Hearing Impairment in South Africa and the Lessons Learned for Planetary Health Genomics: A Systematic Review

Abstract

Hearing impairment (HI) is a silent planetary health crisis that requires attention worldwide. The prevalence of HI in South Africa is estimated as 5.5 in 100 live births, which is about 5 times higher than the prevalence in high-income countries. This also offers opportunity to drive progressive science, technology and innovation policy, and health systems. We present here a systematic analysis and review on the prevalence, etiologies, clinical patterns, and genetics/genomics of HI in South Africa. We searched PubMed, Scopus, African Journals Online, AFROLIB, and African Index Medicus to identify the pertinent studies on HI in South Africa, published from inception to April 30, 2021, and the data were summarized narratively. We screened 944 records, of which 27 studies were included in the review. The age at diagnosis is ∼3 years of age and the most common factor associated with acquired HI was middle ear infections. There were numerous reports on medication toxicity, with kanamycin-induced ototoxicity requiring specific attention when considering the high burden of tuberculosis in South Africa. The Waardenburg Syndrome is the most common reported syndromic HI. The Usher Syndrome is the only syndrome with genetic investigations, whereby a founder mutation was identified among black South Africans (MYO7A-c.6377delC). GJB2 and GJB6 genes are not major contributors to nonsyndromic HI among Black South Africans. Furthermore, emerging data using targeted panel sequencing have shown a low resolution rate in Black South Africans in known HI genes. Importantly, mutations in known nonsyndromic HI genes are infrequent in South Africa. Therefore, whole-exome sequencing appears as the most effective way forward to identify variants associated with HI in South Africa. Taken together, this article contributes to the emerging field of planetary health genomics with a focus on HI and offers new insights and lessons learned for future roadmaps on genomics/multiomics and clinical studies of HI around the world.

Introduction

Hearing impairment (HI) is a sensory condition, whereby an individual cannot hear sounds softer than 25 dB for children and 30 dB for adults in their better hearing ear (the ear they hear best with) (WHO, 2020). HI affects ∼1.57 billion people (Haile et al., 2021) and disabling HI affects ∼433 million people worldwide (WHO, 2020). Disabling HI affects ∼49.66 million people in Sub-Saharan Africa (WHO, 2018c) and has a prevalence of 5.5 per 1000 live births in South Africa (Swanepoel et al., 2009). The South African prevalence was reported about 12 years ago and needs an update.

HI may be classified according to the age of onset of HI: the HI is congenital (present at birth), prelingual (developed before the child acquired language), or postlingual (developed after the child acquires language) (Schrijver, 2004; Tekin et al., 2001). It is also classified according to the number of ears affected, either one ear (unilateral HI) or both ears (bilateral HI) (Schrijver, 2004; Tekin et al., 2001). It may also be classified according to the type of HI (sensorineural, conductive, or mixed HI) or the etiology of HI (genetic, environmental, or unknown etiology) (Schrijver, 2004; Tekin et al., 2001). Furthermore, HI may be classified according to whether it is associated with other clinical manifestations, as in syndromic HI, or not, as in nonsyndromic HI (Schrijver, 2004; Tekin et al., 2001).

Genetic factors account for ∼50% of congenital HI in high-income countries, among which 70% of HI is nonsyndromic (Schrijver, 2004); and ∼70% to 80% of the nonsyndromic HI is due to autosomal recessive inheritance (Chung et al., 1959; Morton, 1991). Environmental factors account for a greater proportion of HI in developing countries, as was indicated in Cameroon (Wonkam et al., 2013). Environmental factors that result in HI may include, but are not limited to, infectious disease, otitis media (collection of fluid in the ear), injury to the head or ear, excessive noise, aging, and/or exposure to ototoxic drugs (Schrijver, 2004; WHO, 2020).

The connexin genes, GJB2 and GJB6, are the most prevalent genes associated with HI in European, Asian, and North American populations (Chan and Chang, 2014; Hutchin et al., 2005; Liu et al., 2002; Najmabadi and Kahrizi, 2014; Pandya et al., 2003). The variations in the connexin genes have, however, been shown to be insignificant in the majority of studied sub-Saharan African populations (Bosch et al., 2014a, 2014b; Tingang Wonkam et al., 2019; Wonkam et al., 2015), including South Africa (Kabahuma et al., 2011). The exception is that of the Ghanaian and Moroccan populations, whereby Ghana has a founder mutation in GJB2 and Morocco has the 35delG mutation in GJB2 present in the populations (Adadey et al., 2019; Gazzaz et al., 2005; Hamelmann et al., 2001a).

South Africa is a multicultural, multiracial [including 80.8% of Black African, 8.8% Colored/Mixed Ancestry, 2.6% Indian/Asian, and 7.8% White (Stats-SA, 2020)], developing country at the southern tip of Africa. It has a population of 55.7 million people (Stats-SA), with a high burden of infectious disease, where an estimated 7.1 to 8.3 million are affected with HIV (WHO, 2019) and 215,000 to 400,000 individuals were infected with tuberculosis (TB) (WHO, 2019) in 2018. Sixty percent of tested TB patients were HIV positive in South Africa in 2017 (WHO, 2018a). However, there is an increasing burden of noncommunicable disease over the past decades (Mayosi et al., 2009), some of which are of genetic origin such as congenital HI. Despite the relatively high incidence of HI in South Africa (Swanepoel et al., 2009), a review of this condition has not been investigated.

This systematic review aims to provide a landscape of HI in South Africa, (1) by examining the diagnostic approaches, prevalence, and etiology, in particular the genetics of HI in the country, and (2) with an eye to inform HI scholarship elsewhere in a context of the emerging field of planetary health genomics.

Materials and Methods

This review is reported in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) statement (Moher et al., 2009).

Ethics approval

This analysis and the research that informed this review were approved by and granted ethics clearance from the Human Research Ethics Committee of the University of Cape Town (HREC: 104/2018). Approval to recruit patients from schools was obtained from the relevant provincial education departments and permission for the recruitment was obtained from the schools from which patients were recruited. Written and informed consent was obtained from individuals 18 years of age and older or from the guardian/parent, with verbal and/or written assent obtained from minor children, including approval to publish identifiable photographs.

Selection criteria

We included observational studies published from inception to April 30, 2021, which report on various aspects of HI in South Africa. These include data on the prevalence, etiologies, clinical profiles, and genetics of HI. For duplicate studies, the most comprehensive and/or initial article with the largest sample size was considered. We excluded qualitative studies, letters to the editor, reviews, and commentaries. Studies with either unavailable full text or missing key data, which could not be accessed after a reasonable request, were excluded.

Search strategy

We searched PubMed, Scopus, and African-specific databases (African Journals Online, AFROLIB, and African Index Medicus) for relevant articles. The keywords used, to search for publications were as follows: “hearing impairment” OR “hearing loss” OR “hearing disorder” OR “hearing disorders” OR “deaf” OR “deafness” AND “South Africa.” In addition, specific researchers active in the field of hearing loss in South Africa were contacted to identify additional sources of information, and additional specific articles were thus added based on their relevance to this review.

Selection of studies

Titles and abstracts obtained from searches were imported into the software EndNote, version X9.3.3, for the removal of duplicates. Regarding our inclusion and exclusion criteria, the initial search and review were performed by one author (N.M.). They screened unduplicated titles and abstracts before reviewing the full text of all selected studies for final inclusion. A second author, a medical geneticist (A.W.), verified that the study screening and selection process were performed correctly. Any disagreement between the two authors was solved through discussion and consensus.

Data extraction process

One researcher (N.M.) used a predesigned data extraction sheet, to summarize data from relevant studies. Extracted data included the following: the last name of the first author, the year of publication, the province(s) where the study was conducted, the settings (hospital, schools, and community), the study design, data collection (prospective vs. retrospective), study population, including demographic information (gender, age, and self-declared “racial background” according to the South African official classification), sample size, and the number of cases of HI (for prevalence studies), methods used to diagnose HI, various classifications of HI: according to the types (sensorineural, conductive, and mixed), levels of HI (mild, moderate, severe, and profound), inheritance patterns, clinical profiles (syndromic vs. nonsyndromic), and data on genetic testing.

For some studies, relevant proportions were calculated from their raw data. Data were summarized narratively. A second researcher (S.M.A.) checked the accuracy of the data extraction process; any discrepancy was resolved through discussion and consensus.

Assessment of methodological quality

The quality of included studies was assessed by two investigators (N.M. and A.W.), with the quality of genetic studies using a risk of bias tool (Q-Genie) developed by Sohani et al. (2015); and for prevalence and other studies, the risk of bias assessment tool developed by Hoy et al. (2012). Discrepancies were solved by discussion and consensus.

Results

The review process

Initially, 939 records and 5 targeted articles were identified through the literature search. Removing 191 duplicates retained 753 records. Two-hundred seventy-seven records were excluded based on their titles and 476 articles were scrutinized, and 449 articles were further excluded. This rendered 27 articles eligible for inclusion in this study (Fig. 1). The characteristics of the studies included in this review are summarized in Table 1.

FIG. 1.

Schematic diagram of the review process.

Table 1.

