Bowdoin Science Journal

Psychology and Neuroscience

The Association Between Tooth Loss and Cognitive Decline

by

Imaging of a vertical (coronal) slice through the brain of an Alzheimer patient (left) compared with a normal brain ( right).
“Could Magnetic Brain Stimulation Help People with Alzheimer’s? | Scientific American.” n.d. Accessed May 4, 2026. https://www.scientificamerican.com/article/could-magnetic-brain-stimulation-help-people-with-alzheimer-rsquo-s/.

Cognitive decline with age is a major concern in medicine and public health. In 2021, the World Health Organization reported 57 million people were affected by dementia worldwide (World Health Organization, 2023). Well-established risk factors include alcohol intake, lower education level, physical inactivity, obesity, and diabetes, and preventive strategies have developed steadily. However, one potential contributor is often overlooked in major dementia research: tooth loss. Galindo-Moreno et al. (2022) examined this relationship through a large-scale analysis of over 100,000 US Americans, making a case for oral health as an underrecognized factor in cognitive decline. 

Edentulism refers to the partial or complete loss of permanent teeth. Edentulism can be caused by a multitude of factors, including biological processes such as caries (tooth decay) and periodontal disease (infection or inflammation of gums and bone), pulpal pathologies (damage to nerves, tissue, and blood vessels in the center of a tooth), trauma, or oral cancer. In addition to biological causes, edentulism can result from factors affecting dental care: patient preference, access to care, treatment options, and health insurance (Felton 2009). A study found 37% of edentulism cases were due to extraction from caries, 29% from periodontal diseases, and 12% due to trauma (Al-Rafee 2020).  

Although oral health care has developed significantly in the last few decades, edentulism remains a prevalent and irreversible condition (Al-Rafee 2020). It can occur at all ages, but the highest incidence occurs between the ages of 75-79 [Figure 1] (Chen et al. 2025). Those most affected by tooth loss typically have a lower socioeconomic standing, which makes health care less affordable and accessible [Figure 2] (Vemulapalli et al. 2024)  

Graph of global incidence and prevalence of edentulism per 100,000 across all ages. Highest incidence rate at ages 75-79. Prevalence per 100,000 gradually increases as age increases.
Figure 1: Global prevalence and incidence rates of edentulism in 2021. Chen, Hui Min, Kuo Shen, Ling Ji, Colman McGrath, and Hui Chen. 2025. “Global and Regional Patterns in Edentulism (1990-2021) With Predictions to 2040.” International Dental Journal 75 (2): 735–43. https://doi.org/10.1016/j.identj.2024.11.022. December 31, 2024: 738

 

Prevalence rate of complete edentulism in US adults 65 years and older across different socio-economic status'. As income increases, the rate of complete edentulism decreases.
Figure 2: Prevalence rate of complete edentulism in US adults 65 years and older according to demographic characteristics: Behavioral Risk Factor Surveillance System 2012-2020. Income level. Vemulapalli, Abhilash, Surendra Reddy Mandapati, Anusha Kotha, Hemanth Rudraraju, and Subhash Aryal. 2024. “Prevalence of Complete Edentulism among US Adults 65 Years and Older.” The Journal of the American Dental Association 155 (5): 399–408. https://doi.org/10.1016/j.adaj.2024.02.002. May 6, 2024: 407

