FAA) is an electroencephalo-graphy (EEG) measure, based on alpha power over the frontal cortex, that quantifies the relative activation of the left hemisphere versus the right [1,2]. A high degree of alpha power over a particular hemisphere actually indicates low activation of that brain region [3]. FAA can be measured reliably [4], and it has been suggested that FAA appears to be a stable trait, that is not sensitive to time, state, or pharmacologic status [5], and its reliability as a marker of major depressive disorder (MDD) has been chal-lenged in some studies [6,7]. When measured across a 1 to 3 year interval in adults, FAA does show a moderate degree of stability, irrespective of history of depression and sex [8]. Consistent with trait-like stability, there is evidence for intergenerational association of FAA between caregivers and their infants [9]. Genetics may contribute to FAA, as an adult female twin study suggested that FAA, as measured at mid-frontal locations (F3 and F4), showed low but significant heritability, where genetic factors accounted for 27% of the observed variance [10].

In the last several decades, a variety of methods have been used to measure and calculate FAA. Perhaps most commonly, the alpha band has been defined as a window of 8−13 Hz, and electrodes F3 and F4 have been used to measure alpha power in the left and right hemispheres, respectively [11-14]. In that case, FAA can be calculated as the natural logarithm of the left hemisphere alpha power subtracted from the natural logarithm of the right (i.e., ln (F4) – ln (F3)) [11,12], although other variants of this formula have been used such as through normalization [1], or instead subtracting the right hemisphere electrode from the left [15,16]. Thus, a positive value for FAA (i.e., ln (right) – ln (left)) indicates greater alpha power over the right hemisphere and therefore greater relative left hemisphere activation, with a negative FAA meaning the opposite. Other frequency bands than 8−13 Hz have also been used for alpha, such as 8−12 Hz [17-19], and in many cases alpha is subdivided into high alpha and low alpha bands [20-23]. Likewise, a variety of frontal electrodes other than F3/4 have been used to calculate FAA, often F7/8, FP1/2, or other sites, and in some cases FAA is averaged across several sites per hemisphere [24-27].

A significant literature, published in the last several decades, has associated differences in frontal lateralization with MDD. It has been suggested that the low alpha band (8−10.5 Hz) in particular is most informative in understanding the potentially negative association between FAA and MDD [28]. Also, while using individualized alpha frequency peaks to compute resting FAA can be reliable and valid approaches, these methods however do not provide an advantage over a nomothetic fixed frequency window (e.g., 8−13 Hz), upon which many studies have relied [29]. Although traditional measures of FAA are often based in EEG measures across several minutes, a future direction comprises measuring transient dynamics, such as per-burst metrics, that unfold on a scale of milliseconds to seconds and can reveal novel information on current or past episodes of MDD [30].

A quantitatively higher (i.e., increased) FAA indicates relatively greater left than right frontal cortex activation, which is associated with enhanced reward-related approach behaviors rather than avoidance or withdrawal, including in the context of bipolar disorder [31-33]. By inference, the left frontal cortex may promote approach instead of behavioral withdrawal [19,31]. A range of studies in the last several decades have indeed suggested that having greater FAA supports enhanced positive affect and protection against MDD, whereas greater right frontal activation (i.e., reduced FAA) is associated with risk for, or currently having, MDD [34-36]. While the hypothesis that elevated right frontal activation is associated with MDD is widely known, a significant number of other studies have instead reported greater left frontal activation in MDD [37,38], or shown that either leftward or rightward bias in FAA can be associated with depression [39,40]. In this brief review, we summarize the literature on leftward or rightward orientation of FAA (i.e., whether FAA is increased or reduced, respectively) in MDD, and find much evidence that MDD is not always characterized by reduced FAA. We also review the limited literature on FAA and monoaminergic neurotransmitters systems (i.e., serotonin, norepinephrine, dopamine), including pharmacologic agents that act on them, where most of these studies have focused on serotonin. When describing individual studies below, FAA was computed using electrodes F3/4 with an alpha band of 8−13 Hz, unless otherwise noted.

