Incidence of periodontal disease in schizophrenia

The incidence of periodontal disease in patients with schizophrenia varies with ethnicity, time, socioeconomic status, psychophysical condition, definition and length of illness and assessment tools (Angelillo et al., Reference Angelillo, Nobile, Pavia, De Fazio, Puca and Amati1995; Tang et al., Reference Tang, Sun, Unqvari and O'Donnell2004; Pihlstrom et al., Reference Pihlstrom, Michalowicz and Johnson2005; Arnaiz et al., Reference Arnaiz, Zumárraga, Díez-Altuna, Uriarte, Moro and Pérez-Ansorena2011; Gurbuz et al., Reference Gurbuz, Alatas, Kurt, Dogan and Issever2011; Teng et al., Reference Teng, Su, Chang and Lai2011). The following three studies – including the current study – among ethnic Chinese populations are considered as examples (Tang et al., Reference Tang, Sun, Unqvari and O'Donnell2004; Teng et al., Reference Teng, Su, Chang and Lai2011). A cross-sectional survey of oral health in central Taiwan using the community periodontal index (CPI) showed that 90% of psychiatric inpatients (schizophrenia: 61%, length of illness: approximately 6 years) had poor periodontal health (Teng et al., Reference Teng, Su, Chang and Lai2011). A dental study in Hong Kong demonstrated that 98.5% of inpatients with chronic schizophrenia (length of illness: 20 years) had poor periodontal health, as assessed using the standardised dental evaluation of the World Health Organization (Tang et al., Reference Tang, Sun, Unqvari and O'Donnell2004). In the current study, the incidence of periodontal disease in patients with newly diagnosed schizophrenia (length of illness: 1 year or less) in Taiwan was 65.7%, as assessed from dentist visit records. Although time, duration of schizophrenia, assessment tools, psychophysical condition and socioeconomic status were different in all three studies, the data crudely illustrated that in terms of duration of schizophrenia, the incidence of periodontal disease in patients with schizophrenia increases from 65.7% after approximately 1 year to 98.5% after 20 years from the schizophrenia diagnosis. The results possibly imply that early preventive measures could prevent the incidence of periodontal disease in approximately one-third of patients newly diagnosed with schizophrenia.

Risk factors for periodontal disease in schizophrenia

Most evidence demonstrates that old age (Angelillo et al., Reference Angelillo, Nobile, Pavia, De Fazio, Puca and Amati1995; Tang et al., Reference Tang, Sun, Unqvari and O'Donnell2004; Arnaiz et al., Reference Arnaiz, Zumárraga, Díez-Altuna, Uriarte, Moro and Pérez-Ansorena2011; Gurbuz et al., Reference Gurbuz, Alatas, Kurt, Dogan and Issever2011; Teng et al., Reference Teng, Su, Chang and Lai2011) and male sex (Gurbuz et al., Reference Gurbuz, Alatas, Kurt, Dogan and Issever2011; Gopalakrishnapillai et al., Reference Gopalakrishnapillai, Iyer and Kalantharakath2012) are associated with significantly higher risks of periodontal disease in patients with schizophrenia. This evidence was obtained from patients with chronic schizophrenia and periodontal disease by using assessment tools such as the CPI. Therefore, we assessed the presence of periodontal disease by identifying dentist visits by patients within 1 year of a schizophrenia diagnosis. Differences between subjects and assessment tools may have led to the identification of young age and female sex as risk factors for treated periodontal disease in our study. In other words, young female patients in the early stage of schizophrenia may have an opportunity or good insights for treatment of periodontal disease with good prognosis. On the contrary, elderly male patients with chronic schizophrenia had poor periodontal health. The findings thus suggest that primary care staff should be more concerned about early prevention of periodontal disease in elderly male patients with schizophrenia because this population may not visit a dentist frequently and their periodontal disease could worsen when they enter the chronic stage of schizophrenia.