Galindo-Moreno et al. proposed multiple pathways by which tooth loss can lead to cognitive decline. Two that play directly into known factors are the “diet and nutrition mechanism” and the masticatory mechanism. The number of teeth and which teeth are present affect what we can eat and how we eat. Mastication — chewing of food (Xu et al. 2008) — is directly influenced by edentulism due to the reduced bite force one can exert with missing teeth or dentures (Galindo-Moreno et al. 2022; Weijenberg et al. 2011). Changes to mastication may impact cognition by decreasing sensory input, which would reduce cell growth and development, impairing the cholinergic neurotransmitter system responsible for regulating memory, muscles, and attention, and reducing the generation of new neurons triggered by exercise (Weijenberg et al. 2011). Mastication additionally restricts our diet and therefore directly plays into the diet and nutrition mechanism. Often, with altered dentition, chewing can be an immense hurdle, for which the solution is a softer yet less nutritious diet.Nutrients such as omega-3 fatty acids, B vitamins, and antioxidants have important neuroprotective properties that help preserve the blood brain barrier, an essential layer that prevents toxins from entering the brain,additionally reducing inflammation, lowering the risk of cognitive decline (Power et al. 2019). Both the masticatory and diet and nutrition mechanisms are intertwined with diabetes and obesity, which are known risk factors for cognitive decline (Galindo-Moreno et al. 2022). 

Another pathway this study mentions is the inflammation/infection mechanism. A leading cause of edentulism is periodontitis, a severe gum infection often driven by the bacterium Porphyromonas gingivalis. This bacterium induces the local release of cytokines, proinflammatory proteins (Galindo-Moreno et al. 2022). Once in the bloodstream, cytokines promote the production of amyloid-β, a peptide whose accumulation is associated with Alzheimer’s disease (Leira et al. 2020). Simultaneously, Porphyromonas gingivalis increases the permeability of the blood-brain barrier (Lei et al. 2023). The heightened permeability of the BBB causes accumulation of overproduced amyloid-β in the brain tissue [Figure 3] (Galindo-Moreno et al. 2022; Leira et al. 2020)  

Amyloid PET scan of patient with Alzheimer's Disease (right), and patient without Alzheimer's (left). Patient with Alzheimer's Disease shows higher detection of Amyloid plaques.
Figure 3: Amyloid PET scan comparison of healthy brain and Alzheimer’s disease. Chapleau, Marianne, Leonardo Iaccarino, David Soleimani-Meigooni, and Gil D. Rabinovici. 2022. “The Role of Amyloid PET in Imaging Neurodegenerative Disorders: A Review.” Clinical Investigation. Journal of Nuclear Medicine 63 (Supplement 1): 13S-19S. https://doi.org/10.2967/jnumed.121.263195.

To investigate the relationship between tooth loss and cognitive decline, the researchers analyzed data from over 100,000 Americans drawn from two large national health surveys, NHIS (2014-2017) and NHANES (2005-2018). The NHIS survey was particularly well-suited for assessing cognitive state, as it included four questions on concentration and memory. However, the survey included only one binary dental question asking whether the participants had a complete dentition or had lost ≥1 teeth. The NHANES survey complemented this with a thorough section on dental records. The exact number and location of lost teeth were documented. However, it assessed cognitive state with only one question on memory and confusion (Galindo-Moreno et al. 2022).   

Their primary statistical tool was multinomial logistic regression, a method used when an outcome has more than two categories. In this case, the categories were cognitive difficulty, ranging from “none” to “some” to “a lot.” By using this model, the researchers simultaneously accounted for other factors known to affect cognitive health, including age, income, education level, depression, anxiety, cardiovascular health, and lifestyle habits such as smoking and exercise, which were included in the health surveys. By modeling these variables together, the researchers could estimate the independent contribution of tooth loss to cognitive decline.  

The results were expressed as odds ratios (ORs), which indicate how much more likely a given outcome is in one group than in a reference group. Here, the reference was a fully toothed person reporting no cognitive difficulties. An OR above 1.0 indicated higher odds of cognitive problems among people with missing teeth. This held true even after the other variables were statistically accounted for. The researchers also used a technique called ROC curve analysis on the NHANES data that included exact tooth counts, allowing them to identify a meaningful threshold below which cognitive risk measurably increased (Galindo-Moreno et al. 2022).  

The researchers found that, overall, the presence of teeth was statistically associated with a better cognitive state. The NHIS data showed that people with edentulism (partial or complete) had an OR > 1 across all cognitive categories, especially memory, even after accounting for other risk factors. This trend was also observed across categories of gender, socio-economic status (SES), education, and cardiovascular risk — all of which negatively impact cognition. Notably, socioeconomic status emerged as one of the strongest predictors, alongside edentulism, reflecting how directly financial circumstances shape access to dental care and, through it, long-term cognitive health. 

