Comorbidities and apolipoprotein E genotypes of patients with mild cognitive impairment in transition to Alzheimer's disease

Abstract

Background

Three common chronic diseases in the elderly: diabetes, hypertension, and hypercholesterolemia, associate with mild cognitive impairment (MCI) and Alzheimer's disease (AD).

Objective

We will examine the association of apolipoprotein E (APOE) ε4 allele, diabetes, hypertension, and hypercholesterolemia (in combination) with the transition of MCI to AD.

Methods

We examine patients from the National Alzheimer's Coordinating Center database from June 2005 to May 2021. AD converted from MCI, stable MCI, and non MCI/AD control subjects were analyzed using Cox proportional hazard models with propensity score weights on matching demographic information and medications prescribed at baseline.

Results

With MCI time of diagnosis as the index date, MCI patients with diabetes and hypertension carried a higher risk of developing AD (HR = 1.17, 95%CI (1.04, 1.31), p = 0.01) compared to MCI patients with a single condition. A similar observation was found among MCI patients with diabetes and hypercholesterolemia (HR = 1.20, 95%CI (1.07, 1.36), p = 0.002). Compared to MCI patients who had a single condition and without APOE ε4 allele, MCI patients with APOE ε4/4 and both diabetes and hypertension have a significantly higher risk of AD onset (HR = 7.6, 95%CI (5.02, 11.5), p < 0.0001). Those with APOE ε3/4 also have a significantly high risk (HR = 2.3, 95%CI (1.92, 2.75), p < 0.0001). Comparable outcomes were found among those with diabetes and hypercholesterolemia.

Conclusions

The combination of diabetes with hypertension or hypercholesterolemia have a significant association with the progression of MCI to AD, and APOE ε4 allele enhances the association of these selected comorbidities in promoting this conversion.

Keywords

Alzheimer's disease diabetes hypercholesterolemia hypertension mild cognitive impairment

Introduction

Recent research underscores the complex interplay between chronic conditions such as diabetes, hypertension, and hypercholesterolemia, and their progression to Alzheimer's disease (AD). A study from the “All of Us” program by the National Institutes of Health¹ examined the relationship between dementia, hypertension, and type 2 diabetes, highlighting that hypertension, in particular, may be a modifiable risk factor for dementia. The study also found that the association between hypertension and dementia varies among different racial and ethnic groups. It was noted that hypertension and diabetes are comorbidities in patients with dementia, with a significant majority of dementia patients who have diabetes also with hypertension. Another study investigated the effects of diabetes and hypertension on dementia risk.² Patients with diabetes followed by hypertension were associated with a significantly higher risk of all-cause dementia and vascular dementia compared to those without subsequent hypertension.² Similarly, patients with hypertension followed by diabetes were associated with significantly higher risks of all-cause dementia, vascular dementia, and other forms of dementia compared to those without subsequent diabetes. Both diabetes and hypertension may independently and cumulatively increase the risk of various types of dementia.³ Furthermore, a previous study reports that the risk factors for dementia may vary between diabetic and non-diabetic subjects.⁴ In patients with both hypertension and hyperlipidemia, the risk of dementia was significantly higher compared to those without these conditions.

Despite these advances, there remains a significant gap in understanding how these conditions impact the disease progression, and whether they worsen the phenotypes among those carrying a genetic risk factor apolipoprotein E (APOE) ε4 allele; different alleles of APOE, ε2, ε3, and ε4, and if they have a distinct impact on the risk and progression of AD.⁵ Of these, APOE ε3 is the most common genotype in the population and APOE ε3/ε3 is typically used as the reference category to describe AD risk. The APOE ε4 allele is the most notable, being the strongest genetic risk factor for sporadic AD; conversely, the APOE ε2 allele is recognized as a protective factor against AD in White and African American but not Asian or Hispanic individuals.^6,7 APOE ε4 exacerbates the accumulation of amyloid-beta plaques in the brain, which are associated with AD pathology.⁸ In addition to neuritic plaques, APOE impacts tau pathology and tau-mediated neurodegeneration, and influences microglial responses to amyloid and tau pathologies.⁵ Genetic studies have shown that different APOE alleles have varied effects on the burden of these plaques and the severity of cerebral amyloid angiopathy (CAA), a condition often associated with AD.⁵ The role of APOE in AD pathogenesis is under intensive investigation, and different APOE alleles can interact with other proteins and processes involved in Alzheimer's pathology.⁹ For instance, APOE ε4 has been linked with a higher risk of dementia in synucleinopathies, a group of neurodegenerative diseases that include Parkinson's disease and dementia with Lewy bodies.¹⁰ Understanding the role of APOE in AD patients with different comorbidities provides information on other neurodegenerative diseases. This may have broader implications for future risk assessment and treatment strategies.