Low income level is a known crucial risk factor for periodontal disease in the general population (Pihlstrom et al., Reference Pihlstrom, Michalowicz and Johnson2005). However, high income level was a risk factor for treated periodontal disease in patients with newly diagnosed schizophrenia – probably because the patients belonged to a high income level class in this study were well aware that dental visits are necessary for maintaining good oral hygiene. The financial burden of dental expenses was limited for the patients in our study because Taiwan's National Health Institutes provide most Taiwanese with basic dental care without large copayments (Teng et al., Reference Teng, Su, Chang and Lai2011). Notably, 2-year history of periodontal disease was the most vital risk factor for periodontal disease in the current study. Therefore, the incidence of periodontal disease increased dramatically from 9% at 2 years before schizophrenia diagnosis to 59.5% in the 1 year after. This increase might be ascribed to the relief that psychiatric treatment can provide to a patient – increasing in a patient's sense of reality and willingness to seek treatment for periodontal disease, leading to more than 6-fold dental visits – albeit with the adverse effects of antipsychotics.

Antipsychotics and periodontal disease in schizophrenia

Little evidence is available about the adverse effects of antipsychotics and periodontal disease in patients with schizophrenia because most previous surveys had small-to-medium samples, and complex patterns of antipsychotics were prescribed to the study subjects (Hede, Reference Hede1995; Gurbuz et al., Reference Gurbuz, Alatas, Kurt, Dogan and Issever2011; Teng et al., Reference Teng, Su, Chang and Lai2011; Gopalakrishnapillai et al., Reference Gopalakrishnapillai, Iyer and Kalantharakath2012; Eltas et al., Reference Eltas, Kartalcı, Eltas, Dündar and Uslu2013). The current paper is among the few reports (Hede, Reference Hede1995; Gopalakrishnapillai et al., Reference Gopalakrishnapillai, Iyer and Kalantharakath2012) that investigate the effects of antipsychotics on periodontal disease in patients with schizophrenia. It was impossible to assess the effect of each antipsychotic; we classified all antipsychotics and other medications into four categories: FGAs, SGAs, anticholinergics and antihypertensives. The findings of the current study illustrated that all four types of medication could accelerate development of periodontal disease, and the order of high to low risk is as follows: antihypertensives, FGAs, SGAs and anticholinergics. Extrapyramidal symptoms often occur with antipsychotics, particularly FGAs, so physicians prescribe co-medications such as anticholinergics or antihypertensives to alleviate these symptoms (Teng et al., Reference Teng, Su, Chang and Lai2011; Gopalakrishnapillai et al., Reference Gopalakrishnapillai, Iyer and Kalantharakath2012; Hu et al., Reference Hu, Chou, Wen, Hsieh, Tsai, Yang, Yang and Lin2016). Unfortunately, anticholinergics and antihypertensives can exacerbate the resultant periodontal disease. Therefore, clinicians should prescribe antipsychotics and other medications to the least extent possible under efficacious pharmacotherapy for schizophrenia.

