Central Auditory Tests to Track Cognitive Function in People With HIV: Longitudinal Cohort Study

Machine learning for predicting cognitive deficits using auditory and demographic factors

IMPORTANCE

Predicting neurocognitive deficits using complex auditory assessments could change how cognitive dysfunction is identified, and monitored over time. Detecting cognitive impairment in people living with HIV (PLWH) is important for early intervention, especially in low- to middle-income countries where most cases exist. Auditory tests relate to neurocognitive test results, but the incremental predictive capability beyond demographic factors is unknown.

OBJECTIVE

Use machine learning to predict neurocognitive deficits, using auditory tests and demographic factors.

SETTING

The Infectious Disease Center in Dar es Salaam, Tanzania.

PARTICIPANTS

Participants were 939 Tanzanian individuals from Dar es Salaam living with and without HIV who were part of a longitudinal study. Patients who had only one visit, a positive history of ear drainage, concussion, significant noise or chemical exposure, neurological disease, mental illness, or exposure to ototoxic antibiotics (e.g., gentamycin), or chemotherapy were excluded. This provided 478 participants (349 PLWH, 129 HIV-negative). Participant data were randomized to training and test sets for machine learning.

MAIN OUTCOME(S) AND MEASURE(S)

The main outcome was whether auditory variables combined with relevant demographic variables could predict neurocognitive dysfunction (defined as a score of <26 on the Kiswahili Montreal Cognitive Assessment) better than demographic factors alone. The performance of predictive machine learning algorithms was primarily evaluated using the area under the receiver operational characteristic curve. Secondary metrics for evaluation included F1 scores, accuracies, and the Youden's indices for the algorithms.

RESULTS

The percentage of individuals with cognitive deficits was 36.2% (139 PLWH and 34 HIV-negative). The Gaussian and kernel naïve Bayes classifiers were the most predictive algorithms for neurocognitive impairment. Algorithms trained with auditory variables had average area under the curve values of 0.91 and 0.87, F1 scores (metric for precision and recall) of 0.81 and 0.76, and average accuracies of 86.3% and 81.9% respectively. Algorithms trained without auditory variables as features were statistically worse (p < .001) in both the primary measure of area under the curve (0.82/0.78) and the secondary measure of accuracy (72.3%/74.5%) for the Gaussian and kernel algorithms respectively.

CONCLUSIONS AND RELEVANCE

Auditory variables improved the prediction of cognitive function. Since auditory tests are easy-to-administer and often naturalistic tasks, they may offer objective measures or predictors of neurocognitive performance suitable for many global settings. Further research and development into using machine learning algorithms for predicting cognitive outcomes should be pursued.

A Central Auditory Test reveals differences between drug treatment regimens in adults living with HIV

OBJECTIVE

This exploratory study examined whether central auditory tests show differences between people living with HIV (PLWH) treated with two predominant antiretroviral drug therapy (ART) regimens.

DESIGN

Cross-sectional.

STUDY SAMPLE

253 PLWH (mean age 39.8 years) from the Shanghai Public Health Clinical Centre, China.

METHODS

The Hearing in Noise Test speech reception threshold (SRT) assessed central auditory function and the Montreal Cognitive Assessment (MoCA) assessed cognition. The relationship between ART regimen and SRT was evaluated with multivariable linear regression incorporating age, HIV duration, and peripheral hearing ability. Multivariable logistic regression was used to ascertain if SRT and ART regimen predicted MoCA impairment.

RESULTS

The two predominant ART regimens differed by one drug (zidovudine or tenofovir). Participants taking the zidovudine-containing regimen had poorer SRT performance (p=.012) independent of age and hearing thresholds. MoCA scores did not differ between drug regimens, but a negative relationship was found between SRT and MoCA impairment (p=.048).

CONCLUSIONS

ART regimens differed in their association with central auditory test performance likely reflecting neurocognitive changes in PLWH taking the zidovudine-containing regimen. Central auditory test performance also marginally predicted cognitive impairment, supporting further assessment of central auditory tests to detect neurocognitive deficits in PLWH.

Gap detection responses modelled using the Hill equation in adults with well-controlled HIV

OBJECTIVE

This study's objective was determining whether gap detection deficits are present in a longstanding cohort of people living with HIV (PLWH) compared to those living without HIV (PLWOH) using a new gap detection modelling technique (i.e. fitting gap responses using the Hill equation and analysing the individual gap detection resulting curves with non-linear statistics). This approach provides a measure of both gap threshold and the steepness of the gap length/correct detection relationship.