In this study, we focus on the association of three common chronic diseases in the US population—diabetes, hypertension, and hypercholesterolemia, with the risk of developing AD among those carrying APOE ε4 allele. A key aspect of this study is to investigate how these chronic conditions, individually and in combination, carry different hazard ratios in the presence of different APOE genotypes.

Methods

Study population and samples

NACC data from June 2005 to May 2021 was obtained on 123,071 subjects with visit data, diagnosis of AD or MCI, cognitive test scores, APOE genotype, and patient reported diagnosis of different comorbidities. Based on this data, we defined our study groups with the following criteria. A patient identified as MCI if he/she was diagnosed with Amnestic MCI or non-Amnestic MCI according to the diagnosis guideline.¹¹ If a patient's record showed a diagnosis with any cognitive impairment (dementia, MCI, or impaired, not MCI) and at least two diagnoses of AD after the first MCI diagnosis (to control for rule out diagnosis), this patient was defined as MCI-to-AD (n = 1330) (Table 1). The inclusion criteria include age>=65 by the end of the study (May 1, 2021), MCI diagnosis before AD diagnosis or any dementia diagnosis. Patients whose first visit date includes the diagnosis of AD or dementia were excluded from the study. Patients who had a normal cognitive test on record after their MCI diagnosis were also excluded from the study.

Table 1.

Descriptive information of three groups of patients.

	Non-MCI/AD		Stable MCI		MCI-to-AD
	N = 1556		N = 1674		N = 1330
	Mean	SD	Mean	SD	Mean	SD
Age	88.0	8.5	86.5	10.1	89.4	8.7
Baseline MMSE Score	28.8	1.4	26.9	2.6	27.2	2.3
	N	%	N	%	N	%
Male	579	37.2	795.0	47.5	664.0	49.9
Female	977	62.8	879.0	52.5	666.0	50.1
Race (white)	1337	85.9	1228.0	73.4	1101.0	82.8
Race (Black)	189	12.2	316.0	18.9	173.0	13.0
Race (Asian)	22	1.4	47.0	2.8	27.0	2.0
Race (Others)	8	0.5	83.0	5.0	29.0	2.2
APOE Genotype
APOE ε4/4	24	1.5	43.0	2.6	98.0	7.4
APOE ε3/4	282	18.1	294.0	17.6	416.0	31.3
APOE ε4/2	30	1.9	18.0	1.1	31.0	2.3
APOE ε3/3	881	56.6	575.0	34.4	563.0	42.3
APOE ε3/2	185	11.9	108.0	6.5	91.0	6.8
APOE ε2/2	10	0.6	7.0	0.4	3.0	0.2
APOE not available	144	9.3	629.0	37.6	128.0	9.6
Comorbidities
Hypercholesterolemia	825	53.0	903.0	53.9	763.0	57.4
Hypertension	858	55.1	1021.0	61.1	757.0	56.9
Thyroid Disease	313	20.2	266.0	16.0	239.0	18.1
Diabetes	182	11.7	319.0	19.1	184.0	13.9
Heart Attack / Cardiac Arrest	92	5.9	144.0	8.6	101.0	7.6
Stroke	38	2.5	172.0	10.3	94.0	7.1
Transient Ischemic Attack	77	5.0	108.0	6.5	88.0	6.7
Alcohol Abuse	44	2.8	115.0	6.9	67.0	5.1
B12 Deficiency	75	4.9	73.0	4.5	55.0	4.2
Congestive Heart Failure	49	3.2	64.0	3.8	40.0	3.0
Seizures	23	1.5	73.0	4.4	35.0	2.6
Parkinson's Disease	12	0.8	146.0	8.8	10.0	0.8

Propensity score matching methodology was used to select non-MCI/AD subjects who had no MCI, dementia or AD during the study time frame (N = 1566). We also extracted patients who were post diagnosed with MCI but no AD or dementia (stable MCI, N = 1674). In total, 4560 patients were selected as the study sample (Table 1).

This study was approved by the VA Bedford Healthcare System Institutional Review Board and all subjects were de-identified.

