SYSTEMATIC REVIEW article

Front. Med., 29 November 2022

Sec. Dermatology

Volume 9 - 2022 | https://doi.org/10.3389/fmed.2022.1054898

Anxiety, depression, and quality of life in children and adults with alopecia areata: A systematic review and meta-analysis

  • MV

    Marije van Dalen 1,2*

  • KS

    Kirsten S. Muller 1

  • JM

    Johanna M. Kasperkovitz-Oosterloo 3,4

  • JM

    Jolanda M. E. Okkerse 2

  • SG

    Suzanne G. M. A. Pasmans 5

  • 1. Department of Pediatric Gastroenterology, Erasmus MC Sophia Children’s Hospital, Rotterdam, Netherlands

  • 2. Department of Child and Adolescent Psychiatry/Psychology, Erasmus MC Sophia Children’s Hospital, Rotterdam, Netherlands

  • 3. Dutch Alopecia Association, ’s-Hertogenbosch, Netherlands

  • 4. IPSO Institutes for Psychosocial Oncology, Almere, Netherlands

  • 5. Department of Dermatology, Center of Pediatric Dermatology, Erasmus MC Sophia Children’s Hospital, Rotterdam, Netherlands

Abstract

Introduction:

Alopecia areata (AA) is a non-scarring hair loss condition, subclassified into AA, alopecia universalis, and alopecia totalis. There are indications that people with AA experience adverse psychosocial outcomes, but previous studies have not included a thorough meta-analysis and did not compare people with AA to people with other dermatological diagnoses. Therefore, the aim of this systematic review and meta-analysis was to update and expand previous systematic reviews, as well as describing and quantifying levels of anxiety, depression, and quality of life (QoL) in children and adults with AA.

Methods:

A search was conducted, yielding 1,249 unique records of which 93 were included.

Results:

Review results showed that people with AA have higher chances of being diagnosed with anxiety and/or depression and experience impaired QoL. Their psychosocial outcomes are often similar to other people with a dermatological condition. Meta-analytic results showed significantly more symptoms of anxiety and depression in adults with AA compared to healthy controls. Results also showed a moderate impact on QoL. These results further highlight that AA, despite causing little physical impairments, can have a significant amount on patients’ well-being.

Discussion:

Future studies should examine the influence of disease severity, disease duration, remission and relapse, and medication use to shed light on at-risk groups in need of referral to psychological care.

Systematic review registration:

[https://www.crd.york.ac.uk/prospero/], identifier [CRD42022323174].

Introduction

Alopecia areata (AA) is a hair loss condition with a lifetime prevalence of 2.1% (). AA has a peak onset between 25 and 29 years old, with a median age at diagnosis of 31 for males and 34 for females. It occurs more frequently in people with a non-white ethnicity (). Males and females appear to be affected equally often (), however research has also reported females to be slightly more likely to experience AA (). AA is typically divided into AA (patchy hair loss), alopecia universalis (AU; total loss of scalp hair), alopecia totalis (AT; total loss of body hair) and alopecia ophiasis (band-like hair loss on the temporal and occipital scalp) ().

Alopecia areata has an unpredictable disease course characterized by relapse and remission (). Full hair regrowth may be observed in 50–80% of patients (, ), but relapse rates of 30–52% have been reported () with around 30% of patients with AA eventually progressing to complete hair loss (). Relapse is more likely in patients with an earlier onset of AA, but is not related to gender, clinical severity and treatment given (). Furthermore, medication often fails to provide sustained hair regrowth ().

There are indications that people with AA experience adverse psychosocial outcomes. Qualitative studies, for instance, have shown that patients reported considerable distress (). Feelings of sadness, insecurity, inadequacy, and self-consciousness (), as well as feelings of depression, anxiety, and suicidal thoughts () were prevalent. The majority of qualitative research highlights that people struggle with everyday activities, such as participating in sports or social events, due to a fear of their appearance being noticed (). The unpredictable nature of AA was also highlighted as a source of distress in particular (, ) and women seem to report more stress and distress than men (, ).

Most quantitative research has focused on anxiety, depression or quality of life (QoL). For anxiety, a meta-analysis including eight studies by Okhovat et al. () showed that people with AA are 2.50 times more likely to experience anxiety. However, it is unclear how the papers were selected and what type of control group was included in the meta-analysis. Other studies have shown that people with AA have a higher chance of being diagnosed with an anxiety disorder than healthy controls (). When the amount of anxiety symptoms of people with AA is compared to people with other dermatological diagnoses mixed results have been found ().

When looking at depression, the aforementioned meta-analysis found that people with AA are 2.71 times more likely to experience depression (). This result is corroborated by other studies reporting people with AA to be more likely to be diagnosed with depression (, ). As for anxiety, it is unclear how people with AA compare to people with other dermatological diagnoses (, ).

A systematic review conducted in 2018 has shown that AA has a considerable impact on QoL (). However, it remained unclear how QoL was related to disease severity (). Furthermore, people with AA were not compared to people with different dermatological diagnoses in this review. More recent research has reported a moderate effect on QoL (), as well as no effect (). Comparisons to people with a different dermatological diagnosis have yielded mixed results. For instance, one study comparing people with AA to people with alopecia androgenetica reported people with AA to have better QoL (), while another study found the opposite result ().

Although previous systematic reviews on psychosocial consequences of AA have been conducted (e.g., , ), it remains unclear how people with AA compare to people without AA or people with a different dermatological diagnosis. In addition, these reviews have not highlighted the psychosocial impact of AA on different age groups (i.e., children or adults). Therefore, the purpose of the current systematic review and meta-analysis was to update and expand previous systematic reviews, as well as describing and quantifying levels of anxiety, depression, and QoL in patients with AA, AU, or AT. We also aimed to explore whether gender or age would influence the amount of anxiety, depression, and QoL experienced by people with AA. We specifically sought to answer the following research question: What is the impact of living with alopecia areata, alopecia totalis or alopecia universalis on levels of anxiety, depression, and quality of life in children and adults? We also wanted to know how levels of anxiety, depression, and QoL of people with AA compared to people with a different dermatological condition and to healthy controls.

Materials and methods

This article was written in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement () and was registered prospectively in the international prospective register of systematic reviews, PROSPERO, registration number CRD42022323174. The protocol was registered with a broad focus on psychological impact of AA, as it was unclear how many papers the search would yield. After selection of relevant papers, a decision was made to focus only on anxiety, depression and QoL and a further nine papers were excluded (see Figure 1).

FIGURE 1

Search strategy

As this article was part of a bigger project for the Dutch Alopecia Association, a broad search focusing on the psychosocial impact of living with AA was conducted by a research librarian on 28 March 2022. The following databases were searched from inception: Embase, Medline, Web of Science Core Collection, Cochrane Central Register of Controlled Trials, PsycInfo, and Google Scholar. The search included terms, both Mesh and free text, related to alopecia and the psychosocial impact, without restrictions on language or publication date. Only published, peer-reviewed papers were used. The full search is displayed in Supplementary material.

Eligibility criteria

Studies were included if they met the following eligibility criteria: (a) studied a sample with AA, AU, and/or AT, (b) reported quantitative data on anxiety, depression or QoL, and (c) the paper was an original research paper. Studies were excluded if they (a) reported no original data (e.g., case-reports, conference abstracts, and systematic reviews), (b) were not written in English, or (c) did not separate AA from other medical diagnoses. No criteria were set for the amount of timepoints in an article (i.e., the article being cross-sectional or longitudinal). In case of a longitudinal intervention study, only the baseline data were included.

Study selection

Studies were selected if they met the inclusion and exclusion criteria. Two reviewers (MD and KM) independently assessed the title and abstract. The interrater agreement was 81.55%. Discrepancies were resolved using consensus. Afterward, the two reviewers independently assessed the full text for eligibility. Interrater agreement for this step was 87.46%. Discrepancies were again resolved using consensus. One of the reviewers (MD) checked the reference list of included articles for additional relevant references. Any references deemed relevant were first screened based on title and abstract. If still relevant, the full-text was read. When the article met the inclusion and exclusion criteria, it was included in the review. Endnote 20 was used to manage references.

