Cognitive performance in adolescents with perinatally-acquired HIV and associations with oral health in Nigeria

Abstract

Objectives:

To compare cognitive performance between adolescents with and without perinatally acquired HIV and to evaluate the association between oral health and cognitive performance.

Methods:

A cross-sectional analysis was conducted using baseline data from a study of 118 age- and sex-matched adolescents (59 PHIV, 59 HUU) aged 9.5–16.5 years in Benin City, Nigeria. Cognitive function was assessed using two tablet-based tools: NeuroScreen and the NIH Toolbox Cognition Battery. Oral health was evaluated via clinical examination using the Decayed, Missing, and Filled Teeth (DMFT) and oral hygiene questionnaires. ANOVA and ANCOVA were used to examine associations between HIV status, oral health, and cognitive outcomes.

Results:

Adolescents with PHIV demonstrated significantly lower cognitive performance compared to their HUU peers, particularly in processing speed, executive function, and language domains. NeuroScreen assessments revealed longer completion times in Trail Making tasks (TM1: 44.8 vs 37.7, adjusted p=0.039; TM2: 51.4 vs 42.5, adjusted p = 0.018; TM3: 25.6 vs 19.8, adjusted p = 0.027) and lower scores in Visual Discrimination (VD2: 19.3 vs 21.9, adjusted p = 0.002) and Number Speed tests (NSd: 53.1 vs 46.1, adjusted p = 0.008) among adolescents with PHIV. In addition, NIH Toolbox results indicated lower scores across all cognitive domains for adolescents with PHIV. Higher DMFT scores were associated with poorer executive function, and this relationship was observed only among adolescents with PHIV in stratified analyses, across both the NeuroScreen and NIH Toolbox assessments.

Conclusion:

Adolescents with PHIV exhibited notable neurocognitive deficits and a tendency toward poorer oral health compared with uninfected peers. Poor oral health correlated with diminished cognitive performance, underscoring the need for integrated care models that address both neurocognitive and oral health in youth living with HIV.

Introduction

Increased access to effective antiretroviral therapy (ART) has led to a growing population of adolescents (10-19 years of age) with perinatally-acquired HIV (PHIV),^1,2 of whom about 89% reside in high HIV-burden African countries.³ Adolescence is associated with unique developmental events, including accelerated physiological, social, sexual and cognitive maturation.⁴ Neurocognitive performance in adolescents living with HIV (ALHIV) may have long-term impacts on viral suppression, adherence to medication, and poor decision-making that further increase HIV transmission risk.

Cognitive deficits associated with perinatal HIV have been well-documented in pediatric populations across various global settings.^5–10 However, comprehensive data on these cognitive impairments are limited in African countries and for ALHIV, who are at a critical developmental stage where cognitive functioning is complex and essential for future social and professional success. Recent tablet-based cognitive assessment technologies, including NIH Toolbox Cognition Battery¹¹ and NeuroScreen, provide promising methods to evaluate cognitive performance in adolescents in African settings.

Several complex factors can influence cognitive performance, including ART history, nutrition, sociodemographic and behavioral factors. The oral-brain axis is majorly unexplored in individuals with HIV. Oral health emerges as an intriguing but understudied area, as ART-associated oral health-related side effects have also been reported. For example, reduced salivary flow among people with HIV taking protease inhibitors and didanosine¹² might predispose them to higher risks of dental caries and other oral diseases.¹³ Chronic inflammation is a hallmark of both oral disease and HIV progression, and may serve as a mediator of neurological complications.¹⁴ Periodontal disease triggers inflammation, oxidative stress, and neuronal damage, which can impact cognitive function. While poor oral health, including periodontal disease and tooth loss, is linked to lower cognitive function in adults, the relationship remains unclear, especially in youth.^15–20

Our study had two objectives: first, to compare cognitive performance of adolescents with PHIV with that of adolescents unexposed and uninfected (HUU) using two tablet-based assessments, and second, to assess associations between oral health and neurocognitive function among them.

Materials and Methods:

Study Participants

This study includes data from the baseline visit of the “Uncovering the Biological Link between Oral and Mental Health in Adolescents Living with HIV” (uBLOoM) study, which enrolled 118 participants aged 9.5-16.5 years, including 59 with PHIV and 59 without HIV. Participants were recruited from the University of Benin Teaching Hospital (UBTH) which is home to existing cohorts.^21–24 Adolescents with PHIV receiving HIV care and treatment at the Special Treatment Center were approached for study recruitment; HUU adolescents were recruited from the immunization center and well-child clinics at UBTH for comparison. Participants in the HUU group were matched by sex and age (+/− 0.5 year) to adolescents with PHIV.

Measurements

Demographics and Medical History:

A questionnaire was administered to collect demographic information, including sex, age, education, school type, and race/ethnicity. Socioeconomic data were also obtained, including maternal education level, maternal employment status, and monthly income. Additional information was collected on antibiotic use, birth history, and history of HPV infection. For adolescents with PHIV, detailed clinical data were collected, including age at initiation of ART, type and duration of ART regimen, and the most recent laboratory measurements (absolute CD4 count and viral load).