Diabetes, hypertension, and hypercholesterolemia

We focus on the association of three chronic diseases (diabetes, hypertension, and hypercholesterolemia) individually in comparison to co-morbid combination of conditions with the occurrence of AD. Subjects are identified with a diagnosis of diabetes before their index date (first MCI diagnosis date; Supplemental Figure 1). Similarly, we defined subjects with hypertension or hypercholesterolemia in a similar fashion. Our focus is on the patients with two chronic conditions as opposed to those patients with one condition. Although self-reported data can be subject to various biases, including reporting bias that may affect the accuracy of the information, evidence of medications prescribed on a daily basis specific for these disease conditions corroborates the diagnosis of diabetes, hypertension, and hypercholesterolemia in the self-reported database. To further control for the confounding effects from the selected sample, we included covariables at the index date which include heart attack/cardiac arrest, congestive heart failure, stroke, transient ischemic attack, Parkinson's disease, seizures, diabetes, hypertension, hypercholesterolemia, B12 deficiency, thyroid disease, alcohol abuse, and a cognitive test score (Mini Mental State Examination, MMSE).¹²

We applied the random forest method, a machine learning approach to screen the variables from the dataset for the models.¹³ MMSE covariate scores in the models provide the adjusted estimation of the effects of the target chronic diseases. Medications also serve as covariates in the models and include: antihypertensive or blood pressure medications and antihypertensive combination therapies (ACEI, antiadrenergic agents, beta blockers, calcium Channel Blockers, diuretics, vasodilators, angiotensin II inhibitors), lipid-lowering medications, nonsteroidal anti-inflammatory medications, anticoagulant or antiplatelet agents, antidepressants, antipsychotics, anxiolytics, sedative, or hypnotic agents, FDA-approved medications for AD (data prior to FDA approval of lecanemab and donanemab), anti-Parkinson agents, estrogen hormone therapies, estrogen + progestin hormone therapies, and diabetes medications.

Analysis approaches

We employ Cox proportional hazard models, accounting for demographic information, comorbidities and baseline medication, to assess the risk these conditions pose to AD development. We examine the combined effect of diabetes, hypercholesterolemia, and hypertension on AD progression, particularly in patients initially diagnosed with MCI. This analysis was done in two groups of subjects: both comprising individuals transitioning from MCI to AD; the first one consisting of non-AD, non-dementia subjects, and the second one consisting of stable MCI subjects.

Patients’ demographic information, APOE, and comorbidities at baseline were extracted. Dependent variables were the time from the index date to the first diagnosis of AD or the end point of the study (May 1, 2021). We conducted doubly robust Cox proportional hazard models with propensity score weights for each of the disease combinations to compare with the corresponding reference group within our study groups. To eliminate outlier subjects’ influence on the model, we applied propensity score stabilization to the propensity score weights. For all patients who had MCI diagnoses, we used their first MCI diagnosis date as the index date. For each patient from the non-MCI/AD group, we applied propensity score matching and used the matched MCI-to-AD patient's index dates. During this process, patients who died or did not have any AD diagnosis before the censoring date were excluded from the analysis.

Our study subjects contain non-MCI/AD subjects, stable MCI and MCI to AD subjects. Statistical models were established to compare the impact of combined conditions on the risk of AD, and their interaction with APOE ε4 allele.

Propensity score weights for the models of all groups were computed based on patient demographic information (age, sex, race). APOE genetic information was included in the model as a covariate comparing the disease combination with a single disease. The comorbidities were included in all the models to control for potential confounding effects. Sensitivity analysis was repeated in a separate model with different propensity score weights to confirm the robustness of the models.

Interaction effects of the combined conditions and APOE ε4 were tested with 7 subgroups based on the different conditions and APOE ε4 genotypes (Supplemental Table 2). Two Cox models with propensity score weights for all groups were applied to examine the different risk of AD onset among these subgroups. In this analysis, we used individual conditions (diabetes, hypertension, and hypercholesterolemia) and a lack of APOE ε4 (no APOE ε4) as the reference group, to explore the different risks to AD onset when combining disease conditions with APOE ε4.