Using ROC curve analysis of the NHANES data, they determined the threshold for cognitive risk to be 20.5 teeth, indicating that a person with fewer than 21 teeth has an increased risk of cognitive decline compared to a fully dentulous person (Galindo-Moreno et al. 2022). Importantly, the study analyzed the NHANES survey and found a gradient effect: the fewer teeth a person had, the worse their cognitive outcomes tended to be, which strengthens the case that the association is meaningful rather than coincidental. Furthermore, a threshold could be determined for each individual tooth category: 5.5, 5.5, 3.5, 4.5, respectively, for molars, premolars, canines, and incisors. The multinomial regression of the NHANES data determined molars had the highest OR. The researchers linked this to the masseter, an important masticatory muscle supported by molars, which may, through its activity, stimulate the release of neurotrophic factors that support brain health. 

The link between edentulism and cognitive decline is still scarcely researched. As of March 2026, there are only 66 results on PubMed, 142 on ScienceDirect, and 148 on Wiley on the correlation between edentulism and cognitive decline. To put this into perspective, there are 2,277 results on PubMed, 18,967 on ScienceDirect, and 10,546 on Wiley on the relationship between diet and cognitive decline. The discussed research article combines two USA national health surveys with diverse samples, NHIS and NHANES, making it one of the largest in scope to date on tooth loss and cognitive decline. Although Galindo-Moreno and his team compellingly demonstrate the correlation, they recognize that their findings cannot answer whether edentulism leads to poorer cognition or rather poor cognition leads to edentulism (Galindo-Moreno et al. 2022, 3498). Some of the issues the researchers faced were the binary assessment of dentition in the NHIS survey, the single question on cognitive condition in the NHANES survey, and the overall lack of records on the reasons for tooth loss (Galindo-Moreno et al. 2022).   

Nevertheless, this study is a step in the right direction. Galindo-Moreno et al. showed that edentulism is correlated with cognition, thereby providing meaningful epidemiological evidence for a relatively young field. Consequently, this study and further research could have great clinical implications for cognitive health, not only in cost-effective treatment and prevention, but also in an important personal impact for those struggling with cognitive impairments and dental hygiene. 

Al-Rafee, Mohammed A. 2020. “The Epidemiology of Edentulism and the Associated Factors: A Literature Review.” Journal of Family Medicine and Primary Care 9 (4): 1841–43. https://doi.org/10.4103/jfmpc.jfmpc_1181_19.  

Chapleau, Marianne, Leonardo Iaccarino, David Soleimani-Meigooni, and Gil D. Rabinovici. 2022. “The Role of Amyloid PET in Imaging Neurodegenerative Disorders: A Review.” Clinical Investigation. Journal of Nuclear Medicine63 (Supplement 1): 13S-19S. https://doi.org/10.2967/jnumed.121.263195.  

Chen, Hui Min, Kuo Shen, Ling Ji, Colman McGrath, and Hui Chen. 2025. “Global and Regional Patterns in Edentulism (1990-2021) With Predictions to 2040.” International Dental Journal 75 (2): 735–43. https://doi.org/10.1016/j.identj.2024.11.022.  

“Dementia.” n.d. Accessed March 27, 2026. https://www.who.int/news-room/fact-sheets/detail/dementia.  

Felton, David A. 2009. “Edentulism and Comorbid Factors.” Journal of Prosthodontics 18 (2): 88–96. https://doi.org/10.1111/j.1532-849X.2009.00437.x.  

Galindo-Moreno, Pablo, Lucia Lopez-Chaichio, Miguel Padial-Molina, et al. 2022. “The Impact of Tooth Loss on Cognitive Function.” Clinical Oral Investigations 26 (4): 3493–500. https://doi.org/10.1007/s00784-021-04318-4.  

