A systematic narrative review of the form, process, and function of sensory hyposensitivity in depression Authors: Christopher F Sharpley, G Lorenzo Odierna, Vicki Bitsika, Christopher B Watson Published: 2026 ABSTRACT AbstractAs well as the key diagnostic symptoms that define Major Depressive Disorder (MDD), there is also evidence that some aspects of sensory sensitivity may decrease concurrently with MDD. However, although various individual studies have reported sensory hyposensitivity as a correlate of depression, most studies have focused upon single senses (i.e., vision, hearing, taste, touch, olfaction) and no comprehensive review of all five senses has been published to date. By applying a systematic narrative review methodology to identify research on loss of each these senses and their association with depression, a total of 51 individual studies and five reviews (of olfaction) were identified and examined for their methodology, major findings, and explanation of those findings. Following that summary of the form of sensory hyposensitivity, the neurobiological processes underlying it are described as being based in peripheral hyposensitivity or dysfunction of central cognitive processes. From these bases, an explanation is provided of how sensory hyposensitivity can be understood from the perspective of providing an ‘escape’ from a noxious environment, thus representing a functional benefit for the depressed individual. Translation of these findings to clinical practice, and avenues for future research, are described.Subject terms: Prognostic markers, Scientific community Introduction Symptoms of Major Depressive Disorder (MDD) MDD [1] is a significant cause of disability and contributor to the global burden of disease [2], and is identified by a list of nine Diagnostic Criteria plus numerous Diagnostic and Associated Features [1] (see Table 1). To qualify for the diagnosis of MDD, a person must present with either Diagnostic Criterion 1 or 2, and at least four of the remaining eight Diagnostic Criteria which together result in clinically significant distress or impairment in social, occupational, or other important areas of functioning [1]. These symptoms must not be attributable to a substance or another medical condition, bereavement, major financial loss, or a natural disaster, and are not better explained by a schizoform disorder [1]. Table 1Diagnostic criteria and Diagnostic Features for Major Depressive Disorder from DSM-5-TR.Diagnostic criteriaaDiagnostic Features1. Depressed mood most of the day, nearly every day (e.g., feels sad, empty, hopeless, appears tearful).Irritability (especially in children or adolescents), persistent anger, angry outbursts, blaming others, exaggerated frustration over minor events. Sad, hopeless, discouraged, “down in the dumps”, feeling “blah” or having no feelings, feeling anxious. Sadness may be replaced by somatic complaints (bodily aches and pains).2. Markedly diminished interest or pleasure in all, or almost all, activities most of the day, nearly every day.Less interested in hobbies, sports or previously-enjoyed activities. Social and/or sexual loss of interest or desire3. Significant weight loss when not dieting or weight gain or decrease or increase in appetite nearly every day.May feel as if they have to force themselves to eat. May crave sweets or other specific foods.4. Insomnia or hypersomnia nearly every day.Initial insomnia less common than middle insomnia (waking during the night) or terminal insomnia (unable to return to sleep). Hypersomnia may be oversleeping at night or during the day.5. Psychomotor agitation or retardation nearly every day observable by others.Agitation: Inability to sit still, pacing, handwringing, pulling or rubbing the skin, clothing or other objects. Retardation: slowed speech, slowed thinking, slow body movements; pauses before answering; soft speech, lack of inflection in speech; reduced speaking frequency and amount, reduced content variation.6. Fatigue or loss of energy nearly every day.Tiredness, often without physical exertion; small tasks (such as dressing) require major effort; efficiency of completing tasks is reduced.7. Feelings of worthlessness or excessive or inappropriate guilt nearly every day.Unrealistic guilt or feelings of worthlessness over minor failures; misinterpretation of neutral or trivial daily events as evidence of major personal defects; may take responsibility for untoward events; delusional sense of responsibility for unfortunate world events; self-blame for illness.8. Diminished ability to think or concentrate, or indecisiveness, nearly every day.Easily distracted; complain of memory difficulties sometimes mistaken for dementia; inability to complete previously-manageable complex intellectual tasks; major drop in grades if studying.9. Recurrent thoughts of death, recurrent suicidal ideation without a specific plan, or a suicide attempt or a specific plan for committing suicide.Thoughts that others would be better off if the depressed person were dead; suicidal thoughts and plans can vary in frequency, intensity and lethality; a desire to give up in the face of perceived insurmountable obstacles, or an intense wish to end a very painful emotional state.At least five of the Diagnostic Criteria must be present via self-report or observation most of the day during the last 2 weeks, representing a change from previous functioning, and including at least Diagnostic Criteria 1 or 2.aThe symptoms must cause clinically significant distress or impairment in social, occupational, or other areas of functioning, must not be due to the direct physiological effects of a substance or a general medical condition. Diagnostic criteria and Diagnostic Features for Major Depressive Disorder from DSM-5-TR. At least five of the Diagnostic Criteria must be present via self-report or observation most of the day during the last 2 weeks, representing a change from previous functioning, and including at least Diagnostic Criteria 1 or 2. aThe symptoms must cause clinically significant distress or impairment in social, occupational, or other areas of functioning, must not be due to the direct physiological effects of a substance or a general medical condition. As is evident from Table 1, these Diagnostic Criteria and Features are heterogeneous, leading to many hundreds of ways in which an individual can qualify for the diagnosis of MDD [3], creating major difficulties for accurate diagnosis. One pathway for improving the MDD-diagnosis process is by ceasing to rely solely on either patient self-reporting or clinical observation, but by complementing those sources of information by relevant biomarker data that is based upon the MDD patient’s psychophysiological state. In support of the development of psychophysiological state measures of MDD, it is noteworthy that, in addition to Diagnostic Criteria and Features, the over 280 million individuals globally who suffer from MDD also experience a wide range of negative effects on their physical health [4–6], functional capacity [7], and quality of life [8, 9]. Thus, MDD may be seen to permeate the psychophysiological states of its sufferers. Identification of salient aspects of those states offers at least a potential pathway towards biomarkers that are relatively independent of patients’ self-report, either on a standardized scale or in response to clinician interviewing. Biomarkers of MDD There is a broad history of the search for biomarkers of MDD, the most evident being an intensive but so far inconclusive focus on genes [10]. Although this research has established the heritability of MDD as about 30% to 50% [11], and a large number of candidate genes have been associated with MDD [12], the impact of each of these is unlikely to be more than small, detracting from a ‘depression gene’ biomarker of realistic value in clinical settings. Commencing with the measurement of catecholamines [13, 14], platelet monoamine oxidase activity [15], and hypothalamus-pituitary-adrenal activity [16], a similar (but inconclusive) research effort has been made to identify peripheral indicators of MDD [17]. Neuroimaging models were applied in the 1980s and initially focused on structural and volumetric indicators of mood disorders [18]. More recently, MRI and associated techniques have been assiduously applied to the measurement of changes in structure and size of specific regions of the brain, deriving from the stress response [19]. For example, the prefrontal cortex and hippocampus have been found to show reductions in volume due to cell apoptosis following chronic stress [20], plus some evidence of dendritic growth in the amygdala [21], but these findings have been confounded by also being found in other mental disorders such as schizophrenia, leaving their value as a biomarker for MDD in question. None of these lines of research has proven conclusive, perhaps due to the heterogeneous nature of MDD, wherein nearly 1500 possible combinations of the Diagnostic Criteria and Features can meet the criteria for MDD [3]. In an attempt to provide ancillary indicators (biomarkers) of MDD so that the specific nature of an individual’s MDD can be more clearly defined, some attention has been given to how changes in the physiological state of the MDD patient contribute to the presentation of the clinical symptomatology of MDD [22]. One aspect of those physiological changes is apparent in the decrease of sensory sensitivity, or sensory hyposensitivity. Sensory hyposensitivity in MDD Sensory hyposensitivity (defined here as a comparatively reduced sensitivity to sensory stimulation in depressed versus non-depressed individuals) has been observed in MDD patients for some time [23]. There are data regarding the primary senses that may be relevant to the clinical diagnosis of MDD, with some studies focusing on vision, others on hearing, olfaction, touch, and taste separately. However, although each of these studies is valuable, they do not provide a summative account of this field. Two overviews by Canbeyli a few years ago [24, 25] of the research supporting a sensory hyposensitivity-MDD connection provided an overview of this association, but there has been no review to date that explicated and tabulated studies from all five senses to allow evaluation of consistency of outcome across the senses, then examined the neurobiological and physiological processes that are involved, and drew these factors together into a cogent explanation of their occurrence. Thus, the purpose of this review was to (a) describe the sensory hyposensitivity-MDD association in each of the separate fields of vision, hearing, olfaction, taste, and touch, (b) trace the psychophysiological pathways that might underly the association between sensory hyposensitivity and MDD, and (c) explain why those changes to sensory sensitivity occur and how they are congruent with the underlying behavioural model of MDD. At this exploratory stage of development, we hypothesized that there may be consistency in findings across sensory domains regarding the effect of MDD on