Results

We analyzed the data extracted from NACC database with the study window of June 2005 to May 2021. Our subjects have an average age of 89.41 (SD = 8.66) for the MCI-to-AD subjects, 86.52 (SD = 10.10) for the stable MCI subjects, and 87.97 (SD = 8.52) for the non-MCI/AD subjects, with 49.92% males in MCI-to-AD subjects, and 47.39% males in stable MCI subjects. Distribution of Caucasians in our samples are: 82.78% in MCI-to-AD, 73.36% in stable MCI, and 85.93% in non-MCI/AD subjects.¹⁰ MCI-to-AD subjects have the highest APOE ε4 at 7.37%, while the Stable MCI subjects are 2.57%, and the non-MCI/AD subjects have the lowest APOE ε4 at 1.54%. This is consistent with APOE's distribution among the general population.¹⁰ The average MMSE score at the index date for the non-MCI/AD subjects was 28.81, while the stable MCI group's average MMSE scores are 26.91, and the MCI-to-AD subjects is 27.15 (Table 1). Patients with MCI (Stable MCI and MCI-to-AD) have a higher prevalence of a number of comorbidities compared to the non-MCI/AD subjects (Table 1).

We assessed the risk among non-MCI/AD subjects and AD subjects who converted from MCI. Second, we focused on patients with MCI diagnosis and assessed their risk among stable MCI and AD subjects who converted from MCI. Patients with the combinations of diabetes and hypertension have a moderate but significantly increased risk of developing AD (HR = 1.13, 95% CI (1.00, 1.27), p = 0.04), compared to those with a single condition (Table 2). Those with comorbid diabetes and hypercholesterolemia did not reveal a significant increase in their risk (HR = 0.95, 95% CI (0.77, 1.17), p = 0.62). Compared to MCI subjects with only diabetes or hypertension, those MCI subjects with both diseases at baseline have a higher risk of developing AD at a later date (HR = 1.17, 95%CI (1.04, 1.31), p = 0.01). Similarly, MCI subjects with both diabetes and hypercholesterolemia at baseline have a higher risk of AD onset (HR = 1.20, 95%CI (1.07, 1.36), p = 0.002), compared to MCI subjects with only one of these two diseases. All the Cox models were adjusted by demographic information, genetic information, and the patient's comorbidities at baseline. Sensitivity analysis with propensity score weights based on the patient's demographic information, APOE, and medication usage at baseline yielded similar conclusions (Supplemental Table 1). Importantly, we found that above HR were lower than those of subjects carrying one or two allele of APOE ε4 allele (Table 2).

Table 2.

Association of diabetes/hypertension/hypercholesterolemia and APOE ε4 allele with the risk to AD using Cox model with propensity score weighting.

		Non-MCI/AD and MCI-to-AD				Stable MCI and MCI-to-AD
Comparison Groups	Parameter	HR	95% CI		p	HR	95%CI		p
Diabetes or Hypertension	Diabetes & Hypertension	1.13	1	1.27	0.0428	1.17	1.04	1.31	0.01
	APOE ε4/4	3.42	2.64	4.44	<0.0001	3.83	3	4.88	<0.0001
	APOE ε3/4	1.59	1.4	1.81	<0.0001	2.19	1.92	2.49	<0.0001
	APOE ε2/4	1.71	1.25	2.34	0.0007	4.85	3.53	6.66	<0.0001
	APOE ε2/3	0.8	0.66	0.97	0.0243	1.86	1.49	2.32	<0.0001
	APOE ε2/2	1	0.26	3.87	0.999	0.73	0.2	2.61	0.63
Diabetes or Hypercholesterolemia	Diabetes & Hypercholesterolemia	0.95	0.77	1.17	0.62	1.2	1.07	1.36	0.002
	APOE ε4/4	2.68	1.99	3.6	<0.0001	4.09	3.23	5.18	<0.0001
	APOE ε3/4	1.69	1.43	2	<0.0001	2.32	2.03	2.64	<0.0001
	APOE ε2/4	1.55	0.95	2.53	0.08	4	2.96	5.42	<0.0001
	APOE ε2/3	0.91	0.65	1.29	0.61	1.75	1.35	2.25	<0.0001
	APOE ε2/2	1.18	0.29	4.78	0.82	0.46	0.12	1.81	0.27

All models were adjusted by age, sex, race, ethnicity, heart attack / cardiac arrest, congestive heart failure, stroke, transient ischemic attack, Parkinson's disease, seizures, weighted by propensity score, which is computed based on demographic information and the diabetes, hypertensions, or hypercholesterolemia disease.

We further explored the interaction of APOE ε4 genotype and their comorbidities on their impact on AD onset. A total of 1885 patients have either diabetes, hypertension or both diseases. There were 1036 patients (non-APOE ε4 carriers) who had either diabetes or hypertension; 435 patients carried APOE ε4 allele but no diabetes and hypertension conditions; 98 patients carried any APOE ε4 allele (APOE ε4/4, APOE ε3/4, or APOE ε4/2) and both diabetes and hypertensions. Using the similar statistical model including propensity score weights with adjusted patient demographics yielded a systematic increase of risks for developing AD when APOE ε4 allele is present with both conditions, compared to those with a single condition (Table 3, Figures 1–4).

