A Longitudinal Dose-Response Curve Between Leisure-Time Physical Activity and the Prevalence of Diabetes Based on the Different Levels of Cognitive Function Among Older Adults

Abstract

This study investigated a dose-response relationship between Leisure-Time Physical Activity participation (LTPA) and the risk of diabetes and a comparison of the risk across different cognitive function groups among older adults. The Health and Retirement Study data were used from 2012 to 2020 (n = 18 746). This study conducted a Cox Proportional Hazard Regression to investigate the Dose-Response Curve between the prevalence of diabetes and the covariates following a level of LTPA participation. The result presented that the Odds Ratio continuously decreased as the level of LTPA participation increased. Among the three cognitive function groups, the high group (OR = .43, P < .05) and the mid group (OR = .71, P < .05) had a larger negative slope coefficient than the low group. This study found that LTPA participation reduces the risk of diabetes and gives evidence for the importance of cognitive function in reducing the prevalence of diabetes.

Keywords

cognitive function diabetes physical activity older adults dose-response curve

Introduction

Diabetes has emerged as a prominent chronic condition affecting older adults that results in significant health complications including heart disease, nerve damage, and risk for dementia and significant expenses related to medical treatment.^1,2 The American Diabetes Association (ADA) has reported that over 26% of older adults suffer from diabetes mellitus and that 1.4 million Americans are diagnosed with diabetes every year.^3–7 Older adults who are diagnosed with diabetes typically experience a reduction in their life expectancy of approximately 5 years, which highlights the severity of the impact of this chronic disease on mortality rates in this vulnerable population.^8–11

Diabetes in older adults has been associated with a heightened risk of several complications and comorbidities including cardiovascular disease, renal disorders, infections, cancers, and an increased mortality rate.^12–15 Thus, based on the statistical data and the observed health challenges and concerns associated with diabetes, researchers have stressed the importance of investigating the prevalence and mortality rates of diabetes and its impact on older adults.^16–19 This finding of this investigation will provide researchers with critically needed data that can be used to design effective preventive strategies that can mitigate the growing burden of this disease.

Recently, a growing body of literature has reported findings related to the epidemiological association between diabetes mellitus and cognitive decline that provided conclusive evidence that diabetes is significantly associated with cognitive decline.^16–19 One extensively supported mechanism involves the escalation of insulin resistance that leads to persistent elevation of blood glucose levels and the risk of diabetes.²⁰ Insulin resistance impacts both cerebrovascular and noncerebrovascular pathways and contributes to cognitive impairment in individuals with diabetes.²¹ Insulin resistance triggers chronic hyperinsulinemia in the brain that results in elevated levels of beta-amyloid peptide-42, the primary constituent of amyloid plaques.^22,23 The presence of amyloid plaques is a defining feature of Alzheimer’s disease, Alzheimer’s Disease and Related Dementias (AD/ADRD), and amnestic cognitive impairment.^24,25 Moran et al ²⁶ reported extensive changes in brain metabolites and structures in patients with diabetes that firmly established the association between diabetes and cognitive impairment.

While several studies have demonstrated that worsening memory and attention-processing function is related to the increasing prevalence of diabetes in older adults, the pathogenetic link between declining cognitive function and the risk of diabetes has not yet been fully elucidated.^17,27 It is important for researchers to identify the changes in the risk of diabetes being followed by cognitive decline and design and implement interventions that effectively prevent the onset of diabetes among older adults.