Data extraction

Data collection was done by one researcher (MD) and checked by another researcher (KM) using a data extraction form. The following data were extracted: type of alopecia, sample size, percentage male, mean age (SD), age range, method involved (questionnaire or interview), main conclusions, mean score (when method is questionnaire), mean prevalence of symptoms/diagnosis, any relevant comparisons between groups (e.g., anxiety symptoms in AA vs. unaffected controls). Authors of papers were contacted when relevant data for meta-analyses was missing.

Quality and risk of bias

Quality and risk of bias were assessed using the relevant NIH quality assessment tool for controlled intervention studies, observational cohort and cross-sectional studies, case-control studies or before-after studies with no control group [National Heart, Lung, and Blood Institute (NIH), 2018] () or the QAVALS (). Questions can be answered with “yes, no or cannot determine/not reported/not applicable responses.” We rated >80% points as good, 60–80% points as fair and <60% as poor quality. Quality assessment was performed independently by two reviewers (MD and KM). Half of the articles were discussed in a consensus meeting, after which the remaining half of the papers was checked by one reviewer (MD).

Data synthesis and statistical analyses

All studies were included in the qualitative synthesis. Meta-analyses were conducted for five or more similar studies. As a high level of between-study heterogeneity was expected, a random-effects model was used to pool effect sizes. The Restricted Maximum Likelihood Estimator (REML) was used to calculate heterogeneity variance (). Means and standard deviations (SDs) of samples were used to compute effect sizes, the standardized mean differences (SMDs), quantified in the form of Hedges’ g (). When means and SDs were not available, medians were transformed to means and SDs as described by Shi et al. (). Publication bias was tested by visual inspection of a contour-enhanced funnel plot () and Egger’s test in case of ≥10 studies. Exploratory meta-regressions were conducted. For each meta-analysis, one model was created with mean age, percentage male, and quality rating as independent variables. The significance level was set to α = 0.05. Analyses were done using the meta package () in RStudio.

Results

Study selection

After removing duplicate records, a total of 1,249 records were retrieved for screening. After title and abstract screening, 280 records were assessed for eligibility. Finally, 93 articles were included for the qualitative synthesis of which 26 articles were also included in the quantitative synthesis. The full selection process is displayed in Figure 1. Overall, 74 papers were of poor quality, 16 papers were of fair quality and 4 papers were of good quality.

A total of 52 papers studied anxiety in children and/or adults with AA. Seven papers () had a combined research group with children and adults (n = 11,007, Mage = 41.78, 43.63% male), eight papers () studied children with AA (n = 398, Mage = 11.85, 47.00% male) and 37 papers (, , , , , , , ) studied adults with AA (n = 88,858, Mage = 40.03, 41.25% male).

For depression, 65 papers were included. Fourteen papers (, , ) looked at children and adults (n = 18.638, mean age = 36.26, 43.44% male), nine papers (, ) studied children (n = 3908, Mage = 11.85, 44.82% male) and 42 papers (, , , , , , ) studied adults with AA (n = 93,047, Mage = 41.69, 40.39% male).

A total of 40 studies investigated QoL in people with AA. Five studies (, , , , ), combined children and adults into one sample (n = 3611, Mage = 31.43, 60.09% male), three studies (, , ) investigated children (n = 258, Mage = 11.50, 47.45% male) and 32 studies (, , , , , , , , , ) investigated adults with AA (n = 5,373, Mage = 41.38, 42.83% male).

Anxiety

The results for anxiety are shown in Table 1.