Figure 1.

Impact of combined diabetes and hypertension with different APOE ε4 allele on AD converted from non-MCI/AD and MCI subjects. All models were adjusted by sex, race, ethnicity, Heart Attack / Cardiac Arrest, Congestive Heart Failure, Stroke, Transient Ischemic Attack, Parkinson's Disease, Seizures, B12 Deficiency, Thyroid Disease, Alcohol Abuse, and weighted by propensity score. APOE ε4/4 carriers with both diabetes and hypertension have the highest hazard ratio of AD onset.

Figure 2.

Impact of combined diabetes and hypertension with different APOE ε4 allele on AD onset among patients with MCI. An MCI patient with a single or double APOE ε4 allele carries a higher hazard ratio compared to those without an APOE ε4 allele. Combination of diabetes and hypertension in MCI patients with APOE ε4 dramatically increased the hazard ratio of AD onset, and the presence of APOE ε2 had a minimum suppressive effect.

Figure 3.

Impact of combined diabetes and hypercholesterolemia with different APOE ε4 allele on AD converted from non-MCI/AD and MCI subjects. Combination of diabetes, hypercholesterolemia, and two APOE ε4 alleles has the highest hazard ratio of AD onset. Carrying a single APOE ε4 allele has a modest but significant increase of hazard ratio of having AD.

Figure 4.

Impact of combined diabetes and hypercholesterolemia with different APOE ε4 allele on AD onset among patients with MCI. An MCI patient without APOE ε4 allele does not have a higher hazard ratio of AD onset when carrying comorbidities of diabetes and hypercholesterolemia. Combination of diabetes and hypercholesterolemia in MCI patients with APOE ε4 dramatically increased the hazard ratio of AD onset.

Table 3.

Interaction of diabetes/hypertension/hypercholesterolemia with APOE ε4 in conversion of MCI to AD.

Interaction with APOE	Parameter	Non-MCI/AD and MCI-to-AD				Stable MCI and MCI-to-AD
Interaction with APOE	Parameter	HR	95% CI		p	HR	95%CI		p
Model 1: Diabetes and Hypertension	APOE ε4/4, Diabetes, Hypertension	12.77	7.79	20.94	<0.0001	7.6	5.02	11.5	<0.0001
	APOE ε3/4, Diabetes, Hypertension	1.68	1.4	2.01	<0.0001	2.3	1.92	2.75	<0.0001
	APOE ε2/4, Diabetes, Hypertension	2.28	1.53	3.41	<0.0001	11.69	7.64	17.89	<0.0001
	Diabetes, Hypertension	1.22	1.06	1.41	0.006	0.94	0.82	1.09	0.43
	APOE ε4	2.14	1.82	2.5	<0.0001	1.9	1.62	2.21	<0.0001
Model 2: Diabetes and Hypercholesterolemia	APOE ε4/4, Diabetes, Hypercholesterolemia	16.71	11.58	24.12	<0.0001	16.91	11.92	23.99	<0.0001
	APOE ε3/4, Diabetes, Hypercholesterolemia	1.34	1.11	1.62	0.0025	2.52	2.11	3.01	<0.0001
	APOE ε2/4, Diabetes, Hypercholesterolemia	2.7	1.81	4.05	<0.0001	4.48	3.11	6.46	<0.0001
	Diabetes, Hypercholesterolemia	1.09	0.93	1.27	0.276	0.88	0.75	1.03	0.11
	APOE ε4	1.94	1.66	2.26	<0.0001	1.85	1.58	2.17	<0.0001

Model 1: reference group includes patients with diabetes or hypertension, without APOE ε4; Model 2: reference group includes patients with diabetes or hypercholesterolemia, without APOE ε4. APOE ε4: includes APOE ε2/4, APOE ε3/4, and APOE ε4/4.

In the absence of APOE ε4 allele, non-MCI/AD subjects and MCI-to-AD subjects with combined diabetes and hypertension have increased risk of AD onset (HR = 1.22, 95% CI (1.06, 1.41), p = 0.006) (Table 3, Figure 1). Consistent with APOE ε4 as the strongest genetic risk factor for AD, all of our subjects with APOE ε4 allele without diabetes, hypertension, or hypercholesterolemia showed increase HR on AD onset (p < 0.001). Among all APOE ε2/4 carriers, we only have 30 non-MCI/AD, 18 stable MCI, and 31 MCI-to-AD subjects (Table 1), and distribution of those APOE ε2/4 carriers with diabetes, hypertension, or hypercholesterolemia is scarce with low statistical power, but they reached statistical significance in differences between subjects with combined conditions compared with a single condition with greater statistical power (Table 3).