Oral Health Assessments:

The Oral Health questionnaire, which was adapted from the WHO Oral Health Survey (5th Edition)²⁵, was administered to collect information on oral hygiene practices, included self-rated oral health (Likert scale) and dental service utilization. Four senior dental registrars conducted the oral examinations. Examiners were trained and re-trained biannually and calibrated against a gold-standard examiner using three pediatric cases. Inter-rater agreement was assessed with unweighted Cohen’s kappa statistic, which was 0.40 or greater.²⁶ Caries was measured via standardized oral clinical examinations to detect cavitated lesions by examiners blinded to HIV status.^27,28 Oral hygiene status was categorized as good, fair, or poor based on the oral hygiene score. Caries detection was assessed at the person-level, and was defined as the sum of cavitated caries lesions to determine the number of decayed, missing, and filled teeth (DMFT index).²⁹

Cognitive Function:

We utilized two tools to assess cognitive function using touchscreen tablets. NeuroScreen (Version 4.4.9)³⁰ has been validated for assessing neurocognitive impairment among people with HIV in the US, Thailand, and South Africa.^31–33 NeuroScreen consists of 12 tests assessing six neurocognitive domains: Trail Making (TM1, TM2, TM3) measures processing speed and executive functioning; Visual Discrimination (VD1, VD2) and Number Speed (NSd) assess processing speed; Number Span (NSnF, NSnB) evaluates working memory; Verbal List Learning (VLL, VLLD) measures learning and memory; and Finger Tapping (FTD, FTND) gauges motor speed.³¹

The NIH Toolbox Cognition Battery (V3 3.10.1)^11,34 was used to assess multiple cognitive domains, including executive function and attention (Flanker Inhibitory Control and Attention Test), cognitive flexibility (Dimensional Change Card Sort Test), working memory (List Sorting Working Memory Test), processing speed (Pattern Comparison Processing Speed Test), episodic memory (Picture Sequence Memory Test), and language ability (Picture Vocabulary Test and Oral Reading Recognition Test). Composite scores were calculated following the NIH Toolbox manual. The Fluid Cognition Composite included the Flanker, Dimensional Change Card Sort, Picture Sequence Memory, List Sorting, and Pattern Comparison tests. The Crystallized Cognition Composite included the Picture Vocabulary and Oral Reading Recognition tests. The Total Cognition Composite was derived as the average of the Fluid and Crystallized composites. Scores from each subtest were age-adjusted and standardized according to NIH Toolbox norms. All assessments were administered in English, which is the primary language of instruction and official communication in the study setting. A detailed description of individual subtests, scoring methods, and score ranges is provided in Appendix Tables S1–S2.

Statistical Analysis

Demographic characteristics, clinical characteristics, and cognitive function test scores for adolescents in PHIV and HUU groups were summarized and compared using Pearson Chi-square tests or Fisher exact tests for categorical variables, and t tests for continuous measures. The association between HIV status, oral hygiene scores, and each of the cognitive function test scores was assessed by conducting ANOVA tests. For the adjusted analysis, ANCOVA was used, accounting for education level, body mass index (BMI), and antibiotic use. Stratified analyses by HIV status were conducted based on priori hypotheses. Sensitivity analyses further adjusted for maternal education and employment status to account for socioeconomic factors. To evaluate whether the association between oral health and cognition differed by HIV status, interaction terms (oral health × HIV group) were included in additional ANCOVA models. Given multiple cognitive outcomes, p-values were corrected for multiple comparisons using the Benjamini–Hochberg false discovery rate (FDR) procedure, with q < 0.05 considered statistically significant. All statistical analyses were conducted using R (version 4.5.1, Vienna, Austria).

Results

We collected data for all 118 uBlOoM study adolescents, including 59 with PHIV and 59 age- and sex-matched HUU participants. As shown in Table 1, the groups did not differ by age or sex. Adolescents with PHIV were more often in primary school (77.1% vs. 55.0%, p = 0.018). Maternal characteristics also differed: no mothers of adolescents with PHIV had education beyond secondary school versus 35.8% in the HUU group (p < 0.001), and they were more likely to hold unskilled/manual jobs (p = 0.021). Mean BMI was slightly lower among adolescents with PHIV (14.7 vs. 15.4, p = 0.032), and recent antibiotic use was more common (72.1% vs. 10.0%, p < 0.001). Oral health indices, including DMFT and oral hygiene scores, were similar between groups. Among adolescents with PHIV, 8.5% received Antiretroviral prophylaxis at birth; mean ART duration was 7.5 years (SD 3.3) with initiation at 4.4 years (SD 3.2). Two-thirds used TDF/3TC/DTG and one-third ABC/3TC/DTG. Most achieved viral suppression (<20 copies/mL, 79.7%), though mean CD4 counts remained lower than in HUU peers (561 vs. 939 cells/μL, p < 0.001).