hyposensitivity. Although the primary focus of research data included in this review is in its translation to the clinical setting, the sensory fields examined here are characterized by studies that rely upon varying sources of information, with some using patient/participant self-reports of sensory hyposensitivity and other studies in which data are collected from EEG and fMRI, which are sources that participants are relatively unaware of in experimental settings. The differentiation of sensory hyposensitivity into sub-categories of self-reported symptoms, neural activity induced by sensory stimulation, or the behavioural ability to detect and discriminate between sensory stimuli has been applied inconsistently in the research reviewed here, producing several sets of data that are not always directly comparable. Importantly, these different forms of sensory hyposensitivity do not correlate strongly [26], although some degree of common sensory hyposensitivity as measured on a standardized scale has been found in depressed versus non-depressed adults [27], suggesting that sensory hyposensitivity of some kind is often a comorbidity of depression, and thus providing a rationale for this review. Because of the differences in data sources of sensory hyposensitivity across the sense fields, each sense (and its specific data sources) is examined separately in the review below. Most specifically, although sensory hyposensitivity may be defined on a number of axes (see above), the focus of this review is upon the general reduction in sensory acuity that may accompany depression. Therefore, the term chosen here to describe this is ‘sensory hyposensitivity’ (defined above), whether it refers to self-reports of reduced sensory sensitivity, peripheral (i.e., sense organ) changes, or phenomena relating to central processing in the brain. Despite each of these having different descriptors at a fundamental scale, the combination of them under the heading of sensory hyposensitivity enables a summary of the data that can inform clinical practice. Method Search strategy A hybrid systematic-narrative methodology was selected to ensure a comprehensive and objective account of the state of play regarding research on sensory hyposensitivity in depression. This approach allows a narrative synthesis of a very broad and complex body of literature while maintaining a transparent article selection process [28]. The search protocols and inclusion/exclusion criteria were drawn from systematic review practices, and a narrative approach was used to analyze the shortlisted articles due to the diverse methodological approaches and varying specificity of research questions across studies. The Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines were followed. A systematic search was performed in December 2025 in the electronic databases Google Scholar, PubMed and SciLit using the search terms “depression”, “sensory sensitivity” and “loss” and “decrease” and “reduction” and “hyposensitivity” plus “vision”, “auditory”, “olfaction”, “taste”, “touch”. A secondary manual search was undertaken in the reference lists of identified articles, previous reviews and meta-analyses. Prior to the application of inclusion/exclusion criteria, two authors (CFS, VB) screened the study titles and abstracts (n = 8900) and conferred for 100% agreement on alignment with the research topic. Figure 1 displays the process for the identification and selection of studies (PRISMA flowchart). Studies were included if they (i) presented data on at least one of the five senses, (ii) made a formal comparison between depressed vs non-depressed participants or reported relevant data on depressed participants only, (iii) used a valid method of identifying and classifying ‘depression’ (e.g., clinician interview, standardized self-report scale), (iv) were reported in a peer-reviewed scientific journal in English, (v) did not have any other disorder as their major focus, and (vi) were not reports solely of interventions using sensory stimulation. PRISMA Flowchart of Search and Selection Process for Studies Examining Sensory Hyposensitivity in Depression. Articles were excluded if they were not written in English, were not original peer reviewed journal articles, did not apply a valid procedure for identifying ‘depression’, did not include sensory hyposensitivity measured by a recognized methodology, and were focused upon patients with other major disorders (e.g., schizophrenia), or the use of sensory stimulation as a treatment for depression. For olfaction, major review papers were examined as well as recent original papers due to the finding of five such reviews being published during the last decade. A synthesis of important study characteristics was undertaken, including location, sample characteristics, MDD diagnostic measure, sensory assessment task, and study findings. Effect sizes were calculated for studies where sufficient data were available. Studies were also carefully examined to determine if a meta-analysis of quantitative data was feasible. Meta-analysis was not attempted for studies of the vision system, auditory system, taste and touch due to the methodological heterogeneity between studies when diagnosing depression and assessing the relevant sensory tasks and outcomes of those tasks [29]. Furthermore, when studies were grouped according to experimental outcome, these sense categories did not meet the threshold for the minimum number of studies required for a valid meta-analytic approach [29]. However, meta-analysis was possible for eight of the 12 identified primary olfactory studies and so correlation coefficients were obtained either directly from the studies or were calculated from the available data. A random-effects meta-analysis model was selected to account for between-study variance. The I2 statistic was estimated to evaluate study heterogeneity. The analysis was performed using IBM Statistical Package for the Social Sciences (SPSS) version 29. Results The search protocol identified 51 papers that reported original research, plus five reviews on olfaction. Research on each of the five senses is described separately below and in tabular format (Table 2). Some studies combined two senses, particularly olfaction and taste. A recent report of a 2013–2014 survey of the olfactory, taste, vision, and auditory sensitivity of 3328 participants in the National Health and Nutrition Examination Survey was the only study found that included four of the five senses. Results from that study indicated that depression scores on the Patient Health Questionnaire-8 were significantly associated with sensory hyposensitivity in depressed versus non-depressed participants, with an increasing risk of depression in participants who reported multiple sensory hyposensitivities [30]. Table 2Studies of sensory hyposensitivity in depression.(a) VisionAuthor, YearPlaceParticipants,1 ageControl Ss2MDD diagnosis3,4,5Visual task6Results7Effect SizeSummaryVasile et al. [166]Boston, USA9, 65–85 yr29, age, sex matchedMelancholia, non-melancholiaVisual Evoked Response (VER)VER abnormality restricted to posterior visual regions on SPM mapV216: p ≤ 0.00 lV368: p ≤ 0.00 lV464: p ≤ 0.00 lV216: Large (d = 2.87, 95% CI [1.87, 3.84]).V368: Large (d = 1.83 95% CI [0.97, 2.68).V464: Large (d = 1.49, 95% CI [0.67, 2.32])Elderly melancholic patients exhibit augmented VER abnormalityJoost et al. [167]Frieburg, Germany20 (11 M) Healthy Ss, 23–67 yrSelf-report Mood state “tiredness”Visual Evoked Potential (VEP)Mood modulated VEP amplitudeInsufficient data to calculate ESMood-influenced changes in lateral geniculate nucleusVasile et al. [168]Boston, USA56,M = 75.5 yr56,M = 65.2 yrMelancholia, non-melancholiaVERVER delay in melancholic, non-melancholic depressed Ss (p < 0.05)Insufficient data to calculate ESVER delay exists in all depressed pts, not only melancholic ptsSwanick et al. [169]Dublin, Ireland15 (7 M),M = 74.2 yr21 (7 M),M = 70.1 yrSelf-report HAM-DFVEPMDD pts. sig. delayed FVEPP2 (p = 0.004)N2 (p =0.08)Insufficient data to calculate ESFVEP abnormalities in depressionFriberg et al. [31]Pittsburgh, USA120 (46 M),M = 41.8 yrNilSelf-report on DSM-IV MDDSelf-report “lights are dimmer”Sig. b/w MDD severity and “dimmer” lights (p = 0.002)Large (r = 0.87)Catecholamine dysfunction in visual pathwayFotiou et al. [37]Thessonaliki, Greece50 (15 M),M = 41.0 yr.16pts with melancholia20 (7 M),M = 36.7 yrClinician,HAM-DPattern reversal VEPNo differences b/w MDD and Control.Melancholic pts. sig. lower transmission velocity than atypical depression pts(p = 0.0001)Large (d = 1.44, 95% CI [0.81, 2.07])Arousal dysregulation in melancholia.Normann et al. [38]Frieburg, Germany40 inpatients74Clinician-Severe Depressive Episode (ICD-10)Reversal VEP checkerboardSig. downregulation of synaptic transmission and cortical activity in depressed Ss compared with ControlsC1 (F = 5.03, p =0.02)N1 (F = 8.26, p = 0.005)C1 = Medium(d = 0.44, 95% CI [0.05, 0.83])N1 = Medium(d = 0.56, 95% CI [0.17, 0.96])Reduced synaptic plasticity in MDD due to long term stress but could be modulated by chronic SSRI application.Friberg et al. [32]Pittsburgh, USA213 (99 M)M = 74.8 yr192 (98 M),M = 74.8 yrSelf-report CESD > 16Self-report “lights are dimmer”MDD pts. sig. more likely to report “dimmer”(χ2 = 15.6, p < 0.0001).Medium (d = 0.40, 95% CI [0.20, 0.59])Serotonergic differences in central nervous systemBubl et al. [39]Frieburg, Germany28 (13 M),M = 37.8 yr21 (age, sex matched)Clinician,Self-report BDI5Contrast discrimination (VEP)MDD pts. had 15% longer discrimination threshold than controls (p = 0.0001)Large (d = 1.44, 95% CI [0.81, 2.07])Disturbed dopaminergic transmission in visual systemBubl et al. [40]Frieburg, Germany40 (sex, age not stated), inpatients40, age and sex matchedClinician,Self-report BDIContrast discrimination (VEP)Sig. r with BDI severity (p < 0.001); no medication effect.Large (r = −0.65)NilBubl et al. [41]Frieburg, Germany14 drawn from reference [19] MDD patients40 from reference [19]Clinician,Self-report BDIContrast discrimination as in [19]Reduced retinal contrast in MDD pts. (p < 0.0001) normalised after successful SSRI.Medium (rs = −0.40)Dopaminergic amacrine cells in retina are central in the contrast detection process, and administration of SSRI may have increased dopamine elsewhere in the brain.Fam et al. [170]Singapore.20 (3 M),44.5 yr20 (3 M),43.7 yrClinician, self-report BDIContrast discrimination (VEP)Sig. r with BDI severity(p < 0.001)Large (r = 0.49)Altered higher visual processing in MDDBubl et al. [42]Frieburg, Germany40, from [19].28 from [19]Clinician, Self-report BDIContrast discrimination (VEP)Reduced VEP, PERG discrimination in MDD patients3% contrast: Medium (rs = −0.44)7%: Medium (rs = −0.37)16%: Medium (rs = −0.32)36%: Small (rs = −0.26)80%: Small (rs = −0.21)Dopaminergic system a likely causal factorSalmela et al. [117]Helsinki, Finland46 (19 M)M = 31.8 yr29 (8 M)M = 32.1 yrClinician,Self-report BDIContrast discrimination (VEP)Reduced contrast suppression in MDD vs controls t35.2 = 2.902, p = 0.006Large (d = 0.68, 95% CI [0.21, 1.16])Due to MDD-linked changes in visual cortex processing.Song et al. [118]Hangzhou, China70 (12 M)M = 23.4 yr52 (18 M)M = 24.5 yrClinician,HAM-DVisual motion suppression (VEP)Impaired visual motion perception in MDD (p = 0.03)Medium (d = 0.39, 95% CI [0.03, 0.76])Reduced GABA concentration in occipital cortex.Du et al. [115]Chongqing, China23 (7 M)M = 30.7 yr27 (9 M)M = 29.7 yrClinician,HAM-DContrast discrimination (VEP)Reduced retinal contrast in MDD pts vs Controls.Sig. r with HAM-D severity(p = 0.0416)Medium (r = 0.43)Increased cortical excitability in MDD pts, due to imbalance between excitatory and inhibitory systems in visual cortexFriedel et al. [171]Frieburg, Germany30 (9 M)> 18 yr42 (13 M)19–65 yrClinician, Self-report BDIContrast discrimination (VEP)Sig. r b/w reduced contrast sensitivity and BDI in MDD pts. and Controls(p = 0.002)Medium (rs = −0.35)Altered dopaminergic transmission in retinal cells during MDD(b) Auditory functionAuthor, YearPlaceParticipants,1 ageControl Ss2MDD diagnosis 3, 4,5,6Auditory taskResultsEffect sizeSummaryMalone et al. [172]London, UK10 (sex not stated),Age not stated.NilClinician,HAM-D, BDIAuditory signal detection before Tt and in remissionAuditory sensitivity improved after Tt (p < 0.0002)Insufficient data to calculate ESMay be due to increased selective attention to stimuli after MDD Tt.Michael et al. [46]Munster, Germany11 (sex not stated),M = 50.6 yr11 (sex not stated),M = 34.8 yrClinician,MADRSAuditory signalAuditory sensitivity reduced in MDD pts.