Increased Prevalence of Obesity/Type 2 Diabetes and Lower Levels of Lithium in Rural Texas Counties May Explain Greater Alzheimer’s Disease Risk

Abstract

Background/Objective:

To compare Alzheimer’s disease (AD) mortality rates and coinciding risk factors in rural and urban Texas populations.

Methods:

155 Texas counties were divided into 73 rural and 82 urban areas using the U.S. Census Bureau definition of rurality. Changes in age-adjusted AD mortality across these counties were calculated using a 7-year aggregation model from 2000–2006 and 2009–2015. Data pertaining to gender, race, education, obesity, diabetes, physical inactivity, and lithium concentrations in tap water were also collected from readily available databases.

Results:

Change in age-adjusted AD mortality was higher in rural counties (9.5±1.4) versus urban (5.9±1.1) over the time period examined. Similarly, obesity (30.2±0.2% ), diabetes (11.0±0.1% ), and physical inactivity (29.4±0.2% ) levels were significantly higher in rural populations compared to urban (29.1±0.2%, 9.7±0.1%, and 26.7±0.3, respectively). In contrast, the percent of population with some college education (40.1±0.7% ) was lower compared to urban (29.4±0.2% and 44.4±0.9%, respectively). Lithium concentrations in tap water was significantly lower in rural counties compared to urban (63.3±8.2 and 33.4±4.7μg/L, respectively). No significant differences were observed among females and however, we did find significant differences in the percent of African American and Hispanics. Correlational analysis uncovered a negative association between education status and AD mortality over time (r = –0.17). Further analysis controlling for physical inactivity, education, and trace lithium concentrations results in a loss of statistical significance.

Conclusions:

AD mortality rates are higher in rural counties when compared to urban counties, and this may be linked to greater physical inactivity, obesity, and diabetes, as well as lower trace lithium levels in tap water.

Keywords

Dementia insulin metabolic disorders physical activity

INTRODUCTION

Alzheimer’s disease (AD), the most prevalent form of dementia, is a neurodegenerative brain disorder characterized by neuroinflammation and the histopathological hallmarks, amyloid-beta senile plaques and neurofibrillary tangles [1, 2]. Progression of this disease typically results in irreparable damage to neurons primarily in the cortex and hippocampal regions, and is associated with a loss of cognitive functions, such as memory, thinking and reasoning, and language [3, 4]. The world’s population is aging rapidly and projections of the number of patients with AD are over 100 million worldwide [5] and will reach 16 million in the United States by 2050 [6]. Age is the most significant risk factor for AD [7], however we recently demonstrated that age-adjusted AD mortality rates have risen over time across several Texas counties, suggesting that other factors independent of age are involved in the development of AD [8]. In our previous study, we found that the prevalence of obesity and type 2 diabetes were both negatively correlated with changes in AD mortality, which supports the recent view of AD as a metabolic disorder [9, 10].

Aside from obesity and type 2 diabetes, rural living has been associated with increased risk of AD and dementia; with early rural living having the strongest effect [11 –17]. Although the exact cause explaining the increased risk in rural dwellers remains unknown, it is of interest to note that several epidemiological studies have consistently demonstrated higher prevalence of obesity in rural populations, with rural residents participating in more ‘pro-obesogenic’ behaviors, including lower intake of fiber and fruits and a higher intake of sugary beverages [14 –17]. Moreover, it was also shown that rural residents were more likely to have impaired glucose metabolism (insulin resistance and type 2 diabetes) and were less likely to meet national physical activity recommendations, both related to increased obesity and AD [14, 17]. Thus, evidence within the literature suggests that the increased risk of AD with rural living may be related to higher prevalence of obesity and type 2 diabetes.

In the present study, we obtained data readily available from online databases in the state of Texas to determine whether rural counties would exhibit higher rates in AD mortality over time and whether this would be associated with a higher prevalence of obesity and type 2 diabetes. To provide further validation to our study, we controlled for other AD risk factors such as physical activity, education status, race, gender, and water lithium content, since we and others have recently shown that trace levels of lithium is negatively associated with AD and dementia [8, 18]. The purpose of our analysis was to determine if the increased age-adjusted AD mortality seen in rural populations is related to primary risk factors such as type 2 diabetes and obesity.