Table 1:

Characteristics of Study Participants at Baseline, by HIV Group

	Total N [%]	PHIV n (%)	HUU n (%)	p-value
Sample size	118 [100.0]	59 [50.0]	59 [50.0]
Enrolment period	25oct2023 – 01mar2024	25oct2023 – 28feb2024	25oct2023 – 01mar2024
Age (years):  Mean {SD}	11.9 {1.0}	11.9 {1.0}	11.8 {0.9}	0.334T
Range	(9.5, 16.5)	(9.5, 16.5)	(9.6, 13.2)
Mother enrolled
No	32 [27.1]	29 [49.2]	3 [5.1]	<0.001 F
Yes	86 [72.9]	30 [50.8]	56 [94.9]
Maternal education level
Above secondary education	19 [23.2]	0	19 [35.8]	<0.001 F
Completed secondary	42 [51.2]	17 [58.6]	25 [47.2]
Below secondary education	21 [25.6]	12 [41.4]	9 [17.0]
Maternal employment status				0.021 F
Clerical	9 [11.0]	0	9 [17.0]
Professional	6 [7.3]	1 [3.4]	5 [9.4]
Skilled manual	8 [9.8]	5 [17.2]	3 [5.7]
Unskilled work	59 [72.0]	23 [79.3]	36 [67.9]
Maternal monthly income:  Mean {SD}	54134.3 {57199.2}	45207.4 {56726.6}	59018.9 {57401.3}	0.298T
Gender
Male	46 [40.0]	23 [40.0]	23 [40.0]	1.000C
Female	72 [60.0]	36 [60.0]	36 [60.0]
Education level
Primary	77 [65.3]	45 [76.3]	32 [54.2]	0.020C
JSS (Junior Secondary School)	41 [34.7]	14 [23.7]	27 [45.8]
Type of school				0.841C
Private	81 [68.6]	40 [67.8]	41 [69.5]
Public	34 [28.8]	17 [28.8]	17 [28.8]
Delivery method
Vaginal	103 [87.3]	55 [93.2]	48 [81.4]	0.094F
C-section	15 [13.7]	4 [6.7]	11 [18.6]
Born prematurely
No	118 [98.3]	58 [96.7]	60 [100.0]	0.496F
Yes	2 [1.7]	2 [3.3]	0
ARV prophylaxis administered at birth
Yes		5 [8.5]
No		54 [91.5]
Mother on ART during pregnancy
Yes		8 [13.6]
No		51 [86.4]
Body mass index - mean {SD}	15.1 {2.0}	14.7 {2.0}	15.5 {2.0}	0.033 T
HIV RNA viral load (copies/ml)
<20		47 [79.7]
20+		12 [20.3]
CD4 (cells/μL)– mean {SD}	749.8 {317.2}	561.0 {216.1}	938.7 {289.4}	<0.001 T
Antibiotics (<30 days)
No	70 [59.3]	17 [29.3]	53 (89.8]	<0.001 F
Yes	48 [40.7]	42 (70.7]	6 [10.2]
ART duration (years) - mean {SD}		7.54 {3.25}
ART age initiation (years)- mean {SD}		4.40 {3.19}
Range		(0.44, 11.79)
antiretroviral therapy (ART) regimen
ABC/3TC/DTG		18 [30.5]
TDF/3TC/DTG		41 [69.5]
Any HPV
No	109 [92.4]	56 [94.9]	53 [89.8]	0.490F
Yes	9 [7.6]	3 [5.1]	6 [10.2]
PHQ-9 (group)
Mild	1 [0.8]	1 [1.7]	0 [0]	1.000F
None – minimal	117 [99.2]	58 [98.3]	59 [100.0]
PHQ-9 (score) - mean {SD}	1.6 {1.0}	1.7 {1.1}	1.6 {0.9}	0.453T
DMFT score - mean {SD}	0.3 {0.8}	0.4 {0.8}	0.3 {0.8}	0.577T
0	97 [82.2]	47[83.1]	50 [84.7]	0.631F
1-4	21 [17.8]	12 [16.9]	9[15.3]
Oral Hygiene score - mean {SD}	1.59 {0.92}	1.65 {0.86}	1.53 {0.98}	0.499T
Good	46 [39.0]	22 [37.3]	24 [40.7]	0.922F
Fair	63 [53.4]	32 [54.2]	31 [52.5]
Poor	9 [7.6]	5 [8.5]	4 [6.8]

N [%] =proportions [row percentages]; n (%) =proportions (column percentages); {SD}=standard deviation;

^T=T test;

^F=Fisher’s exact test;

^C=Pearson Chi-square test; DMFT=Decayed, Missing, and Filled Teeth; HPV=Human papillomavirus; PHQ-9=Patient Health Questionnaire 9.

Figure 1 presents the NeuroScreen Test Performance scores, highlighting group differences in cognitive processing speed, executive function, and memory between adolescents with PHIV and their HUU peers. The TM task showed that participants with PHIV performed worse, with higher scores in TM1 (p=0.046), TM2 (p=0.020), and TM3 (p=0.027), indicating slower task completion and reduced executive function. In the VD task, adolescents with PHIV performed significantly worse than their HUU peers, particularly on VD2 (p = 0.003), suggesting reduced visual processing efficiency. The NSd score was significantly higher in the participants with PHIV (p=0.012), indicating slower cognitive processing. In contrast, no significant differences were observed between the NSnF, NSnB, VLL, or VLLD. Similarly, FTD and FTND, which measure motor function, did not show significant performance gaps between participants with PHIV vs HUU.

Figure 1:

Boxplot of Neuroscreen Test Performance by HIV Group

Figure 2 presents the results of the NIH Toolbox Test Performance. Adolescents with PHIV exhibited lower scores across nearly all domains. Participants with PHIV had significantly lower ORR and PV scores, reflecting language deficits, with greater variability in the HUU group. Similarly, participants with PHIV consistently scored lower in Fluid Cognition tasks, highlighting deficits in executive function and cognitive flexibility. The LSWM scores were also lower in participants with PHIV, suggesting impairments in working memory capacity. The Total Cognition Composite score further confirmed these disparities, as participants with PHIV exhibited substantially lower median scores and a more compressed distribution compared with HUU peers.

