Tooth Loss Associated with Physical and Cognitive Decline in Older Adults Georgios Tsakos, PhD, Richard G. Watt, PhD, Patrick L. Rouxel, PhD, Cesar de Oliveira, PhD, and Panayotes Demakakos, PhD
OBJECTIVES: To examine the effect of total tooth loss (edentulousness) on decline in physical and cognitive functioning over 10 years in older adults in England. DESIGN: Secondary data analysis. SETTING: English Longitudinal Study of Ageing, a national prospective cohort study of community-dwelling people aged 50 and older. PARTICIPANTS: Individuals aged 60 and older (N = 3,166). MEASUREMENTS: Cognitive function (memory) was measured using a 10-word recall test. Physical function was assessed using gait speed (m/s). Generalized estimating equations were used to model associations between baseline edentulousness and six repeated measurements of gait speed and memory from 2002–03 to 2012–13. Models were sequentially adjusted for time, demographic characteristics, socioeconomic status, comorbidities, health behaviors, depressive symptoms, and anthropometric measurements and mutually adjusted for gait speed or memory. RESULTS: Edentulous participants recalled 0.88 fewer words and were 0.09 m/s slower than dentate participants after adjusting for time and demographics. Only the latter association remained significant after full adjustment, with edentulous participants being 0.02 m/s slower than dentate participants. In age-stratified analyses, baseline edentulousness was associated with both outcomes in fully adjusted models in participants aged 60 to 74 but not in those aged 75 and older. Supplementary analysis indicated significant associations between baseline edentulousness and 4-year change in gait speed and memory in participants aged 60 to 74; the former was fully explained in the fully adjusted model and the latter after adjusting for socioeconomic status.
From the Department of Epidemiology and Public Health, University College London, London, UK. Address correspondence to Dr. Georgios Tsakos, Department of Epidemiology and Public Health, UCL, 1–19 Torrington Place, London WC1E 6BT, UK. E-mail: [email protected] DOI: 10.1111/jgs.13190
JAGS 63:91–99, 2015 © 2014, Copyright the Authors Journal compilation © 2014, The American Geriatrics Society
CONCLUSION: Total tooth loss was independently associated with physical and cognitive decline in older adults in England. Tooth loss is a potential early marker of decline in older age. J Am Geriatr Soc 63:91–99, 2015.
Key words: physical function; cognitive function; tooth loss; edentulous; longitudinal studies
ral disease is a significant public health problem.1 In particular, dental caries and periodontal diseases are highly prevalent chronic conditions that have a major effect on quality of life and require costly lifelong dental treatment.2 In older adults, the loss of all natural teeth (edentulousness) is far less prevalent than in previous decades but is still relatively common. In the recent national oral health survey in Britain, 15% of adults aged 65 to 74, 30% of those aged 75 to 84 and 47% of those aged 85 and older were edentulous.3 Tooth loss in middle-aged and older adults is due to a variety of factors, with caries being the primary clinical cause and advanced progressive periodontal disease also being a contributing factor.4,5 In the last 20 years, a great amount of research has focused on the potential link between oral diseases and morbidity and mortality. Most of the research has assessed the association between periodontal disease and risk of diabetes mellitus, cardiovascular disease, and cardiovascular disease mortality.6–9 A number of population cohort studies with follow-up periods varying from 4 to 57 years have also demonstrated independent relationships between tooth loss and morbidity, onset of disability, and premature mortality.10–17 The potential pathway linking tooth loss and systemic diseases is thought to be through impaired masticatory function leading to poor nutritional status18–20 or inflammatory responses associated with past periodontal disease and oral infections.21 Tooth loss can be considered an early marker of decline and frailty in older age.22 Studies have shown an association between tooth loss and cognitive impairment and the onset of dementia in older populations.23–26 In a
O
0002-8614/15/$15.00
92
TSAKOS ET AL.