METHODS

2.1 Data acquisition

We obtained the age-adjusted AD mortality rates (per 100,000) for each Texas county between 2000–2006 and between 2009–2015 from the Center for Disease Control (CDC) Wonder’s Compressed Mortality Database using the code ‘G30’. Several risk factors were tested to determine whether they could contribute to the observed differences in AD mortality rates. These risk factors include: gender, race, education, and the prevalence of obesity, diabetes and physical inactivity. Females, Hispanics, and African Americans are known to be at a greater risk for developing AD [19, 20], and we obtained the percent of population represented by females, Hispanics, and African Americans within each county from 2011–2015 from the Census Bureau’s Population Estimates Program. Individuals with a low education status are also at increased risk for the development of AD [21], and the percentage of adults having some post-secondary education (2011–2015) was obtained from the American Community Survey. Obesity, type 2 diabetes, and physical inactivity have been linked to AD development [5 , 23], and the prevalence of adult (>20 years) obesity and diabetes, as well as estimates on physical inactivity was obtained from the National Diabetes Surveillance system. Physical inactivity was defined as the percentage of adults reporting no leisure-time physical activity (i.e., running, calisthenics, golf, gardening, or walking for exercise). Obesity was defined as having a body mass index >30 kg/m². Respondents were considered to have diagnosed diabetes if they responded, “yes” to the question, “Has a doctor ever told you that you have diabetes?”. Gestational diabetes was excluded from the dataset. The surveillance system does not distinguish between types of diabetes, but because type 2 diabetes accounts for 90–95% of the cases of diagnosed diabetes, trends in type 2 diabetes are likely to be similar to trends documented by the surveillance system, and therefore referred to herein as type 2 diabetes. Previously we determined that there is a negative relationship between lithium levels in water and AD mortality [8]. It was therefore important that this factor be included in our analysis, and as such cationic lithium concentrations in the public water supply were obtained from the Texas Water Development Board Groundwater Database, averaged, and then log-transformed for each county as previously described [8]. For obesity, type 2 diabetes, and physical inactivity, data were obtained and averaged from 2011–2015.

2.2 Change in AD mortality rates

Changes in AD mortality rates were calculated by subtracting the rate obtained between 2000–2006 from those obtained between 2009–2015 as previously described [8]. We chose to obtain rates using a 7-year aggregation to increase the amount of counties with reliable AD mortality rates while maintaining the magnitude with which the AD mortality rates have changed [8]. As per the CDC Wonder Database, a death rate based on fewer than 20 deaths has a relative standard error of 23% or more, and is therefore considered statistically unreliable. In total, 155 counties had reliable age-adjusted AD mortality rates from both 2000–2006 and 2009–2015 time periods.

2.3 Rural and urban county segregation

We used the U.S. Census Bureau’s rural definition, which identifies a county as rural if most of its population (>50% ) are living within rural areas [24]. Although the U.S. Census Bureau separates counties that are mostly rural (50–99.9% of population in rural areas) from those that are completely rural (100% of population in rural areas), we chose to combine these counties to maximize the number of counties identified as rural. Data regarding the percent of population living in rural areas for each Texas county from 2011–2015 was obtained from the Census Bureau’s Population Estimated Program. Of the 155 counties with reliable age-adjusted AD mortality rates from both 2000–2006 and 2009–2015 time periods, 73 were identified as rural and 82 were identified as urban.

2.4 Statistics

All values are means±standard error (SE). All comparisons between rural and urban counties were done using a Student’s t-test. Correlational analyses were done using a Pearson’s correlation model. Partial correlations were done to assess the associations between obesity/diabetes with changes in AD mortality while controlling for education, race, physical inactivity, and trace lithium levels. Statistical significance was set to p≤0.05, and SPSS Statistical Software (IBM, NY, USA) was used to perform all statistical tests.

RESULTS

3.1 Rural counties have experienced larger increases in AD mortality over time and have a higher prevalence of obesity and diabetes

After segregating the 155 counties with reliable AD mortality rates into either rural or urban, our results show that those counties with more than 50% of their population living in rural areas are experiencing almost double the increases in AD mortality rates over time (Fig. 1). In addition, we found that rural counties had a greater prevalence of obesity (Fig. 2B) and diabetes (Fig. 2C) compared with urban counties.