It is well established that Leisure-Time Physical Activity (LTPA) is one of the most effective chronic disease prevention strategies for diabetes and the most important factor in promoting the health and wellbeing of older adults with diabetes.^28,29 Previous findings presented evidence that LTPA contributes to increased cardiorespiratory fitness, enhanced vigor, improved glycemic control, reduced insulin resistance, optimized lipid profile, blood pressure reduction, and the sustainability of weight loss.^30–32 Clinical studies have reported positive effects of LTPA on glycated h Hemoglobin A1C (HbA1C),^33,34 triglycerides (TG), and cholesterol levels in individuals with diabetes compared to non-exercise comparison groups.³⁵ In addition, research has provided evidence that LTPA such as aerobic and exercise activities is associated with a 26% reduction in the prevalence of diabetes³⁶ and a 33% reduction in HbA1C levels, which are leading indicators of diabetes in older adults. For example, cohort studies have demonstrated an association between regular physical activity, moderate to high cardiorespiratory fitness, and reduced risk of diabetes and mortality in both type 1 and type 2 diabetes populations.^37–39 Some studies have investigated the relationship between the level of LTPA participation and the risk of diabetes and found that older adults who frequently participated in LTPA were more likely to reduce the risk of diabetes onset.^34,38,39 For example, engaging in a minimum of 150 minutes per week of aerobic physical activity, such as walking or jogging, and incorporating resistance exercises like weight training has been proven to be effective in reducing HbA1C levels and lowering the risk of diabetes. Moreover, moderate to vigorous levels of LTPA have a more significant impact on decreasing HbA1C and the risk of diabetes than lower-intensity LTPA.^33,34

While a substantial number of clinical studies have reported that LTPA participation provides benefits that reduce the risk of diabetes and related diabetic symptoms,^37–40 there is currently no longitudinal evidence of the relationship between different levels of LTPA participation and the prevalence of diabetes in older adults. Though prior studies investigated how diabetes is associated with cognitive function and have provided evidence that individuals with diabetes are likely to experience cognitive decline,^16–19 little research has been conducted to examine the reverse relationship between different levels of cognitive function and the prevalence of diabetes.

In summary, the aim of this study is to: (a) investigate the longitudinal dose-response relationship between LTPA participation and the risk of diabetes, and (b) compare the risk of diabetes in three different cognitive function groups (i.e., low, mid, and high). The hypotheses that guided this investigation included:

Hypothesis 1

Higher levels of LTPA participation reduces the risk of diabetes.

Hypothesis 2

Higher levels of cognitive function reduces the risk of diabetes.

Methods

Data

This study used the Health and Retirement Study (HRS) data from 2012 to 2020 that has been biannually released since 1992 by the University of Michigan. The data includes a wide range of information about the health status of older adults and the demographic characteristics, pension, housing, and job history of individuals over 45 years old. The HRS uses not only a paper-based survey but also interviews (e.g., telephone, in-person) to gain proxy-measured and biofeedback data. While a similar survey format has been used since initial data collection, the HRS has modified the instruments, questionnaires, and domains they have used over time. We employed data from 2012 to 2020 as each dataset in that timeframe is comprised of identical domains and instruments.

The HRS has used their own data collection protocol to follow up with study respondents. The respondents in the survey participant pool have been identified with Person Identification Numbers (PN) and a Household Number (HH). Changes in family structure such as births, deaths, or divorces are recorded in the HH. For instance, if PN1 and PN2 in HH-1 experience a divorce, they are redesignated as PN1 in HH-1 and PN2 in HH-2. Similarly, if an PN1 in HH-1 has a second marriage, the HH-1 has PN1, PN2.

Study Sample

This study tracked participant samples from 2012 to 2020. Figure 1 shows the overview of the longitudinal follow-up of the study sample populations. The study samples were extracted using the following protocol: (a) the year 2012 as the baseline, (b) track from 2012 to 2020, (c) participants must be over 50 years old, and (d) exclude respondents who had null values in a dependent variable during the period. The total study sample of this study following this protocol was 18 746.

Figure 1.

Flow Chart of the Longitudinal Follow-up of Study Participants.

Measurements

Dependent Variables

Diabetes

The onset of diabetes was screened by a questionnaire item: “Has a doctor told you that you have diabetes?” to which respondents answered either “Yes” or “No.” Each answer was coded as 1 = ‘Yes’ and 0 = ‘No’. The HRS provides information about the time of diabetes onset using an item: “In what year was your diabetes first diagnosed?”