Figure 2:

Boxplot of Age-Adjusted Standard Scores in NIH Toolbox Test Performance by HIV Group

As shown in Table 2, adolescents with higher DMFT scores (poorer oral health) exhibited poorer performance on several NeuroScreen tasks. Mean TM2 scores were higher (worse performance) in the high-DMFT group compared to the low-DMFT group (59.6 vs. 44.2; adjusted p=0.001), with the most pronounced difference observed in participants with PHIV (69.7 vs. 46.7; adjusted p<0.001). TM3 scores were significantly higher in the HUU group with elevated DMFT (26.1 vs. 18.6; adjusted p=0.039), suggesting impairment in complex cognitive tasks. Additionally, VD2 scores and NSnB scores were lower in the high DMFT group, indicating deficits in visual discrimination and working memory, respectively. VLL scores were significantly lower in adolescents with higher DMFT (8.5 vs. 9.2; adjusted p=0.028). In the NIH Toolbox assessments, similar patterns were observed. Among participants with PHIV, Total Cognition Composite scores were significantly lower in those with higher DMFT scores (49.5 vs. 61.2; adjusted p=0.039). Executive function and attention, assessed via DCCS and Flanker tasks, were also lower in participants with PHIV with poorer oral health (DCCS: 79.4 vs. 87.2; adjusted p=0.021; Flanker: 61.8 vs. 73.5; adjusted p=0.009). The Fluid Composite score was also lower in the high-DMFT group among all participants (75.4 vs. 82.7; adjusted p=0.032). These findings suggest a consistent association between poorer oral health and lower cognitive performance. In stratified analyses, this relationship was observed only among adolescents with PHIV, primarily in tests assessing executive function, specifically the TM2 from the NeuroScreen and the DCCS and Flanker tasks from the NIH Toolbox.

Table 2:

NeuroScreen Test Performance and NIH Toolbox Cognition Battery Performance by DMFT Score by HIV Status

DMFT Score	Total N [%]				PHIV N [%]				HUU N [%]
	0	1-4	p-value	p-value (Adjusted)	0	1-4	p-value	p-value (Adjusted)	0	1-4	p-value	p-value (Adjusted)
Sample size	97 [82.2]	21 [17.8]			47 [79.7]	12 [20.3]			50 [84.7]	9 [15.3]
Neuroscreen score
TM 1 – mean {SD}	39.22 {17.65}	50.61 {24.73}	0.015 A	0.012 †	42.46 {18.33}	54.03 {24.83}	0.075A	0.064†	36.18 {16.60}	46.04 {25.31}	0.137A	0.129†
TM 2 – mean {SD}	44.20 {17.99}	59.63 {28.18}	0.002 A	0.001 †	46.70 {16.63}	69.74 {30.87}	<0.001 A	<0.001 †	41.85 {19.06}	46.16 {17.80}	0.532A	0.521†
TM 3 – mean {SD}	21.92 {14.45}	26.35 {14.35}	0.205A	0.200†	25.41 {18.71}	26.56 {12.67}	0.842A	0.839†	18.64 {7.58}	26.07 {17.14}	0.036 A	0.039 †
VD 1 – mean {SD}	11.76 {4.80}	9.95 {3.35}	0.103A	0.097†	11.64 {5.90}	9.58 {3.29}	0.252A	0.261†	11.88 {3.51}	10.44 {3.57}	0.265A	0.239†
VD 2 – mean {SD}	21.03 {4.77}	18.62 {4.58}	0.036 A	0.023 †	19.68 {4.73}	17.92 {4.46}	0.248A	0.190†	22.30 {4.49}	19.56 {4.82}	0.100A	0.085†
NSd – mean {SD}	48.43 {15.07}	55.11 {15.25}	0.069A	0.051†	51.34 {16.46}	60.10 {15.25}	0.101A	0.091†	45.70 {13.22}	48.45 {13.20}	0.568A	0.543†
NSnF – mean {SD}	3.71 {1.05}	3.48 {1.12}	0.360A	0.349†	3.62 {1.09}	3.08 {1.16}	0.142A	0.135†	3.80 {1.01}	4.00 {0.87}	0.580A	0.580†
NSnB – mean {SD}	1.30 {1.25}	0.71 {0.72}	0.041 A	0.039 †	1.26 {1.22}	0.83 {0.72}	0.259A	0.240†	1.34 {1.29}	0.56 {0.73}	0.082A	0.084†
VLL – mean {SD}	9.22 {1.18}	8.52 {1.78}	0.030 A	0.028 †	9.17 {1.43}	8.17 {2.17}	0.058A	0.055†	9.26 {0.90}	9.00 {1.00}	0.435A	0.438†
VLLD – mean {SD}	4.38 {1.50}	4.43 {1.16}	0.892A	0.892†	4.19 {1.61}	4.08 {1.44}	0.833A	0.835†	4.56 {1.37}	4.89 {0.33}	0.480A	0.486†
FTD – mean {SD}	210.11 {31.54}	208.48 {25.23}	0.824A	0.818†	211.72 {32.90}	207.00 {23.94}	0.643A	0.619†	208.60 {30.47}	210.44 {28.22}	0.867A	0.861†
FTND – mean {SD}	189.87 {26.14}	179.19 {24.71}	0.090A	0.070†	189.45 {22.61}	174.83 {26.89}	0.060A	0.032 †	190.26 {29.30}	185.00 {21.59}	0.610A	0.595†
NIH Toolbox Age-Adjusted Standard Score
ORR – mean {SD}	79.42 {34.83}	69.67 {36.43}	0.251A	0.187†	65.89 {27.61}	52.50 {22.71}	0.127A	0.113†	92.14 {36.35}	92.56 {39.72}	0.975A	0.973†
PV – mean {SD}	54.74 {15.83}	55.67 {20.58}	0.819A	0.818†	50.66 {14.16}	46.00 {13.42}	0.309A	0.315†	58.58 {16.49}	68.56 {22.00}	0.118A	0.124†
Crystallized Composite – mean {SD}	66.86 {27.82}	61.38 {29.88}	0.421A	0.371†	55.47 {19.73}	46.83 {15.72}	0.166A	0.163†	77.56 {30.16}	80.78 {33.94}	0.773A	0.761†
DCCS – mean {SD}	89.27 {11.60}	84.95 {11.59}	0.132A	0.125†	87.21 {10.56}	79.36 {7.51}	0.024 A	0.021 †	91.20 {12.31}	91.78 {12.38}	0.897A	0.896†
Flanker – mean {SD}	76.38 {14.23}	69.62 {14.96}	0.053A	0.043 †	73.49 {13.89}	61.75 {12.88}	0.010 A	0.009 †	79.14 {14.13}	80.11 {10.71}	0.846A	0.842†
LSWM – mean {SD}	91.63 {14.36}	89.33 {16.49}	0.520A	0.507†	87.67 {13.27}	84.67 {10.77}	0.472A	0.459†	95.26 {14.49}	95.56 {21.07}	0.958A	0.959†
PC – mean {SD}	91.06 {12.13}	87.33 {10.36}	0.194A	0.188†	88.36 {10.85}	87.67 {11.49}	0.846A	0.843†	93.60 {12.82}	86.89 {9.28}	0.140 A	0.140 †
PSM Form A – mean {SD}	90.96 {12.21}	85.95 {11.19}	0.095A	0.090†	87.45 {13.16}	87.33 {11.99}	0.978A	0.978†	93.90 {10.60}	83.88 {10.29}	0.016 A	0.016 †
Fluid Composite – mean {SD}	82.74 {13.83}	75.42 {13.72}	0.038 A	0.032 †	78.90 {13.56}	70.45 {13.82}	0.074A	0.065†	85.96 {13.34}	82.25 {10.93}	0.460A	0.454†
Total Cognition Composite – mean {SD}	70.19 {21.70}	60.47 {23.91}	0.084A	0.057†	61.22 {17.36}	49.45 {14.94}	0.046 A	0.039 †	77.69 {22.26}	75.63 {26.42}	0.813A	0.806†