Summary of Articles Reviewed for Inclusion

First author's name (year)	Province	Study setting	Study design	Data collection	Study populations	Female (%)	Mean age (years)	Age range (years)	Sample size	Diagnostic tool
Appana (2016)	KwaZulu-Natal	Hospital	Longitudinal	Prospective	MDR-TB patients	48	34	18–56	52	PTA
Beighton (1979)	NR	NR	Case report	Retrospective	Family presenting with a rare syndrome	NR	NR	NR	41
Beighton (1993)	NR	Research	Case report	Retrospective	Patients identified from an Usher syndrome survey	NR	NR	NR	3	NR
Bosch (2014a)	Eastern Cape	School	Cross-sectional	Retrospective	Students from a school of the deaf	NR	NR	NR	25	NR
Butler (2013)	Free State	Hospital	Cross-sectional	Retrospective	Children referred for HI screening	NR	3.4^a	1 months–5.9 years	260	PTA
Chaya (2018)	Eastern Cape	Hospital	Case report	Observational	A child presenting with an undiagnosed progressive condition	0	NA	11 months	1	ABR
Clarke (2006)	NR	Hospital	Case Report	Retrospective	Family presenting with BOR	NR	NR	NR	3	NR
Eksteen (2019)	Western Cape	Preschool	Cross-sectional	Retrospective	Children attending preschool	50.5	NR	4–7	8023	HearScreen, PTA
Els (2018)	Gauteng	Hospital	Cross-sectional	Observational	OME patients	49.7	25.54^b;6.75	2–12	109	PTA and POT
Ghafari (2020)	Western Cape	Hospital	Cohort study	Retrospective	MDR-TB patients	43	34.9^a	NR	102	PTA
Kabahuma, (2011)	Limpopo	School	Cross-sectional	Retrospective	Students two schools for the deaf	65.9	NRr	5–21	182	PTA
Kabahuma, (2021)	Limpopo	NR	Cross-sectional	Retrospective	Congenital HI individuals from Limpopo province	NR	NR	NR	94	PTA
Levin (1966)	NR	Hospital	Case Report	Observational	Two brothers presenting with Pendred syndrome	0	9	7100	2	Audiometry
Louw (2018)	Gauteng	Clinic	Cross-sectional		Patients at clinic	74	41.2	16–97	1084	PTA
Potgieter (2014)	Gauteng	Hospital	Cross-sectional	Prospective	Sclerosteosis patients	60	32	10–72	10	PTA, DPOAEs and ABR
Rappoport (1970)	Gauteng	Hospital	Case Report	Retrospective	Patient identified in hospital	23	NR	NR	NR	Tuning fork
Roberts (2015)	NR	NR	Cross-sectional	Retrospective	Retinal degenerative disease patients presenting with Usher syndrome	23	NR	NR	NR	NR
Roberts (2020)	NR	NR	Cross-sectional	Retrospective	Retinal degenerative disease patients presenting with rod-cone dystrophy, HI, and renal dysfunction	NR	NR	NR	4	NR
Sandstrom (2020)	Gauteng	Clinic	Cross-sectional	Prospective	Patients at community clinics	71	52	20–88	63	HearScreen, PTA
Spracklen (2017)	Western Cape	Hospital	Cross-sectional	Retrospective	Cancer patients treated with cisplatin	28.8	48^a	14–75	222	PTA
Swanepoel (2007)	Gauteng	Hospital	Cross-sectional	Retrospective	Newborn babies	NR	0	0	6241	TEOAE
Tekendo-Ngongang (2019)	Western Cape	Hospital	Cross-sectional	Retrospective	Patients with Noonan syndrome	NR	4.5^a	1 month–51 years	26	NR
Tiedt (2013)	Free State	Hospital	Cross-sectional	Prospective	Children with CSOM	47.7	4.6^a	1–12	86	PTA
Tshifularo (2013)	Gauteng	Hospital	Cross-sectional	Prospective	HIV patients	58.8	31.5	1–76	153	NR
Viljoen (1983)	Western Cape	Hospital	Case Report	Retrospective	The family that was seen at the genetic clinic	62.5	21	11–49	8	PTA
Winship, (1992)	Western Cape	School and Genetic Clinic	Cross-sectional	Retrospective	Children diagnosed with Waardenburg syndrome in a study at school	NR	NR	NR	101	NR
Yaniv (1987)	Eastern Cape	Hospital	Case report	Observational	Patients with tuberculous otitis media	38.7	18	9 months–67 years	31	NR

Median age given instead of mean age.

Mean age given of male participants and female participants separately, with male participants' mean age indicated first.

ABR, auditory brain response; BOR, Branchio-oto-renal syndrome; CSOM, chronic suppurative otitis media; DPOAEs, distortion product otoacoustic emission; MDR-TB, multidrug-resistant tuberculosis; NA, not applicable; NR, not reported; POT, pneumatic otoscopy and tympanometry; PTA, pure-tone audiometry; TEOAE, transient evoked otoacoustic emissions; HearScreen is a hearing screen mobile application developed by the hearX group.

Prevalence and newborn screening

The prevalence of HI in developing countries is estimated to be 6 in 1000 live births (Olusanya and Newton, 2007), but this is a general estimate, and it is not specific to South Africa. Swanepoel et al. (2007) partially resolved the lack of prevalence data by reporting, over 4 years, on the universal newborn hearing screening (UNHS) program that ran at a private hospital in Gauteng. Of the 13,799 births at the hospital, only 6241 newborns were screened for HI (Swanepoel et al., 2007). For the first 22 months of the project, the screening was subsidized by the hospital, whereas in the succeeding 26 months, the parents were responsible for paying for the screening. Within the first 22 months of UNHS, there was a 75% uptake in the UNHS program, and this was followed by a drop to 20% in the succeeding 26 months (Swanepoel et al., 2007).

The UNHS starts with an initial hearing screening for the babies and those who failed were rescreened (Swanepoel et al., 2007). Two hundred nineteen infants, of 694 that failed the initial screening, were rescreened in the hospital and 19 presented with HI that required diagnostic testing (Swanepoel et al., 2007). The authors estimated a 3 in 1000 prevalence of HI at birth, in the private sector, after taking into consideration the newborns who did not return to the hospital for the rescreen (Swanepoel et al., 2007). Swanepoel et al. (2009) furthered the work to determine an estimated prevalence of 5.5 in 1000 births, after taking into consideration the public health care system (Swanepoel et al., 2009).

Age of diagnosis

The median age of diagnosis for children presenting with HI, at a tertiary public hospital in Bloemfontein, was shown to be 3.7 years by Butler et al. (2013). The median age of the first appointment at the hospital was 3.4 years and there was a median delay of 49 days between the first appointment at the hospital and diagnosis (Butler et al., 2013).

The median age of diagnosis for children presenting with HI in a private audiology practice in Bloemfontein was shown to be 2.24 years by Butler et al. (2015). The median age of diagnosis was significantly different between patients seen in the private practice and patients seen in the tertiary public hospital (Butler et al., 2015). The patients who were screened at birth had a median age of diagnosis of 1.25 years, whereas patients who were not screened at birth had a median age of diagnosis of 3.01 years (Butler et al., 2015). The differences in the median age of diagnosis based on newborn screening are, however, not significantly different (Butler et al., 2015).

Technology and techniques used to detect HI

Community health workers used the hearScreen mobile application, from the hearX group, to screen 8023 children in preschool, for HI, in Khayelitsha and Mitchells Plain in the Western Cape, as reported by Eksteen et al. (2019). Of the 8023 children, 2313 failed the initial test and 435 failed their retest (Eksteen et al., 2019). Three-hundred eighty-nine children, of the 435 who failed the retest, were screened by the project audiologist and 124 children were referred for diagnostic evaluation (Eksteen et al., 2019). Only 94 of the 124 children attended their diagnostic evaluations, of which 54 were diagnosed with HI (Eksteen et al., 2019).

The hearTest application, from the hearX group, and Pure Tone Average (PTA) were used to screen 58 individuals for HI after 5 individuals were excluded from the study (Sandstrom et al., 2020). The application was shown to have a 90.6% sensitivity to detect HI above 40 dB (Sandstrom et al., 2020). The HI threshold varied between 0.9 and −5.4 dB between the application and PTA in the patients, with the average time for the HI test on the application being 512 sec (Sandstrom et al., 2020).

Etiology: acquired HI

The most frequent cause of acquired HI was postnatal complications such a meningoencephalitis, streptomycin-induced HI, and middle-ear diseases (Beighton et al., 1991). Meningitis accounted for 230 cases, maternal rubella accounted for 143 cases, severe infections accounted for 118, and jaundice accounted for 66 cases of environmental HI in the 765 children with acquired HI (Beighton et al., 1991).

Otitis media with effusion, chronic suppurative otitis media, and tuberculous otitis media

Otitis media with effusion (OME) was studied, in children admitted for either an adenoidectomy or adenotonsillectomy in the otorhinolaryngology department of an academic hospital in Pretoria, by Els and Olwoch (2018). The prevalence of OME was 11.9% and 22.9% bilaterally and unilaterally, respectively, in a group of 102 children (Els and Olwoch, 2018), whereas another study of 136 children had an OME prevalence of 16.5% (Biagio et al., 2014). The OME reported by Els and Olwoch (2018) resulted in a mean hearing loss of 19.8 dB (Els and Olwoch, 2018), whereas Biagio et al. (2014) had not reported any audiometry for study participants.

Thirty-one patients presented with tuberculous otitis media, a rare form of chronic suppurative otitis media (CSOM), at the Ear, Nose and Throat department of a hospital in the Eastern Cape, what was then the Ciskei, between 1984 and 1985 (Yaniv, 1987). Twenty-six patients were found to have conductive HI of varying degrees, with 5 patients also presenting with sensorineural HI (Yaniv, 1987). Six patients could not be tested, but one patient was noted to have bone sequestra (i.e., a piece of necrotic bone detached from the healthy tissue) and stapes lying loose in the middle ear (Yaniv, 1987). The study found that, of the 31 patients, only 14 patients (47%) had a healthy second ear (Yaniv, 1987).

Tiedt et al. (2013) examined CSOM in 86 children from the Free State. In contrast to Yaniv (1987), 68.6% of patients presented with unilateral CSOM and 31.4% were bilateral (Tiedt et al., 2013). Audiometry was performed in 46 patients, a total of 66 ears, with HI present in 44 ears (66.7% of ears tested), with the median for the average HI being 38.3 dB (Tiedt et al., 2013).

A study to determine the trends of head, neck, and ENT manifestation of HIV-infected patients looked at 153 patients, at an academic hospital in Pretoria, was undertaken by Tshifularo et al. (2013). The study determined that 14.29% of patients presented with sensorineural HI (Tshifularo et al., 2013). The study also indicated that the common otological manifestation was CSOM, which affected 27% of patients, and 15.58% of their patients were affected with OME (Tshifularo et al., 2013).

Medication ototoxicity

Cisplatin-induced ototoxicity

In 222 cancer patients treated with cisplatin, ototoxicity was observed at rates between 39.2 and 66.7%, depending on the ototoxicity grading scale used, according to Spracklen et al. (2017). An earlier study had indicated a 55.1% incidence of cisplatin-induced ototoxicity in the South African population, of which 62.7% was bilateral HI (Whitehorn et al., 2014). Spracklen et al. (2017) used the grading scale indicated by Konrad-Martinet et al. (2005), Chang et al. (2010), (Chang and Chinosornvatana, 2010), and the U.S. Department of Health and Human Services (2017).

Ototoxicity was shown to be associated with increased cisplatin dosage, alone, and with cisplatin dosage and rs6721961in NFE2L2, on all three grading scales following correction for multiple testing (Spracklen et al., 2017). SNP rs316019 in SLC22A2 was significant when using the Chang grading scale (Chang and Chinosornvatana, 2010; Spracklen et al., 2017).

