Lakhasly

Recurrent otitis media and behaviour problems in middle
childhood: A longitudinal cohort study
Ali AH Altamimi ,1,2,3 Monique Robinson,1 Eman MA Alenezi,1,2,4 Tamara Veselinovi
c,1,5 Robyn SM Choi1,5,7 and 1,2,6,7
Christopher G Brennan-Jones
School of 1Telethon Kids Institute, 2Medicine, 5Human Sciences, The University of Western Australia, 6Audiology Department, Perth Children’s Hospital, 7School of Allied Health, Faculty of Health Sciences, Curtin University, Perth, Western Australia, Australia, Faculty of 3Life Sciences, and 4Allied Health Sciences, Kuwait University, Kuwait City, Kuwait
Aim: To investigate the long-term effects of early-life recurrent otitis media (OM) and subsequent behavioural problems in children at the age of 10 years. Methods: Data from the Raine Study, a longitudinal pregnancy cohort, were used to categorise children into those with three or more episodes of OM (rOM group) and those without a history of recurrent OM in the first 3 years of life (reference group). The parent report Strengths and Difficul- ties Questionnaire was used to assess child behaviour at the age of 10 years. Parental questionnaires were used to report past and present diagno- ses of various mental health and developmental conditions, including attention, anxiety, depression, learning, and speech-language problems. Multiple linear and logistic models were used to analyse the data and were adjusted for a fixed set of key confounding variables.
Results: The linear regression analysis revealed significant, independent associations between a history of recurrent OM and higher Strengths and Difficulties Questionnaire scores, including total, internalising, externalising, emotional, attention/hyperactivity and peer problems subscales. Logistic regression analyses revealed an independent increased likelihood for children in the rOM group to have a diagnosis of attention, anxiety, learning and speech-language problems.
Conclusion: Children at 10 years of age with an early history of recurrent OM are more likely to exhibit attentional and behavioural problems when compared to children without a history of recurrent OM. These findings highlight the association between early-life recurrent OM and later behavioural problems that may require professional allied health-care interventions.
Key words: attention; behavioural development; developmental outcome; otitis media.
doi:10.1111/jpc.16518
What is already known on this topic
1 Recurrent otitis media (OM) is common in early childhood.
2 Children with recurrent otitis media can experience long-term
conductive hearing loss.
3 Previous studies have found evidence linking OM-related hearing
loss with long-term developmental problems including behav- iour, language, and mental health.
What this paper adds
1 Further evidence suggests an association between early-life recurrent OM and behaviour problems at the age of 10 years.
2 Parental reports indicate children with an early-life recurrent OM
are at an increased risk of having long-term independent diagno- ses of attention, language, anxiety and learning problems.
Middle ear inflammation, also known as otitis media (OM), is a term that refers to a spectrum of inflammatory conditions and is one of the most common conditions in early childhood.1 A dif- ferentiator between acute OM (AOM) and OM with effusion (OME) is the presence of signs of active infection, such as pain and fever, which are observed in the former and typically absent

Conflict of interest: None declared. Accepted for publication 2 November 2023.
in the latter. However, both forms typically result in effusion in the middle ear, leading to various degrees of conductive hearing loss.2 These symptoms can be prolonged if OM persists or recurs, and may expose children to an increased risk of developmental problems. For example, previous studies that explored the rela- tionship between a history of recurrent and/or persistent OM and behavioural problems reported increased internalising (i.e., inner-directed) and externalising (i.e., outer-directed) behav- ioural problems with emotional, social, attention, and hyperactiv- ity being amongst the most commonly reported problems.3–8 These problems persisted into the early adolescent years, suggesting a potential long-term impact even when any OM- related symptoms have presumably resolved. However, the limi- tations and methodological designs in previous studies do not

provide sufficient evidence to support the relationship between OM and behavioural problems, the extent of these problems, and whether they are of clinical significance. In other words, behav- ioural screening tools are designed to provide information regard- ing present behavioural status but do not confirm behavioural problem diagnoses. Therefore, it is crucial to understand the nature of these associations, if any, as behavioural problems in early childhood may interfere with a child’s academic achieve- ment.9 This study aimed to investigate: (i) whether an early his- tory of recurrent OM can result in behavioural problems in children at 10 years of age; and (ii) the likelihood of these chil- dren receiving a clinical diagnosis for mental health or a develop- mental disorder from a health professional.
Methods
The Raine Study participants
A total of 2900 pregnant women (Gen 1) at 16–20 gestation weeks were recruited in the Raine Study between May 1989 and November 1989 at King Edward Memorial Hospital (KEMH) and
other private clinics in Perth, Western Australia (WA). Gen 1 par- ticipants were required to speak English proficiently to under- stand the implications of participation, had to expect to deliver their child at KEMH and continue residency in WA for follow-up purposes. A total of 2868 children in Gen 2 were comprehen- sively followed up and assessed for various health and develop- ment aspects, including OM and behavioural development. The follow-up process involved surveys that were completed by Gen 1 participants prior to each follow-up visit. These were then verified on the day of the follow-up visit of Gen 2 participants by a research nurse or assistant, who also performed additional assessments. The Raine Study design has been detailed elsewhere.10
Ethical considerations
The human ethics committee at KEMH and Princess Margaret Hospital approved participation and follow-up for the Raine Study. Written informed consent was obtained from all Gen 1 participants at the time of recruitment. Written informed consent was also obtained at each follow-up phase. All Gen 2 participants further consented when they were 18 years of age.