^A=Anova test;

^†=ANCOVA, adjusted for education level, BMI, and antibiotic use; DCCS=Dimensional Change Card Sort; DMFT=Decayed, Missing, and Filled Teeth; Flanker=Flanker Inhibitory Control and Attention; FTD=Finger Tapping: Dominant; FTND=Finger Tapping: Nondominant; HPV=Human Papillomavirus; LSWM=List Sorting Working Memory; NSd=Number Speed; NSnB=Number Span Backward; NSnF=Number Span Forward; ORR=Oral Reading Recognition; PC=Pattern Comparison Processing Speed; PHQ-9=Patient Health Questionnaire 9; PSM Form A=Picture Sequence Memory Form A; PUFA=Pulpal Involvement, Ulceration, Fistula, and Abscess; PV=Picture Vocabulary; TM=Trail Making; VD=Visual Discrimination; VLL=Verbal List Learning; VLLD=Verbal List Learning Delayed;

Table 3 examines the relationship between oral hygiene and cognitive function. Although no significant differences were found across the three oral hygiene groups, participants with poor hygiene generally performed worse on NeuroScreen tasks, particularly in the TM assessments. Among participants with PHIV, TM1 scores were higher (worse) in the poor-hygiene group than in the good-hygiene group (57.4 vs 45.8; adjusted p = 0.648), with a similar but nonsignificant pattern for TM2 and TM3. However, in the VD2 task, participants with PHIV and poor hygiene performed significantly worse than those with good hygiene, with performance improving as oral hygiene status improved (p=0.034; adjusted p=0.017). In the NIH Toolbox results, Total Cognition Composite scores declined as oral-hygiene status worsened, reaching marginal significance (unadjusted p = 0.092; adjusted p = 0.085).