Lei, Shuang, Jian Li, Jingjun Yu, et al. 2023. “Porphyromonas Gingivalis Bacteremia Increases the Permeability of the Blood-Brain Barrier via the Mfsd2a/Caveolin-1 Mediated Transcytosis Pathway.” International Journal of Oral Science15 (January): 3. https://doi.org/10.1038/s41368-022-00215-y.  

Leira, Yago, Álvaro Carballo, Marco Orlandi, et al. 2020. “Periodontitis and Systemic Markers of Neurodegeneration: A Case–Control Study.” Journal of Clinical Periodontology 47 (5): 561–71. https://doi.org/10.1111/jcpe.13267.  

Power, Rebecca, Alfonso Prado-Cabrero, Ríona Mulcahy, Alan Howard, and John M. Nolan. 2019. “The Role of Nutrition for the Aging Population: Implications for Cognition and Alzheimer’s Disease.” Annual Review of Food Science and Technology 10 (1): 619–39. https://doi.org/10.1146/annurev-food-030216-030125. 

Vemulapalli, Abhilash, Surendra Reddy Mandapati, Anusha Kotha, Hemanth Rudraraju, and Subhash Aryal. 2024. “Prevalence of Complete Edentulism among US Adults 65 Years and Older.” The Journal of the American Dental Association 155 (5): 399–408. https://doi.org/10.1016/j.adaj.2024.02.002.  

Weijenberg, R. A. F., E. J. A. Scherder, and F. Lobbezoo. 2011. “Mastication for the Mind—The Relationship between Mastication and Cognition in Ageing and Dementia.” Neuroscience & Biobehavioral Reviews 35 (3): 483–97. https://doi.org/10.1016/j.neubiorev.2010.06.002.  

World Health Organization. 2023. “Dementia” Fact Sheets. https://www.who.int/news-room/fact-sheets/detail/dementia 

Xu, W. L., J. E. Bronlund, J. Potgieter, et al. 2008. “Review of the Human Masticatory System and Masticatory Robotics.” Mechanism and Machine Theory 43 (11): 1353–75. https://doi.org/10.1016/j.mechmachtheory.2008.06.003. 

Pupil Mimicry Strengthens Infant-Parent Bonding

by

Sometimes, it is a wonder how something so small can connect human beings on a deeper level, but that is what pupil mimicry does. Pupil mimicry describes the changes in pupil size that occur in both participants during eye contact, which can help with social bonding. It also reflects the different cognitive and emotional processes that occur during eye contact and socialization, such as showing social interest (Aktar et al., 2020). When pupil size synchronously dilates, meaning the pupil expands when eye contact is made, there is a promotion of trust and bonding between the two responders. The opposite is true for synchronous pupil constriction, which diminishes a positive social bond between the two responders.  

Pupil mimicry is an old, robust phenomenon (Prochazkova et al., 2018) that is modulated by oxytocin. This evolutionarily conserved neuropeptide acts as a hormone and neurotransmitter and facilitates social bonding (Aktar et al., 2020). Pupil mimicry has been observed in monkeys and chimpanzees, where it also increases trust and social familiarity (Kret et al., 2014). This effect occurs in infants as well, which suggests that pupil mimicry may help facilitate bonding between infants and their parents. But how can scientists measure that? 

It has been shown that young infants can differentiate between their own-race faces and other-race faces. Therefore, scientists hypothesized that infants would have quicker pupillary responses to pupils belonging to the same race as their parents when compared to other races. Researchers Aktar, Raijmakers, and Kret conducted a study with three aims to test this hypothesis: 

  1.  Do infants’ pupils react to dynamic videos of eyes with pupil sizes that change realistically? 
  2. Do parents and infants have the same speed in matching pupil size? 
  3. Do both the parents’ and infants’ pupils have differing rates of pupil mimicry between own-race faces and other-race faces

For the first aim, infants and parents watched black-and-white dynamic videos of same-race models (Dutch male and female) that had constricting, static, or dilating pupils while their pupillary reactions were tracked (Figure 1; Aktar et al., 2020). For the second and third aims, infants and parents watched black-and-white dynamic videos with two races: Dutch for the same-race category and Japanese for the other-race category (Aktar et al., 2020). The researchers compared the parents’ pupil mimicry speed to the infants’. 