TABLE 1

ReferencesCountryYearN% maleAge (M, SD)% AA, AT, AUControlsMeasuresConclusionsQuality score (%)
Pediatric and adult samples
Ataseven et al. ()TurkeyNR4372.123.42 (11.41)NR30 healthy controlsHAM-AAA more symptoms of anxiety than healthy controls303
Chu et al. ()Taiwan2000–20095,11749.2NRNR20,468 healthy controlsICD-9 codesAA diagnosed with anxiety more often than controls803
Kökcam et al. ()TurkeyNR17NR26.47 (12.2)NR11 vitiligo, 20 healthy controlsSCL-90-RAA more symptoms of anxiety than healthy controls, no differences with vitiligo203
Marahatta et al. ()NepalAugust 2015–July 20167553.329.40 (9.90)NRNoBAI89.0% very low anxiety, 8.0% moderate anxiety, 0% severe anxiety45.834
Singam et al. ()USA2002–20125,605 hospitalized patients38.342.2 (NR)NRHospitalized patients without AA (N unknown)ICD-9 codesAA diagnosed with anxiety more often than controls453
Talaei et al. ()IranApril–July 20052433.3325.38 (8.32)NR24 healthy controlsSCL-90-RNo significant difference with controls703
Vélez-Muñiz et al. ()MexicoMarch 2017–February 201832 child, 94 adults41NR92.9% patchy AA, 3.2% AT, 1.6% ophiasis, 1.6% AUNoHADSFor adults: 19.1% heightened anxiety/depression, 34.1% no anxiety/depression504
Pediatric samples
Altunisik et al. ()TurkeyNR2729.611.9 (3.3)85.19% AA, 14.81% AU30 dermatology patientsK-SADS-PL; SCARED; STAI-CNo difference with controls on questionnaires or diagnoses. 51.8% of AA patients had at least 1 anxiety diagnosis653
Andreoli et al. ()Italy1997–2000176NRNRNRNoDiagnosis by psychologist16% diagnosis generalized anxiety disorder, 8% social anxiety disorder254
Bilgiç et al. ()TurkeyNR7455.4112.1 (2.8)NR65 healthy controlsSTAI-CAA more state anxiety than controls. Children, but not adolescents more trait anxiety than controls.653
Díaz-Atienza and Gurpegui ()SpainNR315212.2 (3.8)51.61% AA, 48.39% AU/AT23 epilepsy, 25 siblingsSTAI-CNo difference on symptoms of anxiety between AA and epilepsy or sibling group653
Erdoğan and Gür ()TurkeyOctober 2018–December 20193154.8312.54 (3.56)100% AA29 vitiligo, 30 healthy controlsRCADS-C; RCADS-PMore social anxiety and total anxiety in AA (child-reported) for HC. More panic disorder and total anxiety in AA (parent-reported) for HC. No differences with vitiligo.603
Ghanizadeh ()IranAugust 2004–November 200614NR11.66 (6.08)NRNoK-SADS-PL7.1% diagnosis social anxiety SAS, 28.6% specific phobia, 7.1% generalized anxiety disorder504
Liakopoulou et al. ()GreeceNR3330.310.5 (0.3)NR30 patients from pediatricianCMASAA higher scores on worry, oversensitivity and concentration403
Reeve et al. ()USANR12NR11.5 (2.9)NRNoDICA-R; RCMAS58.33% with any anxiety disorder diagnosis37.54
Adult samples
Aghaei et al. ()IranNR4044.835.2 (9.2)NR40 healthy controlsBAIMore symptoms of anxiety in AA patients than controls353
Alfani et al. ()ItalyNovember 2009–October 20107345.235.2 (9.2)61.7% AA,
26.0% AT,
12.3% AU
73 healthy controlsClinical interview; MMPI-2More anxiety in AA patients than controls353
Altinöz et al. ()TurkeySeptember 2011–October 2012305033.3 (8.9)NR30 urticaria, 39 healthy controlsHADSMore anxiety in AA patients than healthy controls. No difference with urticarial.403
Annagur et al. ()TurkeyNR7365.7527.66 (7.79)100% AA78 healthy controlsSCL-90No difference in symptoms of anxiety353
Atış et al. ()TurkeyNR395933.5 (11.6)NR46 vitiligo, 46 healthy controlsHADSAA more anxiety than healthy controls. No difference with vitiligo.203
Baghestani et al. ()IranNR687235.4 (7.6)100% AA68 healthy controlsHAM-AAA more symptoms of anxiety than healthy controls603
Bain et al. ()UKNR3923.0743.15 (12.43)NR23 PsA; 26 healthy controlsHADS2More anxiety in less severe AA and shorter disease duration303
Balieva et al. ()13 European countriesNovember 2011–February 20133333.342.8 (14.1)NR1,359 healthy controlsEQ-5D-3LAA 4 times higher chance of anxiety/depression than controls653
Brajac et al. ()Croatia1995–19994537.7840.24 (13.01)100% AA45 benign scalp lesionsSTAIAA more symptoms of anxiety than healthy controls603
Bukharia et al. ()IndiaNR1004854% 15–30 years, 46% 31–50 yearsNR100 TE, 100 healthy controlsHAM-A36.84% of AA and 43.94% of TE heightened anxiety453
Cakirca et al. ()TurkeyMarch–December 20173375.826.33 (6.08)NR33 healthy controlsHADSAA more symptoms of anxiety than healthy controls303
Colon et al. ()USAApril 1985–October 1987312935.70 (10.23)74% AA, 23% AT, 42% AU1NoDISLifetime prevalence generalized anxiety disorder 39%, specific phobia 23%, panic disorder 13%33.334
Conic et al. ()USA2005–201458431.535.54 (19.28)94.7% AA, 2.05% AT, 3.25% AU172 SDDiagnoses in patient fileNo difference with SD. 13.70% of AA has any diagnosis of anxiety353
Cordan Yazici et al. ()TurkeyNR4360.533.80 (10.02)95.35% AA, 4.65% AT53 healthy controlsHADSNo significant differences between AA and controls253
Devar ()IndiaNR30100NRNR30 TV, 30 healthy controlsTMASAA more symptoms of anxiety than healthy controls, no difference with TV503
Endo et al. ()JapanJune 2009–August 201012233.138.3 (16.5)NRNoSTAIAnxiety not related to disease severity and disease duration56.255
Gallo et al. ()ItalyNR1637.545.95 (13.25)NRNoBSIAA more symptoms of anxiety than norm group39.296
Güleç et al. ()TurkeyMarch 2001–January 20025265.3831.53 (12.61)94.23% AA, 3.65% AU, 1.92% AT52 healthy controlsBAINo differences AA and controls253
Karia et al. ()IndiaNR5060.027.76 (NR)NR50 psoriasis, 50 healthy controlsDSM-IV-TR diagnosis4% of AA any anxiety disorder diagnoses. More often than healthy controls, less often than psoriasis.603
Kim et al. ()South Korea2002–20137,70651.954.6% 20–39, 39.4% 40–59, 6.1% 60+NR30,824 without AAICD-10 codesAA higher risk of anxiety disorder diagnosis than controls653
Kose et al. ()TurkeyNR1810021.3 (NR)NRNoSTAIPositive correlation between anxiety and depression or hopelessness50.007
Macbeth et al. ()UKJanuary 2009–December 20185,43545.938.93 (14.35)NR21,470 healthy controlsDiagnoses in patient file3.24% of AA and 0.24% of healthy controls had anxiety disorder diagnoses803
Rajoo et al. ()AustraliaNR83NR40.95 (13.24)NRNoDASS-2166.3% reported extreme symptoms of anxiety54.174
Ruiz-Doblado et al. ()SpainNR3215NRNRNoSCAN22.2% diagnosis generalized anxiety disorder, 7.4% social phobia37.54
Russo et al. ()ItalySeptember 2016–September 20172733.337.55 (10.37)NR80 AGA, 36 TESTAI; SPSNo differences in trait anxiety or social anxiety. AA less social phobias than AGA and TE503
Şahiner et al. ()TurkeyAugust 2009–July 2010414932.9 (10.5)NR30 psoriasis, 50 healthy controlsBAIAA more symptoms of anxiety than healthy controls, no difference with psoriasis203
Sayar et al. ()TurkeyNR3110023.8 (2.5)NR40 healthy controlsSTAIAA more state and trait anxiety553
Sellami et al. ()TunisiaMarch–July 2010504832.92 (11.81)NR50 healthy controlsHADSAA more symptoms of anxiety than healthy controls453
Senna et al. ()USAJanuary 2011–December 201868,1213940.3 (17.8)98.1% AA, 1.3% AT, 0.6% AUNoICD-9 and ICD-10 codes8.4% had an anxiety disorder45.834
Sorour et al.
()
EgyptNR20858.65NRNR1,042 dermatology patientsDSM-5 interview19.71% of AA had anxiety diagnosis, no effect of gender. No difference with psoriasis. Less symptoms than acne, vitiligo, urticaria, and atopic dermatitis.553
Tan et al. ()ChinaDecember 2012–August 20131685034.5 (11.5)88.1% AA, 11.9% AT/AU100 healthy controlsSCL-90-RAA more symptoms of anxiety and phobic anxiety than controls41.675
Titeca et al. ()13 European countries37NRNRNRNR1,359 healthy controls, 20 AGAHADSAA more symptoms of anxiety than healthy controls and AGA703
Tzur Bitan et al. ()Israel201841,05562.939.97 (13.61)NR41,055 healthy controlsICD-9 codesAA higher risk of anxiety disorder than controls803
Willemsen et al. ()BelgiumSeptember 2006–August 2009212441.95 (13.79)33% patchy AA,
14% ophiasis, 29% AT, 24% AU
NoSCL-90AA more symptoms of anxiety than norm group54.177
Willemsen et al. ()BelgiumApril 1999–April 20042835.7133.421.43% AA, 21.43% ophiasis, 28.57% AU, 3.57% ATNoSCL-90AA more symptoms of anxiety than norm group50.007
Yoon et al. ()South KoreaJanuary 2015–February 20161,20352.1239.45 (12.21)NRNoBAI10.1% symptoms of anxiety, 4.2% severe symptoms of anxiety41.674
Yu et al. ()ChinaOctober 2013–December 201413041.531.78 (10.34)NR212 AGAS-ASNo significant differences703

Results for anxiety.

AA, alopecia areata; AGA, alopecia androgenetica; AU, alopecia universalis; AT, alopecia totalis; NR, not reported; PsA, psoriatic arthritis; SD, seborrheic dermatitis; TE, telogen effluvium; TV, tinea versicolor.

1Some patients had multiple episodes, with different forms of alopecia. Hence, the total is higher than 100%.

2This questionnaire was not administered to the control group.

3As measured by the NIH Quality Assessment of Case-Control studies.

4As measured by the NIH Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies.

5As measured by the QAVALS ().

6As measured by the NIH Quality Assessment of Controlled Intervention Studies.

7As measured by the NIH Quality Assessment Tool for Before-After (Pre-Post) Studies with no Control Group.

Children and adults

Three studies with a total of 5,665 patients with AA, reported that people with AA experienced more symptoms of anxiety and were diagnosed with anxiety more often than healthy controls (). One smaller study (n = 24) () did not find a difference in the amount of symptoms of anxiety between people with AA and healthy controls.

When people with alopecia were compared to people with another (dermatological) condition, studies found that people with AA were diagnosed with an anxiety disorder more often than other hospitalized patients in general (), but no differences were found for people with vitiligo ().

One study without a control group () found that 19.1% of the adults with alopecia reported heightened symptoms of anxiety or depression. The same study also reported that 34.1% did not experience any symptoms of anxiety or depression. In the other study without a control group 89.0% of people with alopecia reported little to no symptoms of anxiety (). Around 8% of people with AA reported moderate symptoms of anxiety.

Children

Of the papers investigating anxiety disorders, one study () reported that over half of the children had an anxiety disorder. However, this study included only 12 children and used the DSM-III-R, which was published in 1987. Two other studies reported that 7.1–16% had a generalized anxiety disorder, 7.1–8% had a separation anxiety disorder and 28.6% had a specific phobia (, ). However, none of the studies specified the number of patients with more than one anxiety disorder. It remains unclear from this data how many children with AA are diagnosed with an anxiety disorder. In a study by Altunisik et al. (), 51.8% of the children was diagnosed with at least one anxiety disorder. This did not differ significantly from children with another dermatological condition.

