Iron deficiency at birth and risk of hidden hearing loss in infants modification by socioeconomic status: mother-newborn cohort in Shenyang, China

We conducted a prospective cohort study on auditory neural maturation at birth among mother-newborns with different iron status and SES. The study was conducted at First Affiliated Hospital of China Medical University, in Shenyang, the capital of Liaoning Province in Northeast China, from November 2013 to May 2017. We choose pregnant women aged 19–35, single births, full-term newborns, without deformities. The information of demographic and clinical profiles hemoglobin (Hb), serum ferritin (SF), hearing screening came from Department of Obstetrics and Otolaryngology, First Affiliated Hospital of China Medical University. All protocols were approved by the medical ethics committee of China Medical University [approval number 62033008]. These enrolled pregnant women signed an information and consent form. Maternal inclusion criteria: the pregnant women were 19–35 years old from Northeast China, and enrolled in late pregnancy (32–36 weeks of pregnancy, monozygous). Maternal exclusion criteria: maternal disease history during pregnancy; early birth or overdue birth, low birth weight or macrosomia, perinatal birth injury, dystocia or history of asphyxia and hypoxia, epilepsy, febrile convulsion and other central nervous system diseases, congenital heart disease and other organic diseases after birth metabolic diseases; recurrent diarrhea, infection and other chronic diseases or acute diseases during examination. Offspring exclusion criteria: according to the evaluation criteria of infant growth and development, in infants of the same sex and age, the weight and height were beyond their mean ± 2 standard deviations (SD); anemia and high sensitivity C-reactive protein; hypothyroidism, hyperlead or zinc deficiency, active rickets and other nutritional diseases, genetic metabolic diseases. We did not accept pregnant women with IDA into the investigation and study, and suggested oral ferric salt treatment. For people who met the inclusion criteria but did not participate in the trial, they still underwent the same medical examination as the subjects; Because the subjects included in the cohort were pregnant women who have not reached the status of IDA, there was no clinical iron supplementation treatment plan. It was suggested to strengthen the diet rich in iron, which was easy to absorb. The sample size calculation was based on \(n\frac{={\left({\mu }_{\alpha }\sqrt{2\overline{p }\overline{q} }+{\mu }_{\beta }\sqrt{{p}_{1}{q}_{1}+{p}_{0}{q}_{0}}\right)}^{2}}{{\left({P}_{1}-{P}_{0}\right)}^{2}}\). We selected ID as the exposure factor and HHL as the outcome indicator. Due to the lack of HHL examination data for newborns, mild hearing loss was used as the outcome for estimation. According to our previous study, the incidence of mild hearing loss (p₀) in neonatal hearing screening was 10%, with an estimated RR of 2. The values of each indicator are: α = 0.05, μ_α = 1.96, β = 0.1, μ_β = 1.28, q0 = 0.90, p₀ = 0.10, p1 = 0.20, q₁ = 0.80, \(\overline{p }\)= (0.10 + 0.20)/2 = 0.15, \(\overline{q }\)= 1-\(\overline{p }\)= 0.85. Assuming a 20% loss to follow-up rate, this study required at least 185 participants in each group. The sample size of this study met the requirements. Thus, we excluded 236 participants including 213 participants who did not meet the inclusion criteria, 12 participants who did not completed hearing screening, 11 participants who hid data information. Base previously sample size calculation [10] and the number of covariates included in the final regression model, we assumed a ratio of 1:3 for ID (n = 206) and normal iron status (NIS, n = 653) in our subject population. As a result, 859 mother-newborns were included in the final analyses.

SES is a comprehensive measure, including income, education level and employment status, which has an important impact on people’s health and disease recovery. Because tertiles SES primarily establish childhood health status, and low family SES can lead to poor neurodevelopment of children [18]. We established a comprehensive scoring system of family SES, which is composed of the education and occupation level of each parent in the family, household income in the past year. Annual household income (unit: RMB) was divided into three grades and scored: low (< 10,000/month, 1 point), middle (10,000 ~ 16,000/month, 2 points) and high (16,000 ~ /month, 3 points), the income criteria was based on economic situation of Northeast China. Occupation was divided into three grades and scored: farmers, manual workers, unemployed people (low, 1 point), businessmen or employees (middle, 2 points), leaders, professionals, or government employees (high, 3 points). Education was divided into three grades and scored: lower than high school (low, 1 point), high school graduation or equivalent (middle, 2 points), university graduation or above (high, 3 points). Then we calculated the total SES scores (3–9 points), and divided families into approximate tertiles of SES status distribution: low (≤ 4 points), middle (5–7 points) and high (≥ 8 points).

The ABR and ECochG were performed and analyzed by the audiologists who skilled in administering electrophysiological measurements to newborns between 48 and 72 h after birth.