Figure 1: Experimental set-up of infants and parents as they observe the stimuli (Aktar et al., 2020).

The researchers confirmed that both infants and parents were able to perform pupil mimicry. They also found that parents had quicker pupil response to dilated or constricted pupils than infants, possibly due to adults being more cognitively advanced than infants. Finally, they concluded that there was no significant difference in pupil mimicry response between own race and other races, but there were slight pupil mimicry delays. The researchers have several explanations for the slight delays. For instance, the infants’ pupils tend to stay dilated when they see a dilated pupil, regardless of race, since infants are still developing their pupil mimicry control. For adults, pupil mimicry tends to take about 2.5 milliseconds longer when given other-race stimuli. This may be from greater cognitive effort used to process other-race faces than own-race faces (Aktar et al., 2020). 

The key finding of the research is race does not affect the participants’ rate of pupil mimicry during emotionally neutral interactions (Aktar et al., 2020). So, though pupil mimicry helps strengthen parent-infant relationships, infants also have the skills to establish trust and awareness with strangers regardless of race. However, when infants are not in a neutral setting, meaning an environment where they feel unsafe and discontent, they are more likely to seek out their parents and less likely to make eye contact (Aktar et al., 2020). That is why, if the infants felt fussy or frightened, the researchers sat the parents right next to them to provide a feeling of safety (Figure 1). Eye contact conveys a great deal of information. Maybe the next time you make eye contact with someone, stare at them to see how their pupil responds to you!

References

Aktar, E., Raijmakers, M. E. J., & Kret, M. E. (2020). Pupil mimicry in infants and parents. Cognition and Emotion, 34(6), 1160–1170. https://doi.org/10.1080/02699931.2020.1732875

Kret, M. E., Tomonaga, M., & Matsuzawa, T. (2014). Chimpanzees and humans mimic pupil-size of conspecifics. PloS one, 9(8), e104886. https://doi.org/10.1371/journal.pone.0104886

Prochazkova, E., Prochazkova, L., Giffin, M. R., Scholte, H. S., De Dreu, C. K. W., & Kret, M. E. (2018, July 16). Pupil mimicry promotes trust through the theory-of-mind network – PNAS. Proceedings of the National Academy of Sciences. https://www.pnas.org/doi/10.1073/pnas.1803916115

Caution in STEM: Inhibition, Intuition, and Counterintuitive Reasoning

by

Imagine you’re on a 1950s game show. The host presents three doors and lays out the rules: Behind one door is a car, and behind the other two are goats. After you choose a door, the host, knowing what’s behind each door, opens one of the remaining two doors, revealing a goat. You have the opportunity to switch. Do you?

This is, of course, the infamous Monty Hall problem. Assuming you prefer the car over the goat, the answer is to always switch, since it will give you double the probability—⅔ rather than ⅓—of winning the car. Here’s an explanation that goes through each possible case (Table 1):

Table 1: Possible outcomes for staying and switching in the Monty Hall problem (Saenen et al., 2018)

If you got it wrong, you’re not alone—between 79% and 87% of adults get it wrong, too (Saenen et al., 2018). But what is behind this phenomenon? Solving unintuitive problems like the Monty Hall problem is thought to require the inhibition of misleading information, such as from prior knowledge or false cues (Dumontheil et al., 2022; Saenen et al., 2018). However, a 2018 study by Brookman-Byrne et al. and a 2022 study by Dumontheil et al. shine a new, more nuanced light on the connection between inhibitory control and (counter)intuition.