This study encompassed samples that consistently provided the same answers on the diabetes diagnosis questionnaire between 2012 and 2020. For instance, if a study participant indicated that they were diagnosed with diabetes in 2014, they were required to report having diabetes in 2016, 2018, and 2020 to be included in the study sample. We only included respondents who were diagnosed with diabetes after 2012 and used the 2020 data to investigate whether the study participants had diabetes. However, this study did not consider the timing of when participants were diagnosed with diabetes between 2012 and 2020, nor did it include information about whether the diabetes diagnosis was type 1 or type 2.

Independent Variable

Leisure-Time Physical Activity

The Psychosocial domain (i.e., LB) was comprised of information about 21 types of leisure activities. This study extracted two items to construct the LTPA variable: “Play sports or exercise” and “Walk 20 minutes or more”.^41,42 A seven point Likert scale was used to assess the level of LTPA participation (1 = ‘Daily’ and 7 = ‘Never relevant’). The scale was reverse coded to use a high score to represent a high level of LTPA participation. We summed LTPA participation each year from 2012 to 2020 to calculate total participation.

Covariates

Cognitive Function The HRS cognition section elicits comprehensive information regarding the cognitive function of older adults via in-person and/or telephone interviews. The HRS has consistently used the telephone Interview for Cognitive Status-27 (TICS-27) for cognitive assessment and collects two types of cognition data using the same instrument: Patient-Reported Outcome (PRO) and Proxy-Measured Outcome (PMO). The HRS interviewer decides whether to use PRO or PMO method based on the responsiveness of the interviewee. If the interviewee can respond reliably to the questionnaire, the interviewer collected data from the interviewee using the PRO method. However, if the interviewee was not reliable and responsive to the questions, the interviewer would use the PMO method that requires support from caregivers. Therefore, PRO data indicates that the respondents are able to answer to the questions independently, and data users can expect PRO data to be reliable. This study employed the PRO method to collect cognitive function data with respondents who were able to engage in the tests without support. These respondents were not considered cognitively impaired, even if they had a low score on each test or had been diagnosed with dementia.

The TICS-27 is comprised of cognitive tests assessing three domains of cognitive function: memory, working memory, and attention and processing speed. The total cognitive function score was derived by adding the results of the three cognitive function tests, with a scoring range of 0 to 27.^43–45 The memory domain was assessed using both recall and delayed recall tests. The interviewer presented 10 random words to the interviewees and asked them to recall as many words as possible immediately in the recall test and then asked them to recall the same 10 words again five minutes later in the delayed recall test (e.g., car, tree, cloud). Each correctly answered word had a one-point value and the total memory score ranged from 0 to 20. The seven-subtraction test was used to measure working memory that requires processing and storing information simultaneously. The respondents were asked to subtract 7 from 100 five times (e.g., 100-7 = 93, 93-7 = 86, 86-7 = 79). Each correct answer was given one point resulting in a total score that ranged from 0 to 5. Attention and processing speed were measured using a counting backward test.^44,45 The respondents were asked to count backward 10 continuous numbers from 20 in two trials (e.g., 20, 19, 18, 17). Each correct trial has one point value and score ranged from 0 to 2.

A TICS-27 score of 6 or below indicates cognitive impairment, whereas a score falling between 7 and 11 suggests cognitive impairment without meeting the dementia threshold, commonly referred to as MCI. Scores ranging from 12 to 27 are deemed to be within the normal cognitive status range as established by previous studies. Therefore, we classified study samples into three cognitive groups (i.e., low, mid, and high) according to their TICS-27 scores. The low group included scores from 0 to 6, the mid group from 7 to 11, and the high group from 12 to 27. The high group was comprised of participants with higher cognitive function than the mid and low groups.

Age and Sex

Effective coefficient estimation with covariates requires predictable outcomes and low correlation with the study variables to avoid multicollinearity risks.⁴⁶ Previous diabetes studies considered covariates such as age, sex, weight, and height.^47–49 Thus, we confirmed a correlation with study variables and put age and sex in 2012 as covariates as they were available in the HRS data.