Multidrug-resistant tuberculosis and kanamycin-induced ototoxicity

Fifty-two patients, from KwaZulu Natal, were included in a study to determine the effects of aminoglycoside treatment, for multidrug-resistant tuberculosis (MDR-TB), over 5 intervals during a 6-month period (Appana et al., 2016). Appana et al. (2016) observed normal hearing in 44% in the right ear and 40% in the left ear of patients before treatment (Appana et al., 2016). Sensorineural HI was present in 50% of patients in the right ear and 54% of patients in the left ear before treatment, with all patients receiving kanamycin (Appana et al., 2016). The hearing profile of all patients changed to show bilaterally clinically significant HI for all patients by the fifth post-treatment session (Appana et al., 2016). Following their fifth post-treatment session, sensorineural HI was present in 94% of patients in their right ear and 96% of patients in their left ear (Appana et al., 2016).

The authors saw a gradual deterioration in the average hearing thresholds following the commencement of treatment when considering the low (125, 250, 500 Hz), mid (1000 and 2000 Hz), high (4000 and 8000 Hz), and ultra-high (10,000 and 12,500 Hz) frequencies (Appana et al., 2016). The deterioration in average hearing thresholds in low frequencies remained within normal limits throughout the course of treatment (Appana et al., 2016).

Similarly, Ghafari et al. (2020) analyzed audiometry data in 102 patients receiving kanamycin for treatment of MDR-TB in Cape Town. The patients were tested for HI before treatment and 4, 8, and 12 weeks after starting treatment (Ghafari et al., 2020). Eighty-four patients (82.4%) developed HI during the course of treatment with kanamycin and the HI was significantly associated with exposure to kanamycin (Ghafari et al., 2020).

Streptomycin-induced ototoxicity

A report by Viljoen et al. (1983) investigated a large, nonconsanguineous family with streptomycin-induced ototoxicity. The HI affected eight family members, of both sexes, and was bilateral, and ranged from moderately severe to very severe sensorineural HI (Viljoen et al., 1983). The ototoxicity susceptibility in the family was determined to be autosomal dominantly inherited (Viljoen et al., 1983) and genetic analysis identified a 1555A to G variation in the mitochondrial DNA as a causative mutation associated with the streptomycin-induced HI (Gardner et al., 1997).

Genetics of HI

Diagnostic surveys of HI in 4452 scholars were undertaken by Beighton et al. (1991) since 1975. The authors were able to ascertain that genetic syndromes accounted for 8% of HI, nonsyndromic HI accounted for 12%, acquired HI accounted for 25% of HI, and unknown causes accounted for 55% in the population (Beighton et al., 1991). Waardenburg syndrome was the most common cause of syndromic HI due to genetics, followed by Treacher-Collins, Sclerosteosis, and Usher Syndrome, among others (Beighton et al., 1991). A handful of genetic syndromes and nonsyndromic genetic cases will be reported below.

Syndromic HI

HI in Sclerosteosis

The audiological profile of patients with Sclerosteosis was studied by Potgieter et al. (2014). The study consisted of 10 individuals, of the total 36 living individuals diagnosed with Sclerosteosis, of which 18 lived in the Gauteng region (Potgieter et al., 2014). All 10 individuals present with HI, which ranged from moderate in 1 individual to profound in 8 individuals (Potgieter et al., 2014). The HI was mixed in all individuals and symmetrical in six participants (Potgieter et al., 2014).

Beighton et al. (1976) had previously reported on the clinical features of Sclerosteosis, in a study that comprised 21 individuals. Conductive HI was indicated to be apparent in patients with Sclerosteosis once the patient started schooling (Beighton et al., 1976). The conductive HI was present bilaterally in all adults and three children in the study (Beighton et al., 1976). The authors indicated that the sensorineural HI may develop in adulthood due to compression of the vestibulocochlear nerve or due to involvement of oval and round windows in the cochlear (Beighton et al., 1976).

The involvement of oval and round windows was observed in CT scans taken in 9 (18 ears) of the 10 patients in the Potgieter et al.'s (2014) study. Abnormalities were noted in 16 ears (89%) in the oval window and in 17 ears (94%) in the round window (Potgieter et al., 2014). The abnormalities observed included windows containing bony overgrowths and/or the windows that have been closed among other observations (Potgieter et al., 2014). The authors, furthermore, noted the narrowing of the internal auditory canal or the closure of the canal in 16 ears (Potgieter et al., 2014).

Brown-Vialetto-van Laere syndrome

The first reported case of Brown-Vialetto-van Laere Syndrome in South Africa was of an 11-month-old baby from the Eastern Cape (Chaya et al., 2018). The case, as reported by Chaya et al. (2018), indicates that the child had normal Apgar scores at birth, but developed progressive dysphagia at 10 weeks and intermittent stridor (Chaya et al., 2018). Symptom progression resulted in constant stridor, persistent dysphagia, regression of motor milestones, ptosis, paradoxical abdominal movements when breathing that were consistent with diaphragmatic paralysis, and bilateral vocal cord palsy at 11 months, and auditory brain stem response indicated that the child had sensorineural HI (Chaya et al., 2018). Genetic analysis indicated a heterozygous pathogenic variant and a variant of unknown significance in SLC5A3, which supported the diagnosis of Brown-Vialetto-van Laere Syndrome (Chaya et al., 2018).

Craniometaphyseal dysplasia

Beighton et al. (1979) investigated a rare syndromic disorder in 41 individuals from a family, of which 15 individuals were determined to be affected. The proband presented with bilateral HI, a prominent forehead, right facial palsy, and left facial weakness (Beighton et al., 1979). Radiological findings indicated cranial thickening with sclerosis and metaphyseal flaring at the knees (Beighton et al., 1979). The symptoms presented in the proband were observed, to varying degrees, in several relatives with a variable presence of HI in the family (Beighton et al., 1979).

Usher syndrome and rod-cone dystrophy

Beighton et al. (1993) identified a rare and potentially novel syndrome, which manifests as sensorineural HI, rod-cone dystrophy, and kidney dysfunction, with the Fanconi lesions resulting in skeletal changes and kidney failure, which may be fatal (Roberts et al., 2020). The patients were identified during a large-scale survey of individuals with Usher syndrome and all patients were of Afrikaans descent (Beighton et al., 1993). The patients studied included two sisters and an unrelated teenage boy; the eldest of the girls and the teenage boy presented with profound sensorineural HI and the younger girl presented with mild sensorineural HI (Beighton et al., 1993). The three children had similar fundal findings (including choroidal depigmentation, a macular scar, and vascular attenuation) and Fanconi-type renal lesions (Beighton et al., 1993). The skeletal changes observed in this condition are indicated in Figure 2A.

FIG. 2.

Images of the characteristics of various syndromic conditions. (A) Indicates the skeletal changes observed in renal dysfunction, rod-cone, and sensorineural HI. The kidney dysfunction results in rickets-like skeletal changes in the condition. Adapted from Beighton et al. (1993) © John Wiley and Sons. (B) Craniofacial features of a child presenting with Noonan syndrome. Adapted from Tekendo-Ngongang et al. (2019) with copyright permission. (C) Fundus photographs of a patient with pathologic variation in USH2, showing reduction of the retinal arteries and bone spicule pigment deposits. Adapted from Hamel (2006), with copyright permission. (D) South African adolescent child presenting with heterochromia iridium, no dystopia canthorum, and a white forelock representative of Type II Waardenburg syndrome, recently observed in our clinic (Written parental consent and verbal assent from minor obtained granting permission to reproduce photograph).

Recent work by Roberts et al. (2020) identified the novel homozygous RRM2B-c.786G>T variant as a plausible disease-causing mutation in two sisters, an unrelated male and an unrelated female affected with renal dysfunction, rod-cone dystrophy, and sensorineural HI. The variant was discovered using whole-exome sequencing on the affected siblings and their carrier parents (Roberts et al., 2020). The two sisters were recruited and reported by Beighton et al. (1993).

Usher syndrome is an inherited HI syndrome characterized by HI and visual impairment due to retinitis pigmentosa (Kimberling et al., 1992). The condition was first reported by Von Graefe (1858) and its inherited nature was emphasized by Usher (1914). Usher Syndrome has three subtypes, of which Usher I has the most severe phenotype.

Usher I is characterized by congenital severe to profound sensorineural HI and vestibular dysfunction and the onset of retinitis pigmentosa in the first decade of life (Kimberling et al., 1992; Van Camp and Smith, 2020). Usher II is characterized by congenital moderate to severe sensorineural HI, absence of vestibular dysfunction, and the onset of retinitis pigmentosa in the first or second decade of life (Kimberling et al., 1992; Van Camp and Smith, 2020; Weston et al., 1996). Usher III presents as progressive sensorineural HI with variable vestibular dysfunction and variable onset/presence of retinitis pigmentosa (Van Camp and Smith, 2020; Weston et al., 1996). The characteristics of retinitis pigmentosa are indicated in Figure 2B.

A study of inherited retinal degenerative disease patients by Roberts et al. (2015) identified a founder mutation that is responsible for a significant proportion of Usher Syndrome among Black South Africans. The mutation in MYO7A was first identified in 2 unrelated individuals and, following screening in another 12 Black Usher patients, it was seen in further 6 Black South Africans (Roberts et al., 2015, 2016). The study had included six Black suspected Usher patients and three Usher patients of mixed ancestry, but the founder mutation was absent within these groups (Roberts et al., 2015). Molecular analysis in a cohort of Usher patients by Roberts et al. (2015) indicated a homozygous deletion of MYO7A, wich affected 6 of 12 (42.86%) Black African individuals. The variation MYO7A c.6377delC was shown to be a founder mutation, whereby the affected patients had the same haplotype (Roberts et al., 2015).

HI in Noonan syndrome

Noonan syndrome is an autosomal dominantly inherited multisystem disorder characterized by congenital heart disease, small stature, ocular hypertelorism (which is abnormally increased distance between the eyes), and skeletal malformations (Roberts et al., 2013). Patients present with distinct craniofacial dysmorphism and the syndrome is clinically heterogeneous (Tekendo-Ngongang et al., 2019). Approximately 10% of patients with Noonan syndrome will present with HI due to sensorineural HI (Ahituv et al., 2000; Roberts et al., 2013). Figure 2C illustrates the craniofacial dysmorphism that may be observed in a child with Noonan syndrome.

A study of 26 patients presenting with Noonan syndrome, by Tekendo-Ngongang et al. (2019), observed that 15% (n = 4) of patients presented with HI. Twenty of the patients were unrelated and 65% of the patient population had at least one cardiovascular abnormality, of which pulmonary stenosis was the most prevalent (Tekendo-Ngongang et al., 2019). The study was able to ascertain causative variations in CBL, PTPN11, and MAP2K1, with the variants segregating with the condition in the respective families (Tekendo-Ngongang et al., 2019).

