Controlling for the effect of education is crucial when investigating the risk factors for cognitive impairment and dementia, as it represents a key component of cognitive reserve; therefore, its influence should be carefully considered. The results of this study suggest that poor periodontal health is associated with poor cognitive performance regardless of education. This is evident from the observation that older individuals with pathological alterations in their periodontium achieved lower scores on the neuropsychological test, suggesting that, compared with healthy individuals, they are more likely to suffer from dementia (Table 3). In both approaches, education alone does not explain these outcomes.
Despite various systematic reviews indicating that poor periodontal health may increase the risk of cognitive impairment and dementia, methodological limitations in the primary studies make it necessary to interpret these findings with caution (1–6). All these reviews remark on the need for adequate control of confounding factors, but only one has directly addressed this issue (4). In that particular review, education was included as a confounding variable in only eight out of fourteen cohort studies. When education was considered, four of these studies maintained the association between periodontal health and cognitive impairment (7–10), while the other four studies did not (11–14). These inconsistencies can be attributed to considerable variations in the treatment of the education variable across studies, with authors often failing to justify the logic behind their chosen categorization. In the present research, education was categorized according to the educational system in Chile during the participants' formative years, and only two categories were used in the models based on mandatory education. We believe that this approach better reflects the participants' real-life educational experiences.
In large-scale population studies, employing gold-standard assessments to identify periodontal disease may not be feasible. Therefore, proxy measures have been widely utilized, ranging from toothbrushing habits (15) to tooth loss (2). Even self-reported measures have shown associations with clinically confirmed periodontal disease (16). In this study, the proxy measure was not only associated with social determinants of health, as expected with any oral health issue (17) but also with two major risk factors for periodontitis: smoking (whose p-value was interpreted as epidemiologically relevant) and diabetes mellitus (18). This suggests that this proxy measure was an effective tool for assessing poor periodontal health. The prevalence of periodontal disease worldwide is estimated to be approximately 50%, while in our study, it reached 58% (including signs of possible disease and unhealthy periodontium). Moreover, the estimated prevalence of unhealthy periodontium is in line with severe periodontitis estimates from previous investigations (11%) (18). Therefore, patients with cases that were identified using this proxy measure likely have more severe forms of periodontal disease, such as periodontitis, where visual signs are prominent even when other clinical signs of periodontal destruction have not been assessed. As a result, it is believed that the evaluations reported in this work have adequately characterized the participants' periodontal health condition despite the measurement method. Nevertheless, one of the main limitations of this study is that periodontal health status was not evaluated using gold-standard clinical examinations, which typically involve a full-mouth examination and periodontal measurements at six sites per tooth (excluding wisdom teeth) (19). Consequently, participants were not diagnosed with gingivitis or periodontitis. However, based on the arguments presented earlier, we believe that our assessment closely resembles that of poor periodontal health, with a significant number of patients with likely periodontitis cases.
Regarding the characterization of cognitive performance, only the ACE-R test was employed. Although other tests exist, the ACE-R test has demonstrated superior performance in detecting cognitive deterioration, particularly in older populations, and exhibits excellent sensitivity in detecting dementia (20). Future studies should consider including clinical assessment tools that ascertain whether individuals themselves or reliable informants report a decline in basic and instrumental activities of daily living instead of assessing cognitive function alone.
Although this design does not rule out reverse causality phenomena, it should be noted that the peak incidence of periodontitis is at approximately 38 years of age (21). Therefore, it is possible to infer that exposure to periodontitis precedes cognitive decline and dementia. Additionally, older people with cognitive impairment often have poor oral health. Caries, periodontal disease, and root debris are more common in people with dementia than in those with normal cognition (22). However, this does not mean that dementia is a risk factor for these oral conditions but rather that the oral condition probably worsens due to difficulties in self-care (including oral hygiene) or access to dental care subsequent to cognitive impairment (23). This underlines the importance of measuring basic activities of daily living longitudinally.
When generalizing these findings, it is important to consider that the MAUCO population is generally less healthy than the general population in Chile. This population segment predominantly relies on the public health system and has a lower socioeconomic status, with approximately 30% residing in rural areas and engaging primarily in agricultural activities (24). However, even though this population is more disadvantaged, an association was found between periodontal condition and cognitive performance, regardless of sociodemographic factors. In addition, the inclusion of this particular cohort is a strength of this study because the participants share very similar living conditions, which allows for a more valid interpretation of the findings, especially with respect to the impact of education. Moreover, the study encompasses more than half of the total older adult population residing in Molina in 2015 (25), and the large sample size ensures high power.
The finding of an association that was independent of education suggests that the observed link between periodontal health and cognitive performance is not solely influenced by cognitive reserve. Although cognitive reserve is an abstract concept that cannot be directly measured (26), in low-income settings characterized by individuals with predominantly low educational levels and unskilled occupations, education can serve as a proxy for cognitive reserve (27), as demonstrated in this cohort. This research is significant because it investigates early- and late-life factors that may impact dementia. Despite the protective effect that education typically confers on cognitive reserve (28), the results of this study suggest that maintaining periodontal health throughout life could play a crucial role in preventing cognitive alterations.