Masking and metacognition: depression and anxiety rates of individuals with hearing loss

During the COVID- 19 pandemic the Centers for Disease Control and Prevention (CDC) and other health organizations recommended that individuals stay at least 6 feet apart and wear masks to prevent illness transmission (Howard et al., 2021; Xu et al., 2020). Though necessary for physical safety, these regulations added a heavy burden to those who were reliant on facial expressions and lip reading for communication, particularly the d/Deaf and hard of hearing population (DHH) (Grote & Izagaren, 2020; Hulzen & Fabry, 2020; Poon & Jenstad, 2022; Roberson et al., 2012; Schlögl & Jones, 2020). DHH rely heavily on lipreading, visual cues, and facial expressions to communicate in their daily lives. Because masks cover the lips and part of the face, DHH experienced a sharp decline in their ability to navigate personal, social, employment, and professional interactions during the pandemic (Grote & Izagaren, 2020; Naylor et al., 2020; Ritter et al., 2022). The inability to communicate successfully in personal and professional interactions diminishes wellbeing and mental health, causing DHH to be more vulnerable to mental health distress (Cruice, 2010; Fellinger et al., 2012; Kvam et al., 2007). In addition, masking interrupts speech signals, causing speech to be less intelligible (Atcherson et al., 2017; Ritter et al., 2022). In a time already marked by high stress, anxiety, and isolation, the preferred (and often mandated) use of masks unintentionally had a negative impact on DHH.

This study quantifies the mental health toll that this diminished capacity to communicate created for DHH in masked conversations. Mental health providers are bounded by ethics and practice protocol that relies on communication to enhance clients’ quality of life, yet their masking may have contributed to decreased quality of life for a percentage of patients/clients. This article sets out to document some of those decreases and provides ideas for best practice in places/spaces/times when masking is needed.

Ongoing research studies assessing depression and anxiety rates for the general population during the pandemic suggest a tripling for both depression and anxiety prevalence (Ettman et al., 2020; Twenge & Joiner, 2020). A systematic review conducted by Lakhan and colleagues (2020) examined the literature between December 2019 and June 2020, identifying prevalence rates of depression as 20% and anxiety as 35% in the general population (N = 113, 285). In a longitudinal study that identified anxiety and depression rates in the general population in six different waves, two waves (5 and 6) in January and April 2021, most closely aligned with the dates of this current study. Wave 5 (N = 7,115) demonstrated rates of depression as 25.5% and anxiety as 23.6%. Wave 6 (N = 7,204) found rates of depression at 23.8% and anxiety as 22.1% (Hajek et al., 2022). In the DHH population, studies are suggesting that face masks hampered communication and correlated with higher levels of depression, anxiety, and stress (Zivkovic Marinkov et al., 2022).

Prior to the pandemic, most studies that assess the prevalence of depression with a hearing loss sample draw from an elderly only (> 70 years of age) sample. However, one study includes a younger sample (age 12–19) along with a sample of age 70 + and found that depression rates for DHH were 11.4% in comparison to 4.9% for those with normal hearing (Li et al., 2014). Using a global sample evaluating anxiety prevalence, Vancampfort et al. (2017) found higher rates of self-reported anxiety (19.1%) from DHH participants than from hearing participants (8.7%).

Metacognition

The stress of communication can affect the lens in which DHH view themselves and their environment. There are several explanations used to understand these complex relationships. Emotional control and planning are developmental tasks that require access to language and auditory cues, thus lack of exposure to auditory signals affects the social and emotional development of DHH (Fellinger et al., 2012). An easily understood example of this is explained by Fellinger et al. (2012): when families lack successful communication strategies with DHH children, the children are more likely to have an impaired quality of life and less social/emotional resources. By looking at this through a metacognition lens, we can better examine how people think about their thinking– which is a key factor to understand when assessing and understanding mental health outcomes (Nordahl et al., 2019). Metacognition is a range of cognitive processes that includes self-awareness and self-regulation, which are essential for learning and problem solving. It refers to being aware of one’s own cognition and the feelings associated with cognitions (Hasson-Ohayon et al., 2020). This awareness can lead to strategies such as planning, monitoring, and evaluating, which aid in regulating both cognitive and emotional experiences (Rhodes, 2019). Studies that examine DHH samples using metacognition have found high associations with cognition and mental health disorders within this population (Booth-Butterfield & Booth-Butterfield, 1994; Dehnabi et al., 2017).