Pendred syndrome

Pendred syndrome, characterized by HI and goiter with the absence of iodine deficiency, was first described by Vaughan Pendred, in two sisters, in 1896 (Pendred, 1896; Reardon et al., 1997). The autosomal recessive inheritance of Pendred syndrome was ascertained by Brain (1927) and the first gene, SLC26A4, associated with Pendred syndrome was first described in 1997 by Everett et al. (1997).

Levin and Klugman (1966) described the case of two brothers with Pendred syndrome in 1966. The 7-year old, younger brother, was admitted into hospital due to swelling in the neck (goiter) (Levin and Klugman, 1966). The child had congenital sensorineural HI and was subsequently diagnosed with Pendred Syndrome. Familial history indicated an elder brother who also had HI and he was, likewise, diagnosed with Pendred syndrome (Levin and Klugman, 1966). The two brothers were treated with 1-thyroxine sodium, of which the younger brother showed greater improvement than his elder brother (Levin and Klugman, 1966).

Branchio-oto-renal syndrome

Branchio-oto-renal syndrome (BOR) was first described in 1976 by Melnick et al. (1976) following the examination of a father and his three affected children. The variability of the phenotypic expression of the condition was noted by Heimler and Lieber (1986), following the study of a large family presenting with BOR and the causative gene, EYA1, was identified in 1997 by Abdelhak et al. (1997). The condition is characterized by abnormalities of the ear (inner, middle, and outer), HI, branchial cleft sinuses and fistulae, lacrimal duct stenosis, and renal abnormalities (Heimler and Lieber, 1986).

Clarke et al. (2006) reported on a prenatal proband presenting with agenesis of the kidneys in a South African Afrikaans family. A previous pregnancy, where the child passed on after birth, had also presented with BOR syndrome, and clinical manifestations were observed in the father of the children (Clarke et al., 2006). The father had a unilateral preauricular pit and HI of unknown etiology (Clarke et al., 2006). A de novo, novel, heterozygous EYA1 727G>T was identified as the causative mutation that originated in the father (Clarke et al., 2006).

Waardenburg syndrome

Waardenburg syndrome was first described by Waardenburg (1951). The syndrome is characterized by sensorineural HI associated with pigmentary abnormalities of the hair, eyes, and skin (white forelock, characteristic blue irises, and depigmentation of the skin) and dystopia canthorum (Van Camp and Smith, 2020; Waardenburg, 1951). The syndrome has four subtypes, whereby Type I and Type II are differentiated by the presence of dystopia canthorum in Type I (Van Camp and Smith, 2020). Type III is a combination of Type I Waardenburg with upper limb abnormalities and Type IV is the combination of Type II with Hirschsprung Disease (Van Camp and Smith, 2020). Waardenburg is inherited, predominantly, in an autosomal dominant pattern (Waardenburg, 1951) and the MITF was the first gene associated with Waardenburg (Tassabehji et al., 1994). Figure 2D illustrates the facial characteristics that may be observed in Waardenburg syndrome.

Rappoport (1970) reported on a 23-year-old woman who was diagnosed with Waardenburg syndrome following a hospital visit. The young woman presented with a white forelock and dystopia canthorum (Rappoport, 1970). The Waardenburg syndrome originated from her grandmother; with the white forelock present in all affected members (Rappoport, 1970). Neither HI nor heterochromia iridium was present in the family members, but some members presented with dystopia canthorum and/or hyperplasia and depigmentation of the medial portion of the eyebrows (Rappoport, 1970).

A retrospective study by Sellars and Beighton (1983) identified 90 children presenting with Waardenburg Syndrome from 3006 HI children in 19 schools. The authors noted the presence of dystopia canthorum in 54% of the children and that a proportion of the children presented with heterochromia iridium in the absence of other clinical manifestations (Sellars and Beighton, 1983). Furthermore, the variability of Waardenburg was noted even among siblings (Sellars and Beighton, 1983)

The possibility of a founder mutation been responsible for Waardenburg Syndrome among the Afrikaans population was examined by de Saxe et al. (1984). The author analyzed genealogy of three Afrikaans families and noted that all three families were related with the Waardenburg phenotype originating from a grandfather who passed it down to his descendants (de Saxe et al., 1984).

Variation in the expression of Waardenburg syndrome was investigated in 68 children attending Schools for the Deaf and 33 individuals, from 7 families, who had attended a genetic clinic staffed by Winship and Beighton (1992). The children, at the schools, all presented with profound sensorineural HI and 89% of the children had variable pigmentation in the eyes (Winship and Beighton, 1992). Bilateral sapphire blue eyes were present in 28 children, 24 children had heterochromia iridium, and 9 children had segmental heterochromia either unilaterally or bilaterally, and none of the children had Hirschsprung disease (Winship and Beighton, 1992).

The variations observed in the children were also observed in the familial cases. Six of the seven families identified presented with type 1 Waardenburg and the seventh family presented with type 1 Waardenburg. HI was present in 6 individuals, of the 33 individuals in the type 1 families, and in the 4 siblings of type 2 (Winship and Beighton, 1992).

A cross-sectional study of a 13-member South African family identified a splice site mutation in PAX3 that segregated with Waardenburg syndrome within the family (Butt et al., 1994). In contrast, a case report detailed the comorbid presence of Waardenburg and tetraphocomelia in a premature neonate that passed away 10 min after birth (Wu et al., 2009). The mother had detailed an uneventful pregnancy, except for swelling and discomfort in the 29th week of pregnancy (Wu et al., 2009). The mother subsequently went into labor 4 days after her discomfort complaint and the neonate was born with gross malformations of the limbs, heterochromia iridium, and a white forelock (Wu et al., 2009). Family history indicated that the mother, grandmother, and uncle of the baby all had white forelocks, but genetic testing was inconclusive (Wu et al., 2009).

Genetic etiology: nonsyndromic HI

The connexin genes have been studied in 25 Black African students from one school of the deaf in the Eastern Cape (Bosch et al., 2014a, 2014b) and 183 Black African students from two schools of the deaf in Limpopo province, by Bosch et al. (2014a) and Kabahuma et al. (2011), respectively. Participants for the study by Kabahuma et al. (2011) underwent audiometric assessments (including tympanometry, transient otoacoustic emissions, and PTA) before inclusion into the study and all participants were determined to have stable sensorineural HI (Kabahuma et al., 2011). The study by Bosch et al. (2014b) of the 25 patients from Eastern Cape had no audiological data, in comparison to the study performed in Limpopo.

Molecular analysis of potentially causative variants in both the Eastern Cape and Limpopo studies yields no causative mutations in either GJB2 or GJB6 (Bosch et al., 2014a, 2014b; Kabahuma et al., 2011). Lebeko et al. (2017) furthered the research for putative causative mutations in 25 patients from the Eastern Cape. The patients were screened for and were negative for variants in MYO7A, CDH23, LOXHD1, OTOF, and SLC26A4 (Lebeko et al., 2017), which had been previously discovered in a targeted sequencing approach (Lebeko et al., 2016).

Kabahuma et al. (2021) published a follow-up study, in 2021, providing molecular analysis for causative mutations in 94 patients, presenting with nonsyndromic HI, from Limpopo. The study identified eight MYO7A variants, of which four variants were novel (Kabahuma et al., 2021). The novel variation p.Thy1780Ser was the most common variation identified, and it was present in four of eight families segregating variations in MYO7A (Kabahuma et al., 2021).

Discussion

The global prevalence of HI is approximately one in five people presenting with HI, which is an estimated 1.57 billion people affected with HI (Haile et al., 2021). HI has been classed as the fourth leading cause of disability worldwide (Wilson et al., 2017), but data on the prevalence in low-income countries are sparse (Haile et al., 2021).

This study is the most comprehensive review of HI in South Africa to date. The research showed that the prevalence of HI has received little attention in South Africa over the last decade, with no study reporting the national prevalence of the condition. The reported prevalence of 5.5 in 1000 live births is much higher than the estimated prevalence of 1.33 in 1000 and 1.86 in 1000 live births in developed countries such as the United States and England, respectively (Morton and Nance, 2006). Researchers, such as Swanepoel (2009), indicated that a large-scale study may be necessary to determine the prevalence of HI in South Africa. This may be aided by the use of smartphone screening applications, similar to the ones used by Eksteen et al. (2019), when taking into consideration the validation done by Sandstrom et al. (2020).

The smartphone application could putatively be used to screen patients for HI at their community clinics and refer patients who fail the screening for diagnostic testing. This screening method may allow for a more accurate determination of the prevalence of HI in South Africa and early detection of HI in both children and adults. This screening method may potentially be a cost-effective approach in screening for HI in underresourced settings and may allow for the early detection of HI in patients.

Early diagnosis of HI is essential in the process of giving appropriate and effective interventions to HI patients (Ching et al., 2017). Cochlear implantation, before 6 months after birth, results in better language development in children with congenital/prelingual HI compared to those who received implants after 24 months (Ching et al., 2017). The implementation of UNHS has helped many countries achieve a reduction in age at diagnoses of HI; an average of 20 months of age of HI diagnoses was observed in Italy (Canale et al., 2006) and 2–5 months in the United States (American Academy of Pediatrics, 2007; D'Aguillo et al., 2019).

Despite the UNSH achievements, some countries still have a high average age of HI diagnosis; 4.1 years in New Zealand (Gruber et al., 2019), 6–11 years in Ghana (Adadey et al., 2019), and 7.4 years in South Korea (Lim et al., 2018). Even though South Africa has a relatively lower age of HI diagnosis (between 2 and 4 years) (Butler et al., 2013, 2015) compared to other African countries, there is a need to reduce the age of diagnosis to few months after birth.

Although South Africa is an emerging country, it has one of the highest rates of wealth inequality (Vorster, 2021). Wealth inequality may play a role in the uptake of newborn hearing screening within the country, as seen by Swanepoel et al. (2007), although early diagnosis is shown to be a cost-effective intervention in addressing the cost of unaddressed HI (WHO, 2017). Universal, subsidized hearing screening for both newborns and other members of the population may mitigate the financial barriers that patients and their families face in seeking a diagnosis.

Determining the barriers to diagnosis of HI and the causes of HI in the South Africa population may allow for better understanding of HI in South Africa. Merugumala et al. (2017) studied the barriers to early diagnosis of HI in an Indian city. One parent noticed that their child was not hearing when the child was 8 to 9 months old, but did not consult with a doctor because their child was not sick, whereas another parent thought the child would start speaking when they were older (Merugumala et al., 2017).