As noted above, increased FAA has been associated with dispositional, trait-like greater approach of rewarding stimuli, as noted in a number of studies (e.g., [41]). For example, in young adults, rewarding stimuli have been associated with greater left frontal alpha power suppression (i.e., increased FAA) relative to punishing stimuli. The opposite activational pattern was observed in the right hemisphere, with similar effects in FAA for electrodes F3/4 and F7/8 [32]. Based on data from a virtual T maze task, greater left frontal brain activation occurred during approach behavior, greater right frontal activation during withdrawal behavior, and bilateral frontal activation occurred during active behavior compared to doing nothing [42].

In the context of bipolar disorder, where reward-related approach may be enhanced during states of hypomania or mania [33], several studies of FAA indicate greater left hemisphere activation. For example, individuals with bipolar disorder, compared to those without the disorder, showed greater left frontal activity (where this study focused on F7/8) when faced with a laboratory-based challenging or potentially rewarding event [33]. In a prospective study of individuals with cyclothymia or bipolar II disorder who were followed for an average of 4.7 years, baseline increased FAA predicted eventual conversion to a more severe phenotype, bipolar I disorder. The authors suggest that since unipolar MDD is often characterized by reduced FAA, hypomania or mania may have an opposing, increased FAA based on elevated approach motivation and reward hypersensitivity [31].

FAA may also not distinguish unipolar depression from bipolar depression, which could indicate pathophysiological similarities between the two depressive conditions as opposed to states of hypomania or mania, although EEG cordance and coherence values may help discriminate the two neuropsychiatric disorders. This study examined alpha in the 8−12 Hz range [43]. Another mood disorder study used a task involving presentation of happy and sad faces and quantified frontal alpha1 (8−10 Hz) asymmetry to distinguish healthy controls (who had increased FAA) from those with unipolar depression, but not from bipolar depression. They calculated FAA by averaging across three left electrodes (F3, F5, F7) versus three right electrodes (F4, F6, F8) [44]. But like Tas et al. [43], this measure did not distinguish unipolar from bipolar depression [44], reinforcing the hypothesis that unipolar and bipolar depression share similar neurophysiology.

A number of studies suggest that reduced FAA is associated with negative affect and MDD, whereas increased FAA supports positive affect and protection from MDD. Resting EEG measurement has shown reduced FAA in individuals with MDD, relative to those without MDD, as indexed with the Beck Depression Inventory [45]. This finding is consistent with the hypothesis that the left frontal region is specialized for positive affect, and the right for negative affect [45]. Another investigation found that female adults exhibiting stable, extremely increased FAA, showed increased generalized positive affect and decreased negative affect, compared to those with extremely reduced FAA [46]. A subsequent study by this group found that increased FAA in female adults may be linked with a self-enhancing regulatory style that could promote lowered risk for psychopathology [47]. It has also been reported in males that positive affective style is associated with increased FAA, as measured at lateral-frontal but not mid-frontal sites [48]. Finally, FAA is associated with choice of affective words, where individuals with greater left-sided resting frontal activity were more likely to select pleasant word pairs than neutral or unpleasant ones. This study focused on anterior frontal (FPF1/2) sites, but the effect was also marginally significant for F3/4 and F7/8 (p < 0.07) [49].

Greater right frontal activation has also been demonstrated in subjects with MDD, relative to healthy controls, or similarly that left-sided activation is reduced [34]. In a study of 163 females, their depression scores correlated negatively with FAA and positively with left frontal alpha power (i.e., reduced left-sided frontal activity; 8−12 Hz) [35]. In a study of 36 subjects, geriatric depression was associated with reduced FAA, although this difference was not statistically significant, and was better observed in the healthy elderly [50]. Individuals with MDD can not only exhibit reduced FAA, but also this is associated with lower physical activity levels, indicating the depressive symptom of psychomotor retardation. FAA was calculated in this study by averaging across FP2-FP1, F4-F3 and F8-F7, with alpha defined as 8−12.5 Hz [36].