Analysis

This study employed a Cox Proportional Hazard Regression to investigate the Dose-Response Curve (DRC) between the prevalence of diabetes and the covariates (i.e., age, sex, cognitive function) based on varying levels of LTPA participation. The Odds Ratio (OR) indicates the probability that the study sample is diagnosed with diabetes. An OR of 1.0 means that the individual has a 100% risk of diabetes and an OR of .70 indicates that, after a specific intervention, the intervention has reduced the risk of diabetes by 30%, indicating a 70% risk of diabetes. The DRC presents the OR as changing coefficients of covariates. The aim of this study was to calculate each covariate coefficient and describe the OR changes as LTPA participation increases. This study not only visualized the changes of the OR as LTPA levels increased, but also compared the OR in three different levels of cognitive function (i.e., low, mid, and high). All statistical analyses were conducted using the SPSS version 29.0 statistical package. The regression equation was as follows:

\begin{array}{l} Odds Ratio (Probability) = \frac{onset of diabetes}{1 - onset of diabetes} = Constant + Age + Sex + Leisure Time Physical Activity \\ OR (low cognitive group) = β_{0} + \{(Age β_{1} + Sex β_{2}) + LTPA β_{3}\} \\ OR (mid cognitive group) = β_{0} + \{(Age β_{1} + Sex β_{2}) + LTPA β_{4}\} \\ OR (high cognitive group) = β_{0} + \{(Age β_{1} + Sex β_{2}) + LTPA β_{5}\} \end{array}

Results

Study respondent demographic data are described in Table 1. The age of respondents ranged from 50 to 109 years old (M = 76.55, SD = 11.82). The sample consisted of 58.3% of females (n = 10 935) and 41.7% males (n = 7811). Almost half of the respondents indicated their marital status as married (54.6%, n = 10 237), while the remainder reported that they were either widowed (22.3%, n = 4175), or divorced (14.8%, n = 2790).

Table 1.

Demographic variables.

Characteristics	n	%
Age
50 to 109 years old (mean = 76.55, SD = 11.82)	18 746	100
Gender
Male	7811	41.7
Female	10 935	58.3
Marital status
Married	10 237	54.6
Living with a partner	13	.00
Separated	778	4.1
Divorced	2790	14.8
Widowed	4175	22.3
Never married	753	4.02

Total n = 18 746.

Table 2 summarizes the study variables, among which 6.10% (n = 1143) of respondents were diagnosed with diabetes between 2012 and 2020. The independent variables were comprised of information from 2012 to 2020. The total amount of LTPA had a mean of 10.52 mean and an SD of 9.93. The total cognitive function score (M = 47.12, SD = 40.17) was categorized into three groups: the high group (27.0%, n = 4995), the mid group (31.9%, n = 5898), and the low group (42.5%, n = 7853).

Table 2.

Descriptive Statistics.

Variables	n	Mean	SD
Dependent variables		.06	.24
Diabetes	1143 (6.10%)
Non-diabetes	17 603 (93.90%)
Independent variables (from 2012 to 2020)
Leisure time physical activity		10.52	9.93
Cognitive function		47.12	40.17
High group	4995 (27.0%)
Mid group	5898 (31.9%)
Low group	7853 (42.5%)

Total n = 18 746.

The omnibus of the model coefficient was tested (Table 3). This test investigates the likelihood ratio of the chi-square statistic vs the null model. A less than a .05 P value indicates that the current model outperforms the null model.⁵⁰ The test revealed a 11 821.82 log-likelihood value (Chi-square = 79.32, df = 4, P < .05).

Table 3.

Omnibus Tests of Model Coefficients.

	Overall (score)			Change from Previous Step			Change from Previous Block
−2 log likelihood	Chi-square	df	Sig	Chi-square	df	Sig	Chi-square	df	Sig
11 821.82	79.32	4	.00	79.82	4	.0	79.82	4	.00

The increase of the OR LTPA participation levels is illustrated in Figure 2. The OR continuously decreased as the level of LTPA participation increased. In the next phase, the OR of diabetes was compared to the three different levels of cognitive function as the level of LTPA participation increased. The low cognitive function group was used as a baseline to compare the OR of diabetes with the mid and high LTPA groups. The equation consisted of three covariates (i.e., parameters): age, sex, and three different cognitive function levels. The OR of diabetes showed a negative slope with increasing age (OR = .99, Wald = 1.14, df = 1, P > .05) and sex (OR = .95, Wald = .40, df = 1, P > .05) but not at significant levels.