These perceptions may not be limited to this Indian city, as in a study pertaining to the perceptions of HI by parents of children with HI in South Africa, 7 of the 11 parents could not identify the cause of HI or attributed the HI to environmental etiology (Gardiner et al., 2019). This calls for thorough engagement with the community, to make the early symptoms of HI and the causes of HI well known, including genetic etiologies. Dispelling incorrect perceptions within the wider population could be a key objective that could aid in lowering the age of diagnosis in South Africa and in other developing countries.

In Africa, the acquired causes of HI require further attention since they account for a large proportion of cases (Adadey et al., 2019; Wonkam et al., 2013). In Ghana and Cameroon, cerebrospinal meningitis, severe malaria, and otitis media were identified among the acquired causes of HI (Adadey et al., 2019; Wonkam et al., 2013). Available data from South African indicate OME as the main cause of acquired HI among children (Biagio et al., 2014; Els and Olwoch, 2018; Yaniv, 1987).

OME is a chronic inflammatory condition, characterized by effusion behind an intact membrane with the absence of the signs and symptoms of acute inflammation (Qureishi et al., 2014). OME is less severe than CSOM, but it is shown by Els and Olwoch (2018) to result in a mean HI of 19.8 dB. The prevalence of HI in the OME patient group was not indicated, although caregivers reported HI in 6.6% of 136 patients in the study by Biagio et al. (2014). CSOM, conversely, resulted in a mean HI of 38.3 dB in the subset of patients who had PTA done (Tiedt et al., 2013).

This means the HI threshold included children where there was no complaint of HI, with children with a complaint of HI (72.7% of tested ears) having a mean HI threshold of 41.7 dB (Tiedt et al., 2013). The work on OME and CSOM is invaluable in understanding the causes of HI in the South African population, but further studies are necessary to identify all underlying causes of HI in the population. Identifying the underlying causes of HI will indicate the proportion of HI that is due to genetic factors and the proportion due to environmental factors, which may allow for better surveillance and management of HI in the population.

In addition, ototoxicity was reported by researchers as possible causes of HI in South Africa. Cisplatin-induced HI was observed in some cancer patients. The HI observed with cisplatin usage has been associated with variants in NFE2L2 and SLC22A2 genes (Chang and Chinosornvatana, 2010; Spracklen et al., 2017). Genetic testing for the variants in NFE2L2 and SLC22A2 may be offered to patients before cisplatin-based therapy. This may assist in determining the risk of developing HI following chemotherapy for patients.

Furthermore, recent clinical trials, in the United States and Canada, have shown sodium thiosulfate as a promising agent to protect against HI induced by the essential component of cancer chemotherapy; cisplatin (Freyer et al., 2017) and lovastatin and atorvastatin were shown to protect against/reduce cisplatin-induced ototoxicity in mice and human adults, respectively (Fernandez et al., 2020, 2021). Putatively, a trial of atorvastatin and sodium thiosulfate may be included in the treatment plans of South African patients undergoing cisplatin-based chemotherapy to mitigate HI due to ototoxicity.

The studies considered for the review also reported HI in patients with MDR-TB (Appana et al., 2016; Ghafari et al., 2020). MDR-TB treatment requires the use of strong drugs, such as an aminoglycoside, which are ototoxic (Jiang et al., 2017; Ruhl et al., 2019) and resulted in HI in 82.4% and 100% of patients in the two reviewed studies (Appana et al., 2016; Ghafari et al., 2020). Aminoglycoside antibiotics induce HI by disrupting the intercellular physiological pathways in sensory cells that take up the drug due to the high rate of aminoglycoside trafficking across the endothelial and epithelial barrier layers in humans (Jiang et al., 2017).

The use of aminoglycosides is necessary for the fight against MDR-TB; however, meta-analysis of 87 studies, comprising 12,030 patients with MDR-TB, analyzed the outcomes associated with different drugs (Collaborative Group for the Meta-Analysis of Individual Patient Data, 2018). The analysis found that kanamycin and capreomycin were associated with worse outcomes in the treatment of MDR-TB (Collaborative Group for the Meta-Analysis of Individual Patient Data, 2018) and both medications are no longer recommended for treatment of MDR-TB by the WHO (2018b). It may thus be necessary to update treatment plans to eliminate kanamycin to treat MDR-TB.

This review takes into regard the genetics of HI, where there was a relatively high amount of literature on syndromic HI gathered. In particular, consanguinity was observed in several syndromic studies reported in Africa (Abdi et al., 2016; Ben-Rebeh et al., 2010, 2016; Ben Said et al., 2012; Boulouiz et al., 2007; Bousfiha et al., 2017; Hmani-Aifa et al., 2009; Riahi et al., 2015) and can be considered to have played an active role in the segregation of recessive conditions within the affected family members. Atipo-Tsiba (2016) discussed the impact of consanguinity in his case report on a 40-year-old man and concluded that “Consanguineous marriages are sometimes the source of rare and often serious genetic disease. Doctors, political and religious leaders should join forces to ban such unions between members of the same family.”

Unions between related individuals are considered null and void, within the South African legal context (Wille et al., 2007). It is, however, feasible that individuals were unaware of their degree of relatedness, when taking into consideration the possibility of founder mutations in the South African population, which are associated with population bottleneck (Roberts et al., 2015). An example would be the three Waardenburg families investigated by de Saxe et al. (1984), where genealogical tracing determined that they were from a common ancestor, although no consanguinity occurred.

Waardenburg syndrome, in the studies observed, is the most common genetic syndrome associated with HI in the South African population, with a similar observation being reported in Cameroon (Wonkam Tingang et al., 2020). This may be due to the syndrome being associated with the clinical features that are easily observable, that is, white forelock and heterochromia (Beighton et al., 1991; Winship and Beighton, 1992). Waardenburg is, however, phenotypically variable and Winship and Beighton (1992) have indicated that careful consideration should take place to differentiate the various forms of Waardenburg.

A shortcoming in these studies on Waardenburg syndrome, similar to the study on rod-cone-dystrophy (Beighton et al., 1993) and craniometaphyseal dysplasia (Beighton et al., 1979), is the absence of underlying causative variations. This is in contrast to the indication of the causative variations for Brown-Vialetto-van Laere Syndrome (Chaya et al., 2018) and Noonan Syndrome (Tekendo-Ngongang et al., 2019). Fortunately, the use of next generation sequencing is rapidly helping to narrow the gap in South Africa, as evidenced by the 2020 work of Roberts et al. (2020).

Unlike South Africa, there are molecular data on syndromic HI in other African settings, mostly from North Africa. Genetic studies in Algeria, Tunisia, and Morocco identified causative mutations in Usher 2 and Usher 1 patients (Abdi et al., 2016; Ben-Rebeh et al., 2016; Boulouiz et al., 2007; Bousfiha et al., 2017; Hmani-Aifa et al., 2009; Riahi et al., 2015). Eighteen unrelated Algerian patients, of which 16 had Usher 1 and 2 had Usher 2, were recruited from a larger cohort of HI patients by Abdi et al. (2016). Seventeen putatively pathogenic mutations were identified within the cohort, whereby 16 patients were homozygous for deleterious mutations and 2 patients were compound heterozygous for deleterious mutations (Abdi et al., 2016).

Eight of the variants were novel and putatively deleterious in the patients. These novel variants were in MYO7A, CDH23, PCDH15, USH1C, USH1G, and USH2A (Abdi et al., 2016). Similarly, a large consanguineous Tunisian family segregated a novel GPR98 mutation that segregated with the USH2C phenotype in the family (Hmani-Aifa et al., 2009). The family, additionally, had members who presented with autosomal recessive retinitis pigmentosa, whereby PDE6B was identified as the gene responsible for the visual impairment (Hmani-Aifa et al., 2009).

Variations in MYO7A, USH1C, and PCDh15 were, likewise, identified as possibly causative variations in four consanguineous Tunisian families, with three of the four identified variations being novel (Ben-Rebeh et al., 2016). And private mutations in MYO7A and USH1C, in four unrelated Tunisian families, were determined to be the causative mutations for the Usher 1 phenotype observed (Riahi et al., 2015). Usher 1B and Usher 2C were identified in two separate studies of two consanguineous Moroccan families (Boulouiz et al., 2007; Bousfiha et al., 2017). In the first family, two homozygous variations in MYO7A were identified, of which c.1687G>A was determined to be the disease-causing variation and p.Y1719C was a polymorphism within the population (Boulouiz et al., 2007). In the second family, compound heterozygous mutations, one nonsense and one missense, in GRP68 were identified as the Usher 2C causative mutations in the family (Bousfiha et al., 2017).

El Bouchikhi et al. (2015) undertook mutation screening for putative causative variations in PTPN11 for Noonan Syndrome, in a North African setting. The study identified 19 participants, of which 16 were reached and had genetic testing performed (El Bouchikhi et al., 2015). Congenital heart disease was present in 17 of the 19 patients and 1 patient presented with sensorineural HI (El Bouchikhi et al., 2015). Known heterozygous causative variations were identified in four patients, of which two patients had the same variation and the other two patients had differing variations (El Bouchikhi et al., 2015).

A novel homozygous frameshift mutation was identified as the causative mutation in a consanguineous Tunisian family by Ben Said et al. (2012), in Pendred Syndrome. The family consisted of six affected family members, whereby diagnosis of Pendred syndrome was based on the HI, enlarged vestibular aqueduct, and goiter (Ben Said et al., 2012). Goiter was present in two of the six affected family members, with five patients presenting with profound HI and one patient presenting with severe HI (Ben Said et al., 2012). Enlarged vestibular aqueduct was present in five patients and the five patients were determined to be homozygous for a c.451delG frameshift mutation in SLC26A4 (Ben Said et al., 2012).

The sixth family member was homozygous for p.E47X in GJB2 and had no goiter or enlarged vestibular aqueduct (Ben Said et al., 2012). Charfeddine et al. (2010) and Ben-Rebeh et al. (2010) identified homozygous p.L445W in three and eight consanguineous Tunisian families, respectively, presenting with Pendred Syndrome. There was variable presence of goiter within the 8 consanguineous families, of which only 3 individuals, of the total 41 individuals, presented with goiter (Ben-Rebeh et al., 2010).

Generally, comparative data on syndromic HI have been rather scarce from other African countries. Visual abnormalities have been studied within the Nigerian population, whereby two separate studies screen school-going individuals for visual impairments (Abah et al., 2011; Onakpoya and Omotoye, 2010). Retinitis pigmentosa, and derivatively Usher Syndrome, was identified in 4/620 (Abah et al., 2011) and 1/156 (Onakpoya and Omotoye, 2010).