Patients with APOE ε3/4 and both diabetes and hypertension have a higher risk of developing AD (HR = 1.68 (95% CI (1.40, 2.01), and HR = 2.30 (95%CI (1.92, 2.75), p < 0.0001, among non-MCI/AD and MCI-to-AD subjects, and stable MCI and MCI-to-AD subjects, respectively) (Table 3, Figures 1 and 2). Presence of a APOE ε3 and a APOE ε4 significantly increases the risk of AD onset when both diabetes and hypercholesterolemia are present in patients, compared to those with a single condition (HR = 1.34, 95% CI (1.11, 1.62, p = 0.0025, and HR = 2.52 (95%CI (2.11, 3.01), p < 0.0001, among non-MCI/AD and MCI-to-AD subjects, and stable MCI and MCI-to-AD subjects, respectively) (Table 3, Figures 3 and 4).

Among non-MCI/AD subjects and MCI-to-AD subjects, patients with double allele of APOE ε4/4 and comorbid diabetes and hypertension have a significantly higher risk of AD onset, with a HR of 12.77 (Table 3, Figure 1); a significantly higher risk of AD onset was also found among stable MCI and MCI-to-AD subjects, reaching a HR of 7.60 (95%CI (5.02,11.5), p < 0.0001) (Table 3, Figure 2).

Among Non-MCI/AD subjects and MCI-to-AD subjects, those with both diabetes and hypercholesterolemia and APOE ε4/4 double allele have a significantly higher risk of AD onset (HR = 16.71, 95% CI (11.58, 24.12), compared to those with a single disease condition (Table 3, Figure 3). A similar risk was found among stable MCI and MCI-to-AD subjects (HR = 16.91, 95%CI (11.92, 23.99), p < 0.0001) (Table 3, Figure 4).

In sum, the visual pattern displays that APOE ε4/4 genotype have a highly significant risk to promote MCI-to-AD conversion in patients with combined diabetes and hypertension (Figures 1 and 2). Similar patterns were observed for diabetes and hypercholesterolemia (Figures 3 and 4).

Discussion

Our findings indicate that patients with a combination of diabetes and hypertension or hypercholesterolemia at baseline demonstrate a higher risk of progressing to AD compared to those with either condition alone. This suggests a significant risk factor in the presence of diabetes, hypertension and hypercholesterolemia for individuals diagnosed with MCI. Previous studies reported the association of comorbid conditions with AD progression and reduced survival time of AD patients.¹⁴ This study focuses on the conversion from MCI to AD among patients with one or two comorbidities of hypertension, hypercholesterolemia, and diabetes, at the intersection of a genetic risk factor of AD, i.e., APOE ε4. This is unique and highly significant in relation to clinical meaningfulness of managing modifiable risk factors for AD.

We found significant associations between the combination of diabetes and hypertension or hypercholesterolemia among MCI patients and the risk of AD onset. We do not suggest a causal inference, based on our outcomes.^15,16 Our results reveal the significant associative interaction of chronic conditions with APOE ε4 allele among MCI patients. The combination of two diseases with exposed patients having APOE ε4/4 allele are at a higher risk compared to patients with no APOE ε4 allele. Our results support a strong association of chronic conditions among MCI patients with any APOE ε4 allele and the conversion from MCI to AD.

Our study, while providing valuable insights into the progression and risk factors of AD, has several limitations. First, there is a lack of medication duration for the three diseases without precise diagnosis dates. The notable limitation is the lack of prescription fill and refill initiation data. Without this information, we cannot determine the precise timeline of medication use, which is crucial for understanding the temporal relationships between drug exposure and AD onset. The absence of dosage information further compounds this issue, Furthermore, we did not study all three-disease combinations (diabetes, hypertension and hypercholesterolemia), since we only had a small “n” (98 patients) who carried any APOE ε4 allele (APOE ε4/4, or APOE ε3/4 or APOE ε2/4) and had comorbid diabetes and hypertensions. We did not have adequate statistical power to analyze those carrying diabetes, hypertension, hypercholesterolemia and different APOE ε4 alleles. In addition, the medication indications for these three conditions may be given for other clinical conditions as well. The lack of Montreal Cognitive Assessment (MoCA) scores to evaluate the MCI-to-AD, stable MCI and non-MCI/AD subjects is another limitation. The MoCA is considered more sensitive than the MMSE for identifying MCI, while both tests are found to be accurate in the detection of AD.¹⁷ Without this, there is a potential for discordance between objective MMSE scores and clinicians’ subjective assessment,¹⁸ which provides a close range of MMSE scores among non-MCI/AD, stable MCI and MCI-to-AD subjects (Table 1). Finally, the small sample size for patients carrying APOE ε2 limits the generalizability of the findings and future work requires extending this analysis to a much larger database, such as the administrative and clinical data within the Veteran Health Administration data sets.