Self-regulatory executive function model

Compelling evidence indicates that psychological distress is associated with how individuals perceive situations (Bailey & Wells, 2015). The self-regulatory executive function model (S-REF) suggests that individuals experience a heightened vulnerability to mental health disorders when they are exposed to stressful situations, particularly as a result of metacognition (Wells & Matthews, 1996). When metacognition is ruminative and negative, it leads to collective responses known as the Cognitive Attentional Syndrome (CAS) (Nordahl et al., 2019; Wells & Matthews, 1996). CAS keeps an individual in an internal threat response, causing them to not see their current stressful situation from a logical or problem-solving perspective. Instead, it causes perpetual worry and decreases cognitive efficiency, as well as impairing a person’s ability to self-regulate (Nordahl et al., 2019). Metacognition has been tested during the pandemic in relationship to anxiety from perceived danger and fear (Aydın et al., 2022), cyberchondria (excessive concerns about physical health due to repeated internet searches) (Seyed Hashemi et al., 2020), and health anxiety (Akbari et al., 2021). Each of these studies found increased anxiety and depression from negative metacognitive beliefs: worrying about worry and lack of control to internal threats. It is not known how metacognition has been impacted by masking in the DHH population.

Purpose of the present study

Based on previous literature and researcher observations, the research questions are:

  1. 1.

    What is the difference in comprehension between masked and unmasked listening situations?

  2. 2.

    What are the rates of depression and anxiety for DHH during the COVID- 19 pandemic?

  3. 3.

    What is the relationship between masking, metacognition, and mental health factors?

Method

A retrospective descriptive exploratory study was conducted to address the research questions and was driven by the S-REF theory. Utilizing an anonymous web-based survey, the research team recruited from nationwide hearing loss support groups on Facebook, from the Speech and Hearing Clinic at the University of Tennessee, and from the personal and professional contacts of the researchers. To ensure appropriate filtering of Facebook groups, the following steps were taken: (1) Groups were identified through targeted searches using keywords such as hearing loss and hearing loss support; (2) Groups required explicit administrator consent for participation; (3) Recruitment posts targeted group members who met the inclusion criteria. Recruitment criteria included: being an adult (18 years or older), having any level of hearing loss, and currently experiencing masked communications. Consent occurred through acknowledging the consent statement and moving forward in the survey. This study received research funds from a student faculty award from the Graduate School at the University of Tennessee. The sample size was determined by a G*power analysis, which determined the minimum sample size should be N = 107. This study was approved by the University of Tennessee Institutional Review Board (#20–06161-XM).

Participants

190 participants responded to the invitation to participate in this study. As an incentive is helpful when recruiting, especially with hard-to-reach populations, a communication kit was mailed to participants who chose to share their address following the completion of the study. The communication kit consisted of a face shield and clear mask (both to be given to a communication partner), an ear saver (to help decrease the burden on the ears from masks and hearing aids), and a button to wear on their masks. The button stated, “I can’t lip read through your mask.” This kit was designed to help facilitate communication when masking.

Survey Instrument

The survey consisted of demographic questions, hearing loss related questions, and several standardized measures of depression and anxiety (detailed below).

Demographic

Participants were asked about their race, gender, age, marital status, annual income, employment status, and education level.

Hearing loss

Participants were asked to self-identify their level of hearing loss, use of hearing aids or cochlear implants, use of sign language, and how often they have masked communications.

Metacognition

To measure the metacognition capacity of the participants, the Metacognition Questionnaire- 30 (MCQ- 30) was selected, which also has a relationship with trait anxiety. The MCQ- 30 assesses unhelpful metacognitions that can influence trait anxiety (Wells & Cartwright-Hatton, 2004). Participants are asked, using a Likert scale of one (do not agree) to four (agree very much), questions such as “I am constantly aware of my thinking” and “my worrying is dangerous for me.” Responses are scored from 30 to 120. The higher the score indicates the more unhelpful the participant’s cognitions are. The MCQ demonstrates good internal consistency with Cronbach’s alpha in the range of 0.72 to 0.93 (Wells & Cartwright-Hatton, 2004). The current study has Cronbach’s alpha of 0.91, demonstrating excellent internal consistency.