Fig.1

Age-adjusted Alzheimer’s disease mortality rates are growing faster over time in rural counties versus urban counties in Texas. ^*p≤0.05, rural n = 73, urban n = 82.

Fig.2

Rural counties have higher prevalence of obesity (A) and diabetes (B) compared with urban counties in Texas. ^*p≤0.05, rural n = 73, urban n = 82.

3.2 AD risk factor assessment in rural and urban Texas counties

We next determined whether differences in several other AD risk factors could contribute to our observed differences in AD mortality rates between rural and urban Texas counties. In this respect, we found that the percent of population that are physically inactive and have some college education was significantly higher and lower, respectively in rural counties versus urban counties (Table 1). With respect to gender and race, we found no significant difference in the percent of population represented by females; however, we did find that the percent of population represented by African Americans was greater in rural counties when compared with urban counties. In contrast, Texas rural counties on average had significantly lower percent of population represented by Hispanics when compared with urban counties (Table 1). Our previous work suggests that tap water lithium concentrations contribute to AD mortality as well as obesity and diabetes [8] and in support of this, here we demonstrate that rural counties have lower lithium concentrations in their public water supply compared with urban counties (Table 1). Altogether, differences in the levels of physical inactivity, education, trace lithium concentrations, and the percentage of African Americans may also contribute to the increased AD risk seen in Texas rural counties. Therefore, our next step was to perform partial correlations between obesity/diabetes and AD mortality, while controlling for physical inactivity, education, trace lithium concentrations, and the percentage of African Americans. Our results show that while obesity/diabetes are significantly correlated with AD mortality independent of the percentage of African Americans, controlling for physical inactivity, education, and trace lithium concentrations resulted in a loss of statistical significance (Table 2).

Table 1

Risk factor comparisons among Texas rural and urban counties

Risk Factor	Urban (mean±SE)	Rural (mean±SE)	p-value
Lithium concentration in tap water (μg/L)	63.3±8.2	33.4±4.7	0.003
% Physically inactive	29.4±0.2	26.7±0.3	<0.0001
% With some college	44.4±0.9	40.1±0.7	0.0002
% Female	49.7±0.3	49.7±0.3	0.95
% African American	7.7±0.7	10.0±0.9	0.05
% Hispanic	36.1±2.4	18.8±1.3	<0.0001

Rural n = 73; urban n = 82, p-value is from a Student’s t-test.

Table 2

Examining the correlations between obesity/diabetes with changes in Alzheimer’s mortality while controlling for education, physical inactivity, % African American, and trace lithium levels

Model	r	p
Obesity
Education	0.10	0.23
% African American	0.17	0.03
Physical inactivity	0.06	0.49
Lithium levels	0.16	0.06
Diabetes
Education	0.14	0.09
% African American	0.19	0.02
Physical inactivity	0.02	0.78
Lithium levels	0.14	0.09

DISCUSSION

Our study questioned whether rural counties in Texas are experiencing faster growth in age-adjusted AD mortality over time when compared to urban counties, and whether this is associated with greater prevalence of obesity and type 2 diabetes. Using a 7-year aggregation model, our results show that changes in age-adjusted AD mortality calculated from 2000–2006 and 2009–2015 are almost 2-fold higher in rural counties when compared to urban counties, and therefore suggestive of a steeper incline in AD. Furthermore, and as expected, Texas rural counties had a greater prevalence of both obesity and type 2 diabetes. However, we also found that other AD risk factors such as low education status, greater physical inactivity, and lower concentrations of lithium in public water were also more prominent in rural counties, and that adjusting for these factors rendered the correlations between obesity/diabetes and changes in AD mortality nonsignificant.

It is possible that the loss in statistical significance observed when controlling for education, physical inactivity, and trace lithium levels may be due to the fact these factors themselves are all correlated with obesity/diabetes. We have recently shown that trace levels of lithium negatively correlates with both obesity and diabetes [8]; and corresponding well with previous studies [25, 26], we have also observed significant correlations with greater physical inactivity/lower education status and higher prevalence of these metabolic diseases (Supplementary Fig. 1). However, it is also important to note that lithium may have a direct effect on brain health and AD protection independent of obesity/diabetes, whereby lithium can act to inhibit GSK3, leading to improved insulin sensitivity and reduced beta-amyloid production and tau hyperphosphorylation [27]. Therefore, when combined with our previous findings [8], our results here suggest that lithium is independently related to obesity, diabetes, and AD mortality.

