Prevalence of dementia risk factors in the Oxford Brain Health Clinic

Table of Contents

Cohort demographics of patients and healthy volunteers

A total of 342 patients attended the OBHC between August 2020 and May 2024. Research uptake was high (see Table 2.); 91.5% (n = 313) consented to their clinical data, including subsequent diagnosis, being used for research purposes. 71.4% of patients consented to be contacted about future research compared to 12.8% in the MAS 2023 audit⁵.

Table 2 Uptake of research by the OBHC patient population.

Table 3 shows the demographic and diagnostic characteristics of 313 patients and 81 healthy volunteers. 54.3% of the OBHC population was male, whereas 44.4% of the healthy volunteers were male; this difference was not significant (X²(1) = 2.51, p = 0.11; Fisher exact: p = 0.13). Of those whose native language was known, 95.4% of patients spoke English as a first language compared to 97.3% of healthy volunteers (X²(1) = 0.48, p = 0.49; Fisher exact: p = 0.17). Of the OBHC patients whose ethnicity was known, 99.0% were white. By comparison, the 2023 MAS audit found that 43% of patients attending memory assessment services were male, 91.6% of patients whose native language was known spoke English as a first language, and 87.9% of those whose ethnicity was known were white⁵.

Table 3 Demographics and diagnoses (ICD-10) breakdown for OBHC patients, healthy volunteers, and the average memory clinic population per the memory assessment service (MAS) audit⁵.

When categorised by diagnosis, the largest proportion of OBHC patients received a dementia related diagnosis (52.4%, n = 150), followed by a diagnosis of MCI (26.9%, n = 77), and the smallest proportion received a non-memory disorder diagnosis or no formal diagnosis (20.6%, n = 59). Compared with national rates, the OBHC had a lower percentage of patients with a dementia-related diagnosis and a higher proportion of patients with MCI and non-memory disorder diagnosis/no formal diagnosis (MAS audit: 70.5% dementia, 17% MCI and 12.5% non-dementia⁵). Among the OBHC dementia subgroup, 86.7% had a diagnosis of dementia in Alzheimer’s disease (F00) and 13.3% had a non-AD dementia diagnosis; compared to 42.6% and 57.4% of patients nationally⁵.

Figure 1 shows the age and ACE-III total score distributions and correlations for the OBHC patient and healthy volunteer populations. The age distributions of OBHC patients (77.7 ± 6.3 years, range: 65–101) and healthy volunteers (76.4 ± 5.8 years, range: 65–92) were similar, with no significant difference in age between the groups (t(132.27) = 1.8, p = 0.07 with small effect size d = 0.21). Most patients and healthy volunteers are aged between 70 and 85 years. By comparison, the mean age of patients attending MAS was 79.7 years (range 35–102), with 93.9% aged 65 years and older⁵.

Compared with healthy volunteers, OBHC patients scored significantly lower on the ACE-III in both the total score (W = 1648.5, p < 0.001, effect size = − 0.61) and memory sub-score (W = 2326, p < 0.001, effect size = − 0.57). 78.3% of patients (n = 235) had a total ACE-III score of 88 or lower, and 58.7% (n = 176) had a total ACE-III score of 82 or lower; in comparison, 6.2% (n = 4) of healthy volunteers had a total score of 88 or lower, and 1.2% (n = 1) had a total ACE-III score of 82 or lower. There was a significant negative correlation between age and the ACE-III total score for both patients (rs = − 0.32, p < 0.001) and healthy volunteers (rs = − 0.33, p = 0.003); the difference in these correlations was not significant (z = 0.06, p = 0.47).

Imaging characteristics

92.7% of OBHC patients completed the full clinical MRI scan, and 54.6% completed the full research MRI scan; 97.5% of healthy volunteers completed the full clinical and research scans. By comparison, the MAS audit disclosed that brain imaging (MRI or CT) was requested for 47.3% of patients (n = 2910), completed for 44.3% (n = 2725) and 13.2% (n = 812) had a MRI scan; the most common reason that the requested scans were not performed was patient decline (40.5%), previous scan (15.1%), not required (7.6%) and contraindication (1.1%); the reason was unknown for 35.7% of the patients. Table 4 shows the imaging characteristics of the OBHC patients from currently available radiology reports.