The studies, however, did not include genetic testing. In Nigeria, BOR was diagnosed in a 4-year-old child who presented with bilateral HI at an ear, nose, and throat specialty hospital (Nasir et al., 2018). The child had bilateral branchial fistulae and preauricular sinuses (Nasir et al., 2018). Her HI was determined to be mixed HI and an ultrasound indicated agenesis of the right kidney (Nasir et al., 2018). The causative variation was determined to be de novo in the child, as both her parents and siblings had no phenotypic manifestations, but genetic testing was unattainable due to the absence of facility at the hospital (Nasir et al., 2018).

Although there are over 120 HI genes globally identified (Van Camp and Smith, 2020), the prevalent genes associated with nonsyndromic HI in South Africa remain unknown. The South African population is genetically diverse, and it is necessary to investigate the contribution of GJB2 and GJB6 at different locations in the country (Bosch et al., 2014a, 2014b; Kabahuma et al., 2011). There was, however, no significant contribution of GJB2 or GJB6 to HI in South Africa (Bosch et al., 2014a, 2014b; Kabahuma et al., 2011). The above observations provide further supporting evidence that the contribution of GJB2 to HI in Africa is negligible, with exceptions from Morocco (Gazzaz et al., 2005; Ratbi et al., 2007), Ghana (Adadey et al., 2019; Brobby et al., 1998; Hamelmann et al., 2001b), Sudan, and Kenya (Gasmelseed et al., 2004).

Furthermore, other HI genes, including REST, CDH23, LOXHD1, OTOF, and SLC26A4, were investigated and found negative within the population (Lebeko et al., 2017; Manyisa et al., 2021). Kabahuma et al. (2021) have pointed to the potential of MYO7A being a prevalent gene within the population. The variations identified within the study, however, need to be ascertained in a larger cohort of unrelated patients presenting with nonsyndromic HI, as was done by Lebeko et al. (2017). The findings, of the abovementioned genetic studies, support the need for large studies in South Africa to understand the genetics of HI within the population and identify prevalent mutations that result in HI in South Africa.

Genetic HI studies within the country may also benefit from modeling the Retinal Degenerative Disease project, which follows patients on a long-term scale (Roberts et al., 2016). The causative mutations that segregate within the family may not necessarily be stagnant and instances where family members carry different mutations may arise, as observed in the family studied by Ben Said et al. (2012), and the presence of both Usher Syndrome and Autosomal recessive retinitis pigmentosa (ARRP) in the family studied by Hmani-Aifa et al. (2009). Long-term follow-up may allow for the larger pedigrees and, coincidently, assist in the study of the genetics of HI within the South African population, likely with the use of whole-exome sequencing.

Conclusions

The study revealed a lack of appropriate prevalence study on HI in South Africa, with the best estimate being 5.5 per 1000 birth. The age at diagnosis is relatively late, around 3 years of age, and the most common environmental factor associated with acquired HI was infection of the middle ear. There were numerous reports on medication toxicity, including kanamycin (treatment used in MDR-TB)-induced ototoxicity, which should deserve specific attention, considering the high burden of TB in South Africa.

The Waardenburg Syndrome is the most commonly reported syndromic HI, but it has not been investigated at the molecular level. Usher Syndrome is the only syndrome that has benefited for intensive genetic study, with identification of a specific founder mutation (MYO7A-c.6377delC), among Black South Africans, and there are emerging data using next generation sequencing on Noonan syndrome as well as Renal dysfunction, rod-cone dystrophy, and sensorineural hearing loss syndrome. GJB2 and GJB6 are not major contributors of nonsyndromic HI among Black South Africans, and the few available data using targeted panel sequencing on limited sample size have shown a very low pick-up rate in known HI genes.

The use of mobile screening approaches in primary health care settings, along with extensive research into the causes of HI in the South African population, is necessary to understand the prevalence and factors that result in HI. Through the study of causes of HI, the preventable cases of HI may be identified, and further work may be done to ease the burden of HI within the population. The use of whole-exome sequencing in multiplex families will be necessary for the better understanding of genetic HI, particularly the nonsyndromic form within South Africa.

Local and Global Impact Statement

HI is a silent planetary health crisis that requires attention within South Africa and internationally. The prevalence of HI in South Africa is estimated to be 5.5 in 1000 live births in South Africa, which is at least five times higher than the prevalence of HI in developed countries. The prevalence of HI within the population requires policy changes to implement systems that will allow for diagnosis and management of HI within the population. There is, however, a dearth of structured information regarding HI within the population and this systematic review remedies this gap by collating key information regarding HI into a single source.

Taken together, this systematic review contributes to the emerging field of planetary health genomics with a focus on HI and offers new insights and lessons learned for future roadmaps on genomics/multiomics and clinical studies of HI around the world.

Footnotes

Authors' Contributions

N.M. performed the literature search and drafted and revised the article; S.M.A, E.T.W., and A.Y. double checked the article selection and performed the quality control analysis. A.W. conceived and supervised the entire project; all authors have read and agreed to the final version of the article, and made a significant intellectual contribution for authorship.

Author Disclosure Statement

The authors declare they have no conflicting financial interests.

Funding Information

This study was funded by the National Institute of Health (NIH, USA), grant number U01-HG-009716 to A.W., and the African Academy of Science and Wellcome Trust, grant number H3A/18/001 to A.W. The funders were not involved in patient recruitment, data generation and analysis, or the decision to publish.

Abbreviations Used

References

Abah

, Oladigbolu

, Samaila

, Merali

, Ahmed

, and Abubakar

. (2011). Ophthalmologic abnormalities among deaf students in Kaduna, Northern Nigeria. Ann Afr Med, 10, 29–33.

Abdelhak

, Kalatzis

, Heilig

, et al. (1997). A human homologue of the Drosophila eyes absent gene underlies branchio-oto-renal (BOR) syndrome and identifies a novel gene family. Nat Genet, 15, 157–164.

Abdi

, Bahloul

, Behlouli

, et al. (2016). Diversity of the genes implicated in Algerian patients affected by Usher Syndrome. PLoS One, 11, e0161893.

Adadey

, Manyisa

, Mnika

, et al. (2019). GJB2 and GJB6 mutations in non-syndromic childhood hearing impairment in Ghana. Front Genet, 10, 841.

Ahituv

, Sobe

, Robertson

, Morton

, Taggart

, and Avraham

. (2000). Genomic structure of the human unconventional myosin VI gene. Gene, 261, 269–275.

American Academy of Pediatrics Joint Committee on Infant Hearing. (2007). Year 2007 position statement: Principles and guidelines for early hearing detection and intervention programs. Pediatrics, 120, 898–921.

Appana

, Joseph

, and Paken

. (2016). An audiological profile of patients infected with multi-drug resistant tuberculosis at a district hospital in KwaZulu-Natal. S Afr J Commun Disord, 63, e1–e12.

Atipo-Tsiba

. (2016). [Consanguineous marriage and morbi-mortality, short literature review based on an exceptional association: Usher syndrome and Von Recklinghausen neurofibromatosis]. Pan Afr Med J, 23, 99.

Beighton

, Bartmann

, Bingham

, and Sellars

. (1993). Rod-cone dystrophy, sensorineural deafness, and renal dysfunction: An autosomal recessive syndrome?. Am J Med Genet, 47, 832–836.

10.

Beighton

, Durr

, and Hamersma

. (1976). The clinical features of sclerosteosis. A review of the manifestations in twenty-five affected individuals. Ann Intern Med, 84, 393–397.

11.

Beighton

, Hamersma

, and Horan

. (1979). Craniometaphyseal dysplasia-variability of expression within a large family. Clin Genet, 15, 252–258.

12.

Beighton

, Viljoen

, Winship

, Beighton

, and Sellars

. (1991). Profound childhood deafness in southern Africa. Ann N Y Acad Sci, 630, 290–291.

13.

Ben-Rebeh

, Grati

, Bonnet

, et al. (2016). Genetic analysis of Tunisian families with Usher syndrome type 1: Toward improving early molecular diagnosis. Mol Vis, 22, 827–835.

14.

Ben-Rebeh

, Yoshimi

, Hadj-Kacem

, et al. (2010). Two missense mutations in SLC26A4 gene: A molecular and functional study. Clin Genets, 78, 74–80.

15.

Ben Said

, Dhouib

, Benzina

, et al. (2012). Segregation of a new mutation in SLC26A4 and p. E47X mutation in GJB2 within a consanguineous Tunisian family affected with Pendred syndrome. Int J Pediatr Otorhinolaryngol, 76, 832–836.

16.

Biagio

, Swanepoel

, Laurent

, and Lundberg

. (2014). Paediatric otitis media at a primary healthcare clinic in South Africa. S Afr Med J, 104, 431–435.

17.

Bosch

, Lebeko

, Nziale

, Dandara

, Makubalo

, and Wonkam

. (2014a). In search of genetic markers for nonsyndromic deafness in Africa: A study in Cameroonians and Black South Africans with the GJB6 and GJA1 candidate genes. OMICS, 18, 481–485.

18.

Bosch

, Noubiap

, Dandara

, et al. (2014b). Sequencing of GJB2 in Cameroonians and Black South Africans and comparison to 1000 Genomes Project Data support need to revise strategy for discovery of nonsyndromic deafness genes in Africans. OMICS, 18, 705–710.

19.

Boulouiz

, Li

, Abidi

, et al. (2007). Analysis of MYO7A in a Moroccan family with Usher syndrome type 1B: Novel loss-of-function mutation and non-pathogenicity of p.Y1719C. Mol Vis 13, 1862–1865.

20.

Bousfiha

, Bakhchane

, Charoute

, et al. (2017). Novel compound heterozygous mutations in the GPR98 (USH2C) gene identified by whole exome sequencing in a Moroccan deaf family. Mol Biol Rep, 44, 429–434.

21.

Brain

. (1927). Heredity in simple goitre. QJM 1927;, 20:303–319.

22.

Brobby

, Muller-Myhsok

, and Horstmann

. (1998). Connexin 26 R143W mutation associated with recessive nonsyndromic sensorineural deafness in Africa. N Engl J Med, 338, 548–550.

23.

Butler

, Basson

, Britz

, De Wet

, Korsten

, and Joubert

. (2013). Age of diagnosis for congenital hearing loss at Universitas Hospital, Bloemfontein. S Afr Med J, 103, 474–475.

24.

Butler

, Ceronio

, Swart

, and Joubert

. (2015). Age of diagnosis of congenital hearing loss: Private v. public healthcare sector. S Afr Med J, 105, 927–929.

25.

Butt

, Greenberg

, Winship

, Sellars

, Beighton

, and Ramesar

. (1994). A splice junction mutation in PAX3 causes Waardenburg syndrome in a South African family. Hum Mol Genet, 3, 197–198.