Several reports have revealed that hyposalivation (xerostomia) may increase the risk of periodontal disease (Wagaiyu and Ashley, Reference Wagaiyu and Ashley1991; Hirotomi et al., Reference Hirotomi, Yoshihara, Ogawa, Ito, Igarashi and Miyazaki2006; Eltas et al., Reference Eltas, Kartalcı, Eltas, Dündar and Uslu2013; Mizutani et al., Reference Mizutani, Ekuni, Tomofuji, Azuma, Kataoka, Yamane, Iwasaki and Morita2015); however, only one of these reports focused on patients with schizophrenia, the findings of which suggested that patients with schizophrenia who had reduced salivary flow rate (hyposalivation) due to antipsychotics were at higher risk of periodontal disease than those with increased salivary flow rate (hypersalivation) (Eltas et al., Reference Eltas, Kartalcı, Eltas, Dündar and Uslu2013) – a result similar to our study. We used the incidence of periodontal disease associated with antipsychotics and other medications as a surrogate because detailed data on the actual saliva flow rate of the patients with schizophrenia who were taking these medications were not available from the NHIRD. The benefit of using periodontal disease as a surrogate is switched to requiring medical attention. Our findings demonstrated that hyposalivation induced by FGA, anticholinergic and antihypertensive is potentially associated with increased risk of periodontal disease; hence, FGA-induced hypersalivation in periodontal disease is considered a protective factor. Although our statistical analyses could not provide sufficient information on the linked pharmacopathology between the adverse effects of saliva and periodontal disease, the data demonstrate how clinicians can reduce periodontal disease caused by the adverse effects of antipsychotics during pharmacotherapy for schizophrenia. The findings suggest that clinicians should avoid iatrogenic adverse effects on saliva while prescribing antipsychotics as far as possible and actively manage the adverse effects and ensure early dental referral. The possible underlying causal pharmacological mechanism must be determined in future studies.

Strengths and limitations of the study

The major strength of the current study was the use of a large population-based cohort that enabled us to evaluate the relationship between antipsychotics and risk factors for periodontal disease in the early stages of schizophrenia. The findings may prove favourable to prevent the development of periodontal disease in patients with schizophrenia, especially the ethnic Chinese population. The well-determined temporal relationship between antipsychotic prescription and the occurrence of periodontal disease was another strength of this study. We obtained not only rigorous illness diagnoses but also correct medication information from the NHIRD. The study sample was identified based on the association between an ambulatory care expenditures database (ICD-9-CM codes and psychiatrists) and the prescription claims database (pharmacotherapy for schizophrenia) to increase the diagnostic accuracy of schizophrenia. We also considered significant covariates, including underlying diseases such as HIV/AIDS, diabetes mellitus, osteoporosis, chronic pulmonary diseases and a history of periodontal disease.

This study had several limitations. First, diagnoses of both periodontal disease and schizophrenia could have been underestimated, for instance, because of Berkson's bias, in which hospital cases and controls in a case–control study can be systematically different from one another because the combination of exposure to risk and disease occurrence increases the likelihood of admission. Hence, we surveyed treated patients with schizophrenia as well as the possible consequences of periodontal disease. Accordingly, treatment-naïve patients may have had only a limited effect on our analysis. Second, patient adherence to medications could not be evaluated because of the prescription claims database in this study. However, medication nonadherence would most likely have resulted in a nondifferentiated misclassification of exposure leading to possible underestimation of actual risk. Third, only treated periodontal disease indicated by the patients’ medical records was considered as the measure of periodontal disease occurrence in this study as opposed to all cases of periodontal disease wherein patients with schizophrenia developed periodontal disease but did not visit a dentist. Fourth, nonavailability of information on dietary, lifestyle and other potential risk factors for periodontal disease, such as illness severity, biochemistry data and patients’ unhealthy lifestyles such as tobacco consumption and poor oral hygiene (Albandar, Reference Albandar2002; Ramon et al., Reference Ramon, Grinshpoon, Zusman and Weizman2003; Scully and Bagan, Reference Scully and Bagan2004; Pihlstrom et al., Reference Pihlstrom, Michalowicz and Johnson2005; Dumitrescu et al., Reference Dumitrescu, Dogaru and Dogaru2008; Chu et al., Reference Chu, Yang, Chou, Chiu and Chi2012; Kossioni et al., Reference Kossioni, Kossionis and Polychronopoulou2012; Thomson et al., Reference Thomson, Sheiham and Spencer2012; Genco and Borgnakke, Reference Genco and Borgnakke2013; Morales-Chávez et al., Reference Morales-Chávez, Rueda-Delgado and Peña-Orozco2014; Hu et al., Reference Hu, Chou, Wen, Hsieh, Tsai, Yang, Yang and Lin2016) – was considered.