Despite the well-known hypothesis that enhanced right frontal activation predicts negative affect and risk for MDD, a number of studies have found no clear relationship between FAA and depression. A study, for example, that compared 20 individuals with MDD and 16 healthy controls, found no significant difference in FAA between groups, while defining alpha as 8−12 Hz and measuring at electrodes F3/4, F7/8, and several other frontal locations [51]. In a larger investigation that compared 135 individuals with comorbid MDD and anxiety with 135 healthy controls, FAA (alpha 8−12 Hz) did not differ between groups, but there was an increase in beta activity within the brain, suggesting generalized neural hyperexcitability in MDD [52]. Although not explicitly about MDD, a study that used a low-cost wearable headset EEG system in 220 adults, found that robust linear regression models did not reveal an association between well-being and FAA, as measured at AF7/8 [53]. Meta-analyses have also, in some instances, yielded mixed results. A meta-analysis of 16 studies (1,883 subjects with MDD, 2,161 controls) found only a statistically non-significant, negligible effect size, suggesting a limited diagnostic value for FAA in MDD [1]. Multiverse analysis of five independent studies found a significant FAA in MDD in only 13 of 270 analyses, where this result is not above chance. Also, of these 13 significant results, 6 favored increased FAA rather than the traditional reduced FAA [54]. It has been suggested that reduced FAA that has been observed in many studies of MDD, is influenced more strongly by comorbid anxiety than by depression [55].

Several studies in children, young adults, or the elderly have also yielded ambiguous results on the potential relationship between reduced FAA and MDD. In an EEG study of 134 Mission Indian children (ages 7−13), no significant relationship was found between FAA (8−12 Hz) and depressed mood or approach behavior, suggesting that FAA may not have the same effects in all age or ethnic groups [56]. A study of 732 twins and their siblings found that FAA is heritable only in young adults (males 32%, females 37%) but not in middle-aged adults. Also, a significant relationship between FAA and depression was only found in young adult females [57]. The data are also mixed in studies of FAA in geriatric depression. A study of elderly subjects with depression (12 with MDD, 8 remitted, 7 controls) found no evidence for group differences in FAA, where FP1/2, F3/4, and F7/8 were considered separately and also collectively [58]. A somewhat larger study of late-life depression that compared 41 individuals with MDD and 44 controls, also showed no significant difference in FAA [59]. Lastly, whereas healthy young female subjects may tend to exhibit a robust FAA (8−16 Hz), the difference between controls and individuals with MDD diminishes and ultimately reverses with increasing age [60].

While some of the above studies suggest a lack of frontal lateralization in MDD, there is also a limited literature reporting a leftward bias (i.e., increased FAA) rather than the traditional rightward one. In a study of 117 subjects with MDD and 120 controls, those with non-melancholic depression surprisingly exhibited increased FAA compared with controls and those with melancholia, where the latter subtype did not differ from controls [37]. In a large dataset of 1,900 subjects suffering from various neuropsychiatric disorders, individuals with MDD exhibited a trend-level tendency toward greater FAA (calculated for FC3/4) instead of the more commonly observed reduced FAA in MDD [38]. Similarly, a systematic review and meta-analysis of 18 studies, 9 of which were homogeneous enough to include for statistical analysis, found a non-statistically significant slight tendency for increased FAA in MDD [2]. Relatively greater left hemisphere activation is also frequently observed in the EEG of individuals at increased risk for suicide, where this effect was found at a variety of parietal and occipital sites in the alpha2 sub-band (9−11 Hz) [61].