26.

Canale

, Favero

, Lacilla

, et al. (2006). Age at diagnosis of deaf babies: A retrospective analysis highlighting the advantage of newborn hearing screening. Int J Pediatr Otorhinolaryngol, 70, 1283–1289.

27.

Chan

, and Chang

. (2014). GJB2-associated hearing loss: Systematic review of worldwide prevalence, genotype, and auditory phenotype. Laryngoscope, 124, E34–E53.

28.

Chang

, and Chinosornvatana

. (2010). Practical grading system for evaluating cisplatin ototoxicity in children. J Clin Oncol, 28, 1788–1795.

29.

Charfeddine

, Mnejja

, Hammami

, et al. (2010). Pendred syndrome in Tunisia. Eur Ann Otorhinolaryngol Head Neck Dis, 127, 7–10.

30.

Chaya

, Zampoli

, Gray

, et al. (2018). The first case of riboflavin transporter deficiency in sub-Saharan Africa. Semin Pediatr Neurol, 26, 10–14.

31.

Ching

, Dillon

, Button

, et al. (2017). Age at intervention for permanent hearing loss and 5-year language outcomes. Pediatrics, 140, e20164274.

32.

Chung

, Robinson

, and Morton

. (1959). A note on deaf mutism. Ann Hum Genet, 23, 357–366.

33.

Clarke

, Honey

, Bekker

, et al. (2006). A novel nonsense mutation in the EYA1 gene associated with branchio-oto-renal/branchiootic syndrome in an Afrikaner kindred. Clin Genet, 70, 63–67.

34.

Collaborative Group for the Meta-Analysis of Individual Patient Data in

MDRTBT

, Ahmad

, Ahuja

, et al. (2018). Treatment correlates of successful outcomes in pulmonary multidrug-resistant tuberculosis: An individual patient data meta-analysis. Lancet, 392, 821–834.

35.

D'aguillo

, Bressler

, Yan

, et al. (2019). Genetic screening as an adjunct to universal newborn hearing screening: Literature review and implications for non-congenital pre-lingual hearing loss. Int J Audiol, 58, 834–850.

36.

De Saxe

, Kromberg

, and Jenkins

. (1984). Waardenburg syndrome in South Africa. Part II. Is there founder effect for type I? S Afr Med J, 66, 291.

37.

Eksteen

, Launer

, Kuper

, Eikelboom

, Bastawrous

, and Swanepoel

. (2019). Hearing and vision screening for preschool children using mobile technology, South Africa. Bull World Health Organ, 97, 672–680.

38.

El Bouchikhi

, Samri

, Houssaini

, et al. (2015). The first PTPN11 mutations in hotspot exons reported in Moroccan children with Noonan syndrome and comparison of mutation rate to the previous studies. Turk J Med Sci, 45, 306–312.

39.

Els

, and Olwoch

. (2018). The prevalence and impact of otitis media with effusion in children admitted for adeno-tonsillectomy at Dr George Mukhari Academic Hospital, Pretoria, South Africa. Int J Pediatr Otorhinolaryngol, 110, 76–80.

40.

Everett

, Glaser

, Beck

, et al. (1997). Pendred syndrome is caused by mutations in a putative sulphate transporter gene (PDS). Nat Genet 17, 411—422.

41.

Fernandez

, Spielbauer

, Rusheen

, et al. (2020). Lovastatin protects against cisplatin-induced hearing loss in mice. Hear Res, 389, 107905.

42.

Fernandez

, Allen

, Campbell

, et al. (2021). Atorvastatin is associated with reduced cisplatin-induced hearing loss. J Clin Investig, 131:e142616.

43.

Freyer

, Chen

, Krailo

, et al. (2017). Effects of sodium thiosulfate versus observation on development of cisplatin-induced hearing loss in children with cancer (ACCL0431): A multicentre, randomised, controlled, open-label, phase 3 trial. Lancet Oncol, 18, 63–74.

44.

Gardiner

, Laing

, Mall

, and Wonkam

. (2019). Perceptions of parents of children with hearing loss of genetic origin in South Africa. J Community Genet, 10, 325–333.

45.

Gardner

, Goliath

, Viljoen

, et al. (1997). Familial streptomycin ototoxicity in a South African family: A mitochondrial disorder. J Med Genet, 34, 904–906.

46.

Gasmelseed

, Schmidt

, Magzoub

, et al. (2004). Low frequency of deafness-associated GJB2 variants in Kenya and Sudan and novel GJB2 variants. Hum Mutat, 23, 206–207.

47.

Gazzaz

, Weil

, Rais

, Akhyat

, Azeddoug

, and Nadifi

. (2005). Autosomal recessive and sporadic deafness in Morocco: High frequency of the 35delG GJB2 mutation and absence of the 342-kb GJB6 variant. Hear Res, 210, 80–84.

48.

Ghafari

, Court

, Chirehwa

, et al. (2020). Pharmacokinetics and other risk factors for kanamycin-induced hearing loss in patients with multi-drug resistant tuberculosis. Int J Audiol, 59, 219–223.

49.

Gruber

, Brown

, Mahadevan

, and Neeff

. (2019). Hearing loss and ophthalmic pathology in children diagnosed before and after the implementation of a Universal Hearing Screening Program. Israel Med Assoc J, 21, 607–611.

50.

Haile

, Kamenov

, Briant

, et al. (2021). Hearing loss prevalence and years lived with disability, 1990–2019: Findings from the Global Burden of Disease Study 2019. Lancet, 397, 996–1009.

51.

Hamel

. (2006). Retinitis pigmentosa. Orphanet J Rare Dis, 1, 40.

52.

Hamelmann

, Amedofu

, Albrecht

, et al. (2001a). Pattern of connexin 26 (GJB2) mutations causing sensorineural hearing impairment in Ghana. Hum Mutat, 18, 84–85.

53.

Hamelmann

, Amedofu

, Albrecht

, et al. (2001b). Pattern of connexin 26 (GJB2) mutations causing sensorineural hearing impairment in Ghana. Hum Mutat, 18, 84–85.

54.

Heimler

, and Lieber

. (1986). Branchio-oto-renal syndrome: Reduced penetrance and variable expressivity in four generations of a large kindred. Am J Med Genet, 25, 15–27.

55.

Hmani-Aifa

, Benzina

, Zulfiqar

, et al. (2009). Identification of two new mutations in the GPR98 and the PDE6B genes segregating in a Tunisian family. Eur J Hum Genet, 17, 474–482.

56.

Hoy

, Brooks

, Woolf

, et al. (2012). Assessing risk of bias in prevalence studies: Modification of an existing tool and evidence of interrater agreement. J Clin Epidemiol, 65, 934–939.

57.

Hutchin

, Coy

, Conlon

, et al. (2005). Assessment of the genetic causes of recessive childhood non-syndromic deafness in the UK—Implications for genetic testing. Clin Genet, 68, 506–512.

58.

Jiang

, Karasawa

, and Steyger

. (2017). Aminoglycoside-induced cochleotoxicity: A review. Front Cell Neurosci, 11, 308.

59.

Kabahuma

, Ouyang

, Du

, et al. (2011). Absence of GJB2 gene mutations, the GJB6 deletion (GJB6-D13S1830) and four common mitochondrial mutations in nonsyndromic genetic hearing loss in a South African population. Int J Pediatr Otorhinolaryngol, 75, 611–617.

60.

Kabahuma

, Schubert

W-D

, Labuschagne

, et al. (2021). Spectrum of MYO7A mutations in an indigenous South African population further elucidates the nonsyndromic autosomal recessive phenotype of DFNB2 to include both homozygous and compound heterozygous mutations. Genes, 12, 274.

61.

Kimberling

, Moller

, Davenport

, et al. (1992). Linkage of Usher syndrome type I gene (USH1B) to the long arm of chromosome 11. Genomics, 14, 988–994.

62.

Konrad-Martin

, Helt

, Reavis

, et al. (2005). Ototoxicity: Early detection and monitoring. ASHA Leader, 10, 1–14.

63.

Lebeko

, Manyisa

, Chimusa

, Mulder

, Dandara

, and Wonkam

. (2017). A genomic and protein-protein interaction analyses of nonsyndromic hearing impairment in Cameroon using targeted genomic enrichment and massively parallel sequencing. OMICS, 21, 90–99.

64.

Lebeko

, Sloan-Heggen

, Noubiap

, et al. (2016). Targeted genomic enrichment and massively parallel sequencing identifies novel nonsyndromic hearing impairment pathogenic variants in Cameroonian families. Clin Genet, 90, 288–290.

65.

Levin

, and Klugman

. (1966). Pendred's syndrome in South African Bantu brothers. S Afr Med J, 40, 759–760.

66.

Lim

, Lim

, Kim

, Choi

, Lee

, and Kim

. (2018). Bony cochlear nerve canal stenosis in pediatric unilateral sensorineural hearing loss. Int J Pediatr Otorhinolaryngol, 106, 72–74.

67.

Liu

, Xia

, Ke

, et al. (2002). The prevalence of connexin 26 (GJB2) mutations in the Chinese population. Hum Genet, 111, 394–397.

68.

Louw

, Swanepoel

, Eikelboom

. Self-Reported Hearing Loss and Pure Tone Audiometry for Screening in Primary Health Care

Clinics

. J Prim Care Community

Health

. 2018 Jan–Dec;9:2150132718803156. doi: 10.1177/2150132718803156.

69.

Manyisa

, Schrauwen

, De Souza Rios

, et al. (2021). A monoallelic variant in REST is associated with non-syndromic autosomal dominant hearing impairment in a South African family. Genes, 12, 1765.

70.

Mayosi

, Flisher

, Lalloo

, Sitas

, Tollman

, and Bradshaw

. (2009). The burden of non-communicable diseases in South Africa. Lancet, 374, 934–947.

71.

Melnick

, Bixler

, Nance

, Silk

, and Yune

. (1976). Familial branchio-oto-renal dysplasia: A new addition to the branchial arch syndromes. Clin Genet, 9, 25–34.

72.

Merugumala

, Pothula

, and Cooper

. (2017). Barriers to timely diagnosis and treatment for children with hearing impairment in a southern Indian city: A qualitative study of parents and clinic staff. Int J Audiol, 56, 733–739.

73.

Moher

, Liberati

, Tetzlaff

, Altman

, and Group

. (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med, 6, e1000097.

74.

Morton

, and Nance

. (2006). Newborn hearing screening—A silent revolution. N Engl J Med, 354, 2151–2164.

75.