When looking at symptoms of anxiety, studies comparing children with AA to healthy controls found mixed results. On the one hand, Bilgiç et al. () reported more state and trait anxiety in children aged 8–12 with AA. They did not find any differences for adolescents aged 12–18. On the other hand, Díaz-Atienza et al. () did not find significant differences when comparing children with AA to their siblings. Erdoğan et al. () found no difference on the Beck Anxiety Inventory (BAI), but found more child-reported separation anxiety and total anxiety and parent-reported panic disorder and total anxiety than healthy controls on the Revised Child Anxiety and Depression Scales (RCADS).

Studies comparing children with AA to children with other (dermatological) conditions found no differences in symptoms of anxiety when comparing to other dermatological conditions (), epilepsy () and vitiligo (). Liakopoulou et al. () found that children scored higher on worry, oversensitivity, and concentration than other patients.

Adults

Eight papers studied the prevalence of anxiety disorders in adults with AA (n = 86,014). These studies reported point prevalence rates of 3.24% (), 4% (), 8.4% (), and 13.70% (). Several papers also reported that people with AA have a higher chance of being diagnosed with an anxiety disorder in comparison to healthy controls (, , , ). Prevalence rates of specific anxiety disorders in people with AA range from 7.4% for specific phobias and 22.2–39% for generalized anxiety disorders (, ). The lifetime prevalence of specific phobia and panic disorder was estimated at 23 and 13%, respectively ().

When looking at symptoms of anxiety, 15 studies compared people with AA (n = 749) to healthy controls (n = 733). These results were combined in a meta-analysis, shown in Figure 2. The results showed that adults with AA reported significantly more symptoms of anxiety than people without AA (g = 0.61, 95% CI [0.48, 0.75], p < 0.001), with a medium to large effect. There was little heterogeneity (I2 = 33.1%, 95% CI [<0.01, 64.0], τ2 = 0.02, 95% CI [<0.01, 0.12]) and visual inspection of the funnel plot showed no indication for publication bias. Egger’s test also did not show indications for a publication bias [t(13) = 0.94, p = 0.363]. Thirteen studies without missing data were included in a meta-regression. The model did not explain any variance in the effect sizes (R2 = <0.01%), with a residual heterogeneity of I2 = 46.59%. Mean age (g = 0.01, p = 0.802, 95% CI [−0.04 to 0.05]), percentage male (g = <−0.01, p = 0.858, 95% CI [−0.01 to 0.01]) and quality score (g = <0.01, p = 0.583, 95% CI [<−0.01 to 0.01]) did not influence study effect sizes.

FIGURE 2

Studies comparing people with AA to people with other (dermatological) conditions showed mixed results. For the majority of studies, no significant differences were found. For instance, no differences were found when comparing to people with chronic urticaria (), vitiligo (, ), seborrheic dermatitis (), tinea versicolor (), alopecia androgenetica and telogen effluvium (), and psoriasis (, ). A smaller number of studies reported that adults with AA experienced more symptoms of anxiety than patients with benign skin lesions () and alopecia androgenetica (, ), but less than people with psoriasis (), acne, vitiligo, chronic urticaria, and atopic dermatitis ().

Three studies compared adults with AA (n = 61) to a norm group. These studies all reported more symptoms of anxiety in adults with AA (, , ).

Depression

The results for depression are shown in Table 2.