Otitis media and behaviour problems
AAH Altamimi et al.
data for the current study was approved by the Raine Study Sci- entific Review Committee.
The present study
Identification of recurrent OM
To determine the history of OM in children in their first 3 years of life, caregivers were asked, ‘has your child ever had middle ear infections (i.e., OM) in his/her life?’, ‘if yes, how many times?’. On this basis, two groups were formed. Children reported to have three or more episodes of OM (range: 3–8) in the first 3 years of life were classified with recurrent OM (rOM group). Children without a history of OM and those with two or fewer reported episodes of OM were considered the reference group.
Outcome measures
The parent report Strengths and Difficulties Questionnaire (SDQ) and a parental questionnaire (PQ) were administered at the 10-year follow-up. The SDQ is a widely used screening tool to assess behavioural strengths and difficulties in children and adoles- cents aged 2–17 years.11 It consists of five subscales that measure emotional and behavioural functioning and include: emotional symptoms, conduct problems, hyperactivity-inattention, and peer problems, in addition to the prosocial subscale that measures per- sonal strengths. The summing of all subscales, except for the prosocial subscale, generates a total SDQ score. Further, inter- nalising scores can be generated by the sum of the emotional and peer subscales, while externalising scores involve the sum of the
conduct and hyperactivity-inattention subscales. Scoring of the SDQ is based on a 3-point Likert scale where ‘not true’ equals 0, ‘somewhat true’ equals 1 and ‘certainly true’ equals 2. Higher scores are indicative of poorer behavioural outcomes. The SDQ is available in self-report, teacher, and parent report versions (www. sdqinfo.org). The current study also used the PQ to obtain informa- tion regarding current and previous diagnoses of various health and developmental problems. Caregivers were asked, ‘Does your child have now, or has your child had in the past, any of the fol- lowing health professional diagnosed medical conditions or health problems?’. These included anxiety, attention, depression, speech and language, and learning problems, amongst other health and medical conditions. This question was answered: (1) no, (2) yes in the past, (3) yes-now, and (4) yes-now and in the past. This study only considered conditions that were persistent at the age of 10 years by dichotomising the answers to numbers 1 and 2 as ‘no’ and the answers to numbers 3 and 4 as ‘yes’.
Statistical analysis
Descriptive statistics were conducted to summarise the character- istics of the study population and the frequency of each categori- cal variable for the SDQ and the PQ. Multiple linear regression models were used for cross-sectional comparisons of both groups and to predict the change in each SDQ subscale and the total, internalising, and externalising scales. The predictor variable used was a history of recurrent OM in the first 3 years of life.

regression was performed to ascertain the effects of recurrent OM on the likelihood of participants being diagnosed with any of the following mental health and developmental problems: anxiety, depression, attention, learning and speech-language. All regres- sion analyses were adjusted for key confounding variables previ- ously identified in the literature, including sex, ethnicity, exposure to passive smoking, family income, day care atten- dance, premature birth and birthweight.12 For all analyses, a P ≤ 0.05 was considered statistically significant. Data were analysed using IBM SPSS Statistics software version (28.0.1.1), Armonk, NY, USA.
Results
A total of 2121 participants had data available regarding the his- tory of OM in the first 3 years of life. Of those participants, 1755 (82.7%) participants were not reported to have recurrent OM, while 366 (17.3%) were reported to have recurrent OM. A total of 1309 out of 2121 (61.7%) had complete SDQ data and infor- mation on confounding variables recorded, with 1078 (83.4%) in the reference group and 231 (17.6%) in the rOM group (Fig. 1). Further, a total of 1637 out of 2121 (77.2%) participants had complete information on the PQ and confounding variables, with 1347 participants (82.3%) in the reference group and
290 participants (17.7%) in the rOM group. A summary of the sociodemographic variables of the participants is listed in Table 1.
The multiple linear regression analysis revealed a statistically significant association between recurrent OM in the first 3 years of life and the total, internalising, externalising, hyperactivity, emotional, and peer subscales after adjusting for all predictor var- iables. This indicates that children in the rOM group demon- strated greater difficulties on these subscales when compared to the reference group. In contrast, there were no significant differ- ences between the groups on the conduct and prosocial subscales (Table 2).
The logistic regression analysis revealed independent increased odds for the rOM group to have a diagnosis in all of the health and developmental conditions listed in Table 3 whilst adjusting for all confounding variables. However, this was only significant for attention, anxiety, learning, and speech and/or language problems.
Discussion
This study investigated whether early-life recurrent OM was associ- ated with behavioural problems at 10 years of age and whether recurrent OM was associated with an increased likelihood of health professional diagnoses for mental health or developmental problems. The results revealed that children with early-life recurrent OM exhibited poorer internalising and externalising behaviour scores than the reference group, as measured by the SDQ. Additionally, these children were more likely to be diagnosed with anxiety, learn- ing, speech and language, and attention-related problems that were present before and at 10 years of age. The findings suggest that the impact of early-life OM can be severe enough to interfere with chil- dren’s development and may result in detrimental long-term anxi- ety, attention, and developmental problems that may require professional allied health-care interventions (e.g., psychologists, speech-language pathologists, etc.).
Previous investigations of behavioural outcomes in the same cohort revealed that behavioural problems were present in the rOM group at ages 5 and 8, which indicates a continuation of problems since early childhood.8 Conjointly, these findings sug- gest that the cumulative history of OM since early childhood is likely a significant contributor to the behavioural problems observed at 10 years of age, which has also significantly impacted children’s cognitive and educational outcomes.