Figure 2.

Cox proportional Hazard regression: Diabetes.

Among the three cognitive function groups (Table 4), the high group (OR = .43, Wald = 75.91, df = 1, P < .05) and the mid group (OR = .71, Wald = 16.30, df = 1, P < .05) had a larger negative slope coefficient than the low group even if these groups had the same cognitive function (Figure 3). In conclusion, the OR of diabetes decreased with increasing LTPA participation levels, and the high cognitive function group was found to have a larger negative OR coefficient than the mid and low cognitive function groups even if they reported the same level of LTPA participation.

Table 4.

Variables in the Equation.

	B	SE	Wald	df	Sig	Exp(B)
Age	−.01	.00	1.14	1	.28	.99
Sex	−.05	.07	.40	1	.52	.95
Low cognition group			75.96	2	.00*
Mid cognition group	−.34	.08	16.30	1	.00*	.71
High cognition group	−.85	.09	75.91	1	.00*	.43

*P < .05.

Figure 3.

Dose-response curve in different cognitive function groups.

Discussion

This study demonstrates a longitudinal dose-response relationship between LTPA participation and the risk of diabetes, and a comparison of the risk of diabetes across different cognitive function groups. This study provides empirical evidence that LTPA participation reduces the risk of diabetes and gives evidence for the importance of cognitive function in reducing the prevalence of diabetes among older adults. This finding suggests that high levels of LTPA participation and cognitive function play an important role in reducing the risk of diabetes among older adults.

Substantial evidence has been presented in the literature that LTPA participation has been associated with reducing the risk of diabetes among older adults and increasing the health and wellbeing of people diagnosed with diabetes.^33,34,40 Previous studies that were conducted using a cross-sectional methodology or in clinical settings reported that LTPA participation is effective in reducing the onset of diabetes among older adults.^35,37–39 Our longitudinal study also indicates that 10 years of LTPA participation was a significant factor in the reduction of the prevalence of diabetes. This longitudinal study extends and supports previous findings.^37–40 that LTPA participation is associated with a reduction in the risk of diabetes and the biomarkers related to diabetes. Our findings provide a cause-effect relationship between LTPA participation and the prevalence of diabetes, and it suggests that health professionals must continue to design and implement LTPA programs for older adults to reduce the onset of diabetes in this population.

As diabetes is recognized as a substantial risk factor for AD/ADRD, individuals diagnosed with diabetes face a 39% increased risk of AD and a 47% increased risk of dementia compared to nondiabetic patients.^24,25 These studies demonstrate the negative effect of diabetes on cognitive function and indicate that older adults with diabetes are likely to experience cognitive decline and an increased risk of AD/ADRD. Prior studies have provided evidence that diabetes is negatively associated with cognitive function among older adults.^16–19 Conversely, this study investigated the relationship between cognitive function and the onset of diabetes and found that the high cognitive function group exhibited a lower OR of diabetes compared to the mid and low cognitive function groups even though they had the same level of LTPA participation. This finding indicates that older adults who possess high levels of cognitive functioning are less likely to contract diabetes as it adds evidence to the literature that improving cognitive function among older adults can reduce the risk of diabetes in older adults that has heretofore been missing.

Overall, this study highlights the importance of LTPA participation as a strategy that can be used to reduce the risk of diabetes among older adults, and that improving cognitive function is essential to reducing the risk of diabetes for older adults. Based on these results, our study makes some practical suggestions on how to reduce the prevalence of diabetes among older adults. A recent systematic review and meta-analysis indicated that short-term (<1 year) multidomain (including two or more nonpharmacological components such as cognitive exercises and LTPA) intervention participation is associated with improvements in cognitive function in older adults when compared with single-domain interventions.⁵¹ This study suggests that the design and implementation of a multidomain program that includes exercise programs and cognitive training (especially in a digital format) should be used to reduce the risk of diabetes in the older adult population.

