Dental Caries in Association With Viral Load in Children Living With HIV in Phnom Penh, Cambodia

Background: Oral health status is known to be associated with overall health among people living with HIV. However, it is unclear whether dental caries is associated with viral load among this population. In particular, dental caries among children living with HIV needs to be better understood, as this can affect their overall health and well-being in future. This study assessed the association between dental caries and viral load among children living with HIV in Phnom Penh, Cambodia. Methods: This cross-sectional study was conducted at the National Paediatric Hospital as a baseline survey of a randomized controlled trial. The study population included children living with HIV aged 3–15 years and their primary caregivers. We collected data on the children’s oral health status by oral examination and the latest HIV viral load data stored in the patients’ information system of the hospital. Multiple logistic regression analysis was conducted to assess association between dental caries and viral load. The cut-off point of undetectable viral load was set at <40 copies/mL. Results: Data from 328 children were included in the analyses. Moreover, 68.3% had an undetectable viral load. The mean number of permanent or deciduous teeth with caries was 7.7 (standard deviation [SD], 5.0). In the regression analysis, dental caries in permanent or deciduous teeth were positively associated with detectable viral load (adjusted odds ratio [AOR]: 1.07, 95% condence interval [CI]: 1.01, 1.14). Having received antiretroviral therapy ≥ 1 year and self-reported excellent adherence to the antiretroviral drug were also negatively associated with detectable viral load. Among children with detectable levels of viral load, dental caries in permanent or deciduous teeth were also positively associated with non-suppression of viral load (>1000 copies/mL) (AOR: 1.12, 95% CI: 1.03, 1.23). Conclusions: Dental caries was associated with viral load status detection among children living with HIV. This nding suggests that oral health status may affect the immune status of the children. Oral health of children living with HIV should be strengthened, and further research is needed to clarify the causal relationship between viral load and oral health status. 0.65) were negatively associated with viral load detection. The model that included children with deciduous teeth (n = 188) showed that dmft was positively associated with viral load detection (AOR: 1.10, 95% CI: 1.01, 1.19). More than one year on ART (AOR: 0.19, 95% CI: 0.07, 0.57) and self-reported excellent adherence to antiretroviral drugs (AOR: 0.21, 95% CI: 0.04, 0.97) were negatively associated with viral load detection. The model that included children with all types of teeth (n = 328) showed that total DMFT/dmft was positively associated with viral load detection (AOR: 1.07, 95% CI: 1.01, 1.14), and that more than one year on ART (AOR: 0.22, 95% CI: 0.09, 0.52) and self-reported excellent adherence to antiretroviral drugs (AOR: 0.16, 95% CI: 0.05, 0.51) were negatively associated with viral load detection.


Background
Globally, AIDS-related deaths decreased by 35%, from 1.7 million in 2004 (the highest rate) to 1.1 million in 2018 [1]. This change implies that people living with HIV now have a longer life expectancy than previously reported [2] and as a result, the number of people living with HIV is higher than ever [1]. Therefore, it is important that they live a healthy life. One important requirement to ensure overall health is to maintain oral health since some chronic diseases (e.g. cardiovascular disease, diabetes) are associated with dental caries or other oral diseases [3][4][5]. Furthermore, severe childhood caries was found to be associated with malnutrition such as iron, vitamin D, calcium, or albumin de ciency [6,7].
The presence of oral lesions indicates HIV progression and an increase in plasma viral load among people living with HIV [8][9][10]. Among these patients, oral hairy leukoplakia and oral candidiasis are suggested to be associated with a decreased viral load [11,12]. However, the association between viral load and dental caries remains unclear, especially in children living with HIV. Childhood is a crucial period in which physical developments determine future oral health status and, consequently, may in uence overall health.
Cambodia is one of the Southeast-Asian countries that have implemented effective HIV prevention and treatment approaches. In 2018, there was a 62% decrease in new HIV infection cases among the general population of all ages, including children, compared with 2010 [1]. Of all people living with HIV, 81% received antiretroviral therapy, and 78% had suppressed viral load in 2018 [1]. However, a high prevalence of dental caries has been reported among uninfected Cambodian children; it was found that approximately 93% at age 6 years and 80% at age 12-13 [13,14] have caries. Children living with HIV are no exception to poor oral health, and the condition is even worse than uninfected children. Thus, these children have a higher number of dental caries than children without HIV infection [15].
This study therefore aimed to assess the association between dental caries and viral load among children living with HIV in Phnom Penh, Cambodia.