various environmental and genetic factors not assessed in this study.9 However, behavioural problems in children may manifest due to various OM-related sequelae, such as reduced communi- cation due to hearing loss,4 reduced quality of life,13 school absence due to illness,14 as well as chronic pain and discomfort.
Children in the rOM group were more likely to have attention and hyperactivity problems as measured by the SDQ compared to the reference group. This finding aligns with other studies that revealed an increased likelihood for children and adolescents with early-life recurrent and/or persistent OM to be inattentive and hyperactive.3,4,6,15 In this study, the severity of these issues was further highlighted by the increased rates of diagnoses of attentional problems, indicating that the attention difficulties experienced by children with a history of recurrent OM may have severely impacted their development and academic achieve- ment to the point of warranting professional intervention. Nota- bly, this study is limited by the lack of information regarding the nature and degree of these diagnoses. However, attempts to explain any association between OM and attention problems may lie in the items of the attention-hyperactivity scale of the SDQ. These items describe characteristics that involve restlessness and a reduced attention span. This is similar to symptoms typically seen in children diagnosed with attention deficit hyperactivity disorder (ADHD) and may also overlap with symptoms seen in children with auditory processing difficulties (APD).16 To this connection, several studies have suggested a potential link between early-life recurrent and/or persistent OM with later ADHD17–19 and APD.20–23 In the current study, however, associa- tions between OM and ADHD could not be supported as the pre- dictive validity of the SDQ to screen for ADHD is controversial, with different studies showing different degrees of sensitivity.24 Nevertheless, our results add to the growing body of literature that links a history of OM with attention and hyperactivity prob- lems. This emphasises the need for future studies to implement more rigorous, validated assessments of attention to differentiate between different types of attentional difficulties, if any, and to determine whether attention problems observed in this popula- tion reflect the involvement of other cognitive processes linked to ADHD and APD. This will potentially allow for more appropriate management strategies and maximise efforts to address their edu- cational needs.
Strengths and limitations
One strength of the current study is the use of data from a population-based sample, which reduces selection bias. We were able to control for a range of relevant confounding variables to allow for a more reliable investigation of any associations between OM and behavioural problems. Investigations of diagno- ses of mental health and developmental problems, as opposed to the sole reliance on behavioural questionnaires, further strength- ened our findings and highlighted the level of impact that recur- rent OM may pose on children and their development. However, the lack of information regarding the timing and magnitude of the diagnosed problems limited our understanding of the nature of these associations. A further limitation is the use of parental reports to document a history of recurrent OM, which may be subject to recall bias and limit the ability to explore further associ- ations between frequency, type and severity of OM on the
measured outcomes. Further, the cross-sectional nature of this study did not allow to control for additional known factors that may contribute to behavioural problems, including the degree and type of hearing loss, in addition to other factors that may explain any associations reported in this study.
Conclusion
The present study showed an association between early-life recurrent OM and poorer behavioural scores at the age of 10 years. Additionally, those children were more likely to have anxiety, attention, learning, and language problems as diagnosed by health professionals. These findings expand on the current knowledge regarding this association and recommend early man- agement and frequent monitoring of OM and developmental out- comes of the affected children to mitigate any potential mental health and developmental sequelae that may be associated with OM. This may be accomplished by incorporating screening tools in clinical settings to allow for early intervention through individualised care plans, potentially reducing the impact of these problems on children in the long term.
Acknowledgements
We would like to acknowledge the Raine Study participants and their families for their ongoing participation in the study and the Raine Study team for study co-ordination and data collection. We also thank the NHMRC for their long-term contribution to funding the study over the last 30 years. The core management of the Raine Study is funded by The University of Western Aus- tralia, Curtin University, Telethon Kids Institute, Women and Infants Research Foundation, Edith Cowan University, Murdoch University, The University of Notre Dame Australia and the Raine Medical Research Foundation. Chris Brennan-Jones was supported by a National Health and Medical Research Council Fellowship (GNT 1142897) and a Western Australian Department of Health Emerging Leader Fellowship. Open access publishing facilitated by The University of Western Australia, as part of the Wiley - The University of Western Australia agreement via the Council of Australian University Librarians. [Correction added on November 18, 2023, after first online publication: CAUL funding statement has been added.