TABLE 2

ReferencesCountryYearN% maleAge (M, SD)% AA, AT, AUControlsMeasuresConclusionsQuality score (%)
Pediatric and adult samples
Ataseven et al. ()TurkeyNR4372.123.42 (11.41)NR30 healthy controlsHAM-D; CDIMore symptoms of depression in AA compared to controls301
Chu et al. ()Taiwan2000–20095,11749.2NRNR20,468 healthy controlsICD-9 codes2.9% AA has depression diagnosis, more often than controls801
Ghajarzadeh et al. ()IranJanuary 2009–January 20101006923.02 (33.4)NR100 psoriasis, 100 vitiligoBDINo difference AA and psoriasis/vitiligo551
Gutierrez et al. ()USA2006–20162,298,432 visits to dermatologist3537.8 (18.04)NRNoICD-9 and ICD-10 codes4.3% of the visits was related to depression58.332
Jagtiani et al. ()IndiaNR3865.825.79 (8.82)NR80 AV, 56 psoriasisBDIAA not significantly different from patients with acne vulgaris or psoriasis551
Kökcam et al. ()TurkeyNR17NR26.47 (12.2)NR11 vitiligo, 20 healthy controlsSCL-90-R; ZSDSAA more symptoms of depression than healthy controls, no difference with vitiligo201
Laitinen et al. ()Finland1987–20161762529.7 (NR)NRNoICD-9 and ICD-10 codes3.98% was diagnosed with depression54.172
Layegh et al. ()IranOctober 2005–May 200673NRNRNR78 AV, 62 psoriasis, 87 vitiligoBDI31.51% minor depression, 23.29% mild depression, 24.66% moderate depression, 20.55% severe depression551
Liu et al. ()USANR91 children, 292 adultsChild: 34.4%, adult: 27.9%Child: 10 (2.92), adult: 41 (15.3)NRNoPHQ-9On average mild symptoms of depression in children and adults20.832
Marahatta et al. ()NepalAugust 2015–July 20167553.329.40 (9.90)NRNoBDI66.7% depressive complaints. No relation to disease severity.45.832
Singam et al. ()USA2002–20125,605 hospitalized patients38.342.2 (NR)NRHospitalized patients without AA (N unknown)ICD-9 codesAA more mood disorders than controls451
Talaei et al. ()IranApril–July 20052433.3325.38 (8.32)NR24 healthy controlsSCL-90-RNo difference AA and controls701
Vallerand et al. ()GBNR6,86143.932.20 (13.50)NR6,137,342 healthy controlsRead codesAA higher chance of depression than controls601
Vélez-Muñiz et al. ()MexicoMarch 2017–February 201832 children, 94 adults41NR92.9% patchy AA, 3.2% AT, 1.6% ophiasis, 1.6% AUNoDSRS-C; HADSChildren: 6.3% symptoms of depression. Adults: 19.1% subclinical depression or anxiety, 34.1% no symptoms of anxiety or depression.502
Pediatric samples
Altunisik et al. ()TurkeyNR2729.611.9 (3.3)85.19% AA, 14.81% AU30 dermatology patientsK-SADS-PL; CDINo difference AA and controls. 14.8% symptoms of depression.651
Andreoli et al. ()Italy1997–2000176NRNRNRNoDiagnosis by psychologist10% dysthymia252
Bilgiç et al. ()TurkeyNR7455.4112.1 (2.8)NR65 healthy controlsCDIAA more symptoms of depression than controls651
Conic et al. ()USA20193,51044.726.2% <10 years, 73.8% 10–18 yearsNR8,310,710 patients without AADiagnoses in patient fileAA diagnosed with depression (2.6%) more often than controls (0.6%)101
Díaz-Atienza and Gurpegui ()SpainNR315212.2 (3.8)51.61% AA, 48.39% AU/AT23 epilepsy, 25 siblingsCDINo differences AA and epilepsy or siblings651
Erdoğan and Gür ()TurkeyOctober 2018–December 20193154.8312.54 (3.56)100% AA29 vitiligo, 30 healthy controlsRCADS-C; RCADS-PAA more depression than healthy controls, no difference vitiligo601
Ghanizadeh ()IranAugust 2004–November 200614NR11.66 (6.08)NRNoK-SADS-PL50% has diagnosis of depression502
Liakopoulou et al. ()GreeceNR3330.310.5 (0.3)NR30 patients from pediatricianCDINo difference AA and controls401
Reeve et al. ()USANR12NR11.5 (2.9)NRNoDICA-R; CDSNo heightened group average37.52
Adult samples
Aghaei et al. ()IranNR4044.835.2 (9.2)NR40 healthy controlsBDIAA more symptoms of depression than controls351
Alfani et al. ()ItalyNovember 2009–October 20107345.225.2 (9.2)61.7% AA,
26.0% AT,
12.3% AU
73 healthy controlsMMPI-2AA patients score above cut-off for depre ssion more often than controls351
Altinöz et al. ()TurkeySeptember 2011–October 2012305033.3 (8.9)NR30 urticaria, 39 healthy controlsHADSAA more symptoms of depression than healthy controls. No difference with urticaria.401
Annagur et al. ()TurkeyNR7365.7527.66 (7.79)100% AA78 healthy controlsSCL-90AA more symptoms of depression than controls351
Atış et al. ()TurkeyNR395933.5 (11.6)NR46 vitiligo, 46 healthy controlsHADSNo differences between AA, vitiligo and healthy controls201
Baghestani et al.
()
IranNR687235.4 (7.6)100% AA68 healthy controlsHAM-DAA more symptoms of depression than controls (OR = 4.48)601
Bain et al. ()UKNR3923.0743.15 (12.43)NR23 PsA; 26 healthy controlsHADS*Depressive symptoms in 18%. Less severe symptoms with higher SALT scores.301
Balieva et al. ()13 European countriesNovember 2011–February 20133333.342.8 (14.1)NR1,359 healthy controlsEQ-5D-3LAA 4 times higher chance of anxiety/depression than controls651
Bashir et al. ()PakistanJanuary–March 20073NRNRNRNoGHQ-12; interview1 person was diagnosed with depression41.672
Bukharia and Jain ()IndiaNR1004854% 15–30 years, 46% 31–50 yearsNR100 TE, 100 healthy controlsHAM-D23.68% AA and 33.33% TE with symptoms of depression452
Cakirca et al. ()TurkeyMarch–December 20173375.826.33 (6.08)NR33 healthy controlsHADSAA more depressive symptoms than controls301
Colon et al. ()USAApril 1985–October 1987312935.70 (10.23)74% AA, 23% AT, 42% AU3NoDISLifetime prevalence depression 39%, dysthymia 16%33.332
Conic et al. ()USA2005–201458431.535.54 (19.28)94.7% AA,
2.05% AT, 3.25% AU
172 SDDiagnoses in patient fileNo difference with control group351
Cordan Yazici et al. ()TurkeyNR4360.533.80 (10.02)95.35% AA, 4.65% AT53 healthy controlsHADSNo difference with controls251
Dai et al. ()TaiwanNR2,12344.831.39 (9.02)NR2,298 siblings, 9,192 healthy controlsICD-9 codes7.87% of AA with MDD diagnoses, 8.22 times higher chance than healthy control. A total of 2.55 higher chance than siblings.851
Devar ()IndiaNR30100NRNR30 TV, 30 healthy controlsBDIAA more symptoms of depression than healthy controls, no difference with TV501
Endo et al. ()JapanJune 2009–August 201012233.138.3 (16.5)NRNoCES-DAA more symptoms of depression than norm group56.254
Gallo et al. ()ItalyNR1637.545.95 (13.25)NRNoBSIAA more symptoms of depression than norm group39.295
Güleç et al. ()TurkeyMarch 2001–January 20025265.3831.53 (12.61)94.23% AA, 3.65% AU, 1.92% AT52 healthy controlsBDINo differences between AA and controls251
Gupta and Gupta ()USANR4524.4444.7 (11.6)NR72 AV, 146 AD, 217 psoriasisCRSDAA less depressive symptoms than AV and psoriasis, no difference with AD151
Karia et al. ()IndiaNR5066.0027.76 (NR)NR50 psoriasis, 50 healthy controlsDSM-IV-TR diagnosis18% AA depression diagnoses. More often than healthy controls, less often than psoriasis.601
Kim et al. ()South Korea2002–20137,70651.954.6% 20–39, 39.4% 40–59, 6.1% 60+NR30,824 people without AAICD-10 codesAA higher chance of depression than controls651
Kose et al. ()TurkeyNR1810021.3 (NR)NRNoBDIOn average subclinical depressive symptoms50.006
Macbeth et al. ()UKJanuary 2009–December 20185,43545.938.93 (14.35)NR21,470 healthy controlsDiagnoses in
patient file
AA higher chance of depression than controls801
Mirza et al. ()USA2002–20121380NRNRNoDiagnoses in
patient file
21.74% has depression diagnosis58.332
Pascual-Sánchez et al. ()SpainNR16045.1 (NR)100% AUNoBDIOn average subclinical depressive symptoms29.176
Rajoo et al. ()AustraliaNR83NR40.95 (13.24)NRNoDASS-2147.0% reported extreme depressive symptoms54.172
Ruiz-Doblado et al. ()SpainNR3215NRNRNoSCAN7.4% depression diagnosis, 7.4% previously diagnosed, but currently free of symptoms37.52
Şahiner et al. ()TurkeyAugust 2009–July 2010414932.9 (10.5)NR30 psoriasis, 50 healthy controlsBDIAA more depressive symptoms than healthy controls, no difference with psoriasis201
Sayar et al. ()TurkeyNR3110023.8 (2.5)NR40 healthy controlsBDIAA more symptoms of depression than controls551
Sellami et al. ()TunisiaMarch–July 2010504832.92 (11.81)NR50 healthy controlsHADSAA more symptoms of depression than controls451
Senna et al. ()USAJanuary 2011–December 201868,1213940.3 (17.8)98.1% AA,
1.3% AT,
0.6% AU
NoICD-9 and ICD-10 codes9.5% had depression diagnosis45.832
Sorour et al. ()EgyptNR20858.65NRNR1,042 dermatology patientsDSM-5 interview19.71% of AA had diagnosis of depression. 24.33% in psoriasis, 55.34% acne vulgaris, 31.47% vitiligo, 43.64% urticarial, and 43.63% in atopic dermatitis551
Tan et al. ()ChinaDecember 2012–August
2013
1685034.5 (11.5)88.1% AA, 11.9% AT/AU100 healthy
controls
SCL-90-RAA more symptoms of depression than controls41.674
Titeca et al. ()13 European countries37NRNRNRNR1,359 healthy controls, 20 AGAHADSAA more symptoms of depression than healthy controls701
Tzur Bitan et al.
()
Israel201841,05562.939.97 (13.61)NR41,055 healthy controlsICD-9 codesAA diagnosed with depression more often than controls801
Willemsen et al.
()
BelgiumSeptember 2006–August 2009212441.95 (13.79)33% patchy AA, 14% ophiasis, 29% AT, 24% AUNoSCL-90AA more symptoms of depression than norm group54.176
Willemsen et al.
()
BelgiumApril 1999–April 20042835.7133.4 (NR)21.43% AA, 21.43% ophiasis, 28.57% AU, 3.57% ATNoSCL-90AA more symptoms of depression than norm group50.006
Yoon et al. ()South KoreaJanuary 2015–February 20161,20352.1239.45 (12.21)NRNoBDI40.9% depressive symptoms. Women more often than men, more symptoms with more severe AA.41.672
Yu et al. ()ChinaOctober 2013–December 201413041.531.78 (10.34)NR212 AGAZSDSNo differences between AA and AGA701

Results for depression.

*This questionnaire was not administered to the control group.

AD, atopic dermatitis; AGA, alopecia androgenetica; AV, acne vulgaris; PsA, psoriatic arthritis; SD, seborrheic dermatitis; TE, telogen effluvium; TV, tinea versicolor.

1As measured by the NIH Quality Assessment of Case-Control studies.

2As measured by the NIH Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies.

3Some patients had multiple episodes, with different forms of alopecia. Hence, the total is higher than 100%.

4As measured by the QAVALS ().

5As measured by the NIH Quality Assessment of Controlled Intervention Studies.

6As measured by the NIH Quality Assessment Tool for Before-After (Pre-Post) Studies with no Control Group.

Children and adults

In terms of diagnoses of depression, 4.3% of the visits to a psychologist by people with AA were related to depression (). The point prevalence varied from 2.9% () to 3.98% (). Different studies reported that people with AA were diagnosed with depressive disorders (, ) and mood disorders in general () significantly more often than healthy controls.

When looking at depressive symptoms, results concerning comparisons to healthy controls are mixed. Two studies, with a combined sample size of 60, reported more depressive symptoms in people with AA (, ), while one study did not find any significant differences (n = 24) ().