In summary, our study offers important insights of diabetes, hypertension and hypercholesterolemia that impact the progression of AD from MCI. Robust data collection methods, objective clinical assessments, comprehensive medical records, and standardized cognitive evaluations are the foundation for future studies to validate the modifiable risk factors contributing to AD progression. Our results indicate the importance of following the comorbidities of those with MCI. In light of recent efforts to re-purpose glucagon-like peptide-1 (GLP-1) receptor agonists as a novel AD therapeutic agent,¹⁹ diabetes, hypertension, and hypercholesterolemia, would be critical morbidities for recruiting subjects for clinical trials and to start new treatment upon FDA approval.

Supplemental Material

sj-docx-1-alr-10.1177_25424823251353209 - Supplemental material for Comorbidities and apolipoprotein E genotypes of patients with mild cognitive impairment in transition to Alzheimer's disease

Supplemental material, sj-docx-1-alr-10.1177_25424823251353209 for Comorbidities and apolipoprotein E genotypes of patients with mild cognitive impairment in transition to Alzheimer's disease by Mingfei Li, Ying Wang, Lewis Kazis, Jiaying Weng and Weiming Xia in Journal of Alzheimer's Disease Reports

Footnotes

Acknowledgements

The views expressed in this article are those of the authors and do not represent the views of the US Department of Veterans Affairs and the US Government. The NACC database is funded by NIA/NIH Grant U24 AG072122. NACC data are contributed by the NIA-funded ADRCs: P30 AG062429 (PI James Brewer, MD, PhD), P30 AG066468 (PI Oscar Lopez, MD), P30 AG062421 (PI Bradley Hyman, MD, PhD), P30 AG066509 (PI Thomas Grabowski, MD), P30 AG066514 (PI Mary Sano, PhD), P30 AG066530 (PI Helena Chui, MD), P30 AG066507 (PI Marilyn Albert, PhD), P30 AG066444 (PI David Holtzman, MD), P30 AG066518 (PI Lisa Silbert, MD, MCR), P30 AG066512 (PI Thomas Wisniewski, MD), P30 AG066462 (PI Scott Small, MD), P30 AG072979 (PI David Wolk, MD), P30 AG072972 (PI Charles DeCarli, MD), P30 AG072976 (PI Andrew Saykin, PsyD), P30 AG072975 (PI Julie A. Schneider, MD, MS), P30 AG072978 (PI Ann McKee, MD), P30 AG072977 (PI Robert Vassar, PhD), P30 AG066519 (PI Frank LaFerla, PhD), P30 AG062677 (PI Ronald Petersen, MD, PhD), P30 AG079280 (PI Jessica Langbaum, PhD), P30 AG062422 (PI Gil Rabinovici, MD), P30 AG066511 (PI Allan Levey, MD, PhD), P30 AG072946 (PI Linda Van Eldik, PhD), P30 AG062715 (PI Sanjay Asthana, MD, FRCP), P30 AG072973 (PI Russell Swerdlow, MD), P30 AG066506 (PI Glenn Smith, PhD, ABPP), P30 AG066508 (PI Stephen Strittmatter, MD, PhD), P30 AG066515 (PI Victor Henderson, MD, MS), P30 AG072947 (PI Suzanne Craft, PhD), P30 AG072931 (PI Henry Paulson, MD, PhD), P30 AG066546 (PI Sudha Seshadri, MD), P30 AG086401 (PI Erik Roberson, MD, PhD), P30 AG086404 (PI Gary Rosenberg, MD), P20 AG068082 (PI Angela Jefferson, PhD), P30 AG072958 (PI Heather Whitson, MD), P30 AG072959 (PI James Leverenz, MD).

ORCID iD

Weiming Xia

Ethical considerations

All studies were performed in accordance with the Guidelines of “the Declaration of Helsinki 1964 and its later amendments”, and studies were approved by the ethics committee Bedford VA Hospital Institutional Review Board.