Depression

The CES-D (Center for Epidemiologic Studies Depression Scale) asks participants how often they have felt depression related factors in the past week (Radloff, 1977). Using a Likert scale from rarely (1) or none of the time to most or all of the time (4), participants ranked the frequency of questions such as “I was bothered by things that usually don’t bother me” and “I felt lonely.” The CES-D indicates a total score of 16 or higher is considered Major Depression Disorder (Santor et al., 1995). Internal consistency of the CES-D has been shown to be excellent with the Cronbach’s alpha ranging from 0.85 to 0.90 (Radloff, 1977). In this current study the Cronbach’s alpha for the CES-D was excellent with α = 0.94.

In addition, the Patient Health Questionnaire (PHQ- 2) screens participants for risk of having major depressive disorder (Löwe et al., 2005; Staples et al., 2019). Scores of 3 or higher on this two-question scale indicate higher risk for major depression disorder. Two questions are asked regarding how often the participant has been bothered in the last two weeks: “little interest or pleasure in doing things” and “Feeling down, depressed or hopeless.” Responses are on a Likert scale from one (not at all) to four (nearly every day). Internal consistency for the PHQ- 2 is good with Cronbach’s alpha demonstrated in previous studies at 0.83 (Löwe et al., 2005). In the current study, the internal consistency was 0.86 demonstrating good internal consistency.

Anxiety

The STAI-T20 (State Trait Anxiety Inventory) measures anxiety in clinical and research settings (Spielberger, 1983). It specifically looks at trait anxiety, which is situational anxiety. Participants are asked to respond to 20 statements on a Likert scale from one (not at all) to four (very much so). Statements range from “I feel calm” to “I feel jittery.” Scores at or above 40 suggest the presence of Generalized Anxiety Disorder. The psychometric properties of the STAI-T are good with internal consistency ranging from 0.86 to 0.95 (Spielberger et al., 1983). In the current study, the Cronbach’s alpha was 0.96, demonstrating excellent internal consistency.

The GAD- 2 (Generalized Anxiety Disorder) screens participants and identifies the risk of having generalized anxiety disorder (Plummer et al., 2016). Scores of 3 or higher on this two-question scale deem the participant at risk of GAD. Two questions are asked regarding how often the participant has been bothered in the last two weeks: “feeling nervous, anxious or on edge” and “not being able to stop or control worrying.” Responses are on a Likert scale from one (not at all) to four (nearly every day). The GAD- 2 has good internal consistency with α = 0.81 (Staples et al., 2019). For this study, the GAD- 2 had a good internal consistency with the Cronbach’s alpha of 0.88.

Communication

To measure communication disruptions, the APHAB (Abbreviated Profile of Hearing Aid Benefit) was identified as the most appropriate measure because of its flexibility (Cox & Rivera, 1992). In its original form, the APHAB asks participants to rate how well they hear in a variety of situations without a hearing aid, and then with a hearing aid. For the purposes of this study, the researchers were able to adapt the questionnaire to indicate how well the participants hear in situations in which people are not masked or in which they are masked. There are four subscales in the APHAB: Ease of Communication (EC), Background Noise (BN), Reverberation (RV), and Aversiveness (AV). As the AV scale was not applicable to masked situations (e.g., “The sound of screeching tires is uncomfortably loud”) it was not used in this study. The three subscales that were used have demonstrated previous varying internal consistency. The EC scale and RV scale have shown to have acceptable consistency (α = 0.74 and 0.73). The BN scale has shown good internal consistency with alpha 0.83 (Cox & Rivera, 1992). For this current study, the Cronbach’s alpha for the EC, BN and RV subscales are respectively 0.89, 0.74, and 0.82.

Data analysis

Statistical analyses were computed using SPSS v. 28 (IBM Corp, 2021). Of the 190 total responses, 154 participants completed initial demographic data. Of those who completed demographic questions, only 122 completed at least one subscale. Thirty-two participants did not finish the first measure (APHAB). All 154 participants who completed demographic questions were included in descriptive statistics analysis. Only those respondents who had completed at least one outcome measure subscale (n = 122) were included in the bivariate and multivariate analyses.