Two studies compared people with AA to people with another (dermatological) condition. They did not find significant differences concerning the amount of depressive symptoms when comparing to people with psoriasis or vitiligo () or people with acne vulgaris, psoriasis or vitiligo ().

Four studies (n = 657) did not use a control group. They found little to no depressive symptoms in 31.5% (), 33.3% (), and 34.1% () of people with AA. According to these studies around 60–65% of people with AA experience at least moderate depressive symptoms.

Children

Three studies (n = 3,700) investigated depressive disorders. A small study of 14 children found 50% of the children to be eligible for a diagnosis of depressive disorder (). Bigger studies reported that 10% of the children were diagnosed with dysthymia () and that children with AA were diagnosed with a depressive disorder more often than other patients ().

Three studies (n = 136) investigated symptoms of depression in comparison to healthy controls. Two studies found more depressive symptoms in children with AA (, ), while one study did not find a significant difference when comparing to unaffected siblings ().

Four studies compared children with AA to children with a different (dermatological) condition. They did not find a difference in depressive symptoms when comparing children with AA to children with other dermatological conditions (), epilepsy (), vitiligo (), and pediatric patients in general ().

One study with 14 children with AA did not use a control group. This study did not find a heightened group average for depressive symptoms ().

Adults

Several studies investigated the prevalence of depressive disorders in adults with AA. One study, conducted in the late 1990s, found a lifetime prevalence of 39% for depression and 16% for dysthymia (). Estimates for point prevalence range from 7.4% (), 9.5% (), 18% (), 21.74% () to 55.29% (). The largest and most recent study found a point prevalence of 9.5% (). Furthermore, adults with AA have a higher chance of being diagnosed with a depressive disorder than healthy controls (, , , ). One study did not find any difference in the number of diagnoses (). There were no differences in the number of diagnoses when comparing to adults with psoriasis or vitiligo () or seborrheic dermatitis ().

Fifteen studies compared adults with AA to healthy controls on the amount of depressive symptoms. These studies were analyzed in a meta-analysis. The results are shown in Figure 3. A total of 749 adults with AA and 724 healthy controls were analyzed. Adults with AA reported significantly more depressive symptoms than the control group (g = 0.73, 95% CI [0.47, 0.98], p < 0.001), with a medium to large effect. There was considerable heterogeneity (I2 = 78.5%, 95% CI [65.2, 86.8], τ2 = 0.20, 95% CI [0.08, 0.62]). Visual inspection of the funnel plot showed no signs of publication bias and Egger’s test was not significant [t(13) = 0.80, p = 0.438]. Thirteen studies without missing data were included in a meta-regression. The model explained very little variance in the effect sizes (R2 = 1.21%) and residual heterogeneity was high (I2 = 77.27%). Mean age (g = 0.04, p = 0.289, 95% CI [−0.03 to 0.12]), percentage male (g = 0.01, p = 0.168, 95% CI [−0.01 to 0.03]) and quality score (g = 0.01, p = 0.265, 95% CI [−0.01 to 0.03]) did not influence study effect sizes.

FIGURE 3

Nine studies used a control group of adults with a different (dermatological) condition to assess the amount of depressive symptoms. The vast majority of the studies did not find any significant differences. For instance, no differences were found when comparing to chronic urticaria (), vitiligo (), telogen effluvium (), tinea versicolor (), atopic dermatitis (), psoriasis (), and alopecia androgenetica (, ). One study found that adults with AA reported less depressive symptoms than adults with acne vulgaris or psoriasis ().

Studies without a control group found that people with AA (n = 183) reported more symptoms of depression than a norm group (, , , ). On average, they reported subclinical symptoms (, ). Estimates of the prevalence rates of people with depressive symptoms were 47.0% () and 40.9% ().

Quality of life

The results for QoL are shown in Table 3.