Both studies had British schoolchildren aged 11-15 undergo a volley of tests assessing their response inhibition (the ability to manage and filter out conflicting information), semantic inhibition (the ability to suppress responses driven by impulse ), vocabulary, reasoning, and working memory. Researchers then had participants complete a set of intuitive (control) and counterintuitive math and science problems. Dumontheil et al. (2022) measured neural activity using fMRI (functional Magnetic Resonance Imaging, an imaging technique which measures blood-oxygen levels to determine which parts of the brain are active) throughout.

Unsurprisingly, researchers consistently found that participants were more accurate and faster in solving intuitive problems than counterintuitive problems. Furthermore, in counterintuitive reasoning, response inhibition predicted response times, whereas semantic inhibition predicted accuracy. Interestingly, however, the only predictors of counterintuitive reasoning ability found in both studies were a more extensive vocabulary and increased age, both of which also predicted response inhibition (Brookman-Byrne et al., 2018; Dumontheil et al., 2022). Given these unexpected findings, neuroimaging results by Dumontheil et al. (2022) were necessary to provide some insight into what goes on in participants’ brains. 

Figure 1: Brain regions showing greater activation for (A) counterintuitive versus control (intuitive) problems, (C) response inhibition versus no response inhibition, and (D) semantic inhibition versus no semantic inhibition (Dumontheil et al., 2022). 

Figure 2: A comparison between areas showing increased activation during counterintuitive reasoning and (A) complex inhibition behavior, and (B) interference control behavior (Dumontheil et al., 2022). 

Since the overlap is limited in Figure 2, researchers concluded that the relationship between inhibitory control and counterintuitive problem solving was not direct (Dumontheil et al., 2022). They posit that the role of inhibition in counterintuitive reasoning may be limited to specific types of inhibition. In particular, semantic inhibition might be a better explanation than just response inhibition (Dumontheil et al., 2022). 

Neurosynth (an fMRI image database) also associates areas activated during counterintuitive reasoning with “working memory,” “calculation,” “symbolic,” “attention,” “visually,” and “spatial,” suggesting that inhibition is not the only factor at play (Dumontheil et al., 2022). They highlight that two areas known as the intraparietal sulcus (IPS) and Brodmann area 7 (BA 7)—which together are responsible for visuo-spatial attention—show increased activation during counterintuitive reasoning, response inhibition, and semantic inhibition (Dumontheil et al., 2022). Hence, they also suggest that visuo-spatial attention may be another factor in counterintuitive reasoning (Dumontheil et al., 2022). 

So what does this mean, practically? For educators, it seems that curriculum design in STEM should not be done in isolation. Given the impact of semantic reasoning, it would be prudent to balance training in purely symbolic reasoning with training in semantic reasoning (e.g., by requiring humanities classes be taken with STEM classes). For cognitive neuroscientists, this research suggests that there may be another dimension to understanding counterintuitive reasoning: the complex causal relationships between visuo-spatial attention, inhibitory control, and counterintuitive reasoning. Indeed, this is a cautionary tale about the importance of inhibition in science itself—causation is difficult to establish, and the most intuitive models in science may not always be right, either.

 

References

Brookman-Byrne, A., Mareschal, D., Tolmie, A. K., & Dumontheil, I. (2018, June 21). Inhibitory control and counterintuitive science and maths reasoning in adolescence. PLoS ONE, 13(6), 1-19. https://doi.org/10.1371/journal.pone.0198973

Dumontheil, I., Brookman-Byrne, A., Tolmie, A. K., & Mareschal, D. (2022). Neural and Cognitive Underpinnings of Counterintuitive Science and Math Reasoning in Adolescence. Journal of Cognitive Neuroscience, 34(7), 1205. https://doi.org/10.1162/jocn_a_01854

Saenen, L., Heyvaert, M., Van Dooren, W., Schaeken, W., & Onghena, P. (2018). Why Humans Fail in Solving the Monty Hall Dilemma: A Systematic Review. Psychologica Belgica, 58(1), 128-158. https://doi.org/10.5334/pb.274