First, descriptive statistics were used to identify sociodemographic characteristics, hearing loss characteristics, and hours a day in masked conversations. Next, to further understand the impact of masking, paired samples t-tests were computed between the APHAB conditions of masked and unmasked communication. Finally, hierarchical multiple linear regression analyses were conducted to examine the association between masking, mental health factors, and metacognition. Three models were computed with each dependent outcome variable (PHQ- 2, CESD-D, GAD- 2, and STAI-T). Model 1 assessed the main effect of masking exposure (IV) on each DV. Model 2 assessed the main effect of metacognition (IV) on each DV. Model 3 assessed the interaction between masking and metacognition (IV) on each DV.

Results

Descriptive characteristics of the study sample are shown in Table 1. Of the 122 respondents who completed the study, 52.4% (n = 63) are aged 55 or older, 71.3% (87) are female, and 91.8% (n = 112) are white. Most participants have at least a 2-year college degree or technical certifications (60.7%, n = 74). Regarding hearing loss, over half of the participants stated they have severe hearing loss (53.3%, n = 65) and exactly half use hearing aids (50.0%, n = 61). A large majority used spoken language, not sign language to communicate (84.4%, n = 103). Further demographic details are provided in Table 1.

Table 1 Descriptive data for participants who completed the study
Full size table

To compare listening comprehension between masked and unmasked communication conditions, paired samples t-tests were conducted in SPSS. Each of the 3 subscales (EC, BN, and RV) showed statistically significant results when comparing masked to unmasked listening situations. In the EC subscale (how much effort is involved in listening in quiet environments), participants could hear in unmasked situations at a statistically significant increase of 25% compared to masked situations, t(123) = 7.096, p <.001. In the BN subscale (listening in the presence of background noise), participants could hear in unmasked situations at a statistically significant increase of 25% compared to masked situations, t(122) = 9.322, p <.001. In the RV subscale (listening in rooms that have reverberation), participants could hear in unmasked situations at a statistically significant increase of 25% compared to masked situations, t(119) = 7.424, p <.001. These results indicate that participants with hearing loss during the COVID- 19 pandemic had 25% less speech and communication understanding when they were in listening situations where others were masked.

Rates of depression and anxiety

To determine the rates of depression and anxiety for participants, scores from the PHQ- 2, CES-D, GAD- 2, and STAI-T were recoded as at risk if the score was above the clinical threshold. Table 1 also indicates the results of each depression and anxiety measure for those at or above threshold cut off scores. For depression, 22% scored at clinically significant levels for the PHQ- 2 and 43% scored at clinically significant levels for the CES-D. Women were more likely to have clinically significant depression levels then men at the rate of 88.9% for the PHQ- 2 and 77.3% for the CESD. The age bracket of 55–64 years of age was more likely than the other age brackets to have depression (22.2% PHQ- 2; 18.9% CESD). Those who self-identified their hearing difficulty as severe were more likely to have depression (63.0% PHQ- 2; 58.5% CESD), although the moderately-severe category also had high rates of depression (22.2% PHQ- 2; 22.6% CESD). For anxiety, 36% of the general sample scored at clinically significant levels for the GAD- 2 and 38.0% scored at clinically significant levels for the STAI-T. For the GAD- 2, 84.1% of those who met clinical levels were women. For the STAI-T clinical levels of anxiety, 80.4% were women. Age and degree of hearing difficulty followed the same trend as depression. Those in the 55–64 age range had the highest rates of anxiety (25.0% GAD- 2; 21.7% STAI-T). The participants who identified their degree of hearing difficulty as severe were more likely to have anxiety (50.0% GAD- 2; 56.5% STAI-T).

Relationship between masking, metacognition, and mental health

To examine the relationship between masking, metacognition, and mental health factors, correlations and multiple hierarchical linear regressions were conducted. For each dependent variable (PHQ- 2, CES-D, GAD- 2, and STAI-T) a multiple linear regression was run with independent variables entered in a sequential order: (1) Hours/day in masked communications (Hours), (2) Metacognition measure (MCQ- 30), and (3) an interaction variable (Hours*MCQ- 30). Results can be found in Tables 2 and 3. For all four dependent variables, the interaction variable (Hours*MCQ- 30) was not a significant predictor of the dependent variables. Therefore model 2, which contained only the main effects of hours in masked conversation and metacognition, was the most appropriate model for each dependent outcome variable.