(p = 0.028)Insufficient data to calculate ESCortical dysfunction in MDDTollkotter et al. [173]Munster, Germany20 (11 M),M = 47.7 yr27 (12 M),M = 32.8 yr)Clinician, MADRSVowel ‘a’, plus sine toneResponse to stimulus delayed in MDD pts. vs Controls(p = 0.019)Large (rs = −0.72)Cortical dysfunction in MDDUekermann et al. [49]Bochum, Germany29 (14 M)M = 37.0 yr29 (13 M)M = 39.0 yrClinician,BDIAffective prosodySig. more errors in naming prose by MDD vs Controls(p < 0.0001)Large (d = 1.25, 95% CI [0.69, 1.82])Executive function dysfunction in MDDAubert-Khalfa et al [47]Marseille, France13 (10 M),M = 38.9 yr30 (10 M),M = 34.9 yrClinician,BDI25 pure tone stimuliAuditory thresholds higher in MDD vs Controls (p < 0.0001)Insufficient data to calculate ESCortical dysfunction in MDDNaranjo et al. [50]Belgium23 (5 M),M = 40.8 yr23 (5 M),M = 39.8 yrClinician,BDIMusical and vocal emotion stimuliMDD pts. showed impaired emotional processing of musical (p < 0.001), vocal stimuli (p < 0.001)Music: Large (d = 1.25, 95% CI [0.62, 1.88])Vocal: Large (d = 2.14, 95% CI [1.48, 2.95])Attentional difficulties in MDDSchwenzer et al. [51]Aachen, Germany12 (6 M),M = 45 yr12 (6 M),M = 46 yrClinicianPitch identification from memorized stimulusFemale MDD pts sig. less accurate in identifying pitch (p = 0.047)Sig. r b/w inaccurate pitch identification and BDI in women only(p < 0.001).Large (d = 1.01, 95% CI [0.16, 1.85])Large (r = 0.80)MDD associated with neurocognitive dysfunction in female MDD, possible serotonergic dysfunctionSchlipf et al. [53]Tubingen, Germany23 (7 M),M = 43.4 yr22 (8 M),M = 44.1 yrClinician,BDIHAM-DPositive, negative, neural adjectivesMDD rated positive words less intensely than Controls(p < 0.01).Sig. r with BDI(p < 0.05)Large (d = 0.89, 95% CI [0.28, 1.51])Medium (r = 0.43).Reduced attention to positive stimuli a function of MDD anhedoniaDoose-Grunefelf et al. [45]Dusseldorf, Germany41 (22 M),M = 36.5 yr41 (22 M),M = 37.6 yrClinician,BDI,HAM-D10 laughs, 10 yawns, 10 screams paired with visual stimuliNo sig. diffs b/w MDD vs Controls for sounds(p = 0.63).No sig. r b/w BDI & sound(p = 0.590)Small (d = 0.11, 95% CI [−0.33, 0.54])Small (r = 0.08)Any diffs b/w MDD and Controls not due to cognitive effects of MDDMu et al. [174]Dalian,China20 (5 M),M = 40.3 yr20 (5 M),M = 40.9 yrClinician,HAM-DMusical tonesMDD sig. poorer at detecting change in musical timbre(p < 0.01)Large (d = 0.84, 95% CI [0.19, 1.49])Poor pre-attention processes in MDDMehboob et al. [48]Karachi, Pakistan120 (60 M),M = 42.8 yrNilTinnitus pts,DASS for ‘depression’Pure tone audiometryNo sig. difference b/w treated vs untreated MDD pts for hearing loss (p = 0.066)Small (d = 0.31, 95% CI [0.68, 0.08])Depression may be caused by hearing lossEstalayo-Gutierrez et al. [175]Madrid, Spain14 (25% M),M = 55 yr86 (25% M),M = 45 yrHAM-DPure tone audiometryMDD sig. lower hearing threshold(p < 0.05)Insufficient data to calculate ESDepression and hearing loss associatedde la Salle et al. [52]Ottawa, Canada18 (5 M),M = 49.7 yr18 (5 M),M = 48.2 yrClinician,MADRS,BDI,DASPaired stimuli to test ‘gating’ responseMDD pts showed sig. reduced gating vs Controls(p 0.026).Sig. r b/w DAS and reduced gating (p < 0.01)Medium (d = 0.79, 95% CI [0.11, 1.47]).Large (r = 0.58)(c1) Reviews of loss in olfactory function in depressionAuthor, YearJournalNumber of studies reviewedResultsSummaryKohli et al. [176]Chemical Senses13Depressed pts had sig. lower olfactory threshold, discrimination, identification than Controls.Stress-induced cytokine activity limits hippocampal neurogenesis and central and peripheral olfactory neuronsCroy et al. [94]J. Neurology35Depression is “often” accompanied by “at least partial” negative alterations in olfactory function.Olfaction is impaired by reduced attention (in MDD pts)Taalman et al. [177]Frontiers in Psychiatry15Nine studies supported sig. lower olfactory function in depressed vs non-depressed, but 6 found no sig. differences. Most common difference was in olfactory thresholdOlfaction central processing is highly integrated with emotion and memory via connectivity between olfactory bulb and amygdala/hippocampus (involved in depression)Rochet et al. [178]Brain Sciences26• Reduced olfactory sensitivity may be a marker of depression• Olfactory identification not altered in depression• Hedonic value of odours did not change in depression• Odour discrimination/intensity assessment not sig. different in depressionConfirms the presence of olfactory impairments in depressed ptsKim et al. [179]Ear, Nose & Throat Journal8Meta-analysis method used.• Olfactory function sig. impaired in depressed pts• Olfactory discrimination not sig. different in depressed pts• Olfactory threshold not sig. different in depressed ptsDepression is accompanied by changes in prefrontal limbic network, which is connected to olfactory regions(c2) Primary articles of olfaction and depressionAuthor, YearPlaceParticipants,1 ageControl Ss2MDD diagnosis3Smell taskResultsEffect sizeSummaryPause et al. [180]Dusseldorf, Germany24 (9 m)M = 48 yr24 (9 M)M = 44 yrClinician,BDITwo graded odorous solutionsSig. inverse r b/w BDI and sensitivity to two stimuli (p =0.013)Small (r = −0.22)Medium (r = −0.34)Dysfunction in olfactory bulbLombion-Pouthier et al. [181]Besancon, France48 (14 M)M = 43 yr58 (12 M)M = 38 yrClinician,BDIOlfactif (olfactory sensitivity)(2 = high sensitivity; 10 = low sensitivity)Sig. lower sensitivity in MDD pts (M = 4.2, SD = 1.38) vs Control (M = 3.66, SD = 1.36)(p = 0.035)Medium(d = 0.39, 95% CI[0.01, 0.78])Dysfunction in orbito-frontal cortexPollatos et al. [182]Munich, Germany48 (14 M)M = 28 yrBDISniffin Sticks(higher score = lower threshold)Sig. inverse b/w BDI score and olfactory sensitivity (p < 0.05)Medium (rs = −0.36)Effect present even in healthy Ss, due to abnormal activity in amygdalaScinska, et al. [183]Warsaw, Poland25 (7 M)M = 67 yr60 (22 M)M = 67 yrGeriatric Depression ScaleSniffin SticksNo sig. correlation b/w GDS score and sensitivity (p > 0.05)Small (r = 0.02)Any possible differences not present in older adultsSwiecicki et al. [184]Warsaw, Poland46 (16 M)M = 37 yr30 (9 M)M = 35 yrClinician,HAM-DSniffin SticksNo sig. sensitivity difference b/w depressed (M = 7.2, SE = 0.3) vs non-depressed (M = 8.1, SE = 0.4)(p > 0.05)Medium (d = −0.43, 95% CI [−0.89, 0.04])Not statedNegoias et al. [33]Dresden, Germany21 (6 M)50 yr21 (6 M)M = 39 yrClinician,CIDISniffin SticksMDD pts. Sig. lower sensitivity (M = 7.56, SD = 2.67) than controls (M = 9.14, SD = 1.89)(p = 0.03)Medium (d = −0.68, 95% CI [−1.31, −0.06])Sig. decreased neurogenesis in olfactory bulb in MDD?Naudin et al. [185]Tours, France18 (6 M)M = 50 yr54 (18 M)M = 49 yrClinician,MADRSSoluble odorantsDepressed (M = 3.1, SD = 2.4) vs non-depressed (M = 2.1, SD = 1.9)(Χ2 = 9.6, p = 0.008)Medium (d = 0.49, 95% CI [−0.05, 1.03])Not statedCroy et al. [186]Dresden, Germany27 (0 M)M = 38 yr28 (0 M)M = 35 yrClinician,BDI, HAM-DSniffin Sticks,P2 latencyNo sig. sensitivity difference b/w MDD pts. (M = 6.8, SD = 2.8) and Controls (M = 7.2, SD = 2.2) for threshold.Small (d = −0.16, 95% CI [−0.68, 0.37])Lower odor sensitivity in MDDRossi et al. [61]Montaþeses, Argentina29 (7 M)M = 65 yr30 (14 M)M = 63 yrBDI, HAM-D,ClinicianSniffin SticksMDD pts. sig. lower sensitivity (M = 4.6, SD = 0.5) than controls (M = 5.6, SD = 0.4), p < 0.05Large (d = −2.22, 95% CI [−2.86, −1.56])Not statedKamath et al. [62]Maryland, USA134 (37 M)M = 50.3 yr72 (42 M)M = 50.5 yrClinicianThreshold Smell TestMDD pts. sig. lower threshold sensitivity (M = −5.4, SD = 1.18) than controls (M = −5.28, SD = 1.44), p = 0.017Small (d = −0.09,95% CI [−0.38, 0.19])Not statedSanna et al. [63]Monserrato, Italy273 (96 M)M = 41 yrBDISniffin SticksNo sig. r b/w sensitivity and depression (p > 0.05)BDI: Small (r = 0.02)Olfactory sensitivity associated with cognitive and emotional central pathways connecting the orbitofrontal cortex, piriform cortex, hippocampal and entorhinal cortex, and the amygdalaHerrman et al. [95]Dresden, Germany21 (9 M)M = 39.3 yr21 (9 M) M = 38.8 yrClinician,BDISniffin SticksMDD pts. lower sensitivity (M = 9.0, SD = 3.2) than controls (M = 10.4, SD = 2.5), but not significant (p = 0.13)Medium (d = −0.48, 95% CI [−1.10, 0.13])No sig. difference b/w dep vs non-dep for basic processing of odor cues(d) Taste perception in depressionAuthor, YearPlaceParticipants,1 ageControl Ss2MDD diagnosis 3, 4,5Taste taskResultsEffect sizeSummarySteiner et al. [65]Israel21 MDD pts18Clinician,HAM-DPerception of sweet, salty, sour, bitterDepressed pts. have reduced sensitivity to all four tastes; sig. - r b/w taste sensitivity and HAM-D scoreInsufficient data to calculate ESNot statedAmsterdam et al. [66]Philadelphia, Pennsylvania36 (17 M),M = 47.0 yr19 (9 M),M = 37.5 yrClinician, HAM-DPerception of seven concentrations of sucroseDepressed pts. rated sucrose as less intense than Controls (p < 0.02)Large (d = 0.71, 95% CI [0.14, 1.28])Carbohydrate intolerance in depressed ptsSteiner et al. [64]Israel21 (8 M),M = 52.8 yr16 (9 M),M = 43.3 yrClinician,BDIPerception of sucrose, citric acid, quinineNo sig. differences b/w depressed vs non-depressed(p > 0.05)Insufficient data to calculate ESInfluence of anhedonia on subcortical processesScinska et al. [187]Warsaw, Poland33 (20 M) community Ss,M = 63.0 yrBDITested for sensitivity to sweet, bitter, sour, saltNo sig. r b/w BDI and taste ratings(p = 0.06)BDI: Medium(r = −0.39)Effects of depression on taste not present in non-clinical samplesDichter et al. [188]Chapel Hill, USA12 (6 M),M = 39 yr)15 (6 M),M = 30.8 yrClinician,HAM-DSweet Taste Test (5 concentrations of sucrose)No sig. r