Rural populations have been shown to experience greater barriers to accessing health care. Residents within rural areas, on average, have less services available (i.e., hospitals and health clinics), have farther distances and increased travel time to receive care, as well experience larger financial burden, in line with a greater likelihood of poverty [28 –30]. These barriers may lead to the unattended risk factors discussed earlier and attribute to the overall poorer health outcomes. More specifically, leading to the greater chronic health conditions like AD that are seen within these populations [16, 31]. In addition, given the importance of lithium to metabolic health and AD risk, our results showing differences in trace lithium levels across rural and urban counties presents another important disparity found in the state of Texas that could place those individuals living in rural areas at a disadvantage. However, the reasons explaining the differences in lithium levels in tap water across the urban and rural counties remain unknown.

Limitations with our study include our ecological study design, which prevents any form of individual associations or assessment of causation. In addition, this study design disallows temporal associations from the findings, that is to say it is difficult to discern in our populations which risk factors preceded others. It is possible as well that individuals moved residency between urban and rural counties in our chosen time frame, thus we cannot say for certain the exact length of exposure to rurality. Lastly, AD mortality rate may not be directly indicative of AD prevalence as previous research has shown AD-associated mortality has been underreported [32]. Nonetheless, our findings are consistent with what is currently known within the literature and should prompt futures studies with more powerful designs (i.e., longitudinal cohort design).

In summary, the findings from this study show that rural counties in Texas are experiencing rapid rises in AD mortality and that this may be related to higher prevalence rates of both obesity and type 2 diabetes in these counties. Although other risk factors, such as physical inactivity, low education, and lower levels of lithium in the public water supply, may also contribute to the differences in AD mortality, all three factors are also correlated with obesity/diabetes. In any event, our results add further support towards the disparities in health seen between rural and urban dwellers and we anticipate that our study will lead to further studies that can provide policymakers with enough evidence to make substantial changes towards reaching equitable healthcare.

Footnotes

ACKNOWLEDGMENTS

We would like to acknowledge the Natural Sciences and Engineering Research Council of Canada (NSERC) for the financial support to fund this project (RGPIN-2017-03904 R.E.K. MacPherson). VAF was supported by NSERC PDF.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Ward

, Himmelstein

, Lancia

, Binder

(2012) Tau oligomers and tau toxicity in neurodegenerative disease. Biochem Soc Trans 40, 667–671.

Selkoe

, Hardy

(2016) The amyloid hypothesis of Alzheimer’s disease at 25 years. EMBO Mol Med 8, 595–608.

Nussbaum

, Ellis

(2003) Alzheimer’s disease and Parkinson’s disease. N Engl J Med 348, 1356–1364.

O’Brien

, Wong

(2011) Amyloid precursor protein processing and Alzheimer’s disease. Annu Rev Neurosci 34, 185–204.

Rolland

, Abellan van Kan

, Vellas

(2008) Physical activity and Alzheimer’s disease: From prevention to therapeutic perspectives. J Am Med Dir Assoc 9, 390–405.

Barnes

, Yaffe

(2011) Accuracy of summary risk score for prediction of Alzheimer disease: Better than demographics alone?. Arch Neurol 68, 268; author reply 268–270.

Korolev

(2014) Alzheimer’s disease: A clinical and basic science review. Med Student Res J 4, 24–33.

Fajardo

, Fajardo

, LeBlanc

, MacPherson

REK

(2017) Examining the relationship between trace lithium in drinking water and the rising rates of age-adjusted Alzheimer’s disease mortality in Texas. J Alzheimers Dis 61, 425–434.

Ahmed

, Mahmood

, Zahid

(2015) Linking insulin with Alzheimer’s disease: Emergence as type III diabetes. Neurol Sci 36, 1763–1769.

10.

de la Monte

, Wands

(2008) Alzheimer’s disease is type 3 diabetes-evidence reviewed. J Diabetes Sci Technol 2, 1101–1113.

11.