ABR was performed with GN Otometrics tester. These subjects slept naturally. We used IHS SmartEP auditory evoked potential meter to record in a soundproof and shielded room. The stimulus signal is a short sound with alternating polarity, with a pulse width of 0.1 ms and a repetition rate of 19.3 times/s. The recording electrode was placed in the forehead hairline, the reference electrode was placed in the ipsilateral mastoid, the contralateral mastoid was grounded, and the local skin was degreased with alcohol cotton ball before placing the electrode. Peripheral auditory sensitivity and nerve conduction function of brainstem auditory pathway were mainly detected. The inter pole resistance is less than 10 k Ω. Filter bandpass from 100 to 1500 Hz. We selected the ABR results of one ear with a shorter V-wave latency to measure the absolute latency of wave I, III, and V, as well as IPLs (I-III, III-V, and I-V waves).

ECochG recorded three kinds of potentials, namely AP and SP. The measurement used alternating polarity of short sound stimulation, tympanic membrane electrode. After cleaning the external auditory canal and tympanic membrane with acetone solution, the electrodes were placed on the surface of the posterior lower part of tympanic membrane for recording. Click stimuli (0.1 ms) of 70 dB nHL was presented at a rate of 7.1/s to the test ear through an insert earphone (ER-3A). Waveforms were collected, the stimulus response was passed through a bandpass filter with a range of 5 to 2000 Hz and take the average value. Then SP/AP was calculated. The above outcome assessments were considered as indicators of auditory nerve maturation at different levels of auditory pathways. Impaired auditory neural maturation was defined as high absolute latency wave V, IPL III-V and SP/AP ratio above the 50th percentile [19].

Through face-to-face surveys of mothers, we collected the demographics and life style of mothers [age, gestational age, hemoglobin (Hb), fasting blood glucose (FBG), body mass index (BMI), activities, smoking, drinking, dietary pattern, cesarean delivery, assisted reproductive technology, hypertension], newborns (SF, birth weight, Apgar score, sex), and children [different symptoms of COVID-19 (ever, some throat, earache, hearing loss, ear tightness, resolved sense of smell, nasal conference, false, muscle soreness, cough)].

We used EpiData 4.0 software (EpiData Association, Denmark) to enter data in the form of double entry, and finally selected 10% of the questionnaires for review. SPSS 20.0 software (IBM SPSS, Inc., Chicago, IL, USA) was used for statistical analysis on the data.

The data of normal distribution were expressed as mean ± SD. The data of the skewed distribution (FBG) was expressed as median (interquartile, IQR: Q1-Q3) or as Logarithm transformation (LGFBG) for mean ± SD. The continuous variables with normal distribution were compared by using two-sample t-test between two groups (ID compared with NIS), by Scheffe test of analysis of variance (ANOVA) among iron status quartile (range) and tertiles SES. Other variables with skewed distribution were compared by using Mann–Whitney U test between two groups (ID compared with NIS) or by using Kruskal–Wallis H test among iron status quartile (range) through tertiles SES. For categorical variables, Chi-square test or Fisher’s exact test was used to compare frequency (percentage, %) of partial demographic characteristics. Pearson correlation and intraclass correlation coefficients (ICCs) were used to assess the variability and reproducibility of neonatal SF and maternal Hb by SES tertiles. Parametric Pearson and Non-parametric Spearman correlation coefficients were calculated between iron status and auditory neural maturation by SES.

Multiple logistic regression model was used to assess the effect estimates [odds ratio (OR) and 95% CI] of SES and ID on children OME, respectively. Children OME after COVID-19 was as dependent variable, and ID was as independent variables after controlling for Birth weight, cesarean delivery and assisted reproductive technology. Also, multiple logistic regression of relationship between possible HHL criteria in newborn as dependent variable and ID as independent variables after controlling above covariates.

To examine whether ID mediates a link between SES and HHL prevalence, mediation analyses were performed by using the addition of Process to SPSS. The combined effects of iron status and SES on OME risk and HHL potential diagnostic markers (higher wave V, IPL III-V, and SP/AP ratio) were examined by comparing six combinations of different SES (low, moderate, high) and iron status (ID and NIS). The combination of NIS and high SES was considered as the reference group. Children adenoid hypertrophy, chronic sinusitis and allergic rhinitis may affect their hearing functions, sensitivity analysis was carried out after excluding children with adenoid hypertrophy, chronic sinusitis and allergic rhinitis. In addition, COVID-19 was the main confounding factor of children OME, a sensitivity analysis was conducted by additionally adjusting different symptoms of COVID-19 (ever, some throat, earache, hearing loss, ear tightness, resolved sense of smell, nasal conference, false, muscle soreness, cough).