prospective study of a sample of older U.S. men, tooth loss was independently associated with lower cognitive scores at 32-year follow-up.27 In a 4-year prospective cohort study of older Japanese adults, those with few teeth and without dentures and those who could not chew well were at significantly greater risk of onset of dementia than those with 20 or more natural teeth and those who could chew any food, respectively,28 and in a recent 5-year follow-up in a cohort of people with type 2 diabetes mellitus, tooth loss was independently associated with dementia and cognitive decline.29 Furthermore, periodontitis has been shown to be a potential risk factor for cognitive decline in an aging cohort,30 although the association between oral health and cognitive functioning may be complex and bidirectional, with worse oral health a risk factor for and a consequence of neurodegenerative processes.31 In that respect, cognitive decline was associated with oral health measures and behaviors in middle-aged adults.32 In contrast, few studies have assessed the association between tooth loss and physical function. In a Danish cohort study, being edentulous was independently associated at 5-year follow-up with onset of physical disability.33 Chewing ability has also been associated with disability in activities of daily living and less physical activity in elderly adults.34 The existing literature assessing the link between tooth loss and decline in cognitive and physical functioning has a number of design, measurement, and analytical limitations. Using a large and nationally representative sample of noninstitutionalized older adults from the English Longitudinal Study of Ageing (ELSA), the effect of total tooth loss on decline in physical and cognitive functioning was examined over 10 years. Whether these associations persisted after controlling for sociodemographic characteristics, physical health, health behaviors, depression, and relevant biomarkers was also assessed.
JANUARY 2015–VOL. 63, NO. 1
JAGS
dental data had not been collected from them, as were individuals who were dentate at baseline but became edentulous later (incident cases of edentulousness, n = 147) and those with missing data at baseline on dental status (n = 57), memory (n = 138), or gait speed (including walking aids users) (n = 580). The analytical sample comprised 3,166 individuals after excluding an additional 26 participants with missing data in other covariates used in the analysis. The National Research Ethics Service has approved ELSA, and all participants have given informed consent.
Assessment of Dental Status Baseline self-reported dental status was measured using the following question: “I would now like to ask a few questions about dental health. Can I just check, do you still have some of your own teeth, or have you lost them all?” Participants were categorized into those with some natural teeth (dentate) and those with no natural teeth (edentulous).
Assessment of Memory (Cognitive Function) Memory was measured using a 10-word recall test.35 Participants were asked to listen to 10 words and immediately recall as many words as they could. A few minutes later and with no prior notice, they were asked again to recall the 10 words they had heard earlier. A summary memory score was derived by counting the number of words participants correctly recalled immediately and after a delay (range 0–20). A poor memory variable was also derived by categorizing participants into two categories: a poor memory category that included all those in the lowest quartile of the memory summary score (≤6 words) and a category that included the rest of the sample that recalled seven or more words.
METHODS Assessment of Gait Speed (Physical Function) Study Population The current study sample was drawn from ELSA, a prospective observational panel study designed to be representative of the community-dwelling population aged 50 and older in England. The ELSA sample was recruited from consenting households that had previously participated in the Health Survey for England (HSE) and included at least one individual aged 50 and older. HSE is an annual crosssectional health examination survey that uses different nationally representative samples each year. The ELSA sample at baseline in 2002–03 comprised 11,391 individuals who had participated in HSE 1998, 1999, or 2001. Follow-up interviews took place in 2004–05, 2006–07, 2008–09, 2010–11, and 2012–13 and health examinations in 2004–05, 2008–09, and 2012–13. Because of the sampling framework, HSE data that predate the ELSA baseline are available for ELSA participants. (These data are known as ELSA Wave 0.) The current study included participants aged 60 and older (n = 7,225) because gait (walking) speed was not measured in those aged 50 to 59. Individuals recruited from consenting households that had participated in HSE 1998 were excluded from the study (n = 3,111) because