Study design and sites
This cross-sectional study was conducted from February to April 2018 as a baseline survey of a randomized controlled trial aiming to improve oral health of children living with HIV at the National Paediatric Hospital, Phnom Penh. Details of the trial have been published elsewhere [16]. This hospital is a tertiary referral facility that provides comprehensive care and treatment for children from across the country, including HIV and dental services and is a major paediatric HIV clinic in Phnom Penh that provides antiretroviral therapy (ART). The children living with HIV receive consultation and get their medicine every two months. Dental care is provided free of cost to children living with HIV in this hospital.

Study population
The target population in this study was children living with HIV and their caregivers. In the case of older children, who visited the hospital and administered their medication on their own, we interviewed only the children. The children were included if they were aged 3-15 years on the day of data collection, had a patient identi cation number at the study site hospital, and were under ART. They were selected from the registry of hospital's ART clinic using age-strati ed random sampling method, Randomization was performed using a computerised algorithm by a data analyst, who was not a primary member of the research team. The caregivers were eligible only if they were ≥ 18 years old and were the primary caretaker of the child.
Sample size for children was calculated based on the number required for the following intervention phase. The sample size set in the study protocol was calculated according to the decayed, missed, or lled permanent teeth (DMFT) score collected in a previous survey among 8-15-year-old children living with HIV [16]. However, after the completion of the baseline survey among the 3-15-year-old children, we obtained accurate DMFT scores for the study population. Therefore, we revised the sample size based on the following indicators: increment of DMFT: 17%, baseline DMFT of children living with HIV: 4.0 (SD: 3.6), power: 80%, alpha: 5%. The nal sample size required was 199 for each group. However, because of the improvement in the prevention of mother-to-child transmission in Cambodia, the number of children living with HIV aged < 8 years was low, and we could not recruit the required sample number. Therefore, 160 children were recruited in each group. In this study, both intervention and control groups of children living with HIV were examined, and thus, in total, 320 children were expected to participate.

Data collection
Two teams each consisting of one dentist and one dental assistant collected data on the children's dental caries status. To ensure accuracy of the examination, one of the dentist's researchers provided a one-day training session on how to assess dental status of patients using guidelines from the World Health Organization (WHO) [17]. Reproducibility of intra-examiner and inter-examiner evaluations was assessed. The dentists checked for DMFT in 10 children and compared the results between the two teams. The consistency rate of the results was > 85%. These data were not included in the main data collected. The total number of carious, missing, and lled permanent teeth was calculated as the DMFT score, and the dmft (decayed, missed, or lled deciduous teeth) score was obtained for deciduous teeth. The overall values of dmft and DMFT were evaluated separately and together by the sum of both scores. The severity of dental caries was expressed based on DMFT/dmft = 0 (no caries) and DMFT/dmft > 0 (presence of caries). If permanent and deciduous teeth were found to occupy the same tooth space, the status of the permanent tooth was recorded according to the WHO guidelines.
The research assistants collected clinical data from the HIV clinic's registered documents, including age, latest viral load within 12 months, ART regimen, and duration of ART. Six research assistants interviewed the caregivers using a structured questionnaire including the child's adherence to antiretroviral drugs developed based on the previously published questionnaires [18,19]. They received one-day training from the rst author to clarify and to improve their understanding of the questionnaires. For the adherence question, if the drug was self-administered by the child, we interviewed the child to obtain accurate information. The question was, 'How would you rate your/your child's adherence over the last month?' The response choices were very poor/poor/fair/good/very good/excellent [18].