b/w HAM-D and taste ratings (p = 0.06)HAM-D: Medium(r = −0.33)Anhedonia in MDD does not extend to subjective responses to sweet tasteHur et al. [189]National survey, USA5275 survey participants,40+yrPHQ-9Self-report of “taste alteration” during last 12 monthsSs who reported reductions to sense of taste also more likely to be MDD (p = 0.0001)Small: OR = 1.77,95% CI [1.78, 3.49]Loss of taste sensitivity associated with MDD(e) Somatosensory touch perception in depressionAuthor, YearPlaceParticipants,1 ageControl Ss2MDD diagnosis 3, 4Somatosensory task5ResultsEffect sizeSummaryDietl et al. [36]Munich, Germany15 (7 M),M = 45.4 yr15 (7M0,M = 44.9 yrClinician,BDI,HAM-D1296 x 0.2 ms duration electro stimuli to left index finger, 9 EEG sitesMDD pts. showed larger EEG response to electro stimuli than ControlsMedium (d = 0.59, 95% CI [−0.13, 1.33])Dysfunction of the anterior cingulum and the frontal cortex in MDD ptsKivrak et al. [78]Kars, Turkey140 community participants59 M, M = 40.5 yrBDIVenipuncture: pain to be self-rated on VASNo sig. r b/w BDI and VAS (p = 0.875)BDI-VAS: Small (r = 0.02)NilKurita et al. [80]Gunma, Japan10 (6 M),M = 39.9 yr10 (6 M),M = 39.3 yrClinician,HAM-D0.1 ms electro stimulus to right medial nerve at wristMDD pts shows higher threshold for stimulus than Controls (p < 0.001)Large (d = 2.18, 95% CI [1.08, 3.29]Dysfunction of somatosensory processing in MDD1M male, 2No history of MDD or neurological disorder, 3Hamilton Depression Scale, 4Center for Epidemiological Studies Depression Scale, 5BDI Beck Depression Inventory, 6Flash Evoked Visual Potential, 7Pattern electroretinogram.1M male, 2No history of MDD or neurological disorder, 3BDI Beck Depression Inventory, 4HAM-D Hamilton Depression Scale, 5MADRS Montgomery-Asperg Depression Rating Scale, 6 DASS Depression, Anxiety, Stress Scale, DAS Dysfunctional Attitudes Scale.1M male, 2No history of MDD or neurological disorder, 3BDI Beck Depression Inventory, CIDI Composite International Diagnostic Interview, MADRS Montgomery-Asberg Depression Rating Scale, 4OI Odor Identification.1M male, 2No history of MDD or neurological disorder, 3BDI Beck Depression Inventory, 4HAM-D Hamilton Depression Scale, 5PHQ-9 Patient Health Questionnaire-9.1M male, 2No history of MDD or neurological disorder, 3BDI Beck Depression Inventory, 4HAM-D Hamilton Depression Scale, 5VAS Visual Analogue Scale. Studies of sensory hyposensitivity in depression. VER abnormality restricted to posterior visual regions on SPM map V216: p ≤ 0.00 l V368: p ≤ 0.00 l V464: p ≤ 0.00 l V216: Large (d = 2.87, 95% CI [1.87, 3.84]). V368: Large (d = 1.83 95% CI [0.97, 2.68). V464: Large (d = 1.49, 95% CI [0.67, 2.32]) 56, M = 75.5 yr 56, M = 65.2 yr 15 (7 M), M = 74.2 yr 21 (7 M), M = 70.1 yr MDD pts. sig. delayed FVEP P2 (p = 0.004) N2 (p =0.08) 120 (46 M), M = 41.8 yr 50 (15 M), M = 41.0 yr. 16pts with melancholia 20 (7 M), M = 36.7 yr Clinician, HAM-D No differences b/w MDD and Control. Melancholic pts. sig. lower transmission velocity than atypical depression pts (p = 0.0001) Sig. downregulation of synaptic transmission and cortical activity in depressed Ss compared with Controls C1 (F = 5.03, p =0.02) N1 (F = 8.26, p = 0.005) C1 = Medium (d = 0.44, 95% CI [0.05, 0.83]) N1 = Medium (d = 0.56, 95% CI [0.17, 0.96]) 213 (99 M) M = 74.8 yr 192 (98 M), M = 74.8 yr MDD pts. sig. more likely to report “dimmer” (χ2 = 15.6, p < 0.0001). 28 (13 M), M = 37.8 yr Clinician, Self-report BDI5 Clinician, Self-report BDI Clinician, Self-report BDI 20 (3 M), 44.5 yr 20 (3 M), 43.7 yr Sig. r with BDI severity (p < 0.001) 3% contrast: Medium (rs = −0.44) 7%: Medium (rs = −0.37) 16%: Medium (rs = −0.32) 36%: Small (rs = −0.26) 80%: Small (rs = −0.21) 46 (19 M) M = 31.8 yr 29 (8 M) M = 32.1 yr Clinician, Self-report BDI 70 (12 M) M = 23.4 yr 52 (18 M) M = 24.5 yr Clinician, HAM-D 23 (7 M) M = 30.7 yr 27 (9 M) M = 29.7 yr Clinician, HAM-D Reduced retinal contrast in MDD pts vs Controls. Sig. r with HAM-D severity (p = 0.0416) 30 (9 M) > 18 yr 42 (13 M) 19–65 yr Sig. r b/w reduced contrast sensitivity and BDI in MDD pts. and Controls (p = 0.002) 10 (sex not stated), Age not stated. Clinician, HAM-D, BDI 11 (sex not stated), M = 50.6 yr 11 (sex not stated), M = 34.8 yr Clinician, MADRS Auditory sensitivity reduced in MDD pts. (p = 0.028) 20 (11 M), M = 47.7 yr 27 (12 M), M = 32.8 yr) Response to stimulus delayed in MDD pts. vs Controls (p = 0.019) 29 (14 M) M = 37.0 yr 29 (13 M) M = 39.0 yr Clinician, BDI Sig. more errors in naming prose by MDD vs Controls (p < 0.0001) 13 (10 M), M = 38.9 yr 30 (10 M), M = 34.9 yr Clinician, BDI 23 (5 M), M = 40.8 yr 23 (5 M), M = 39.8 yr Clinician, BDI Music: Large (d = 1.25, 95% CI [0.62, 1.88]) Vocal: Large (d = 2.14, 95% CI [1.48, 2.95]) 12 (6 M), M = 45 yr 12 (6 M), M = 46 yr Female MDD pts sig. less accurate in identifying pitch (p = 0.047) Sig. r b/w inaccurate pitch identification and BDI in women only (p < 0.001). Large (d = 1.01, 95% CI [0.16, 1.85]) Large (r = 0.80) 23 (7 M), M = 43.4 yr 22 (8 M), M = 44.1 yr Clinician, BDI HAM-D MDD rated positive words less intensely than Controls (p < 0.01). Sig. r with BDI (p < 0.05) Large (d = 0.89, 95% CI [0.28, 1.51]) Medium (r = 0.43). 41 (22 M), M = 36.5 yr 41 (22 M), M = 37.6 yr Clinician, BDI, HAM-D No sig. diffs b/w MDD vs Controls for sounds (p = 0.63). No sig. r b/w BDI & sound (p = 0.590) Small (d = 0.11, 95% CI [−0.33, 0.54]) Small (r = 0.08) Dalian, China 20 (5 M), M = 40.3 yr 20 (5 M), M = 40.9 yr Clinician, HAM-D MDD sig. poorer at detecting change in musical timbre (p < 0.01) 120 (60 M), M = 42.8 yr Tinnitus pts, DASS for ‘depression’ 14 (25% M), M = 55 yr 86 (25% M), M = 45 yr MDD sig. lower hearing threshold (p < 0.05) 18 (5 M), M = 49.7 yr 18 (5 M), M = 48.2 yr Clinician, MADRS, BDI, DAS MDD pts showed sig. reduced gating vs Controls (p 0.026). Sig. r b/w DAS and reduced gating (p < 0.01) Medium (d = 0.79, 95% CI [0.11, 1.47]). Large (r = 0.58) • Reduced olfactory sensitivity may be a marker of depression • Olfactory identification not altered in depression • Hedonic value of odours did not change in depression • Odour discrimination/intensity assessment not sig. different in depression Meta-analysis method used. • Olfactory function sig. impaired in depressed pts • Olfactory discrimination not sig. different in depressed pts • Olfactory threshold not sig. different in depressed pts 24 (9 m) M = 48 yr 24 (9 M) M = 44 yr Clinician, BDI Small (r = −0.22) Medium (r = −0.34) 48 (14 M) M = 43 yr 58 (12 M) M = 38 yr Clinician, BDI Olfactif (olfactory sensitivity) (2 = high sensitivity; 10 = low sensitivity) Sig. lower sensitivity in MDD pts (M = 4.2, SD = 1.38) vs Control (M = 3.66, SD = 1.36) (p = 0.035) Medium (d = 0.39, 95% CI [0.01, 0.78]) 48 (14 M) M = 28 yr Sniffin Sticks (higher score = lower threshold) 25 (7 M) M = 67 yr 60 (22 M) M = 67 yr 46 (16 M) M = 37 yr 30 (9 M) M = 35 yr Clinician, HAM-D No sig. sensitivity difference b/w depressed (M = 7.2, SE = 0.3) vs non-depressed (M = 8.1, SE = 0.4) (p > 0.05) 21 (6 M) 50 yr 21 (6 M) M = 39 yr Clinician, CIDI MDD pts. Sig. lower sensitivity (M = 7.56, SD = 2.67) than controls (M = 9.14, SD = 1.89) (p = 0.03) 18 (6 M) M = 50 yr 54 (18 M) M = 49 yr Clinician, MADRS Depressed (M = 3.1, SD = 2.4) vs non-depressed (M = 2.1, SD = 1.9) (Χ2 = 9.6, p = 0.008) 27 (0 M) M = 38 yr 28 (0 M) M = 35 yr Clinician, BDI, HAM-D Sniffin Sticks, P2 latency 29 (7 M) M = 65 yr 30 (14 M) M = 63 yr BDI, HAM-D, Clinician 134 (37 M) M = 50.3 yr 72 (42 M) M = 50.5 yr Small (d = −0.09, 95% CI [−0.38, 0.19]) 273 (96 M) M = 41 yr 21 (9 M) M = 39.3 yr Clinician, BDI Clinician, HAM-D 36 (17 M), M = 47.0 yr 19 (9 M), M = 37.5 yr 21 (8 M), M = 52.8 yr 16 (9 M), M = 43.3 yr Clinician, BDI No sig. differences b/w depressed vs non-depressed (p > 0.05) 33 (20 M) community Ss, M = 63.0 yr No sig. r b/w BDI and taste ratings (p = 0.06) BDI: Medium (r = −0.39) 12 (6 M), M = 39 yr) 15 (6 M), M = 30.8 yr Clinician, HAM-D HAM-D: Medium (r = −0.33) 5275 survey participants, 40+yr Small: OR = 1.77, 95% CI [1.78, 3.49] 15 (7 M), M = 45.4 yr 15 (7M0, M = 44.9 yr Clinician, BDI, HAM-D 140 community participants 59 M, M = 40.5 yr 10 (6 M), M = 39.9 yr 10 (6 M), M = 39.3 yr Clinician, HAM-D 1M male, 2No history of MDD or neurological disorder, 3Hamilton Depression Scale, 4Center for Epidemiological Studies Depression Scale, 5BDI Beck Depression Inventory, 6Flash Evoked Visual Potential, 7Pattern electroretinogram. 1M male, 2No history of MDD or neurological disorder, 3BDI Beck Depression Inventory, 4HAM-D Hamilton Depression Scale, 5MADRS Montgomery-Asperg Depression Rating Scale, 6 DASS Depression, Anxiety, Stress Scale, DAS Dysfunctional Attitudes Scale. 1M male, 2No history of MDD or neurological disorder, 3BDI Beck Depression Inventory, CIDI Composite International Diagnostic Interview, MADRS Montgomery-Asberg Depression Rating Scale, 4OI Odor Identification. 1M male, 2No history of MDD or neurological disorder, 3BDI Beck Depression Inventory, 4HAM-D Hamilton Depression Scale, 5PHQ-9 Patient Health Questionnaire-9. 1M male, 2No history of MDD or neurological disorder, 3BDI Beck Depression Inventory, 4HAM-D Hamilton Depression Scale, 5VAS Visual Analogue Scale. Of note is the choice of dependent variables used in these studies. Some are self-report (e.g., under Table 2 (a), Friberg et al. [31, 32], in which research participants were asked to indicate if “the lights in my surroundings seem dimmer than usual” (p. 114) which reflects participants’ self-perception). By contrast, all the other studies in Table 2 (a) relied on the Visual Evoked Potential or the Visual Evoked Response, which are measures of neurological function based upon EEG data collected from the occipital region and which reflect dysfunction between the retina and that region. Thus, these are different kinds of data regarding visual sensory hyposensitivity and may only be compared with difficulty, although we have reported effect sizes where possible. Similar differences exist in the variables used in studies of depression and auditory hyposensitivity, where auditory hyposensitivity was measured via neural data in some studies but via participants’ self-reports of auditory pitch identification in other studies. Many studies of taste perception, olfaction, and somatosensory touch perception were similarly based on participants’ self-reports, although olfaction studies sometimes included measurement of the olfactory bulb via MRI [33, 34] because the size of the olfactory bulb has been found to directly correlate with olfactory sensitivity [35], and one study measured EEG responses to electrostimulation of the left index finger [36]. Reduced sensory sensitivity: form Visual system All but two studies regarding vision in MDD included healthy control participants, matched by age and sex, and the age ranges (years) of the samples (MDD and control) ranged from mid-20’s to mid-70’s. Diagnosis of MDD was performed by trained clinician in 11 of the 17 studies, with self-reports from valid and reliable instruments in the remaining studies. A range of places and samples is evident, so the