Table 2 Correlations for depression, anxiety, metacognition, and hours masked
Full size table
Table 3 Masking, metacognition, mental health factors regression
Full size table

For PHQ- 2, only the main effect was significant (p =.001) while hours in masked conversations was not (p =.960). This indicates that hours in masked conversations did not relate to PHQ- 2, however, as the MCQ- 30 score increased, so did the PHQ- 2. For CES-D the main effects of hours in masked listening situations and MCQ- 30 scores were significant. As hours in masked conversations increased, CES-D scores increased (p =.039) and as MCQ- 30 scores increased, CES-D scores increased (p =.001).

A similar pattern was demonstrated for the GAD- 2 and STAI-T. For GAD- 2 only the main effect of MCQ- 30 (p =.001) was significant. The number of hours in masked listening situations did not have a relationship with the GAD- 2 (p =.575). For STAI-T, the main effect of the number of hours in masked conversations significantly predicted an increase in score (p =.024). The main effect of MCQ- 30 was also significant, indicating that as MCQ- 30 increased, so did the STAI-T score (p = < 0.001).

Discussion

This study sought to understand the differences between masked and unmasked listening comprehension and the relationship of metacognition, masking, and mental health for DHH adults. When the listening environment went from unmasked to masked communications, there was a 25% decrease of listening comprehension. As the number of hours in masked conversations increased, the rates of anxiety and depression also increased. These results are consistent with other studies that have documented the ways the COVID- 19 pandemic exacerbated barriers due to masking and social distancing, causing an increase in stress. (Kaya et al., 2021; Zivkovic Marinkov et al., 2022)

Depression and anxiety rates varied based on the measure that was used. Depression rates were 21.4% (n = 27) as measured by the PHQ- 2 and 44.5% as measured by the CES-D. Anxiety rates were 35.2% (n = 44) as measured by the GAD- 2 and 38% (n = 46) as measured by the STAI-T. Multiple studies took place during the COVID- 19 pandemic to measure depression and anxiety rates. Although this study is not a true comparison, due to smaller sample size and not yielding prevalence estimates, the trend in this study suggests higher rates of depression and anxiety among participants with hearing loss than in the general hearing public. In their longitudinal study, Hajek et al. (2022) identified anxiety and depression rates in the general population in six different waves (Wave 5 demonstrated rates of depression as 25.5% and anxiety as 23.6%; Wave 6 found rates of depression at 23.8% and anxiety as 22.1%).

Of those who met the clinical thresholds for depression, approximately 80% of them were women. This is also true for anxiety. Regarding age, those who were between the ages of 55–64 had higher rates of depression and anxiety than all the other age brackets. This trend is consistent with other studies that measured depression and anxiety during the pandemic (Hajek et al., 2022). Those who self-reported their degree of hearing difficulty as moderately-severe or severe experienced higher rates of depression and anxiety than those who self-reported their hearing loss as mild or moderate. As masking distorts speech sounds, those with more severe hearing difficulties may rely on facial expressions and access to lip reading more than those with mild hearing difficulty.

The Cognitive Attentional Syndrome (CAS), when activated by metacognitive beliefs, causes chronic negative thinking, rumination, and a hyperfocus on perceived threats (Bailey & Wells, 2015; Mohammadkhani et al., 2022; Wells & Matthews, 1996). The activated CAS maintains attention on intrusive thoughts in masked communication, such as “What if I can’t hear?” and can lead to avoidant coping strategies such as thought suppression, disengaging from conversation, or delaying needed tasks that involve communication. This pattern contributes to intolerance, distress, and other negative emotions, which strongly impacts depression and anxiety (Aydın et al., 2022; Bailey & Wells, 2015; Mohammadkhani et al., 2022). The current study found that negative metacognitions (perception of the threat of masked communication) had a stronger relationship with both depression and anxiety than the actual condition of being in masked conversations.

DHH people experience multiple communication barriers in their daily lives that have an impact on their mental health (Fellinger et al., 2005; Recio-Barbero et al., 2020). During the pandemic, DHH had the dual challenge of mitigating the risk of contracting the virus and accommodating the increase of communication barriers (Garg et al., 2021). This study demonstrated an elevated risk of depression and anxiety as hours in masked conversations increased. However, metacognitive beliefs appear to play a bigger role in depression and anxiety than masking itself.