TABLE 3

ReferencesCountryYearN% maleAge (M, SD)% AA, AT, AUControlsMeasuresConclusionsQuality score
(%)
Pediatric and adult samples
Ghajarzadeh et al. ()IranJanuary 2009–January 20101006923.02 (33.4)NR100 psoriasis, 100 vitiligoDLQI; SF-36AA better QoL than psoriasis. No difference with vitiligo. On average moderate effect on QoL.551
Liu et al. ()USANR91 children, 292 adultsChild: 34.4%; adult: 27.9%Child: 10 (2.92); adult: 41 (15.3)NRNoCDLQI; DLQI; FDLQIChildren, adults, and family members have moderate effect on QoL. Worse QoL related to more depressive symptoms.20.832
Park et al. ()South KoreaNR4027.530.0% 10–19 years, 17.5% 20–29, 17.5% 30–39, 17.5% 40–49, 17.5% 50+NRNoSkindex-29Symptoms, emotions, and total score very little impairment. Functioning mild impairment37.52
Vélez-Muñiz et al. ()MexicoMarch 2017–February 201832 children, 94 adults41NR92.9% patchy AA, 3.2% AT, 1.6% ophiasis, 1.6% AUNoCDLQI; DLQIChildren small impairment on QoL. Adults moderate effect. No differences for gender, disease duration, and disease severity.502
Pediatric samples
Bilgiç et al. ()TurkeyNR7455.4112.1 (2.8)NR65 healthy controlsPedsQL-P; PedsQL-CLess QoL on child and parent reports. Less psychosocial QoL on parent reports.651
Erdoğan and Gür ()TurkeyOctober 2018–December 20193154.8312.54 (3.56)100% AA30 healthy controls; 29 vitiligoCDLQIAA worse QoL than vitiligo601
Putterman et al. ()USAApril 2017–July 201815343.7911.0 (4.8)NRNoCDLQI; FDLQI; QLCCDQOn average small effect on child QoL, moderate effect for family members. Worse QoL for more disease severity and worse emotional QoL for higher age.502
Adult samples
Abedini et al. ()IranOctober 2013–October 201417664.2331.39 (9.05)NRNoDLQIPatients with mild AA moderate effect on QoL, patients with severe AA very large effect on QoL.
Patients with more severe AA reported worse QoL on: symptoms and feelings, daily activities, leisure, personal relationships, work and school, treatment, and the total score.
502
Abideen et al. ()IndiaNR606533.9 (9.3)NRNoDLQI30% no effect on QoL, 55% small effect, 6.7% moderate effect, 8.3% very large effect20.832
Al-Mutairi and Eldin ()KuwaitAugust 2002–July 20092,962 (300 for DLQI)65.0258.03% between 21 and 40 yearsNR300 healthy
controls
DLQINo difference between males and females or disease duration. Worse QOL for more severe alopecia401
Andersen et al. ()DenmarkNR1,4943351.3 (16.0)NRNoDLQI; EQ-5D-5L75% no effect on QoL. On average small effect.41.672
Atış et al. ()TurkeyNR395933.5 (11.6)NR46 healthy controls,
46 vitiligo
DLQIOn average moderate effect, no difference with vitiligo201
Balieva et al. ()13 European countriesNovember 2011–February 20133333.342.8 (14.1)NR1,359 healthy controlsEQ-5D-3LNo significant difference for mobility, self-care, activity, and pain/discomfort651
de Hollanda et al. ()BrazilJanuary 2011–October 20123737.8435.89 (11.59)NR49 healthy controlsSF-36AA score lower on mental health, role emotional and social functioning. No differences for vitality, bodily pain, general health, physical functioning, and role physical.551
Dubois et al. ()FranceNR6035.0040.1 (15.2)NRDermatologic conditions and healthy controls from literatureSF-36; SkindexLower scores on role-physical, general health, vitality, social functioning, role-emotional, and mental health41.672
Endo et al. ()JapanJune 2009–August 201012233.138.3 (16.5)NRNoSF-8Average scores on physical and mental functioning56.253
Essa et al. ()EgyptJanuary–June 201517NRNRNR500 healthy
controls
Skindex-16No difference AA and dermatological conditions. AA worse QoL than healthy controls.41.673
Fayed et al. ()EgyptFebruary 2015–January 2016417826.68 (4.49)NRNoDLQI0% no effect on QoL, 4.9% small effect, 29.3% mild effect, 29.3% moderate effect, 36.6% very large effect504
Gonul et al. ()TurkeyNR5655.429.34 (8.13)92.86% AA, 7.14% AT82 AGAHairdex; TQLAA better QoL than AGA on total, emotions, functions, symptoms, and self-confidence. No difference on stigmatization and TQL.601
Güleç et al. ()TurkeyMarch 2001–January 20025265.3831.53 (12.61)94.23% AA, 3.65% AU, 1.92% AT52 healthy controlsSF-36AA worse QoL on vitality and mental health than controls. AA higher QoL than healthy controls on social functioning.251
Han et al. ()USAAugust 2018–November 201914126.243.3 (15.6)76.6% AA, 13.5% AU, 9.9% ATNoAASISMore stress is related to lower QoL41.672
Jankovic et al. ()SerbiaApril 2012–June 20136026.737.35 (14.26)NR110 psoriasis,
66 AD; 140 OM
DLQI; SF-36; Skindex-29AA better QoL than psoriasis. Partially better QoL than AD and OM.41.672
Karia et al. ()IndiaNR506627.76 (NR)NR50 psoriasis, 50 healthy controlsWHOQOL-BREFAA higher QoL than psoriasis and healthy controls601
Lai et al. ()AustraliaNR3619.441 (14.5)41.7% patchy,
25.0% AT,
33.3% AU
NoAASIS; aQoL-8DNo difference with norm group755
Liu et al. ()USANR3053.338.00 (21.80)NRNoSkindex-16No difference between males and females29.174
Masmoudi et al. ()TunisiaMarch–July 2010504832.92 (11.81)NR50 healthy controlsSF-36AA worse scores on mental health, role emotional, social functioning, general health and total mean score. No significant differences for physical functioning, role physical and bodily pain. No relation QoL and disease severity.551
Nasimi et al. ()IranAugust 2017–August 20181006529.24 (8.31)NRNoAA-QLI; DLQIOn average very large effect. Males better QoL than females.20.833
Nijsten et al. ()ItalyNR46NRNRNR151 AV;
76 psoriasis,
54 SD; 27 vitiligo; 100 nevi
Skindex-2917.4% in worst category for total, 21.7% in worst category for emotions551
Öztürkcan et al. ()TurkeyJanuary–February 20043NRNRNR16 CD, 6 psoriasis, 3 urticaria, 16 TP; 35 AVDLQIOn average small effect41.673
Qi et al. ()ChinaJanuary 2010–July 20126985038.8 (12.0)82.5% patchy, 17.5% AT/AUNoDLQIOn average moderate effect on QoL54.172
Reid et al. ()USAMarch–November 2009230NRNR33 TE, 41 AGA, 7 unknown alopeciaSkindex-16No differences on QoL for different alopecia types551
Russo et al. ()ItalySeptember 2016–September 20172733.337.55 (10.37)NR80 AGA, 36 TEDLQIFemales worse QoL than men. No differences with AGA or TE.501
Sampogna et al. ()ItalyNR5NRNRNRDermatological conditionsScalpdex;
Skindex-29
Average impact on symptoms, emotions, and functioning551
Sanclemente et al. ()ColombiaNR11NRNRNRDermatological conditionsSkindex-29Median score indicates moderate effect on total score, symptoms, emotions, and functioning601
Senna et al. ()USA201925949.439.1 (13.6)NRNoSkindex-16Worse QoL for longer disease duration, higher disease severity, and females41.672
Temel et al. ()TurkeyNR504630.92 (10.92)84% AA, 6% AT,
10% AU
50 AV; 50 vitiligoDLQIOn average moderate effect on QoL. No difference with AV or vitiligo.401
Titeca et al. ()13 European
countries
37NRNRNRNR1,359 healthy controls, 20 AGADLQIWorse QoL than AGA701
Willemse et al. ()Multiple countriesNR2431137.9 (13.0)NRNoDLQIOn average no effect on QoL. No differences for gender or disease severity. Worse QoL for shorter disease duration.54.172
Willemsen et al. ()BelgiumSeptember 2006–August 2009212441.95 (13.79)33% patchy, 14% ophiasis, 29% AT, 24% AUNoSF-36; Skindex-17SF-36: average physical functioning, below average mental functioning in comparison to norm group
Skindex: moderate effect on psychosocial functioning, less physical symptoms in comparison to norm group
54.174
Yoon et al. ()South
Korea
January 2015–February 20161,20352.1239.45
(12.21)
NRNoSkindex-2930.3% impaired QoL, 9.9% severely impaired QoL. Females worse QoL than men.41.672
Yu et al. ()ChinaOctober 2013–December 201413041.531.78
(10.34)
NR212 AGADLQIAA worse QoL than AGA on total, symptoms and feelings, daily activities, and leisure. AA scored higher on treatment. No differences for work and school and personal relationships.701
Unknown age
Jun et al. ()South
Korea
March 2012–February 2017161NRNRNR380 AGASkindex-29AA worse scores than AGA on functioning and better scores on symptoms. No differences for emotions and total score.451
Shi et al. ()USANR5322750% older
than 40
52% AU/AT, 46.6% with 100% hair lossNoDLQI; Skindex-16On average moderate effect on QoL. Moderate effect on emotions, symptoms, and functioning20.832

Results for quality of life.

AD, atopic dermatitis; AGA, alopecia androgenetica; AV, acne vulgaris; CD, contact dermatitis; HS, hidradenitis suppurativa; NF1, neurofibromatosis, type 1; NR, not reported; OM, onychomycosis; SD, seborrheic dermatitis; TE, telogen effluvium; TP, tinea pedis.

1As measured by the NIH Quality Assessment of Case-Control studies.

2As measured by the NIH Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies.

3As measured by the QAVALS.

4As measured by the NIH Quality Assessment Tool for Before-After (Pre-Post) Studies with no Control Group.

5As measured by the NIH Quality Assessment of Controlled Intervention Studies.

Children and adults

People with AA reported worse QoL than people with psoriasis, but there was no difference with vitiligo (). On average, people reported a small (, ) or moderate (, , ) impact on their QoL.

Children

Children with AA reported more impaired QoL than healthy controls (, ). In a study without a control group children with AA reported a small effect on their QoL ().

Adults

Fourteen studies used the Dermatology Life Quality Index (DLQI) to assess disease-specific QoL in 3,978 adults with AA (, , , , , , , , , , , ). These studies were included in a meta-analysis, shown in Figure 4. The total scores of the DLQI can be interpreted as follows: 0–1 = no effect on patient’s life, 2–5 = small effect, 6–10 = moderate effect, 11–20 = very large effect, 21–30 extremely large effect (). Results from the meta-analysis showed that people with AA reported a weighted average of 6.67 (95% CI [5.54, 7.81]), which is a moderate effect. However, there was very high heterogeneity amongst studies (I2 = 98.9%, 95% CI [98.5, 99.0], τ2 = 4.25, 95% CI [2.07, 12.29], p < 0.001). Results should thus be interpreted with extreme caution. Meta-regressions were run on 11 studies without missing data. The model explained 62.89% of the variance in the data, but still included a substantial amount of heterogeneity (I2 = 89.56%). Mean age was negatively related to DLQI scores (g = −0.28, p = < 0.001, 95% CI [−0.44 to −0.12]). Studies with a higher mean age had lower DLQI scores and thus less impaired QoL. The same was true for the quality ratings of studies (g = −0.08, p = 0.009, 95% CI [−0.13 to −0.02]), where studies with a lower quality rating reported higher DLQI levels. The percentage male (g = −0.04, p = 0.200, 95% CI [−0.11 to 0.02]) was not significantly related to DLQI scores.

FIGURE 4

As two studies did not provide clear data on their sample and may have included children (, ), we conducted a sensitivity analysis to assess whether this influenced the results. Twelve studies (, , , , , , , , , , , ) with 3,346 people were included. The mean DLQI was unchanged (M = 6.68, 95% CI [5.33, 8.02]) and heterogeneity remained high (I2 = 98.8%, 95% [98.4, 99.0], τ2 = 5.14, 95% [2.38, 16.35]).