Participants able to walk were asked to walk a distance of 8 feet at their usual pace from a standing point twice. The interviewer recorded the time of the two successive walks using a stopwatch. The gait speed variable was derived by dividing distance (in meters) by mean time (in seconds). If only one time recording was available, it was used as a measure of gait speed instead of mean time. Extreme time values (
OBJECTIVES: To examine the effect of total tooth loss (edentulousness) on decline in physical and cognitive functioning over 10 years in older adults in England. DESIGN: Secondary data analysis. SETTING: English Longitudinal Study of Ageing, a national prospective cohort study of community-dwelling people aged 50 and older. PARTICIPANTS: Individuals aged 60 and older (N = 3,166). MEASUREMENTS: Cognitive function (memory) was measured using a 10-word recall test. Physical function was assessed using gait speed (m/s). Generalized estimating equations were used to model associations between baseline edentulousness and six repeated measurements of gait speed and memory from 2002–03 to 2012–13. Models were sequentially adjusted for time, demographic characteristics, socioeconomic status, comorbidities, health behaviors, depressive symptoms, and anthropometric measurements and mutually adjusted for gait speed or memory. RESULTS: Edentulous participants recalled 0.88 fewer words and were 0.09 m/s slower than dentate participants after adjusting for time and demographics. Only the latter association remained significant after full adjustment, with edentulous participants being 0.02 m/s slower than dentate participants. In age-stratified analyses, baseline edentulousness was associated with both outcomes in fully adjusted models in participants aged 60 to 74 but not in those aged 75 and older. Supplementary analysis indicated significant associations between baseline edentulousness and 4-year change in gait speed and memory in participants aged 60 to 74; the former was fully explained in the fully adjusted model and the latter after adjusting for socioeconomic status.
From the Department of Epidemiology and Public Health, University College London, London, UK. Address correspondence to Dr. Georgios Tsakos, Department of Epidemiology and Public Health, UCL, 1–19 Torrington Place, London WC1E 6BT, UK. E-mail: [email protected] DOI: 10.1111/jgs.13190
JAGS 63:91–99, 2015 © 2014, Copyright the Authors Journal compilation © 2014, The American Geriatrics Society
CONCLUSION: Total tooth loss was independently associated with physical and cognitive decline in older adults in England. Tooth loss is a potential early marker of decline in older age. J Am Geriatr Soc 63:91–99, 2015.
Key words: physical function; cognitive function; tooth loss; edentulous; longitudinal studies
ral disease is a significant public health problem.1 In particular, dental caries and periodontal diseases are highly prevalent chronic conditions that have a major effect on quality of life and require costly lifelong dental treatment.2 In older adults, the loss of all natural teeth (edentulousness) is far less prevalent than in previous decades but is still relatively common. In the recent national oral health survey in Britain, 15% of adults aged 65 to 74, 30% of those aged 75 to 84 and 47% of those aged 85 and older were edentulous.3 Tooth loss in middle-aged and older adults is due to a variety of factors, with caries being the primary clinical cause and advanced progressive periodontal disease also being a contributing factor.4,5 In the last 20 years, a great amount of research has focused on the potential link between oral diseases and morbidity and mortality. Most of the research has assessed the association between periodontal disease and risk of diabetes mellitus, cardiovascular disease, and cardiovascular disease mortality.6–9 A number of population cohort studies with follow-up periods varying from 4 to 57 years have also demonstrated independent relationships between tooth loss and morbidity, onset of disability, and premature mortality.10–17 The potential pathway linking tooth loss and systemic diseases is thought to be through impaired masticatory function leading to poor nutritional status18–20 or inflammatory responses associated with past periodontal disease and oral infections.21 Tooth loss can be considered an early marker of decline and frailty in older age.22 Studies have shown an association between tooth loss and cognitive impairment and the onset of dementia in older populations.23–26 In a
O
0002-8614/15/$15.00
92
TSAKOS ET AL.