Other studies of FAA have found evidence for both rightward and leftward lateralization in their pool of subjects with MDD or tendencies toward depression. MDD can indeed be characterized by greater variance in FAA than in control subjects, rather than a systematic shift toward the left or right, measured here separately at AF3/4 and F3/4, and with these two pairs of sites combined [39]. In a study of 21 subjects, those with a recent history of MDD tended to have either a marked leftward or rightward shift in alpha activity (FP1/2) in the frontal region [40]. While a study of 67 subjects with MDD found generally reduced FAA in depression versus healthy controls, there was increased FAA in MDD accompanied by suicidal ideation versus MDD without this ideation. This study used an alpha band of 8−12 Hz with analysis focused on F7/8 [17]. In a study of young adults, females who previously had childhood depression had reduced FAA relative to controls, whereas men with previous childhood depression had increased FAA, measured at 7.5−12.5 Hz [62]. In a study of adolescents (12−17 years) with MDD, there was paradoxically greater FAA in depression, whereas in the healthy controls higher depression scores were associated with reduced FAA. Here, alpha comprised the band from 7.5−13.5 Hz, and was calculated collectively from 17 frontal electrodes in each hemisphere [63]. However, another study of adolescent (13−18 years) depression found reduced FAA in individuals with current MDD, compared with healthy controls. In this study, alpha was defined as the band from 7.5−13.0 Hz, and averaged across six electrodes per hemisphere [24]. In another study of MDD, suicidal behavior in individuals with reduced FAA was positively correlated with the overall alpha band (8−12 Hz) as well as low alpha (8−10 Hz) asymmetry indices, whereas this behavior in individuals with increased FAA was positively related with high alpha (10−12 Hz) asymmetry, measured at F7/8 [15].

Clinical use of repetitive transcranial magnetic stimulation (rTMS), where many of these studies have focused on pharmacologically treatment-resistant depression and in many cases stimulate left or bilateral dorsolateral prefrontal cortex, may also shed light on FAA and the neural circuitry underlying MDD [64,65]. One of these studies reinforces the potentially trait-like qualities of FAA, since even when treatment with rTMS resulted in symptomatic remission in all 8 subjects with MDD, FAA (alpha1 [8−11 Hz], alpha2 [11−13 Hz]; averaged across three frontal electrodes per hemisphere) did not change significantly [66]. A number of other studies have, however, found that rTMS does alter FAA, although not always in a manner that reinforces the hypothesis that reduced FAA underlies MDD. For example, Price et al. [67] found evidence that a decrease in depressive symptomatology after multiple sessions of rTMS may be weakly associated with a reduction in FAA rather than an increase. Other studies of rTMS have, however, found that an increase in FAA after repeated treatments is associated with amelioration of depression. One of these studies used alpha 8−11.75 Hz, measured at 5 frontal electrodes per hemisphere centered around F3/4; the other used 8−12 Hz averaged across three frontal electrodes per hemisphere [64,68]. Two other publications failed to find systematic leftward or rightward shifts in FAA after treatment of MDD with rTMS, that were different from control subjects or responders [69,70], perhaps consistent with both rightward and leftward bias in FAA within a given population of subjects with MDD. Thus, a very limited literature on these topics in rTMS is consistent with the hypothesis that either increased or reduced FAA can be associated with MDD.

Studies of the hypothalamic-pituitary-adrenal (HPA) axis stress hormone, cortisol, may also be relevant to understanding frontal lateralization. Acute treatment of healthy male subjects with cortisol resulted in reduced FAA and inhibition of approach motivation, where this effect was maximal at F7/8 [25]. It has been suggested that cortisol is functionally opposed to norepinephrine in vivo [71], so one possibility is that this effect of acute cortisol on FAA was in part mediated by inhibition of noradrenergic signaling. In undergraduates exposed to exam stress, EEG measures taken before the exam revealed that reduced FAA was associated with a higher cortisol level after awakening, where this effect was present at F7/8 [26]. During the Trier Social Stress Test, in another study, greater cortisol response was associated with reduced FAA, thus predicting a stronger physiological stress response [12]. In contrast, a higher cortisol awakening response has been associated with increased FAA, measured at FC1/2 and FC3/4, during the resting state [72].