Author contributions

Mingfei Li: Conceptualization; Data curation; Investigation; Methodology; Project administration; Resources; Supervision; Writing – original draft; Writing – review & editing.

Ying Wang: Data curation; Investigation; Writing – review & editing.

Lewis Kazis: Investigation; Writing – review & editing.

Jiaying Weng: Investigation; Writing – review & editing.

Weiming Xia: Conceptualization; Investigation; Project administration; Resources; Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by National Institute on Aging R01 AG063913 (WX).

Declaration of conflicting interests

Weiming Xia is an Associate Editor of this journal but was not involved in the peer-review process of this article nor had access to any information regarding its peer-review.

Data availability statement

The data supporting the findings of this study are available within the article.

Supplemental material

Supplemental material for this article is available online.

References

Nagar

Pemu

Qian

, et al. Investigation of hypertension and type 2 diabetes as risk factors for dementia in the all of us cohort. Sci Rep 2022; 12: 19797.

Yen

F-S

Wei

JC-C

Yip

H-T

, et al. Diabetes, hypertension, and the risk of dementia. J Alzheimers Dis 2022; 89: 323–333.

Zuin

Roncon

Passaro

, et al. Metabolic syndrome and the risk of late onset Alzheimer's disease: an updated review and meta-analysis. Nutr Metab Cardiovasc Dis 2021; 31: 2244–2252.

Fan

Y-C

Hsu

J-L

Tung

H-Y

, et al. Increased dementia risk predominantly in diabetes mellitus rather than in hypertension or hyperlipidemia: a population-based cohort study. Alzheimers Res Ther 2017; 9: 7.

Yamazaki

Zhao

Caulfield

, et al. Apolipoprotein E and Alzheimer disease: pathobiology and targeting strategies. Nat Rev Neurol 2019; 15: 501–518.

Choudhury

Ramanan

Boeve

. Apoe ɛ4 allele testing and risk of Alzheimer disease. JAMA 2021; 325: 484–485.

Belloy

Andrews

Le Guen

, et al. APOE Genotype and Alzheimer disease risk across age, sex, and population ancestry. JAMA Neurol 2023; 80: 1284–1294.

Raulin

Doss

Trottier

, et al. Apoe in Alzheimer's disease: pathophysiology and therapeutic strategies. Mol Neurodegener 2022; 17: 72.

Serrano-Pozo

Das

Hyman

. APOE And Alzheimer's disease: advances in genetics, pathophysiology, and therapeutic approaches. Lancet Neurol 2021; 20: 68–80.

10.

Jin

Sonoustoun

, et al. APOE4 Exacerbates α-synuclein seeding activity and contributes to neurotoxicity in Alzheimer's disease with Lewy body pathology. Acta Neuropathol 2022; 143: 641–662.

11.

NACC Uniform Data Set Researchers Data Dictionary. National Alzheimer’s Coordinating Center: National Alzheimer’s Coordinating Center, 2015.

12.

Folstein

McHugh

. “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12: 189–198.

13.

Breiman

. Random forests. Mach Learn 2001; 45: 5–32.

14.

Rajamaki

Hartikainen

Tolppanen

A-M

. The effect of comorbidities on survival in persons with Alzheimer’s disease: a matched cohort study. BMC Geriatr 2021; 21: 173.

15.

Kezios

Hayes-Larson

. A clarification on causal questions: we ask them more often than we realize. Am J Public Health 2018; 108: e4.

16.

Hernán

John

Healy

. . A second chance to get causal inference right: a classification of data science tasks. Chance 2019; 32: 42–49.

17.

Pinto

TCC

Machado

Bulgacov

, et al.

Is the Montreal Cognitive Assessment (MoCA) screening superior to the Mini-mental state examination (MMSE) in the detection of mild cognitive impairment (MCI) and Alzheimer's disease (AD) in the elderly?

Int Psychogeriatr 2019; 31: 491–504.

18.

Morin

Wang

, et al. Clinical staging of Alzheimer's disease: concordance of subjective and objective assessments in the Veteran's affairs healthcare system. Neurol Ther 2022; 11: 1341–1352.

19.

Cummings

Atri

Feldman

, et al. Evoke and evoke+: design of two large-scale, double-blind, placebo-controlled, phase 3 studies evaluating efficacy, safety, and tolerability of semaglutide in early-stage symptomatic Alzheimer's disease. Alzheimers Res Ther 2025; 17: 14.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.05 MB