Statistical analyses
The data were analysed descriptively to assess the distribution of the variables. Subsequently, dental caries was classi ed into dental caries in permanent teeth, dental caries in deciduous teeth, and dental caries in all teeth. We assessed the association of viral load, the dependent variable, with dental caries, age, sex, duration of ART, and adherence to antiretroviral drugs. The age, sex, and duration of ART variables were included following the model used in a previous study which examined the association between DMFT and CD4 count [20]. The variable of adherence to antiretroviral drugs was also included because it is related to viral load level in most cases [21]. For all participants, the independent variables were rst examined for association with the viral load ('detected'/'undetected') de ned with a cut-off point of < 40 copies/mL, according to the detection limit of tests. We also examined the association with viral non-suppression, 1000 copies/mL, which is the threshold for treatment failure [22] among only those who had a viral load detected. For bivariate analyses, we used the Chi-square test or Fisher's exact test if a count in one cell was smaller than 5, for categorical variables. We used Student's t-tests for continuous variables. Further, we applied the multiple logistic regression analysis and p < 0.05 was used to indicate statistical signi cance. All data analyses were performed using IBM SPSS, version 24.0 (SPSS Inc., Chicago, IL, USA).

Ethical consideration
Ethical approval was obtained from the National Ethics Committee for Health Research, Ministry of Health, Cambodia (289NECHR) and the Research Ethics Committee of Kyushu University (29067). We obtained written informed consent from the caregivers before data collection and assent to participate in the study from the children. Their participation was voluntary, and con dentiality was maintained.

Results
In total, 337 children living with HIV participated in our study. However, data from nine children were excluded owing to missing values. Finally, data from 328 children were included for the analyses. Regarding the interview responses, 100 were from the children only and 228 were from both child and caregiver.
General characteristics Table 1 shows the characteristics of participant children. Of all the children, 48.2% were female. The mean age of the children was 10.8 years (SD 3.0). The mean duration of ART was 6.3 years (SD 3.6). Moreover, 68.3% children had undetectable viral load. Among children with a detectable viral load, the load ranged up to approximately 5,012,000 copies/mL. Of the participants, 93.0% had at least one permanent tooth (age range: 6 to 15 years old), 57.3% had at least one deciduous tooth (age range: 3 to 15 years old). Of them, 50.3% had both permanent and deciduous teeth (age range: 6 to 15 years old). The mean values for the DMFT and dmft were 4.0 (SD 3.6) and 7.0 (SD 4.9), respectively. The mean number of teeth with caries including permanent and deciduous teeth was 7.7 (SD 5.0). Only 5.8% were free from caries. For adherence to antiretroviral drugs, no one answered, 'very poor,' and the most frequent answer was 'good' (51.2%).  Table 2 shows the bivariate and multivariate associations between different factors and viral load detection. The bivariate analyses indicated that total DMFT/dmft was positively associated with viral load detection (OR: 1.07, 95% CI: 1.02, 1.12). Being on ART for more than one year (OR: 0.19, 95% CI: 0.08, 0.43) and self-reported excellent adherence to antiretroviral drugs (OR: 0.24, 95% CI: 0.08, 0.70) were signi cantly associated with an undetectable viral load. negatively associated with viral load detection. The model that included children with all types of teeth (n = 328) showed that total DMFT/dmft was positively associated with viral load detection (AOR: 1.07, 95% CI: 1.01, 1.14), and that more than one year on ART (AOR: 0.22, 95% CI: 0.09, 0.52) and self-reported excellent adherence to antiretroviral drugs (AOR: 0.16, 95% CI: 0.05, 0.51) were negatively associated with viral load detection. Table 3 shows the association between dental caries and viral non-suppression (> 1000 copies/mL). The bivariate analyses showed that dmft (OR: 1.12, 95% CI: 1.01, 1.25) and total of DMFT/dmft (OR: 1.10, 95% CI: 1.02, 1.19) were positively associated with viral non-suppression. ART, antiretroviral therapy; DMFT, decayed, missed, or lled permanent teeth; dmft, decayed, missed, or lled deciduous teeth † 10 years old was set as cut-off point as it is the median of mixed dentition. ‡ set 1 year as cut-off point: most likely period attaining viral RNA thresholds in low-and middle-income countries [23], § viral load was a binary variable with cut-off point of log3.00 (= 1000 copies/mL) * p < 0.05; **p < 0.01 Among children with detectable viral load and who had deciduous teeth (n = 68), dmft was positively associated with viral non-suppression (AOR: 1.22, 95% CI: 1.05, 1.41). Among all children with detectable viral load (n = 104), the analyses indicated that total of DMFT/dmft was positively associated with viral nonsuppression (AOR: 1.12, 95% CI: 1.03, 1.23).