Mayeux

, Stern

(2012) Epidemiology of Alzheimer disease. Cold Spring Harb Perspect Med 2 doi:.

12.

Profenno

, Porsteinsson

, Faraone

(2010) Meta-analysis of Alzheimer’s disease risk with obesity, diabetes, and related disorders. Biol Psychiatry 67, 505–512.

13.

Walker

, Harrison

(2015) Shared neuropathological characteristics of obesity, type 2 diabetes and Alzheimer’s disease: Impacts on cognitive decline. Nutrients 7, 7332–7357.

14.

Trivedi

, Liu

, Probst

, Martin

(2013) The metabolic syndrome: Are rural residents at increased risk?. J Rural Health 29, 188–197.

15.

Trivedi

, Liu

, Probst

, Merchant

, Jhones

, Martin

(2015) Obesity and obesity-related behaviors among rural and urban adults in the USA. Rural Remote Health 15, 3267.

16.

Befort

, Nazir

, Perri

(2012) Prevalence of obesity among adults from rural and urban areas of the United States: Findings from NHANES (2005-2008). J Rural Health 28, 392–397.

17.

Jackson

, Doescher

, Jerant

, Hart

(2005) A national study of obesity prevalence and trends by type of rural county. J Rural Health 21, 140–148.

18.

Kessing

, Gerds

, Knudsen

, Jorgensen

, Kristiansen

, Voutchkova

, Ernstsen

, Schullehner

, Hansen

, Andersen

, Ersboll

(2017) Association of lithium in drinking water with the incidence of dementia. JAMA Psychiatry 74, 1005–1010.

19.

Viña

, Lloret

(2010) Why women have more Alzheimer’s disease than men: Gender and mitochondrial toxicity of amyloid-beta peptide. J Alzheimers Dis 20 (Suppl 2), S527–533.

20.

Anderson

, Bulatao

, Cohen

(2004) Critical Perspectives on Racial and Ethnic Differences in Health in Late Life. The National Academies Collection: Reports funded by National Institutes of Health, Washington, DC.

21.

Sharp

, Gatz

(2011) Relationship between education and dementia: An updated systematic review. Alzheimer Dis Assoc Disord 25, 289–304.

22.

Naderali

, Ratcliffe

, Dale

(2009) Obesity and Alzheimer’s disease: A link between body weight and cognitive function in old age. Am J Alzheimers Dis Other Demen 24, 445–449.

23.

Whitmer

(2007) Type 2 diabetes and risk of cognitive impairment and dementia. Curr Neurol Neurosci Rep 7, 373–380.

24.

Ratcliffe

, Burd

, Holder

, Fields

(2016) ACSGEO-1, U.S. Census Bureau Washington, DC.

25.

Ansari

(2009) Effect of physical activity and obesity on type 2 diabetes in a middle-aged population. J Environ Public Health 2009, 195285.

26.

Suwannaphant

, Laohasiriwong

, Puttanapong

, Saengsuwan

, Phajan

(2017) Association between Socioeconomic Status and Diabetes Mellitus: The National Socioeconomics Survey, 2010 and 2012. J Clin Diagn Res 11, LC18–LC22.

27.

Freland

, Beaulieu

(2012) Inhibition of GSK3 by lithium, from single molecules to signaling networks. Front Mol Neurosci 5, 14.

28.

Pathman

, Ricketts

3rd, Konrad

(2006) How adults’ access to outpatient physician services relates to the local supply of primary care physicians in the rural southeast. Health Serv Res 41, 79–102.

29.

Brems

, Johnson

, Warner

, Roberts

(2006) Barriers to healthcare as reported by rural and urban interprofessional providers. J Interprof Care 20, 105–118.

30.

Goins

, Williams

, Carter

, Spencer

, Solovieva

(2005) Perceived barriers to health care access among rural older adults: A qualitative study. J Rural Health 21, 206–213.

31.

Russ

, Batty

, Hearnshaw

, Fenton

, Starr

(2012) Geographical variation in dementia: Systematic review with meta-analysis. Int J Epidemiol 41, 1012–1032.

32.

James

, Leurgans

, Hebert

, Scherr

, Yaffe

, Bennett

(2014) Contribution of Alzheimer disease to mortality in the United States. Neurology 82, 1045–1050.