Table 3:

NeuroScreen Test Performance and NIH Toolbox Cognition Battery Performance by Oral Hygiene Score Status and Stratified by HIV Status

Oral Hygiene Score Status	Total					PHIV					HUU
	Good	Fair	Poor	p-value	p-value (Adjusted)	Good	Fair	Poor	p-value	p-value (Adjusted)	Good	Fair	Poor	p-value	p-value (Adjusted)
NeuroScreen score
TM 1 - mean {SD}	41.30 {20.24}	40.47 {19.59}	46.41 {15.28}	0.732A	0.722†	45.80 {24.53}	42.16 {17.53}	57.42 {8.67}	0.666A	0.648†	37.17 {14.65}	38.73 {21.67}	32.66 {8.27}	0.927A	0.926†
TM 2 - mean {SD}	44.68 {21.90}	47.65 {19.18}	53.63 {27.41}	0.234A	0.221†	49.67 {25.45}	49.97 {17.75}	68.05 {29.11}	0.246A	0.244†	40.12 {17.37}	45.25 {20.57}	35.61 {9.68}	0.734A	0.726†
TM 3 - mean {SD}	20.53 {10.41}	24.12 {17.23}	23.94 {10.31}	0.245A	0.239†	23.25 {12.81}	26.42 {21.20}	31.22 {7.20}	0.333A	0.327†	18.05 {6.99}	21.75 {11.73}	14.83 {3.91}	0.634A	0.641†
VD 1 - mean {SD}	11.22 {3.51}	11.73 {5.39}	10.56 {3.81}	0.940A	0.939†	10.73 {3.52}	12.06 {6.69}	8.00 {2.92}	0.841A	0.843†	11.67 {3.52}	11.39 {3.69}	13.75 {1.71}	0.618A	0.598†
VD 2 - mean {SD}	21.17 {4.25}	20.46 {5.14}	18.67 {5.02}	0.168A	0.134†	20.64 {4.29}	18.97 {4.82}	15.80 {4.02}	0.034 A	0.017 †	21.67 {4.24}	22.00 {5.07}	22.25 {3.86}	0.758A	0.750†
NSd - mean {SD}	47.40 {13.81}	51.39 {16.44}	48.58 {13.46}	0.358A	0.322†	49.28 {15.64}	55.21 {17.42}	56.65 {13.04}	0.190A	0.175†	45.68 {11.96}	47.45 {14.60}	38.49 {3.50}	0.745A	0.728†
NSnF - mean {SD}	3.65 {0.82}	3.70 {1.14}	3.56 {1.59}	0.974A	0.973†	3.45 {0.91}	3.66 {1.12}	2.80 {1.79}	0.658A	0.654†	3.83 {0.70}	3.74 {1.18}	4.50 {0.58}	0.570A	0.571†
NSnB - mean {SD}	1.22 {1.21}	1.17 {1.16}	1.22 {1.48}	0.921A	0.920†	1.27 {1.32}	1.22 {1.07}	0.40 {0.55}	0.260A	0.242†	1.17 {1.13}	1.13 {1.26}	2.25 {1.71}	0.351A	0.358†
VLL - mean {SD}	9.00 {1.45}	9.27 {0.95}	8.33 {2.45}	0.772A	0.770†	9.00 {1.75}	9.19 {1.18}	7.40 {2.97}	0.270A	0.271†	9.00 {1.14}	9.35 {0.66}	9.50 {1.00}	0.128A	0.125†
VLLD - mean {SD}	4.65 {1.06}	4.16 {1.71}	4.67 {0.71}	0.314A	0.312†	4.59 {1.10}	3.84 {1.85}	4.40 {0.89}	0.274A	0.284†	4.71 {1.04}	4.48 {1.50}	5.00 {0.00}	0.893A	0.895†
FTD - mean {SD}	213.52 {30.58}	207.27 {30.20}	208.78 {32.82}	0.372A	0.355†	215.41 {34.53}	207.13 {28.81}	213.60 {33.83}	0.552A	0.525†	211.79 {27.09}	207.42 {32.05}	202.75 {35.50}	0.498A	0.477†
FTND - mean {SD}	192.39 {26.33}	184.63 {26.11}	188.67 {24.25}	0.253A	0.222†	191.64 {26.36}	182.88 {23.30}	186.80 {16.53}	0.326A	0.270†	193.08 {26.86}	186.45 {29.01}	191.00 {34.52}	0.538A	0.518†
NIH Toolbox Age-Adjusted Standard Score
ORR - mean {SD}	77.41 {30.49}	77.70 {37.21}	79.00 {46.10}	0.917A	0.904†	65.77 {23.76}	64.72 {30.05}	41.80 {4.76}	0.200A	0.187†	88.08 {32.47}	91.10 {39.53}	125.50 {21.17}	0.170A	0.138†
PV - mean {SD}	49.59 {14.45}	48.95 {13.12}	55.67 {17.31}	0.471A	0.469†	47.00 {10.59}	52.28 {16.32}	45.20 {9.07}	0.578A	0.586†	59.75 {15.43}	58.77 {17.16}	72.50 {31.46}	0.479A	0.484†
Crystallized Composite - mean {SD}	64.35 {24.56}	66.52 {29.35}	69.22 {38.60}	0.590A	0.547†	54.27 {16.80}	55.63 {21.41}	39.00 {0.00}	0.345A	0.344†	73.58 {27.13}	77.77 {32.39}	107.00 {23.42}	0.113A	0.098†
DCCS - mean {SD}	90.02 {13.02}	88.10 {10.65}	83.89 {10.79}	0.151A	0.144†	87.73 {12.78}	85.23 {8.22}	80.00 {11.42}	0.140A	0.134†	92.13 {13.15}	90.97 {12.08}	88.75 {8.92}	0.595A	0.590†
Flanker - mean {SD}	75.18 {15.16}	75.48 {14.57}	72.89 {12.02}	0.825A	0.816†	71.45 {16.93}	71.38 {12.97}	67.80 {13.44}	0.723A	0.716†	78.74 {12.61}	79.71 {15.12}	79.25 {6.85}	0.841A	0.837†
LSWM - mean {SD}	92.29 {16.40}	90.29 {13.10}	92.33 {17.79}	0.693A	0.683†	87.29 {16.11}	88.09 {10.89}	79.40 {4.83}	0.472A	0.459†	96.67 {15.69}	92.55 {14.89}	108.50 {13.63}	0.780A	0.785†
PC - mean {SD}	91.76 {11.66}	89.94 {12.56}	86.67 {6.93}	0.220A	0.214†	90.64 {10.02}	87.06 {11.88}	85.00 {6.52}	0.178A	0.174†	92.79 {13.12}	92.90 {12.75}	88.75 {7.80}	0.726A	0.727†
PSM Form A - mean {SD}	91.36 {12.82}	88.98 {11.73}	90.75 {11.91}	0.499A	0.491†	90.67 {13.88}	84.93 {11.54}	88.50 {15.15}	0.265A	0.265†	91.96 {12.08}	92.90 {10.68}	93.00 {9.38}	0.765A	0.765†
Fluid Composite - mean {SD}	83.07 {14.09}	80.72 {14.13}	78.25 {13.64}	0.287A	0.268†	80.50 {16.26}	75.89 {11.92}	68.75 {12.50}	0.097A	0.087†	85.30 {11.81}	85.23 {14.71}	87,75 {6.08}	0.837A	0.835†
Total Cognition Composite - mean {SD}	68.70 {21.26}	68.26 {22.24}	69.13 {30.46}	0.979A	0.977†	61.90 {19.83}	58.79 {15.66}	42.50 {7.00}	0.092A	0.085†	74.61 {21.09}	77.10 {24.00}	95.75 {15.00}	0.183A	0.171†