Limitations and future directions

In this study, two main limitations are important to discuss, yet they do not detract from the usefulness of this research. One, the APHAB was originally created as a pencil and paper study, not for electronic use. The questions and responses are in a side-by-side matrix. While completing this measure on a desktop works well, it is not effective when a mobile device is used to complete it as it is difficult to see the comparison category on a small screen. Thus, a lot of participants stopped taking the survey at this point or they skipped ahead. While the missing data is unfortunate, there are still enough participants with complete data to have the necessary power for statistical analyses. Future research using the APHAB for masked/unmasked conditions should consider using paper or a desktop computer to collect the data.

Another limitation is that the sample was drawn from individuals who are seeking help for their hearing loss, whether in a social media support group or in the speech and hearing clinic. Thus, there is potential risk for bias in the sample, as those who have hearing loss but who do not seek support were not included in the study. However, the study does not seek to generalize the findings to all individuals with hearing loss. This study is limited in its scope due to utilizing convenience sampling methods. Convenience sampling cannot yield prevalence estimates (Jager et al., 2017). If masking becomes a requirement in a future pandemic, a general population study on masked communication should include questions on hearing loss and help seeking behavior.

Implications for clinical practice

To our knowledge, this is the first study to demonstrate the role of metacognition in masking related depression and anxiety. Results confirm that the communication barrier of masking does contribute to the negative mental health impact of those with hearing loss. This study, however, highlights the role of metacognition as a stronger predictor of depression and anxiety than the number of hours in masked conversations. This finding suggests that DHH individuals could benefit from becoming aware of their thoughts, and how their thoughts impact their response to a communication stressor such as masking.

To address negative metacognitions, multiple different treatment options have been found helpful. These include metacognitive therapy, metacognitive training, or metacognitively oriented integrative psychotherapies (Philipp et al., 2020). Metacognitively oriented psychotherapy includes components of cognitive behavioral therapy, psychodynamic, recovery-based interventions, and humanistic approaches (Gordon-King et al., 2018; Philipp et al., 2020). Perhaps more importantly, mental health professionals can provide a space for DHH to discuss and process their thoughts and feelings surrounding communication challenges.

While the stressful days of masking and isolation are mostly behind us now, the COVID 19 pandemic can be used as a learning edge for all of us. We can learn from this to better position ourselves to provide care that is accessible to all clients when/if future mask widespread masking is used. Further, many professionals have maintained masking practices as a result of the pandemic and need to be aware of how masking may inhibit or even discriminate against DHH seeking access to services.

Mental health professionals pride themselves on being effective communicators. As such, we must learn from this research to create and maintain better practices for working with DHH clients. Further, we may not always know who, among our clients, are DHH. To ensure that we develop effective therapeutic relationships when masking is preferred (or mandated) mental health professionals should consider asking every client what their communication preferences are. The clinician can offer these suggestions if needed:

  1. 1)

    Use of a clear mask or face shield to allow lips and facial expressions to be seen.

  2. 2)

    Use of technology (that is often free) on either the client’s phone, or the clinician’s electronic device. Technology includes AI captioners on platforms such as Zoom or PowerPoint and AI captioning apps easily found in app stores.

  3. 3)

    Reducing background noise in the communication space.

  4. 4)

    Rearranging furniture for conducive communication practices (e.g., moving the clinician’s chair so that the light from a window is not blocking their face).

  5. 5)

    Using a white board or other visual means that are inexpensive and easily seen.

  6. 6)

    Allowing the client to record their session to provide the opportunity to catch missed information.

  7. 7)

    Offering an online meeting or session in place of in-person and remove your face mask.

Conclusion

The rates of depression and anxiety were higher in this DHH sample than they were in the general population during the pandemic. This study showed a relationship between masked communication in depression/anxiety and a relationship between metacognition and depression/anxiety. Both masked communications, and the stressful thoughts of facing masked communications, have an impact on mental health.

Being ethical clinicians requires us to include accessibility in our mental health practice. Accessible service provision can only be done through intentional planning and implementation. Making slight changes can help us as clinicians shoulder some of the communication responsibility, which will decrease some of the barriers that DHH routinely encounter. Learning from this research and implementing changes to increase the communication accessibility in our practices will help us with masking now and prevent any further harm in future masking scenarios.