possible confound of sociogeographical limitation on results is unlikely. Apart from two studies that used depressed patients’ self-reports of the dimming of lights around them [31, 32], most research on vision applied the Visual Evoked Potential/Response, in which a set of squares of opposite colors (black/white) are reversed e.g., [37, 38], or a test of ability to detect contrast changes in visual stimuli e.g., [39, 40]. Patients diagnosed with MDD were significantly more likely to report or exhibit vision hyposensitivity in some form across most studies: either that lights were dimmer, or that their threshold for detecting contrast was higher than non-MDD control participants. The heterogeneity in measured vision stimuli and target outcome variables prevented application of meta-analytic procedures; however, 16 studies reported 21 statistically significant estimates of effect, of which nine had a large effect size, 10 were medium, and two were small. All effect estimates were in the same direction (i.e., MDD participants had vision hyposensitivity compared to control participants). Two papers deserve further comment. Bubl et al. [41] found that depressed patients whose retinal contrast sensitivity was lower than healthy controls also showed increases in visual sensitivity after 3 months of antidepressive medication. Of note, the reestablishment of normal visual sensitivity in formerly depressed patients was not found in those patients whose depression was not alleviated by the medication. The consistency of this finding across 17 studies, with three different kinds of visual stimuli, as well as the reversibility of visual impairment upon administration of antidepressants, argues that there is a reliable association between decrements in vision and depression. Secondly, the same laboratory reported evidence taken from EEG recordings that depression-linked visual hyposensitivity was present at the retina and also the visual cortex [42]. Because the visual cortex is heavily innervated by the dopamine system, those authors argued for the relevance of the dopamine system’s reward function in vision and depression, particularly anhedonia (one of the two required symptoms of MDD: Table 1). The comparative relevance of various neurotransmitter systems is discussed below. Auditory system Emotional state can influence perception of sound [43], even within 20 ms of delivery of speech sounds [44], providing a prima facie argument for the association between depression and auditory hyposensitivity. Ten of the 13 studies in Table 2, section b) were performed in Europe, with two from Asia and one from Canada, so generalizability to the global population is somewhat limited. Eleven studies included sex-matched healthy controls, but age of MDD and control participants was not always congruent. Eleven studies diagnosed MDD via clinician interviews, and most studies complemented that with a valid and reliable self-report instrument. All but one [45] of these 13 studies reported one or more statistically significant finding that depressed patients had poorer auditory discrimination than healthy control participants. This outcome was consistent across different auditory stimuli, including an auditory tone stimulus e.g., [46–48], naming of prose [49], emotional processing of music [50], identification of pitch [51], and inhibition of an irrelevant stimulus [52]. Twelve studies reported 14 statistically significant effect estimates of which nine had a large effect size, two medium, and three small. Explanations of these findings ranged from simple statements that hearing loss may initiate depression [48], loss of attention to positive stimuli because of dopamine dysfunction in MDD [53], to the effects of depression-induced rumination [52]. Olfactory system The olfactory system has a long history of research in the context of MDD, perhaps due to historical observations by J. B. Watson published in 1907 which claimed that olfactory bulbectomized rats displayed behaviors reminiscent of those exhibited by depressed individuals, such as increased irritability [54], p. 51. A more recent review also found that anhedonia and hyperactivity were observed in rats that had undergone olfactory bulbectomy [55, 56]. Though the interpretation of bulbectomy as a model for MDD remains controversial, these findings (alongside many others) contributed significantly to establishing a connection between the olfactory system and MDD. Because of the relatively large number of reviews published on the olfaction-depression association, the first part of section Table 2c (c1: olfaction) differs from the previous sections of Table 2 by presenting the findings of five major reviews published during the last decade as a means of understanding the current state of knowledge regarding olfaction and depression. The reviews described in Table 2, c1, evaluated 41 individual studies, with some overlap between reviews, and generally concluded that elevated levels of depression were associated with olfactory hyposensitivity, although there was some difference noted in the type of olfactory function measured, i.e., threshold, identification, or discrimination, and unipolar MDD was not separated from bipolar MDD, nor some other forms of depression (e.g., Seasonal Affective Disorder), thus challenging the overall conclusions from those reviews in terms of the MDD-olfactory sensitivity focus of this review. Due to the lack of consensus regarding olfactory identification and discrimination reported in these reviews, only those individual studies that examined at least olfactory threshold (i.e., ability to detect an odor) are presented in Table 2, section c2, on the basis that reduced sensitivity to odors is a parallel to reduced visual acuity. If, as is hypothesized in this review, sensory hyposensitivity is the ‘target’ variable relating to depressed status, then olfactory threshold remains the salient outcome variable, although olfactory identification and discrimination may also have been included by those researchers. The studies summarized under Table 2, section c2 present data on olfactory hyposensitivity reported by depressed patients defined as suffering from MDD compared to non-MDD controls. Table 2, c2 summarizes data from studies published during the last 25 years in order to provide a perspective on the most recent research in this field. Olfactory threshold (or sensitivity) was the most common variable measured in the 12 individual studies, with depressed participants usually demonstrating a reduced threshold compared to non-depressed control participants. By contrast, olfactory discrimination and identification were less likely to show this effect in depressed participants, and there was a degree of ambiguity in the results across studies, perhaps due to methodological differences in how these olfactory variables (i.e., threshold, identification, discrimination) were measured, participants selected (those with poorer olfaction may have not volunteered for research), and the underlying confound of whether participants experienced sensory hyposensitivity as a result of being depressed or had become depressed by experiencing olfactory hyposensitivity [57]. Seven of the 12 olfactory studies reviewed reported a statistically significant lower sensitivity in individuals with MDD, with a small to medium average effect size. Eight of the 12 studies had sufficient methodological consistency and available data to be included in the random-effects meta-analysis (see Fig. 2). This meta-analysis demonstrated an overall effect size of r = −0.27, which was statistically significant (p = 0.02). However, there was high heterogeneity (I2 = 0.81) in the results, indicating that differences between MDD and control groups were not just due to random chance, but likely stemmed from factors such as varying study designs, populations, or interventions. Meta-Analysis of the Association Between MDD and Olfactory Threshold Sensitivity. Taste Studies in the taste section of Table 2 were representative of various places and cultures and ranged in age from 39 to 52 yr for MDD patients, and similarly for control participants. Four of the six studies compared MDD patients with healthy controls, with MDD being diagnosed by clinician interview (4 studies) or self-report using a validated scale (5 studies). Taste is a difficult variable to assess for its association with depression in the same way as has been done for vision, hearing, and olfaction, because it is confounded with olfactory cues from the food before and while it is in the mouth, and there is a bidirectional relationship between food and mood (i.e., eating can influence mood, just as mood may influence eating). Additionally, some antidepressant medications can distort taste perception [58], although about 40% of depressed patients also report having an unpleasant taste in their mouths regardless of prior treatment [59]. Contrary to these findings, when a depressive mood was induced by watching unpleasant video clips, there was some evidence of a change in taste perception [60]. Overall, four studies [61–64] reported statistically significant estimates of effect; however, only two [62, 64] provided sufficient data to calculate an effect size, which were large and small respectively. Two studies [65, 66] reported effect estimates nearing significance (p = 0.06) with a corresponding medium effect size. These findings suggest the presence of sensory hyposensitivity in MDD patients compared to non-depressed control participants. It is important to frame any research studies about the association between taste perception and depression in the light of eating habit change, and gain/loss of weight that are recognized as a Diagnostic Criterion for MDD. Clearly, there is an interaction between taste and depression, but its direction is unclear at the moment, although the effect of depression-induced anhedonia has been suggested as a possible pathway. Although several studies have reported on the effect of depression on preferences for fat, sweet foods, carbohydrates and chocolate [67–69], there are not as many studies of the association between loss in taste sensitivity and depression as there are for vision or hearing hyposensitivity. Of note, a recent meta-analysis of 17 studies of the association between taste and depression in patients with schizophrenia concluded that, due to inconsistency in methodology across studies, no conclusions could be drawn at that time [70], perhaps reflecting the finding shown in Table 2 that varying methodologies have produced varying results. Taste may also be indirectly influenced by the texture of food, as in avoidant/restrictive food intake disorder (ARFID) where food may be rejected on the basis of aversions to specific tastes, textures, or smells [1]. The overlap between taste and somatosensory hyposensitivity, and their contribution to depression, has been documented in comparative suicide rates between ARFID and non-ARFID participants [71], perhaps mediated by a disgust of certain food textures [72]. Touch There are many studies