Limitations

There are several limitations inherent to this study that can affect its application to future research. Considering the nature of the secondary data sets used, our research did not include the types of diabetes (e.g., type 1 or type 2) due to the lack of availability of this information. As different types of diabetes may affect the design and specific contents of an LTPA program, future studies are needed to investigate which type of diabetes are best addressed with LTPA and cognitive function training. Furthermore, this study did not consider the timing of participants' diabetes diagnoses between 2012 and 2020. As the time of diabetes diagnosis may affect the relationship between the prevalence of diabetes and LTPA, future researchers will be able to control the timeframe for how long the study samples have been diagnosed with diabetes. In addition, this study specified three cognitive function groups based on aggregate MoCA scores. Each category of cognitive function can be associated with LTPA participation and the onset of diabetes among older adults. Investigating the relationship between each cognitive category and LTPA participation provides more insightful information to researchers. Last, there are confounders that may affect the association between diabetes and LTPA participation such as comorbidities and other related factors (e.g., cardiovascular disease, heart function, and limitations to mobility and LTPA access). Future research is needed to investigate how these comorbidities and related factors are associated with the association between the onset of diabetes and LTPA participation.

Footnotes

Data Availability Statement

The datasets generated and analyzed during the current study are available at: https://hrs.isr.umich.edu/

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was carried out with the support of the ‘R&D Program for Forest Science Technology (Project No. 2021387D10-2323-0101),’ provided by the Korea Forest Service (Korea Forestry Promotion Institute).

Ethical Statement

ORCID iDs

Jungjoo Lee

Junhyoung Kim

References

International Diabetes Federation . IDF Diabetes Atlas. 10th ed.; 2020. Available from www.diabetesatlas.org

World Health Organization . Global report on diabetes. Published April 21, 2016. Available from: https://www.who.int/publications/i/item/9789241565257

American Diabetes Association . Statistics about Diabetes. Published July 28, 2022. Available from: https://diabetes.org/about-us/statistics/about-diabetes

American Diabetes Association . Economic costs of diabetes in the U.S. in 2017. Diabetes Care. 2018;41(5):917-928

American Diabetes Association . Insulin affordability survey. Vault Consulting, LLC. Published May 22, 2018. http://main.diabetes.org/dorg/PDFs/2018-insulin-affordability-survey.pdf

Saeedi

Salpea

Karuranga

et al. Mortality attributable to diabetes in 20-79 years old adults, 2019 estimates: results from the International diabetes Federation diabetes Atlas, 9th edition. Diabetes Res Clin Pract. 2020;162:108086

World Health Organization . Use of Glycated Hemoglobin (HbA1c) in the Diagnosis of Diabetes Mellitus: Abbreviated Report of a WHO Consultation; 2011. Available from https://apps.who.int/iris/handle/10665/70523

Campbell

Newton

Patel

Jacobs

Gapstur

. Diabetes and cause-specific mortality in a prospective cohort of one million US adults. Diabetes Care. 2012;35(9):1835-1844

Gregg

Wang

et al. Changes in diabetes-related complications in the United States, 1990–2010. N Engl J Med. 2014;370(16):1514-1523

10.

Gordon-Dseagu

Shelton

Mindell

. Diabetes mellitus and mortality from all-causes, cancer, cardiovascular and respiratory disease: evidence from the health survey for England and Scottish health survey cohorts. J Diabetes Complications. 2014;28(6):791-797

11.

Rao Kondapally Seshasai

Kaptoge

Thompson

et al. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med. 2011;364(9):829-841

12.

Aune

Norat

Leitzmann

Tonstad

Vatten

. Physical activity and the risk of type 2 diabetes: a systematic review and dose–response meta-analysis. Eur J Epidemiol. 2015;30:529-542

13.

Boniol

Dragomir

Autier

Boyle

. Physical activity and change in fasting glucose and HbA1c: a quantitative meta-analysis of randomized trials. Acta Diabetol. 2017;54:983-991

14.

Fang

Lin

Huang

Chuang

. Correction: Fang et al. In Vivo rodent models of type 2 diabetes and their usefulness for evaluating flavonoid bioactivity. Nutrients. 2023;15(13):2881.