DESIGN

The relationship between the correct identification rate at each gap length was modelled using the Hill equation. Results were analysed using a nonlinear mixed-effect regression model.

STUDY SAMPLE

45 PLWH (age range 41-78) and 39 PLWOH (age range 38-79) were enrolled and completed gap detection testing.

RESULTS

The likelihood ratio statistic comparing the full regression model with the HIV effects to the null model, assuming one population curve for both groups, was highly significant (p < 0.001), suggesting a less precise relationship between gap length and correct detection in PLWH.

CONCLUSIONS

PLWH showed degraded gap detection ability compared to PLWOH, likely due to central nervous system effects of HIV infection or treatment. The Hill equation provided a new approach for modelling gap detection ability.

Peripheral Auditory Function in Tanzanian Children Living With HIV With Clinically Normal Hearing

IMPORTANCE

Despite normal audiometry, adults living with HIV have lower distortion product otoacoustic emissions (DPOAEs) compared with HIV-negative controls, but the degree of these differences in children living with HIV is unknown. If subclinical auditory deficits are present, results could affect developmental outcomes in children living with HIV (CLWH).

OBJECTIVE

To compare DPOAEs and auditory brainstem responses (ABR) between 2 age- and sex-matched groups of younger children with normal audiometry, 1 infected with HIV and the other uninfected.

DESIGN, SETTING, AND PARTICIPANTS

Cohort study in an infectious disease center in Dar es Salaam, Tanzania. Participants included 340 Tanzanian children aged 3 to 9 years with clinically normal hearing, type A tympanograms bilaterally, and air-conduction thresholds of 20 dB HL or less from 0.5 to 8 kHz. Participants in the cohort repeated testing approximately every 6 months (approximately 2.2 sessions per participant) for a total of 744 total observations. Data were analyzed from March 2020 to January 2022.

MAIN OUTCOMES AND MEASURES

DPOAE amplitudes from 1.5 to 8 kHz using an f2 to f1 ratio of 1.2 and L1/L2 values of 65/55 dB sound pressure level and click-evoked ABR using a slow (21.1/s) and fast (61.1/s) click rate.

RESULTS

A total of 141 CLWH (70 female participants [49.3%]; mean [SD] age, 7.24 [1.67] years) and 199 HIV-negative individuals (99 female participants [49.7%]; mean [SD] age, 7.26 [1.44] years) participated in the study. The groups did not differ significantly in age, static immittance, or air-conduction thresholds. HIV status was independently associated with approximately 1.4 dB (95% CI, -3.28 to 0.30 dB) to 3.8 dB (95% CI, 6.03 to -1.99 dB) lower DPOAE amplitudes at 6 and 8 kHz bilaterally and 0.28 μV (95% CI, 0.01 to 0.33 μV) lower ABR wave V amplitudes in the right ear.

CONCLUSIONS AND RELEVANCE

Consistent with previous findings in young adults, CLWH had slightly, but reliably, lower DPOAEs and ABR wave V amplitudes than HIV-negative controls. The magnitude of these differences was small, but results suggest an early and consistent association between HIV infection or treatment and outer hair cell and auditory brainstem responses in children as young as 3 years. These subclinical changes suggest tracking both auditory function and development outcomes in CLWH is warranted.

The Relationship Between Central Auditory Tests and Neurocognitive Domains in Adults Living With HIV