of the benefits of affective touch upon depression, with several recent major meta-analyses confirming the results of individual studies e.g., [73–75]. Affective touch activates C-tactile unmyelinated fibres, a particular group of mechanoreceptors in hairy skin that have a low threshold response to a slow and gentle touch at an optimum speed of between 10 and 100 mm/sec. These fibres conduct signals at a slower speed than myelinated Aβ fibres (which process discriminative touch), are rated as consistently more pleasant than strokes at faster or slower speeds, and are associated with the release of oxytocin and dopamine [76]. However, although the pleasure associated with affective touch is diminished in depressed patients [77], the central theme of this review is focused upon sensory experiences that are more appropriately inclusive of non-affective touch as an index of sensory hyposensitivity in depression. Referred to as the ‘somatosensory’ system, this form of touch includes pressure upon the skin, pain inflicted on the skin, vibration and temperature, body movement and proprioception, which are registered via mechanical, chemical and thermal pathways of the skin. Table 2 (somatosensory section) presents the details of the very few studies that were identified as investigations of the association between this form of touch and depression, from three nations, and comparing MDD vs control participants in two studies. Diagnosis of MDD was via clinical interview and self-report in two studies, and via self-report alone in one study. Although the study by Kivrak et al. [78] that relied upon participants’ self-reports of pain found no significant differences between depressed and non-depressed participants’ sensory hyposensitivity, Dietl et al. [36] found that their sample of depressed patients exhibited larger EEG-based response to electrostimulation of the left finger than non-depressed participants, leading those authors to refer to the possible role of dysfunction of the anterior cingulum in reducing the rate of habituation to a stimulus, previously observed in depressed patients [79]. Kurita et al. [80] reported a delayed magnetoencephalography response (suggestive of a dysfunction in central somatosensory processing) to nerve stimulation at the wrist compared to non-depressed participants, implying basic level consistency with the data report by Dietl et al. [36] in hypothesizing hypoactivation as the underlying central information-processing dysfunction of depressed patients, a finding supported by two statistically significant estimates of effect across two studies [36, 79] with effect sizes of medium and large respectively. It should be noted that most of the studies that examined the somatosensory system and depression did so as an adjunct to the study of Post-Traumatic Stress Disorder (PTSD) and the somatosensory system e.g., [81–83]. Alternately, some studies used self-reports of sensory sensitivity to gauge somatosensory changes in depression rather than direct stimulation of the skin [84, 85], but it was the latter procedure that was examined here. Perhaps reflective of a link between somatosensory responses and depression, somatosensory stimulation has been found to reduce depression when delivered as a multi-week treatment [86]. Summary of review findings Based on the statistically significant effect estimates and corresponding effect sizes described in Table 2 and mentioned in the separate sections for each sense, it is possible to summarise the overall findings from all senses as indicative of consistent evidence of reduced sensory sensitivity across 57 effect sizes reported in 46 individual studies. Although there were more studies reported for some senses than others, there were no effect sizes that contradicted the overall finding of reduced sensory sensitivity during depression. It is of interest to conjecture as to the processes involved in this widespread association, first by a discussion of the central (brain) versus peripheral (sense organs) origins, and second by examining how sensory hyposensitivity can be understood from an adaptive behaviour standpoint. Sensory hyposensitivity: process Although some studies reported in Table 2 have referred to a general loss in cognitive function associated with MDD, so that sensory inputs are processed less effectively centrally, each sense has a different pathway from the sense organ to the projection area in the brain (Fig. 3). These pathways represent possible points at which depression may interfere with transmission and processing of sensory signals. Evidence of peripheral and central dysfunction in sensory pathway processes of individuals with major depressive disorder. A diagrammatic summary describing peripheral measures of sensory organ function and central measures of cortical function that have been explored in major depressive disorder. Coloured arrows represent afferent sensory processing pathways for smell (green), sight (blue), touch (purple), taste (red), and hearing (red). Black arrows indicate valency of change in patients with major depressive disorder versus healthy controls. Question marks indicate a lack of relevant studies have been performed comparing patients with major depressive disorder versus healthy controls. A1 primary auditory area, Amy amygdala, CN cochlear nucleus, ERP Event related potential, G1 primary gustatory area, NTS nucleus tractus solitarius, O1 primary olfactory area, S1 primary somatosensory area, SST somatostatin, Thal thalamus, V1 primary visual area. Peripheral sensory organ dysregulation Evidence of reduced responses from sensory organs in patients with MDD suggests that diminished sensory sensitivity has a significant origin in the periphery. Reduced visual electroretinogram amplitude in MDD patients is the strongest evidence for this, since these recordings reflect processes occurring solely in the retina and not in any other visual pathway-related structures in the brain (Fig. 3) [40–42]. Dysfunction of the retina, specifically, in MDD is further supported by evidence of retinal thinning in patients that appears to correlate with symptom severity [87, 88]. Though the electrical activity of other sensory organs has not been directly explored in humans, some evidence exists in rodent research to suggest that chronic stress and depression reduced the functioning of the olfactory mucosa, arguing for a peripheral factor in the depression-sensory hyposensitivity association [89]. In humans, several studies have reported that the olfactory bulb is smaller in MDD patients, correlating with olfactory hyposensitivity [33, 34, 90, 91]. Although the olfactory bulb itself does not contain olfactory receptor neurons, it does receive synaptic connections from them and serves as the major conduit through which olfactory information is conveyed to higher brain centers (Fig. 3). In terms of causal pathways, evidence of the effect of impaired olfactory bulb functionality upon the amygdala indicates a possible pathway from olfactory bulb dysfunction to depression [92] and is congruent with arguments that the salience of odors is related to their emotional significance [93]. Of note, the olfactory bulb has been described as the evolutionary ancestor of the limbic system, having made a major contribution to human survival by providing information regarding mate choice, detection of fear and joy in the body odor of others [94], thus establishing olfaction as occupying a place of major importance in human daily functioning and plausibly influencing mood. Additional to impairment in the olfactory bulb, inability to respond to odors has been reported via relatively smaller P2 and P3-1 amplitudes in EEG studies of chemosensory processing of odors among depressed versus non-depressed humans [92]. Confirming data come from an fMRI study showing that, although BOLD data did not differ between depressed and non-depressed participants in the primary olfactory areas (olfactory bulb, piriform cortex, entorhinal cortex, amygdala), depressed patients showed a significantly lower activation in the secondary activation areas (orbitofrontal cortex, insula, thalamus, hippocampus) [95]. Further challenging a simple explanation of olfaction and depression is the recent suggestion that olfactory hyposensitivity is more pronounced in patients with recurrent depression than those suffering from a single episode of MDD [96]. Thus, although the association between olfactory sensory hyposensitivity and depression is plausibly well-established, there is clearly much more to be done to understand how that association occurs, particularly its possible dose effect (i.e., is more frequent/more severe depression more likely to be accompanied by olfactory hyposensitivity). Comparable results have been observed for the gustatory system. Diffusion tensor imaging has been applied to study the afferent white matter tract originating from the nucleus tractus solitarius, which carries gustatory information to higher brain centres (Fig. 3). The authors of that study reported that patients with MDD exhibited decreased white matter integrity of this tract, which is consistent with a peripheral origin for gustatory hyposensitivity in depression [97]. Shared peripheral dysfunction across multiple sensory modalities may serve as support for the monoamine hypothesis of MDD. Modulation by monoamines is one factor shared by most sensory organs despite their otherwise disparate cellular composition, anatomy and afferent pathways into the brain. Dopamine-releasing cells have been found within the retina [98], olfactory mucosa [99], cochlea [100] and skin [101], all of which have been found to express dopamine receptors [98, 99, 102, 103]. Likewise, serotonin-releasing cells, or cellular processes, exist in the retina [104], olfactory bulb [105], cochlea [106], tongue [107] and skin [108], which also express serotonin receptors [104, 109–112]. To date, no study has investigated the levels of monoamine neurotransmitters, or their receptors, within the sensory organs of patients with MDD, despite evidence of markedly reduced levels of monoamine neurotransmitters within the serum of MDD patients [113, 114]. Future studies focused on mapping the expression of monoamines and their receptors within sensory organs of individuals with MDD may yield insightful and explanatory results. Central hyperexcitability across sensory modalities Measurements of brain function in MDD patients often reveal changes to scalp-measured electrical waveforms originating in the cortex. This is especially true for regions of the cortex that are associated with sensory processing. The visual cortex has been investigated by studying visually evoked potentials, which are a relatively accessible measure of compound cortical activity in response to a visual stimulus. Individuals with MDD exhibit visually evoked potentials with larger amplitudes [115] and altered plasticity [38] relative to healthy controls, both of which are consistent with increased cortical excitability. Interestingly, similar patterns have been observed