Various serotonergic and dopaminergic genes may play an important role in lateralization of frontal acti-vation. In a study in which healthy subjects viewed an aversive film consisting of scenes of real injury and death, individuals homozygous for the short allele of the serotonin transporter-linked polymorphic region (5-HTTLPR) exhibited a clear, rightward shift of dorsolateral frontal activation, where alpha was defined as 8−12 Hz. This neural response, which was not strongly present in carriers of the long allele, may indicate a greater tendency for short/short individuals, who may have greater synaptic serotonin, to exhibit self-regulation related to avoidance moti-vation. Dopaminergic signaling, studied here with the catechol-O-methyltransferase (COMT) Val158Met polymor-phism, may in contrast be associated with self-regulation related to approach motivation [18]. In a study of the 5-HTTLPR in 70 children (ages 4−6), individuals homozygous for the short allele showed reduced FAA. In contrast, children homozygous for the long allele exhibited a consistently increased FAA, and heterozygotes had similar left and right frontal activation. In this study, alpha was analyzed in 7−11 Hz, and two pairs of clusters of electrodes were studied, one pair centered at F3/4 and the other at F1/2 [27]. The 5-HT1A receptor gene is another prominent mediator of serotonergic signaling. In college-age subjects, variation in the 5-HT1A gene modulated trait-like EEG asymmetry, regardless of history of MDD. Also, individuals who were homozygous for the 5-HT1A risk allele for depression exhibited significantly reduced FAA at a variety of electrode sites, including F1/2, F5/6, F7/8, but not F3/4 (where p < 0.14 for this latter electrode pair) [73].

Studies of tryptophan depletion, where this amino acid is a metabolic precursor in serotonin biosynthesis, may also provide information on frontal lateralization. In healthy volunteers with a positive family history of MDD, acute tryptophan depletion, which decreases systemic serotonin levels, increased self-ratings of depressed mood and reduced FAA, measured at both the low (8−10.5 Hz) and high alpha (10.5−13 Hz) bands [20]. In a related study by this group, which investigated individuals with remitted MDD who were smokers, acute tryptophan depletion increased self-ratings of depressed mood and also reduced FAA in the low alpha (8−10.5 Hz) band at F7/8. These effects were not evident in non-smokers [21]. Pedaling exercise, which can boost serotonergic (and probably noradrenergic) signaling, resulted in an increase in FAA in the alpha1 (7.5−10 Hz) band at FP1/2. Increases in urinary serotonin positively correlated with degree of FAA in this study [74]. These three studies collectively suggest, in contrast to some publications reviewed earlier, that serotonin is more involved in left rather than right frontal activation, although this relationship may be modulated by the presence of smoking and MDD. They also underscore limitations of considering FAA as trait-like in all circum-stances. Since acute exercise is associated with boosting of noradrenergic signaling [75,76], the findings of Ohmatsu et al. [74] may suggest that norepinephrine activates left frontal cortex, particularly if the boosting of norepineph-rine exceeds that of serotonin. But perhaps consistent with the acute tryptophan studies above, a pharmacologic investigation of female subjects with MDD who were also experiencing menopausal syndrome showed reduced FAA, and a single dose of the SSRI citalopram increased FAA in the alpha2 band) [77].