Objective: Tests requiring central auditory processing, such as speech perception-in-noise, are simple, time efficient, and correlate with cognitive processing. These tests may be useful for tracking brain function. Doing this effectively requires information on which tests correlate with overall cognitive function and specific cognitive domains. This study evaluated the relationship between selected central auditory focused tests and cognitive domains in a cohort of normal hearing adults living with HIV and HIV- controls. The long-term aim is determining the relationships between auditory processing and neurocognitive domains and applying this to analyzing cognitive function in HIV and other neurocognitive disorders longitudinally. Method: Subjects were recruited from an ongoing study in Dar es Salaam, Tanzania. Central auditory measures included the Gap Detection Test (Gap), Hearing in Noise Test (HINT), and Triple Digit Test (TDT). Cognitive measures included variables from the Test of Variables of Attention (TOVA), Cogstate neurocognitive battery, and Kiswahili Montreal Cognitive Assessment (MoCA). The measures represented three cognitive domains: processing speed, learning, and working memory. Bootstrap resampling was used to calculate the mean and standard deviation of the proportion of variance explained by the individual central auditory tests for each cognitive measure. The association of cognitive measures with central auditory variables taking HIV status and age into account was determined using regression models. Results: Hearing in Noise Tests and TDT were significantly associated with Cogstate learning and working memory tests. Gap was not significantly associated with any cognitive measure with age in the model. TDT explained the largest mean proportion of variance and had the strongest relationship to the MoCA and Cogstate tasks. With age in the model, HIV status did not affect the relationship between central auditory tests and cognitive measures. Age was strongly associated with multiple cognitive tests. Conclusion: Central auditory tests were associated with measures of learning and working memory. Compared to the other central auditory tests, TDT was most strongly related to cognitive function. These findings expand on the association between auditory processing and cognitive domains seen in other studies and support evaluating these tests for tracking brain health in HIV and other neurocognitive disorders.

Peripheral Auditory Function in Young HIV-Positive Adults With Clinically Normal Hearing

OBJECTIVE

Little is known about peripheral auditory function in young adults with HIV, who might be expected to show early evidence of hearing loss if HIV infection or treatment does affect peripheral function. The goal of this study was to compare peripheral auditory function in 2 age- and gender-matched groups of young adults with clinically normal hearing with and without HIV.

STUDY DESIGN

Matched cohort study with repeated measures.

SETTING

Infectious disease center in Dar es Salaam, Tanzania.

METHODS

Participants included HIV-positive (n = 38) and HIV-negative (n = 38) adults aged 20 to 30 years who had clinically normal hearing, defined as type A tympanograms, air conduction thresholds ≤25 dB HL bilaterally from 0.5 to 8 kHz, and distortion product otoacoustic emissions (DPOAEs) >6 dB above the noise floor bilaterally from 1.5 to 8 kHz. Participants were tested multiple times over 6-month intervals (average, 2.7 sessions/participant) for a total of 208 observations. Primary outcome measures included tympanograms, air conduction audiograms, DPOAEs, and click-evoked auditory brainstem responses.

RESULTS

HIV groups did not significantly differ in age, static immittance, or air conduction thresholds. HIV-positive status was independently associated with approximately 3.7-dB lower DPOAE amplitudes from 2 to 8 kHz (95% CI, 1.01-6.82) in both ears and 0.04-µV lower (95% CI, 0.003-0.076) auditory brainstem response wave I amplitudes in the right ear.

CONCLUSION

Young adults living with HIV have slightly but reliably smaller DPOAEs and auditory brainstem response wave I amplitudes than matched HIV-negative controls. The magnitude of these differences is small, but these results support measuring peripheral auditory function in HIV-positive individuals as they age.

Nonverbal cognitive assessment of children in Tanzania with and without HIV

Children living with HIV can experience cognitive difficulties. Most neuropsychological tests have been constructed in Western languages, meaning they may not be appropriate for use in non-Western settings. To address this, we used an entirely nonverbal measure of cognitive ability in a sub-Saharan African sample. For this cross-sectional analysis, 316 children (162 HIV+ and 154 HIV-, ages 3-9) completed the Leiter-3 as part of a larger study in Dar es Salaam, Tanzania. Statistical tests included analysis of covariance and multiple linear regression to account for environmental variables. HIV+ children performed worse than HIV - controls on two composite scores: Nonverbal IQ (p < .001) and Processing Speed (p < 0.001). Similar trends were observed on core subtests. Multiple linear regression models revealed that age, socioeconomic status, and school attendance predicted all Leiter-3 test composites. Critically, the addition of HIV status to the models improved prediction of Nonverbal IQ (∆R² = 0.03, p = .001) and Processing Speed (∆R² = 0.06, p < .001). Children living with HIV performed worse than HIV- controls on most Leiter-3 measures. While age, SES, and school attendance predicted Leiter-3 performance, HIV status improved prediction capabilities when added to the model. The Leiter-3 may offer a viable measure of cognitive ability in non-Western settings that can be used in its original form without translation.