across other sensory modalities when investigating event related potentials. Electroencephalogram recordings of auditory stimulus-evoked waveforms revealed higher rP50 and lower dP50 scores, which are indicative of reduced sensory gating [52]. As with the visual cortex, the auditory cortex of individuals with MDD exhibits increased amplitude of the auditory event related potential [116] and altered plasticity [46] consistent with hyperexcitability. Mentioned above (section 4.1) there is some argument that olfactory hyposensitivity may be associated with central dysfunction as well as peripheral dysfunction. Measurement of somatosensory evoked potentials elicited by electrical stimuli applied to the fingers has revealed enhanced amplitude of long latency responses [36], and median nerve stimulation produced prolonged latency of the P60m component of the somatosensory evoked field [80]. It has been hypothesised that cortical hyperexcitability for each sensory modality occurs via alterations in GABAergic circuitry in MDD patients [115, 117]. Indeed, evidence of reduced GABA concentration in the visual cortex of depressed individuals has been observed across multiple independent studies [118–121], as well as in studies of schizophrenia [122]. Moreover, the auditory cortex exhibits decreased density of GABAergic neurons [123], and the amygdala (which is a major recipient of input from the olfactory pathway (Fig. 3)) contains fewer somatostatin-positive GABA interneurons [124]. When considering evidence of sensory hyposensitivity (Table 2) and reduced event related potential amplitudes in patients with MDD (Fig. 3), hyperexcitability of the sensory cortices can be arguably explained as a process of gain enhancement in an attempt to maintain function globally across the sensory pathway. In this light, cortical changes in MDD can be interpreted as homeostatic, or adaptive, in response to blunted sensory input. More work mapping the excitability of components along each sensory pathway is needed to confirm this interpretation but doing so may help interpret the role of well-reported changes in MDD, such as those in the prefrontal cortex (reviewed in [20]). Reduced sensory sensitivity: function Depression is ‘adaptive’ Table 2 and the preceding sections of this article report on numerous studies of sensory hyposensitivity that may accompany depression. A few of those studies have also hypothesized about the processes that support that finding, such as loss in anhedonia-linked dopamine, or dysfunctions of specific sensory projection areas in the brain. These are valuable conjectures on how sensory sensitivity might be reduced in depression, but they do not address the underlying question of why sensory sensitivity should be reduced in depression. It is reasonable to search for some advantage inherent in these losses of sensory acuity during depressive periods. This theme is often referred to as an ‘evolutionary’ model of depression, and has been argued at length for some time as an explanation of depression that does not necessarily assume that MDD and its symptoms are always a ‘disease’ [125–127] but may confer some advantage to the depressive sufferer. This interpretation of depression as evolutionary has been criticized because it does not follow the definition of evolutionary behaviour. That is, according to Darwin [128] and others [129], evolutionary processes may be seen when different traits occur between members of a population, and those traits are inherited by the next generation, and they confer a survival or reproductive advantage to those offspring. In such cases, the traits will be naturally selected and will become more common in the population via evolutionary processes [130]. Of particular importance to the current discussion, the trait cannot be developed (e.g., in response to environmental stressors) during the parent’s lifetime (that would be Lamarckism, not ‘evolutionary’ in the current Darwinian sense). Following this model, to be an evolutionary-based behaviour, depression in parents would (a) somehow enhance their chances of reproducing successfully, (b) be passed on to their children, and (c) enhance their children’s chances of also reproducing successfully. Both (a) and (c) are unlikely, and so the definition of depression as ‘evolutionary’ must be questioned. Behavioural withdrawal However, there is one immediate benefit that the depressed person may experience and which may provide an adaptive advantage in the current environment [131–134]. By applying Bolling, Kohlenberg and Parker’s [132] statement that “Identified features of the depressive syndrome are all in the spectrum of a withdrawal from contact with the world and the consequences of activity” (p. 127), it is possible to conceive the MDD Diagnostic Criteria of sadness, anhedonia, sleep and appetite change and cognitive disturbances as part of a “biological response pattern …identified as the conservation-withdrawal response to excesses or deficits of stimulation” (p. 127). Because it reduces the effect of noxious stimuli in the environment, this withdrawal from aversive stimuli may be described as “adaptive” [126]. Additionally, because depressed people believe that they have no real control over the noxious stimuli they are experiencing, they are left with a single response that will reduce their distress—withdrawal from the environment that is causing the unpleasantness. This immediate benefit is what Darwin (1876/2002, p. 51) referred to when he commented that “But pain or suffering of any kind, if long continued, causes depression and lessens the power of action; yet it is well adapted to make a creature guard itself against any great or sudden evil”. Thus, while depressive withdrawal behaviour may confer an advantage of some kind to the sufferer, that advantage does not necessarily enhance the likelihood of the depressive person’s offspring reproducing successfully. Unfortunately, maintenance of depressive withdrawal behaviour also entails the depressed individual losing access to sources of social reinforcement. Thus, a circular process is established in which withdrawal from an aversive environment becomes both self-reinforcing and self-defeating [134, 135]. Because these initial withdrawal responses are unlikely to be advantageous in the longer term, they do not qualify as ‘evolutionary’, but rather as (short-termed) ‘adaptive’ [136–138]. As such, behavioural withdrawal serves a short-term function for the depressed individual [131, 133, 134, 139]. This behavioural withdrawal may be nonverbal and/or verbal [140], vocal [141], involve social interaction behaviour [142], and is consistent with some of the major Diagnostic Criteria for MDD, such as increase in sleeping, loss of appetite, social withdrawal, and lack of interest or pleasure in previously valued activities (anhedonia). Because of the established causal relationship between chronic stress and depression [143, 144], the development of MDD symptomatology might be expected to follow a train of stressful events as well as occur following a major acute stressor. In either case, the individual is required to adjust to a severe change in their environment or adapt to an ongoing series of smaller, but demanding, chronic stressors. Such adjustment necessitates changes in attitude, lifestyle, and relationships, all of which call upon the physiological and psychological resources of the individual. Meanwhile, the stressor(s) remain in place. Thus, the individual may be caught within an uncontrollable spiral of ‘stressor—adjustment—loss of personal resources—stressor’. Under these conditions, exhaustion may occur, which might only be relieved by active behavioural withdrawal from the stressful environment. This account of the function of depression is not an isolated view. For example, catatonia is a very severe form of mental disorder indeed, requiring monitoring to reduce the risk of self-harm [1]. This rare disorder often occurs comorbidly with MDD and has been explained as having immediate short-term self-protective benefits for the organism (i.e., stillness in the face of unavoidable threat from predators can assist in evading death) [145]. This phenomenon of ‘freezing’ is common among mammals as a response to a predator and has functional value in that it shows submission and hence may cause the predator to cease attacking [146, 147]. Freezing may also function because many predators’ actions are triggered by movement, and they may lose interest when the prey remain motionless for some time, thus conveying a survival advantage upon the frozen prey [148]. Additionally, some predators do not eat dead meat and may refrain from eating prey that appears to be dead (i.e., immobile) [149]. Similar freezing behaviour has been reported in humans when faced with rape [150] and airline disasters [151], and has been described as an evolutionary-based fear response [145]. Behavioural withdrawal from their noxious environment has other advantages for the depressed person. As explained by Andrews et al. [152], withdrawing from social interactions also provides valuable time for the depressed individual to reflect upon their problems and perhaps find a solution. Depressed persons may also elicit assistance from partners and close others [145, 153], receive intense nurturing in the form of psychotherapy [154], experience reduction of the organic unpleasantness associated with depression via antidepressant medication [155], have a reduction of workload via formal sick-leave [156], and receive wider social support in the form of counselling and welfare personnel. Thus, reporting that “I am depressed” may confer some advantages to the chronically stressed person. It is important to acknowledge that all of the symptoms of MDD are not necessarily unpleasant (e.g., social withdrawal for ‘peace and quiet’, sleeping, removal of the need to attend workplaces, and a general social forgiveness for failure to strive within a competitive society), and that the symptoms that are unpleasant might be profitably viewed as psychophysiological responses to an uncontrollable stressful environment. In this way, the symptoms of MDD might be viewed as (for example) vomiting and diarrhea after food poisoning; that is, as physical responses by the body to rid itself of a pathogen. Support for this perspective comes from the work by Hart et al. [157] and others [158]. Additionally, it has been suggested that this kind of sickness behaviour may also help the sufferer build tolerance against a pathogen (or depression-inducing stressor), in the same way as fever may protect tissues by a heat shock response in the organism [158]. In the discussion above, the emphasis has been upon subjective withdrawal behaviour, i.e., instigated by the individual as a chosen path to reduced aversive environmental stimuli. However, some of the studies reviewed in Table 2 also include central processing dysfunctions that may be considered as not usually within the subjective control of depressed people. In that way, those centrally induced or mediated withdrawal