^A=Anova test;

^†=ANCOVA, adjusted for education level, BMI, and antibiotic use; DCCS=Dimensional Change Card Sort; Flanker=Flanker Inhibitory Control and Attention; FTD=Finger Tapping: Dominant; FTND=Finger Tapping: Nondominant; HPV=Human Papillomavirus; LSWM=List Sorting Working Memory; NSd=Number Speed; NSnB=Number Span Backward; NSnF=Number Span Forward; ORR=Oral Reading Recognition; PC=Pattern Comparison Processing Speed; PHQ-9=Patient Health Questionnaire 9; PSM Form A=Picture Sequence Memory Form A; PUFA=Pulpal Involvement, Ulceration, Fistula, and Abscess; PV=Picture Vocabulary; TM=Trail Making; VD=Visual Discrimination; VLL=Verbal List Learning; VLLD=Verbal List Learning Delayed;

Sensitivity analyses adjusting for maternal education and employment yielded consistent results (Appendix Table S3).

After false-discovery-rate (FDR) correction for multiple comparisons, none of the associations remained statistically significant, although the direction and magnitude of effects were unchanged (Appendix Table S4).

Interaction analyses including oral-health × HIV-status terms indicated that the association between oral health and cognitive performance did not differ significantly by HIV status (Appendix Table S5).

Discussion

This study examined neurocognitive function and oral health among adolescents with PHIV compared with age- and sex-matched HUU peers, revealing three key findings: 1) adolescents with PHIV demonstrated lower performance across several cognitive domains, particularly executive function and processing speed, as measured by both the NeuroScreen and NIH Toolbox assessments; 2) poor oral health, as measured by caries burden with DMFT scores, was associated with worse cognitive performance, particularly in executive function tasks; and 3) the oral health-cognition associations were primarily observed among adolescents with PHIV, suggesting greater vulnerability in this group. These results suggest that HIV contributes to both neurocognitive impairments and poorer oral health outcomes, emphasizing the need for integrated care approaches that address both domains.

Consistent with prior studies demonstrating neurodevelopmental vulnerability in youth with perinatally acquired HIV^{10,32,35–37}, adolescents with PHIV performed worse on tasks measuring executive function and processing speed, including Trail Making, Visual Discrimination, and Number Speed. These domains were also among those most impaired in the Adolescents and Adults Living with Perinatal HIV (AALPHI) Cohort in England³⁶, suggesting that deficits in higher-order cognitive skills may represent a persistent neurodevelopmental consequence of perinatal HIV, even in youth receiving ART. Lower scores in language and fluid cognition domains on the NIH Toolbox further highlight broad cognitive impacts in this population.

Although mean DMFT and oral hygiene scores were slightly higher among adolescents with PHIV, group differences were modest and not statistically significant. The direction of the difference aligns with prior research reported by our group^21,38–40 and others^41–44, which may be linked to immune suppression, medication effects, and reduced access to preventive dental care. Adolescents with PHIV also reported higher antibiotic use, possibly due to greater susceptibility to bacterial infections from incomplete immune recovery despite ART, or clinicians’ lower threshold for empirical antibiotic treatment.^45,46

One of our study’s novel findings is the observed association between oral health outcomes and neurocognitive performance. Higher DMFT scores were associated with worse performance on NeuroScreen tasks, particularly Trail Making and Visual Discrimination, indicating slower processing and reduced executive control. Adolescents with higher DMFT also had lower working memory (NSnB) and verbal learning (VLL) scores. In the NIH Toolbox assessments, adolescents with poorer oral health scored lower on the Total Cognition Composite, as well as executive function measures (DCCS, Flanker). These findings align with a broader literature documenting oral–brain connections across the lifespan, including studies linking periodontal disease and tooth loss to poorer cognitive outcomes in children and older adults.^20,47–55 These findings suggest that oral health and neurocognitive outcomes are interconnected, potentially through shared inflammatory and immune pathways.^15–20 While causation cannot be inferred, these consistent patterns across age groups underscore oral health as a potential indicator of broader neurodevelopmental well-being.