Discussion
In this study, we found that dental caries, duration of ART, and ART adherence were associated with the levels of viral load. To the best of our knowledge, this is the rst study to reveal the association between dental caries and viral load among children living with HIV.
The undetectable viral load rate of 68.3% in our study was similar to that reported in following previous studies. In the intervention study which aimed to examine the effectiveness of ART among Cambodian children living with HIV, the undetectable viral load rate was 81% [24]. One meta-analyses that included seven studies on children living with HIV from 2010 and later estimated the undetectable viral load rate (< 40 copies/mL level) at 72.7% (95% 62.6-82.8) after six months on rst-line ART [25].
In our study, the high occurrence of dental caries in permanent or deciduous teeth was associated with a detectable viral load. Furthermore, a high number of dental caries in permanent or deciduous teeth was associated with viral non-suppression. However, this nding contrasted with a study in the US conducted by Moscicki et al. who found no association between viral load and oral health indicators among children and adolescents aged 7-16 years [26]. However, most of the participants in that study had permanent teeth, and their results do not completely contradict those of our study: we too did not nd an association when we included only children with permanent teeth. However, the reason for the lack of this association is unclear. To address this, the child's immature immunologic status or the status of cariogenic capacity during development will have to be studied. For adults living with HIV, some studies suggested an association between dental caries and advanced viral load level. In one US study, Baqui et al. demonstrated that the majority of adults with HIV with a high viral load had a DMFT score > 20 [8]. However, the correlation was not signi cant. Some studies, including our previous study that focused on children 8 to 15 years old only, demonstrated a positive association between dental caries and CD4 + cell count among children living with HIV [20,27]. This may suggest a correlation between dental caries and viral load, as CD4 + cell count has a negative correlation with viral load. With respect to the association between viral load and dental caries, people living with HIV in the advanced stages of AIDS experience xerostomia. It is often diagnosed due to a lack of salivary ow among people living with HIV in advanced immune suppression stages [28]. The lack of salivary ow could be one of the causes of dental caries [28,29]. However, the causal relation is not clear, and the in uence of poor oral health status on AIDS-related diseases cannot be ruled out based on our results.
Our study also demonstrates that other factors such as duration of ART and adherence to ART were associated with the viral load. Similar to our research, longer duration on ART and high adherence to ART drugs were found to be associated with viral suppression or a lower level of viral load in different studies [25,[30][31][32][33]. This association implies the effectiveness of ART in improving the immune system. In many cases, the self-reported adherence rate is often biased due to improper recalling and social desirability [34]. However, the positive association identi ed in our study may suggest that participant responses are close to reality.
This study has some limitations. A cross-sectional design was employed, and there was no control group. Therefore, this study does not prove causality. In addition, as the data were collected from only one hospital, it may not be generalized. However, our study site hospital covers the largest number of children living with HIV in the country and is the main referral hospital for children living with HIV in Phnom Penh. Thus, the ndings in this study could be applicable to the general population of children living with HIV in Cambodia.

Conclusions
In conclusion, this study revealed a signi cant association between HIV viral load and dental caries among children living with HIV in Phnom Penh.

Declarations
Ethics approval and consent to participate Ethical approval was obtained from the National Ethics Committee for Health Research, Ministry of Health, Cambodia (289NECHR) and the Research Ethics Committee of Kyushu University (29067). We obtained written informed consent from the caregivers before data collection and assent to participate in the study from the children. Their participation was voluntary, and con dentiality was maintained.

Consent for publication
Not applicable.
Availability of data and materials