responses may be likened to the example given above of vomiting and diarrhea to rid the body of harmful pathogens. Other examples include increase in body temperature to necrose infectious bacteria, initiation of mucus discharge from the nasal passages to flush pathogens, and the phagocytic action of high levels of cortisol on invasive species within cells [159]. None of these are usually within voluntary control of the individual and so may be termed as ‘centrally mediated’, similar to the changes in EEG signals and MRI volume data reported in some of the studies reviewed in Table 2. Although the issue of ‘which system initiates change’ is of great interest, for the purposes of this discussion it is sufficient to classify centrally mediated and voluntary-induced behavior similarly in terms of their adaptive outcome. Discussion Major findings Based on the rationale presented above for conceiving depression as a behavioural withdrawal process, the sensory hyposensitivity reported in Table 2 may be understood as part of that process, enabling the depressed individual to receive less vivid or intense sensory information from their (noxious) environment and focus instead on self-reflection and reserving their personal resources. Although not consistently investigated as often as vision, investigations in the other senses largely concur with those from vision research in finding that depression is very commonly associated with sensory hyposensitivity. These effects are present despite the format of depression diagnosis (i.e., clinician interview, participant self-report), are found in various national and cultural settings, across both biological sexes, in a range of ages, and despite several different methods of testing sensory acuity, as well as being confirmed by ES for each sense, and the meta-analysis of olfactory data. Thus, sensory hyposensitivity may be accepted as having potential to be an associated feature of MDD. Tracing the neurological pathways from peripheral sensing to central processing suggests that both ends of that connection may be involved in sensory hyposensitivity, and that there may be some differences across senses in the precise nature of that connection, although that could also be an outcome of the kinds of methodologies used. Extension of those methodologies to include comparisons of peripheral versus central processing components is a potentially advantageous avenue for future research, as would be the focus upon variation in sensory hyposensitivity during depression. Some studies have reported an improvement in sensory acuity following treatment for depression, but the precise mechanism for that improvement remains unexplained in the kind of detail that might influence common treatment practice for depression. Similarly, detection of sensory hyposensitivity prior to the full onset of MDD represents an opportunity to test the predictive value of sensory hyposensitivity, and to connect that sensory hyposensitivity with chronic stressor experience. Further to that, the kinds of stressors that precede sensory hyposensitivity in depression deserve attention to provide awareness of whether experiencing a specific chronic stressor is likely to lead to sensory hyposensitivity. For example, some time ago Keller and Nesse [138, 160, 161] found that different stressor categories (which they called “precipitants”, p. 27) were associated with six subgroups of depressive symptoms (e.g., social losses were associated with crying and arousal; failure to reach a goal was linked with fatigue and pessimism). However, not all assumed stressors have negative effects upon those who experience them, and some even have reportedly positive effects thus challenging the validity of grouping many stressors into exclusive categories. Investigation of specific stressors as precipitants of MDD symptoms for individual patients might be more informative. More recently, the link between specific stressors and particular MDD symptoms has been evaluated within a functional analysis model in which depressive symptoms are understood to be ‘adaptive’ [162], as postulated in this review. Clinical implications In terms of the direct implications this review has for clinical practice, it behooves clinicians to at least enquire about depressed patients’ sensory hyposensitivity as an accompanying aspect of the depression experience. If thought advisable, explaining the ‘function’ of this sensory hyposensitivity might help patients understand this aspect of their symptomatology, although that is not a necessary step to reap at least some benefits from helping patients understand how their physical senses are responding to the chronic stress that is often a precursor to depression. There may be a case for instituting sensory training for some patients as a possible means of reversing some MDD symptoms. Limitations and future directions As in all research, this review has some limitations which argue for further research. Some of the senses have been investigated for their connection with depression in only a few studies and, unlike vision, hearing, and olfaction, need further replication and expansion before firm conclusions can be drawn. Only studies reported in English were included in the review section of this article, and there may be other studies in different languages that hold the keys to further understanding of the sensory hyposensitivity-depression association. Due to the very wide range of instruments and scores obtained from sensory tasks, plus the differences in number of studies of various senses, it was not realistic to undertake a full meta-analysis, although those techniques were applied to the olfactory studies. However, where possible, effect sizes were calculated for all studies and were able to contribute to summaries of outcomes. Perhaps explaining that variability in instrumentation, the differentiation of sensory sensitivity into sub-categories of self-reported symptoms, neural activity induced by sensory stimulation, or the behavioural ability to detect and discriminate between sensory stimuli was made above, and future research could focus on those separate aspects of sensory hyposensitivity. The inconsistency in the choice of dependent variables reported in the studies reviewed reflects the different sensory, neural, or behavioral processes that have been focused on in the research reviewed here, producing several sets of data that are not always directly comparable. As was noted earlier in this review, these different forms of sensory hyposensitivity do not correlate strongly [26], but some degree of sensory hyposensitivity has been found in depressed versus non-depressed adults [27], arguing that there is a common underlying process that may be neural. Identifying that common neural process is logically a major next step for research into the association between depression and sensory hyposensitivity. Our search procedures were restricted to actual measures of MDD patients’ sensitivity to sensory stimuli, and did not include studies of the way that individuals process signals from their environment as an underlying trait, such as described in Dunn’s [163] Sensory Profile analysis which classifies respondents according to their ability to modulate sensory information, process it, and adapt their behavior to their environment, usually measured via a questionnaire about how they achieve those goals [164]. Although this form of sensory processing has been studied in depression [27], it is not congruent with the methodology used to measure sensitivity to sensitivity to sensory stimuli measured used in the studies reviewed in Table 2. The literature searches undertaken here were based on depression as the key variable. However, there may be justification in searching other disorders for their association with sensory hyposensitivity because of the comorbidity of depression with other disorders such as anxiety, PTSD, chronic pain, or neurodevelopmental conditions to identify the pathways between depression and sensory hyposensitivity. Similarly, depression is not homogeneous, and (as mentioned in section 1.1) may be diagnosed in nearly 1500 different combinations of the Diagnostic Criteria and Features for MDD [3] as set out in the DSM-5-TR [1]. There are also well-recognized subgroupings of these MDD symptoms that have been formed into depression ‘subtypes’ [165]; research into the sensory hyposensitivity accompanying these subtypes would be of value. These steps were beyond the scope of this review, which aimed to examine depression per se and the accompanying sensory hyposensitivity that may have occurred in participants whose primary diagnosis was MDD. Participant recruitment is always a potentially biassing confound but this was not universally reported in studies reviewed here. As a future methodological step, the comparison between (for example) volunteers vs students vs depressed patients vs sensory hyposensitivity patients would be valuable. Finally, although the behavioural withdrawal model was applied to explain sensory hyposensitivity, there may be other theoretical models that could also help explain this phenomenon. Conclusion Despite these limitations, and the current lack of conclusive data regarding all senses and the association between sensory hyposensitivity and depression, this review is the first to bring together all the extant literature on change in the five senses as experienced by depressed individuals, and to demonstrate the generally common presence of sensory hyposensitivity among depressed participants. As well as representing an intriguing neuropsychological issue that deserves further research attention, these summarized findings argue for the inclusion of information about patients’ sensory hyposensitivity in clinical assessment protocols. By clinicians raising the possibility of sensory hyposensitivity, depressed patients may also be more likely to expand upon the effect that has upon their emotional state, whether it is beneficial (by reducing the intensity of the stressor) or disturbing (in patients’ inability to perceive their environment as they used to), and what they might do about it. These are all therapeutically valuable and may contribute to the patients’ gradual movement towards greater self-management. Author contributions CFS conceptualized and planned the study, compiled the review of studies, co-wrote the summaries of study findings, drafted the initial ms and contributed to the final version; GLO wrote the section on peripheral and central processes, compiled Fig. 2, and contributed to the final version; VB wrote the sections on behavioural withdrawal and clinical implications and contributed to the final version; CBW co-identified the studies, analysed the effect sizes, developed Figs. 1 and 3, co-wrote the summaries of study findings, and contributed to the final version. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Data availability No new data were collected for this article. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement No new data were collected for this article.