15.

Gregg

Cheng

Srinivasan

et al. Trends in cause-specific mortality among adults with and without diagnosed diabetes in the USA: an epidemiological analysis of linked national survey and vital statistics data. Lancet. 2018;391(10138):2430-2440

16.

Antal

McMahon

Sultan

et al. Type 2 diabetes mellitus accelerates brain aging and cognitive decline: Complementary findings from UK Biobank and meta-analyses. Elife. 2022;11:e73138

17.

Willmann

Brockmann

Wagner

et al. Insulin sensitivity predicts cognitive decline in individuals with prediabetes. BMJ Open Diabetes Res Care. 2020;8(2):e001741

18.

Zhao

Jian

Hao

Rao

. Hsa_circ_0054633 in peripheral blood can be used as a diagnostic biomarker of pre-diabetes and type 2 diabetes mellitus. Acta Diabetol. 2017;54:237-245

19.

Zheng

Ley

. Global aetiology and epidemiology of type 2 diabetes mellitus and its complications. Nat Rev Endocrinol. 2018;14(2):88-98

20.

Palta

Carlson

Crum

et al. Diabetes and cognitive decline in older adults: the ginkgo evaluation of memory study. J Gerontol A Biol Sci Med Sci. 2017;73(1):123-130

21.

Matsuzaki

Sasaki

Tanizaki

et al. Insulin resistance is associated with the pathology of Alzheimer disease: the Hisayama study. Neurol. 2010;75(9):764-770]

22.

Klunk

. Amyloid imaging as a biomarker for cerebral β-amyloidosis and risk prediction for Alzheimer dementia. Neurobiol Aging. 2011;32 Suppl 1:S20-S36

23.

Pantoni

. Cerebral small vessel disease: from pathogenesis and clinical characteristics to therapeutic challenges. Lancet Neurol. 2010;9(7):689-701

24.

Lin

Kuo

. Diabetes and the risk of multi-system aging phenotypes: a systematic review and meta-analysis. PLoS One. 2009;4(1):e4144

25.

Willette

Bendlin

Starks

et al. Association of insulin resistance with cerebral glucose uptake in late middle–aged adults at risk for Alzheimer disease. JAMA Neurol. 2015;72(9):1013-1020

26.

Moran

Bundy

Becker

DiMeglio

Gitelman

Goland

Greenbaum

Herold

Marks

Raskin

Sanda

Schatz

Wherrett

Wilson

Krischer

Skyler

Type 1 Diabetes TrialNet Canakinumab Study Group Pickersgill

de Koning

Ziegler

Böehm

Badenhoop

Schloot

Bak

Pozzilli

Mauricio

Donath

Castaño

Wägner

Lervang

Perrild

Mandrup-Poulsen

AIDA Study Group . Interleukin-1 antagonism in type 1 diabetes of recent onset: two multicentre, randomised, double-blind, placebo-controlled trials. Lancet. 2013;381(9881):1905-1915

27.

Koekkoek

Kappelle

van den Berg

Rutten

Biessels

. Cognitive function in patients with diabetes mellitus: guidance for daily care. Lancet Neurol. 2015;14(3):329-340

28.

Bullard

Trinh

Mackenzie

Mullen

. A systematic review and meta-analysis of adherence to physical activity interventions among three chronic conditions: cancer, cardiovascular disease, and diabetes. BMC Publ Health. 2019;19(1):636-711

29.

Zhao

Zhang

Ding

et al. Gut bacteria selectively promoted by dietary fibers alleviate type 2 diabetes. Science. 2018;359(6380):1151-1156

30.

Chudyk

Petrella

. Effects of exercise on cardiovascular risk factors in type 2 diabetes: a meta-analysis. Diabetes Care. 2011;34(5):1228-1237

31.

Colberg

Sigal

Yardley

et al. Physical activity/exercise and diabetes: a position statement of the American Diabetes Association. Diabetes Care. 2016;39(11):2065-2079

32.