Using FAA and other EEG measures to predict response to a particular antidepressant drug in a given individual experiencing MDD, is a feasible approach to treatment that is of growing interest to the field [78]. Increased FAA (in low alpha 8−10 Hz, measured at F5/6) in MDD has paradoxically been associated with a better response to the antidepressants escitalopram and vortioxetine, where both drugs boost synaptic serotonin. Also in this study, reduced FAA (F7/8) was associated with higher melancholia scores in MDD [16]. In a cross-validation study, analysis of females with MDD revealed that greater FAA prior to antidepressant treatment, which focused on selective serotonin reuptake inhibitors (SSRIs) or serotonin-norepine-phrine reuptake inhibitors (SNRIs), was associated with a better clinical outcome eight weeks later [79]. In a study of 1,008 subjects with MDD and 336 healthy controls, no differences in FAA were found between the two groups as a whole. However, increased FAA in women was associated with favorable response to the SSRIs escitalopram and sertraline [13]. In 34 females with MDD, subdividing the alpha band into sub-bands (lower alpha1, lower alpha2, upper alpha; where these frequency bands were individualized for each subject) did not improve the sensitivity of FAA in predicting MDD, as no significant difference in this measure was observed between MDD and control subjects, as measured at F3/4, F7/8, or F(3,7)/(4,8). However, the seven individuals from this study who had been taking antidepressants (SSRI = 4, SNRI = 3) in a stable manner exhibited increased FAA based on data from the lower alpha1 band at electrodes F7/8 [80]. That study may indicate that greater baseline left frontal activation is associated with a therapeutic reaction to serotonergic antidepressants, with those individuals choosing to continue taking those drugs. Changes in FAA in the low alpha band, in response to treatment of mild or moderate depression with the antidepressants fluoxetine (SSRI) or maprotiline (noradrenergic boosting drug), have indeed been associated with a therapeutic response to these medications [81]. A study of MDD that examined responses to happy versus sad faces, before and after a week of antidepressant treatment, found a group × emotion × time interaction for the alpha1 (8−10 Hz) FAA, measured with mean of F3 and F5 versus F4 and F6. In this result, antidepressant responders showed a relatively greater response to happy than sad faces after a week of treatment, whereas this pattern was not altered in nonresponders [22].

While several studies in the previous paragraph support the hypothesis that greater baseline FAA predicts favorable response to serotonergic boosting antidepressants, there are conflicting data on this topic as well as studies suggesting baseline FAA does not predict antidepressant response. For example, in subjects suffering from MDD, SSRI treatment in individuals with a positive theta + alpha asymmetry (i.e., greater left hemisphere frontal activation) was associated with worsening suicidal ideation during the first 4 weeks of treatment [82]. In a study of treating MDD with the SSRI escitalopram, norepinephrine-dopamine reuptake inhibitor (NDRI) bupropion, or both drugs, treatment responders exhibited high frontal alpha2 (10.5−13 Hz) power at baseline, but no association with alpha2 asymmetry and response emerged [83]. In a study of 103 subjects with MDD that aimed to predict response to antidepressant treatment, there was no difference between responders and nonresponders in baseline FAA (either in alpha1 [8−10 Hz] or alpha2 [10−12 Hz]) or change in FAA at week 1, although there was a decrease in prefrontal activity at week 1. There were also no EEG differences based on SSRI versus SNRI treatment [23]. Thus, there are exceptions to the hypothesis that greater baseline FAA predicts favorable response to serotonergic antidepressants.

Pharmacologic studies of dopaminergic agents suggest that this neurotransmitter system, like serotonin and norepinephrine, may also modulate FAA. In individuals with schizophrenia, dopaminergic D2 receptor blocking antipsychotic drugs can, relative to healthy controls, normalize irregularities in FAA (7.5−12.5 Hz) in a state-dependent manner where a verbal task was compared to a spatial task. This effect, however, was strongest for electrodes located over central regions of the brain [84]. In a behavioral experiment in which male subjects approached female experimenters, increased FAA (8−12 Hz) was associated with behavioral approach of experimenters rated as most attractive. Administration of the dopamine D2 receptor blocker sulpiride reversed this association, and further it was shown that the COMT Val158Met polymor-phism, a genetic variant known to modulate prefrontal dopamine levels, was also associated with FAA [19]. In a task that manipulated positive emotion, extraversion was positively associated with emotional stability-flexibility and FAA favoring left hemisphere activation under placebo conditions, whereas the dopaminergic blocking drug sulpiride reversed these associations [14]. One possibility is that dopamine may facilitate trait-like left hemisphere frontal activation and associated approach behaviors, at least under some conditions, and blocking dopaminergic D2 receptors may counteract these processes.