To assess whether HIV status modified the association between oral health and cognition, we examined stratified models and formal interaction terms. Although the formal interaction between HIV status and oral health was not statistically significant (Appendix Table S5), the pattern of results from stratified analyses (Tables 2–3) suggested that the associations between poorer oral health and lower cognitive performance were more pronounced among adolescents with PHIV. This pattern may reflect heightened vulnerability to systemic inflammation or immune-related neurodevelopmental disruptions in this group. Sensitivity analyses adjusting for maternal education and employment (Appendix Table S3) yielded directionally consistent but attenuated results, indicating partial mediation by socioeconomic context. After adjusting for multiple comparisons using the false discovery rate (Appendix Table S4), no associations remained statistically significant, though the pattern of effects was stable across tests. These findings suggest that while socioeconomic and statistical corrections reduce significance, the observed oral–cognition associations are robust in direction and biologically plausible.

Several mechanisms may underlie the neurocognitive and oral health challenges observed among adolescents with PHIV. Persistent immune activation and neuroinflammation, even in virally suppressed individuals, can alter neuronal signaling and synaptic pruning during adolescence^37,56–58, while simultaneously exacerbating oral inflammatory processes. Chronic oral inflammation can increase systemic inflammatory markers (IL-6, TNF-α, CRP) that cross the blood-brain barrier and activate neuroinflammatory cascades. This process may be amplified by HIV-related immune dysregulation and increased intestinal/oral permeability. Additionally, oral dysbiosis, altered oral microbiome composition documented in previous studies by our group and others^21,22,59 may contribute to systemic inflammation and metabolic changes affecting brain function. Certain antiretrovirals (e.g., ritonavir) have limited CNS penetration, potentially allowing localized viral persistence that further contributes to cognitive vulnerability. Evidence linking missing teeth or reduced masticatory efficiency to poorer cognition^60,61 suggests additional functional pathways, although causality remains uncertain.⁶² Collectively, these mechanisms indicate that adolescents with PHIV may be especially susceptible to oral–brain interactions, reinforcing the need for integrated neurological and dental care.

This study has several strengths, including the use of a well-matched control group, which minimizes the influence of confounding factors such as age and sex; for example, males, on average, performed better in psychomotor domains in similar HIV studies.⁶³ Furthermore, the inclusion of participants with similar educational backgrounds helped reduce the potential impact of education on cognitive performance. These design features, combined with the use of comprehensive neurocognitive assessment tools, allowed for a thorough evaluation of cognitive function across multiple domains.

It is important to note some limitations. Our findings may not be generalizable to all adolescents with PHIV, especially those from different cultural, educational or socioeconomic backgrounds. The cross-sectional design of this study precludes causal inferences about the relationship between oral health and cognition. Our sample size, while adequate for detecting moderate effect sizes, may have limited statistical power for robust subgroup analyses (observed with attenuation of associations after FDR correction) and likely influenced the distribution of scores on tests like VLL. The cognitive assessment tools used in this study have not been formally validated in adolescents living in Nigeria, raising potential concerns about cultural relevance and score interpretability. Moreover, the normative data used for age-adjusted scoring were based on U.S. reference populations, which may not fully reflect regional cognitive baselines.

Our study highlights the substantial neurocognitive and oral health needs of adolescents with PHIV and provides actionable insights for clinical practice and research. By demonstrating consistent links between oral health indicators and cognitive performance, particularly in executive function, our findings suggest that routine dental assessment could serve as an early, non-invasive signal for broader neurodevelopmental concerns in this population. These results support the integration of oral health screening into HIV care settings and underscore the value of multidisciplinary models that coordinate dental, neurocognitive, and educational support. Rather than simply identifying gaps, our findings point toward concrete next steps: incorporating oral health into adolescent HIV care, testing targeted interventions, and advancing mechanistic research to better understand and ultimately mitigate neurodevelopmental challenges faced by youth living with PHIV.

List of Acronyms

ALHIV

adolescents living with HIV

ART

antiretroviral therapy

ASS

Age-Adjusted Standard Score

CSS

Change Sensitive Score

DCCS

Dimensional Change Card Sort

DMFT

Decayed, Missing, and Filled Teeth

Flanker

Flanker Inhibitory Control and Attention

FTD

Finger Tapping: dominant

FTND

Finger Tapping: nondominant

HPV

Human papillomavirus

NIHTB-CB

NIH Toolbox Cognition Battery

NSd

Number Speed

NSnB

Number Span Backward

NSnF

Number Span Forward

ORR

Oral Reading Recognition

PC

Pattern Comparison Processing Speed

PHQ-9

Patient Health Questionnaire 9

PHIV

perinatally-acquired HIV

PSM Form A

Picture Sequence Memory Form A

PV

Picture Vocabulary

TM

Trail Making

VD

Visual Discrimination

VLL

Verbal List Learning

VLLD

Verbal List Learning Delayed