prospective study of a sample of older U.S. men, tooth loss was independently associated with lower cognitive scores at 32-year follow-up.27 In a 4-year prospective cohort study of older Japanese adults, those with few teeth and without dentures and those who could not chew well were at significantly greater risk of onset of dementia than those with 20 or more natural teeth and those who could chew any food, respectively,28 and in a recent 5-year follow-up in a cohort of people with type 2 diabetes mellitus, tooth loss was independently associated with dementia and cognitive decline.29 Furthermore, periodontitis has been shown to be a potential risk factor for cognitive decline in an aging cohort,30 although the association between oral health and cognitive functioning may be complex and bidirectional, with worse oral health a risk factor for and a consequence of neurodegenerative processes.31 In that respect, cognitive decline was associated with oral health measures and behaviors in middle-aged adults.32 In contrast, few studies have assessed the association between tooth loss and physical function. In a Danish cohort study, being edentulous was independently associated at 5-year follow-up with onset of physical disability.33 Chewing ability has also been associated with disability in activities of daily living and less physical activity in elderly adults.34 The existing literature assessing the link between tooth loss and decline in cognitive and physical functioning has a number of design, measurement, and analytical limitations. Using a large and nationally representative sample of noninstitutionalized older adults from the English Longitudinal Study of Ageing (ELSA), the effect of total tooth loss on decline in physical and cognitive functioning was examined over 10 years. Whether these associations persisted after controlling for sociodemographic characteristics, physical health, health behaviors, depression, and relevant biomarkers was also assessed.
JANUARY 2015–VOL. 63, NO. 1
JAGS
dental data had not been collected from them, as were individuals who were dentate at baseline but became edentulous later (incident cases of edentulousness, n = 147) and those with missing data at baseline on dental status (n = 57), memory (n = 138), or gait speed (including walking aids users) (n = 580). The analytical sample comprised 3,166 individuals after excluding an additional 26 participants with missing data in other covariates used in the analysis. The National Research Ethics Service has approved ELSA, and all participants have given informed consent.
Assessment of Dental Status Baseline self-reported dental status was measured using the following question: “I would now like to ask a few questions about dental health. Can I just check, do you still have some of your own teeth, or have you lost them all?” Participants were categorized into those with some natural teeth (dentate) and those with no natural teeth (edentulous).
Assessment of Memory (Cognitive Function) Memory was measured using a 10-word recall test.35 Participants were asked to listen to 10 words and immediately recall as many words as they could. A few minutes later and with no prior notice, they were asked again to recall the 10 words they had heard earlier. A summary memory score was derived by counting the number of words participants correctly recalled immediately and after a delay (range 0–20). A poor memory variable was also derived by categorizing participants into two categories: a poor memory category that included all those in the lowest quartile of the memory summary score (≤6 words) and a category that included the rest of the sample that recalled seven or more words.
METHODS Assessment of Gait Speed (Physical Function) Study Population The current study sample was drawn from ELSA, a prospective observational panel study designed to be representative of the community-dwelling population aged 50 and older in England. The ELSA sample was recruited from consenting households that had previously participated in the Health Survey for England (HSE) and included at least one individual aged 50 and older. HSE is an annual crosssectional health examination survey that uses different nationally representative samples each year. The ELSA sample at baseline in 2002–03 comprised 11,391 individuals who had participated in HSE 1998, 1999, or 2001. Follow-up interviews took place in 2004–05, 2006–07, 2008–09, 2010–11, and 2012–13 and health examinations in 2004–05, 2008–09, and 2012–13. Because of the sampling framework, HSE data that predate the ELSA baseline are available for ELSA participants. (These data are known as ELSA Wave 0.) The current study included participants aged 60 and older (n = 7,225) because gait (walking) speed was not measured in those aged 50 to 59. Individuals recruited from consenting households that had participated in HSE 1998 were excluded from the study (n = 3,111) because
Participants able to walk were asked to walk a distance of 8 feet at their usual pace from a standing point twice. The interviewer recorded the time of the two successive walks using a stopwatch. The gait speed variable was derived by dividing distance (in meters) by mean time (in seconds). If only one time recording was available, it was used as a measure of gait speed instead of mean time. Extreme time values (