The studies reviewed in this section and the previous sections reinforce the hypothesis that the widely studied alpha frequency band of 8−13 Hz is not always homogeneous with respect to FAA in affective disorders, as well as modulation of FAA by monoaminergic systems. In other words, sub-bands such as alpha1 and alpha2 often show differential effects on FAA in these studies [21,22,44,74]. Likewise, the above studies in some cases show variability in FAA results at electrodes F3/4 versus other electrode pairs or clusters [16,21,26,48]. Future studies of FAA may continue to address the specificity of the effects of various sub-bands of alpha, as well as diverse electrode sites, underscoring that no single method (including the frequency range when a single band is used for alpha) for measuring FAA is being applied in all ongoing studies.

This publication has focused on the relationship between FAA, MDD, and the monoaminergic systems. In recent years, and even dating back beyond that, serotonergic (and more generally, monoaminergic) hypotheses of mood disorders have been vigorously challenged in the scientific literature [85-87]. While theories of MDD and bipolar disorder based solely on monoamines are almost certainly an oversimplification, including due to the involvement of other underlying molecular pathways in these disorders [88,89], one possibility is that the monoaminergic systems interact with many of these pathways. Thus, it is possible that the monoamines may nonetheless be primary players in the neurophysiology of unipolar MDD and bipolar disorder. The widely demonstrated efficacy of monoaminergic antidepressants (MAOis, tricyclics, SSRIs, SNRIs, NDRIs) over many decades of use clinically, although not effective in all cases, reinforces the hypothesis that even if the monoamines do not form the etiological basis of most cases of MDD, their modulation by these drugs can indeed treat these disorders, in many instances. Even ketamine, which is thought to modulate glutamatergic signaling such as via N-methyl-D-aspartate (NMDA) receptor blockade, appears to acutely boost noradrenergic signaling, which could play a role in its therapeutic effects [90-92].

A number of the studies reviewed above suggest that rightward frontal lateralization (i.e., reduced FAA) does not characterize all cases of MDD. We suggest here a revised hypothesis: there may be two principal electrophysiological subtypes of MDD, one characterized by reduced FAA and high serotonin (and relatively low norepinephrine), the other by increased FAA and high norepinephrine (as well as relatively low serotonin). In this scenario, the high serotonin type may be a variant of melancholic depression, and the high norepinephrine type a variant of atypical depression [71]. An overarching principle of this hypothesis is that extreme activation of either the right or left frontal cortex promotes MDD and its symptomatology, consistent with a number of the studies reviewed above [39,40]. This hypothesis is also consistent with SSRIs or other serotonergic antidepressants being preferentially therapeutic in individuals with elevated baseline FAA [13,16,79], and also with serotonergic trans-mission promoting genotypes having reduced FAA [18,27]. The hypothesis is also consistent with high left frontal activation and its associated increased FAA promoting suicidality [17,61]. It may also be consistent with bipolar disorder and its potentially elevated noradrenergic signaling increasing FAA [31,33], where one possibility is that elevated noradrenergic signaling may characterize both depression and hypomania/mania in bipolar disorder, which is differentially coupled with dopamine in these opposing mood states [93,94]. Since cortical control of the sympathetic nervous system, where norepinephrine is a major output molecule, is biased toward the right hemisphere [95], this lateralization could represent a compensatory mechanism for norepinephrine preferentially activating the left hemisphere and possibly inactivating the right. While this simplified model of FAA and depressive subtypes put forth here may not characterize all cases or states of depression, it is a largely testable hypothesis that may have implications for precision medicine and targeted treatment of individuals with MDD via particular pharmacologic agents or combinations of agents.

No potential conflict of interest relevant to this article was reported.