Five studies compared 188 adults with AA to healthy controls (, , , , ). There seemed to be no difference on physical functioning (, , ). However, adults with AA had more impaired mental (, , ) and overall () functioning.

Thirteen studies compared adults with AA to adults with another (dermatological) diagnosis (, , , , , , , ) and found very mixed results. On the one hand, no differences were found when comparing to adults with vitiligo (, ), alopecia androgenetica and telogen effluvium () and acne vulgaris (). On the other hand, people with AA reported better QoL than people with psoriasis () or alopecia androgenetica (). In yet other studies, people with AA reported worse QoL than people with alopecia androgenetica (, ).

Unknown samples

Two studies did not report whether they studied children or adults (, ). They found a moderate impact on QoL (). When comparing to alopecia androgenetica, people with AA scored higher on subscales functioning and lower on symptoms (). No differences were found for emotions and total score.

Discussion

In this systematic review and meta-analysis we aimed to provide an overview of the current literature on anxiety, depression and QoL in people with AA. Results showed that people with AA experienced adverse psychosocial consequences in all three domains. Results also point to more diagnoses of anxiety and depression, as well as more symptoms of anxiety and depression, compared to healthy controls.

Meta-analytic results showed that people with AA experience more symptoms of anxiety and depression than healthy controls. With a medium to large effect for both meta-analyses, we can conclude that this constitutes a clinically relevant effect. Our results were unable to shed light on which patients are at risk for experiencing symptoms of anxiety or depression as average age, percentage male and quality of the studies did not explain variance in the effect sizes. While the same studies were included in both meta-analyses, we found high heterogeneity for depression but not for anxiety. The range for effect sizes is much larger in depression than anxiety, however it is unclear where this originates from.

Meta-analytic results also showed that people with AA experience a moderate impact on their QoL. We were able to include around 3,800 patients in this meta-analysis, which makes it likely that our results generalize to other adults with AA. However, as we found very high heterogeneity, the moderate impact of AA on QoL is unlikely to be true for everyone with AA. Subgroups may exist based on variables that were not studied in the current meta-analysis, such as severity of disease, medication use, duration of disease or other variables.

Results concerning people with AA compared to people with other dermatological diagnoses were mixed for anxiety, depression, and QoL. However, the majority of the studies seems to point to people with AA experiencing the same amount of anxiety, depression, and impairment of QoL as people with other diagnoses. So, even though patients with AA do not experience physical symptoms that people with other dermatological diagnoses may experience, such as pain or itching (), their QoL is comparable.

While we did not directly compare age groups, some observations can be noted. Firstly, for all three domains more studies were included for adults than for children. Hence, conclusions for adults can be made with more certainty. Both for anxiety and depression results of children with AA compared to healthy controls were mixed, while results for adults showed that adults with AA experienced more symptoms of anxiety and depression than healthy controls. As the mean age of the studies with children was 11.85, it is possible that symptoms of anxiety do not develop before puberty or adulthood, when appearance and peer relations become more important. This is corroborated by other studies showing more appearance-related distress in puberty (). For QoL only three studies were found for children, so direct comparisons are hard to make.

Overall, our results are in line with a previous meta-analysis finding positive associations between AA and experiencing (symptoms of) anxiety or depression (). In addition, we have shown that adults with AA experience more symptoms of anxiety and depression than healthy controls. Our results concerning QoL are also in line with Toussi et al. (), who found diminished QoL in children and adults with AA. More specifically, we found diminished QoL in mental wellbeing but not necessarily in physical wellbeing. This is slightly unsurprising, as AA is associated with little physical impairment. Despite this, qualitative studies have shown that losing one’s hair has a considerable impact on mental health (, ).

It is also noteworthy that many studies included patients that were referred to a dermatologist. This could introduce a selection bias, where those who experience less psychological complaints are less likely to visit a dermatologist. However, large studies on primary care databases also reiterate that patients with alopecia are diagnosed with anxiety and depression more often than patients without alopecia (, ), with a diagnosis of depression preceding the diagnosis of AA for some patients ().

This systematic review also has some strengths and limitations. A particular strength is the thorough literature search conducted. A formal search was created by a librarian, yielding 1,249 unique records. With this thorough search it is highly likely that no relevant articles were missed in the search process.

Despite the thorough literature search, we could only include a limited amount of studies in a quantitative analysis. For instance, we did not have enough data to disentangle psychological wellbeing in separate forms of AA (i.e., areata, universalis, or totalis) or how psychological wellbeing was related to disease severity or disease duration. Another limitation is that the included studies did not look at the remitting and relapsing course of AA specifically. Most studies were cross-sectional and longitudinal studies were often designed to look at a medical or psychological intervention. Qualitative research has highlighted that the unpredictable nature of AA can lead to feelings of anger or stress (), but this has not been studied quantitatively. Hence, it remains unclear how the remitting and relapsing course of AA influences psychological wellbeing. A third limitation is that the included studies did not provide data on medication use. Inclusion criteria were often unclear when it came to participants’ medication use and medication use was often omitted from reporting in the outcome data. We do know that medical treatments often fail to provide sustained hair regrowth and may lead to substantial side effects (). Hence, it remains unclear whether the pros of medication use outweigh the cons.

Another limitation is that the goal of the included studies did not always line up with the goal of the current systematic review. For instance, this review also included questionnaire validation studies () and baseline data of randomized controlled trials (). The data was therefore approached in a different manner than the original authors intended. This may impede the strength of the current conclusions. However, as the intention of this review was to provide a thorough overview of the current literature, minimal limitations were set for the inclusion of different types of papers.

Based on these limitations, future studies should aim to study AA longitudinally and investigate the influence of disease severity, disease duration, disease status (inactive, remission, or relapse) and medication use on psychological wellbeing. These results would provide useful insights on potential at-risk groups in need of referral to psychological care.

The results of the current study highlight the impact of AA on psychological wellbeing. Clinicians treating people with AA should therefore be aware of the impact and refer to psychological care if needed. This could be accomplished through regular screening, for instance as part of value-based healthcare (), or through the physician checking in on people’s mental health during outpatient clinic appointments.

Conclusion

In summary, we have shown that living with AA has important consequences for psychological wellbeing. People with AA experienced worse psychological outcomes than healthy controls and comparable psychological outcomes compared to people with other dermatological diagnoses. Important challenges lay ahead on how to treat AA, both psychologically as well as medically.

Statements

Data availability statement

The datasets analyzed as well as analysis scripts for this study can be found on the Open Science Framework (OSF): https://osf.io/fxt7p/.

Author contributions

MD: conceptualization, methodology, formal analysis, writing—original draft, visualization, and project administration. KM: formal analysis and writing—review and editing. JK-O, JO, and SP: writing—review and editing. All authors contributed to the article and approved the submitted version.

Funding

This work was funded by the Dutch Alopecia Association (grant number n/a) with unrestricted funding.

Acknowledgments

The authors thank Elise Krabbendam from the Erasmus MC Medical Library for developing the search strategies.

Conflict of interest

Author JK-O is a board member of the Dutch Alopecia Association (volunteer position). The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmed.2022.1054898/full#supplementary-material

References

Summary

Keywords

alopecia, alopecia areata, psychosocial functioning, anxiety, depression, quality of life, meta-analysis

Citation

van Dalen M, Muller KS, Kasperkovitz-Oosterloo JM, Okkerse JME and Pasmans SGMA (2022) Anxiety, depression, and quality of life in children and adults with alopecia areata: A systematic review and meta-analysis. Front. Med. 9:1054898. doi: 10.3389/fmed.2022.1054898

Received

27 September 2022

Accepted

11 November 2022

Published

29 November 2022

Volume

9 - 2022

Edited by

Andrew Robert Thompson, Cardiff and Vale University Health Board, United Kingdom

Reviewed by

Alina Constantin, British Association of Dermatologists, United Kingdom; Andrew Messenger, Sheffield Teaching Hospitals NHS Foundation Trust, United Kingdom

Updates

Copyright

*Correspondence: Marije van Dalen,

This article was submitted to Dermatology, a section of the journal Frontiers in Medicine

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Outline

Figures

Cite article

Copy to clipboard


Export citation file


Share article

Article metrics