Wing

Goldstein

Acton

et al. Behavioral science research in diabetes: lifestyle changes related to obesity, eating behavior, and physical activity. Diabetes Care. 2001;24(1):117-123

33.

Umpierre

Ribeiro

PAB

Schaan

Ribeiro

. Volume of supervised exercise training impacts glycaemic control in patients with type 2 diabetes: a systematic review with meta-regression analysis. Diabetologia. 2013;56:242-251

34.

Liubaoerjijin

Terada

Fletcher

Boulé

. Effect of aerobic exercise intensity on glycemic control in type 2 diabetes: a meta-analysis of head-to-head randomized trials. Acta Diabetol. 2016;53:769-781

35.

Sluik

Buijsse

Muckelbauer

et al. Physical activity and mortality in individuals with diabetes mellitus: a prospective study and meta-analysis. Arch Intern Med. 2012;172(17):1285-1295

36.

Anderson

Kerr

Thames

Xiao

Cohen

. Electronic health record phenotyping improves detection and screening of type 2 diabetes in the general United States population: a cross-sectional, unselected, retrospective study. J Biomed Inform. 2016;60:162-168

37.

Gregg

Gerzoff

Caspersen

Williamson

Narayan

KMV

. Relationship of walking to mortality among US adults with diabetes. Arch Intern Med. 2003;163(12):1440-1447

38.

Van Dam

Liu

. Diet and risk of type II diabetes: the role of types of fat and carbohydrate. Diabetologia. 2001;44:805-817

39.

Stampfer

Solomon

et al. The impact of diabetes mellitus on mortality from all causes and coronary heart disease in women: 20 years of follow-up. Arch Intern Med. 2001;161(14):1717-1723

40.

Diabetes Canada Clinical Practice Guidelines Expert Committee Sigal

Armstrong

Bacon

et al. Physical activity and diabetes. Can J Diabetes. 2018;42:S54-S63

41.

Lee

. Psycho-social correlates of leisure-time physical activity (LTPA) among older adults: a multivariate analysis. Eur Rev Aging Phys Act. 2020;17(1):6-7

42.

Lee

Kim

. Different levels of leisure walking and Mental health among older adults with mild cognitive impairment. J Aging Phys Act. 2023;31:841-848

43.

Ciesielska

Sokołowski

Mazur

Podhorecka

Polak-Szabela

Kędziora-Kornatowska

. Is the montreal cognitive assessment (MoCA) test better suited than the mini-mental state examination (MMSE) in mild cognitive impairment (MCI) detection among people aged over 60? Meta-analysis. Psychiatr Pol. 2016;50(5):1039-1052

44.

Fisher

Hassan

Rodgers

Weir

. Health and Retirement Study Imputation of Cognitive Functioning Measures: 1992-2010. Ann Arbor: Institute for Social Research, Survey Research Center, University of Michigan; 2013.

45.

Ofstedal

Fisher

Herzog

. Documentation of Cognitive Functioning Measures in the Health and Retirement Study. Ann Arbor: University of Michigan; 2005.

46.

Strawbridge

Wallhagen

. Self-rated health and mortality over three decades. Res Aging. 1999;21(3):402-416

47.

Vagelatos

Eslick

. Type 2 diabetes as a risk factor for Alzheimer's disease: the confounders, interactions, and neuropathology associated with this relationship. Epidemiol Rev. 2013;35(1):152-160

48.

van Manen

van Dijk

Stel

et al. Confounding effect of comorbidity in survival studies in patients on renal replacement therapy. Nephrol Dial Transplant. 2007;22(1):187-195

49.

Zhang

Mei

Zhang

et al. Risk for newly diagnosed diabetes after COVID-19: a systematic review and meta-analysis. BMC Med. 2022;20(1):444-510

50.

Doornik

Hansen

. An omnibus test for univariate and multivariate normality. Oxf Bull Econ Stat. 2008;70:927-939

51.

Hafdi

Eggink

Hoevenaar-Blom

et al. Design and development of a mobile health (mHealth) platform for dementia prevention in the prevention of dementia by mobile phone applications (PRODEMOS) project. Front Neurol. 2021;12:733878