Morton

. (1991). Genetic epidemiology of hearing impairment. Ann N Y Acad Sci, 630, 16–31.

76.

Najmabadi

, and Kahrizi

. (2014). Genetics of non-syndromic hearing loss in the Middle East. Int J Pediatr Otorhinolaryngol, 78, 2026–2036.

77.

Nasir

, Ladan

, Bemu

, and Jibrin

. (2018). Branchiootorenal syndrome: A case report. Niger Postgrad Med J, 25, 60–62.

78.

Olusanya

, and Newton

. (2007). Global burden of childhood hearing impairment and disease control priorities for developing countries. Lancet, 369, 1314–1317.

79.

Onakpoya

, and Omotoye

. (2010). Screening for ophthalmic disorders and visual impairment in a Nigerian school for the deaf. Eur J Ophthalmol, 20, 596–600.

80.

Pandya

, Arnos

, Xia

, et al. (2003). Frequency and distribution of GJB2 (connexin 26) and GJB6 (connexin 30) mutations in a large North American repository of deaf probands. Genet Med, 5, 295–303.

81.

Pendred

. (1896). Deaf-mutism and goitre. Lancet, 148, 532.

82.

Potgieter

, Swanepoel

, Heinze

, Hofmeyr

, Burger

, and Hamersma

. (2014). An auditory profile of sclerosteosis. J Laryngol, Otol, 1–9.

83.

Qureishi

, Lee

, Belfield

, Birchall

, Daniel

. (2014). Update on otitis media—Prevention and treatment. Infect Drug Resist, 7, 15–24.

84.

Rappoport

. (1970). A coloured family showing features of Waardenburg's syndrome. S Afr Med J, 44, 412–413.

85.

Ratbi

, Hajji

, Ouldim

, Aboussair

, Feldmann

, and Sefiani

. (2007). [The mutation 35delG of the gene of the connexin 26 is a frequent cause of autosomal-recessive non-syndromic hearing loss in Morocco]. Arch Pediatr, 14, 450–453.

86.

Reardon

, Coffey

, Phelps

, et al. (1997). Pendred syndrome—100 years of underascertainment?. QJM, 90, 443–447.

87.

Riahi

, Bonnet

, Zainine

, et al. (2015). Whole exome sequencing identifies mutations in Usher syndrome genes in profoundly deaf Tunisian patients. PLoS One, 10, e0120584.

88.

Roberts

, Allanson

, Tartaglia

, and Gelb

. (2013). Noonan syndrome. Lancet, 381, 333–342.

89.

Roberts

, George

, Greenberg

, and Ramesar

. (2015). A founder mutation in MYO7A underlies a significant proportion of Usher syndrome in indigenous South Africans: Implications for the African Diaspora. Invest Ophthalmol Vis Sci, 56, 6671–6678.

90.

Roberts

, Goliath

, Rebello

, et al. (2016). Inherited retinal disorders in South Africa and the clinical impact of evolving technologies: How human genetics came to SA. S Afr Med J, 106, 33–37.

91.

Roberts

, Julius

, Dawlat

, et al. (2020). Renal dysfunction, rod-cone dystrophy, and sensorineural hearing loss caused by a mutation in RRM2B. Hum Mutat, 41, 1871–1876.

92.

Ruhl

, Du

, Wagner

, et al. (2019). Necroptosis and apoptosis contribute to cisplatin and aminoglycoside ototoxicity. J Neurosci, 39, 2951–2964.

93.

Sandstrom

, Swanepoel

, Laurent

, Umefjord

, and Lundberg

. (2020). Accuracy and reliability of smartphone self-test audiometry in community clinics in low income settings: A Comparative Study. Ann Otol Rhinol Laryngol, 129, 578–584.

94.

Schrijver

. (2004). Hereditary non-syndromic sensorineural hearing loss: Transforming silence to sound. J Mol Diagn, 6, 275–284.

95.

Sellars

, and Beighton

. (1983). The Waardenburg syndrome in deaf children in southern Africa. S Afr Med J, 63, 725–728.

96.

Services USDOHaH. (2017). United States Department of Health and Human Services. Common Terminology Criteria for Adverse Events (CTCAE).

97.

Sohani

, Meyre

, De Souza

, et al. (2015). Assessing the quality of published genetic association studies in meta-analyses: The quality of genetic studies (Q-Genie) tool. BMC Genet, 16, 50.

98.

Spracklen

, Vorster

, Ramma

, Dalvie

, and Ramesar

. (2017). Promoter region variation in NFE2L2 influences susceptibility to ototoxicity in patients exposed to high cumulative doses of cisplatin. Pharmacogenomics J, 17, 515–520.

99.

Stats-SA. (2020). Department: Statistics South Africa. http://www.statssa.gov.za/, Accessed January 8, 2022.

100.

Swanepoel

, Ebrahim

, Joseph

, and Friedland

. (2007). Newborn hearing screening in a South African private health care hospital. Int J Pediatr Otorhinolaryngol, 71, 881–887.

101.

Swanepoel

, Storbeck

, and Friedland

. (2009). Early hearing detection and intervention in South Africa. Int J Pediatr Otorhinolaryngol, 73, 783–786.

102.

Swanepoel

. (2009). Early detection of infant hearing loss in South Africa. S Afr Med J, 99, 158–159.

103.

Tassabehji

, Newton

, and Read

. (1994). Waardenburg syndrome type 2 caused by mutations in the human microphthalmia (MITF) gene. Nat Genet, 8, 251–255.

104.

Tekendo-Ngongang

, Agenbag

, Bope

, Esterhuizen

, and Wonkam

. (2019). Noonan syndrome in South Africa: Clinical and molecular profiles. Front Genet, 10, 333.

105.

Tekin

, Arnos

, and Pandya

. (2001). Advances in hereditary deafness. Lancet, 358, 1082–1090.

106.

Tiedt

, Butler

, Hallbauer

, et al. (2013). Paediatric chronic suppurative otitis media in the Free State Province: Clinical and audiological features. S Afr Med J, 103, 467–470.

107.

Tingang Wonkam

, Chimusa

, Noubiap

, Adadey

, Jv

, and Wonkam

. (2019). GJB2 and GJB6 mutations in hereditary recessive non-syndromic hearing impairment in Cameroon. Genes (Basel), 10, 844.

108.

Tshifularo

, Govender

, and Monama

. (2013). Otolaryngological, head and neck manifestations in HIV-infected patients seen at Steve Biko Academic Hospital in Pretoria, South Africa. S Afr Med J, 103, 464–466.

109.

Usher

. (1914). On the inheritance of retinitis pigmentosa, with notes and cases. https://bjo.bmj.com/content/bjophthalmol/26/7/289.full.pdf, Accessed January 8, 2022.

110.

Van Camp

, and Smith

. (2020). Hereditary Hearing Loss Homepage. https://hereditaryhearingloss.org/, Accessed January 8, 2022.

111.

Viljoen

, Sellars

, and Beighton

. (1983). Familial aggregation of streptomycin ototoxicity: Autosomal dominant inheritance?. J Med Genet, 20, 357–360.

112.

Von Graefe

. (1858). Exceptionelles verhalten des gesichtsfeldes bei pigmententartung der netzhaut. Von Graefes Arch Ophthalmol 4, 250—253.

113.

Vorster

. (2021). South Africans have become poorer over the last 6 years: Government. Business Tech.

114.

Waardenburg

. (1951). A new syndrome combining developmental anomalies of the eyelids, eyebrows and nose root with pigmentary defects of the iris and head hair and with congenital deafness. Am J Hum Genet, 3, 195–253.

115.

Weston

, Kelley

, Overbeck

, et al. (1996). Myosin VIIA mutation screening in 189 Usher syndrome type 1 patients. Am J Hum Genet, 59, 1074–1083.

116.

Whitehorn

, Sibanda

, Lacerda

, et al. (2014). High prevalence of cisplatin-induced ototoxicity in Cape Town, South Africa. S Afr Med J, 104, 288–291.

117.

WHO. (2017). Global costs of unaddressed hearing loss and cost-effectiveness of interventions: A WHO report, 2017. World Health Organization. https://apps.who.int/iris/handle/10665/254659, Accessed January 8, 2022.

118.

WHO. (2018a). Global Health Observatory data repository (Global Health Observatory data repository). https://www.who.int/data/gho, Accessed January 8, 2022.

119.

WHO. (2018b). Rapid Communication: Key changes to treatment of multidrug- and rifampicin-resistant tuberculosis (MDR/RR-TB). https://www.who.int/tb/publications/2019/WHO_RapidCommunicationMDR_TB2019.pdf, Accessed January 8, 2022.

120.

WHO. (2018c). WHO global estimates on prevalence of hearing loss. pp. 1—15. https://www.who.int/pbd/deafness/WHO_GE_HL.pdf, Accessed January 8, 2022.

121.

WHO. (2019). Global Health Observatory data repository (HIV/AIDS). https://www.who.int/data/gho, Accessed January 8, 2022.

122.

WHO. (2020). Deafness and hearing loss. https://www.who.int/news-room/fact-sheets/detail/deafness-and-hearing-loss, Accessed January 8, 2022.

123.

Wille

, Du Bois

, and Bradfield

. (2007). Wille's Principles of South African Law. Cape Town, South Africa, Juta and Company. Ltd.

124.

Wilson

, Tucci

, Merson

, and O'donoghue

. (2017). Global hearing health care: New findings and perspectives. Lancet, 390, 2503–2515.

125.

Winship

, and Beighton

. (1992). Phenotypic discriminants in the Waardenburg syndrome. Clin Genet, 41, 181–188.

126.

Wonkam

, Bosch

, Noubiap

, Lebeko

, Makubalo

, and Dandara

. (2015). No evidence for clinical utility in investigating the connexin genes GJB2, GJB6 and GJA1 in non-syndromic hearing loss in black Africans. S Afr Med J, 105, 23–26.

127.

Wonkam

, Noubiap

N, Djomou

, Fieggen

, Njock

, and Toure

. (2013). Aetiology of childhood hearing loss in Cameroon (sub-Saharan Africa). Eur J Med Genet 56, 20—25.

128.

Wonkam Tingang

, Noubiap

, Jv

, et al. (2020). Hearing impairment overview in Africa: The case of Cameroon. Genes (Basel) 11. Electronic publication, 11:233. doi: 10.3390/genes11020233.

129.

, Wainwright

, and Beighton

. (2009). Tetraphocomelia with the Waardenburg syndrome and multiple malformations. Clin Dysmorphol, 18, 112–115.

130.

Yaniv

. (1987). Tuberculous otitis media: A clinical record. Laryngoscope, 97, 1303–1306.