Table 4 Summary of completed clinical and research MRI scans (n) and main findings from radiology reports.

DMT eligibility

To explore the proportion of patients who may be eligible for DMTs, we calculated potential eligibility for Lecanemab on the basis of the CLARITY AD trial¹¹ inclusion/exclusion criteria as the first treatment for AD licenced for use in Great Britain by the Medicines and Healthcare Products Regulatory Agency (MHRA). This includes diagnosis, age, cognition, BMI, depressive symptoms, and MRI exclusion criteria based on radiology reports (for more detail, see “Statistical analysis” section). 75.4% of the OBHC population did not meet the trial eligibility criteria, with 149 patients excluded based on age and diagnosis only and an additional 81 excluded based on cognition, BMI, depressive symptoms, and imaging characteristics. Consequently, 24.6% of patients in the OBHC population may be eligible for further investigation into the Lecanemab trial criteria, including exploration of brain amyloid pathology.

Risk prevalence

Compared with healthy volunteers, OBHC patients tended to have a greater prevalence of potentially modifiable dementia risk factors, and the associations between risk prevalence and population were statistically significant for depressive symptoms (PHQ9; X²(1) = 5.92, p = 0.015; Fisher exact: p = 0.015), physical activity (SALS; X²(1) = 16.30, p < 0.001; Fisher exact: p < 0.001) and sleep (PSQI; X²(1) = 5.61, p = 0.017; Fisher exact: p = 0.016), with a greater presence of categorical “high risk” factors in patients than in healthy volunteers.

Figure 2 shows the prevalence of potentially modifiable dementia risk factors in the OBHC patient population. 16.7% (n = 49) of OBHC patients reported moderate-severe symptoms of depression on the PHQ9, compared to 5.5% (n = 4) of healthy volunteers. 54.2% (n = 169) of OBHC patients and 27.8% (n = 20) of healthy volunteers reported engaging in less than 30 min of moderate-intensity physical activity per week and 15.9% (n = 47) of OBHC patients and 8.6% (n = 7) of healthy volunteers had a BMI over 30, indicating obesity. 8.1% (n = 23) of OBHC patients reported no formal qualifications compared to 4.6% (n = 3) of healthy volunteers, 1.0% (n = 3) of OBHC patients and 0% of healthy volunteers reported consuming 6 or more drinks on a single occasion daily or almost daily, and 10.5% (n = 32) of OBHC patients compared to 1.5% (n = 1) of healthy volunteers reported an inappropriate sleep duration of less than 5 or more than 10 h per night.

Self-reported, known family history of dementia was available for 162 OBHC patients and 72 healthy volunteers; 43.8% of patients and 52.8% of healthy volunteers had a known history of dementia; however, patient and heathy volunteer data was collected using different methods. ApoE genotyping was available for 179 OBHC patients, of whom 43.0% (n = 77) had at least 1 ApoE Ɛ4 allele (35.7% heterozygous and 7.3% homozygous); ApoE genotyping was not available for the healthy volunteers.

Risk prevalence across diagnoses

There was a statistically significant difference in age across diagnoses (F(3, 363) = 20.3, p < 0.001). Bonferroni post hoc analysis revealed that patients with a dementia diagnosis (mean = 80.2, n = 150) were significantly older than healthy volunteers (mean = 76.4, n = 81, p < 0.001), patients with an MCI diagnosis (mean = 76.0, n = 77, p < 0.001), and those who received non-memory disorder diagnoses or no formal diagnosis (mean = 74.1, n = 59, p < 0.001). There was also a statistically significant difference in the ACE-III total score across diagnostic groups (H(3) = 246.1, p < 0.001); post-hoc comparisons showed all interactions were significant.

There was a significant difference in depressive symptoms across diagnostic groups (H(3) = 16.38, p < 0.001), with lower total PHQ9 scores in healthy volunteers (mean = 2.9) than in those who received no dementia-related diagnosis (mean = 6.2). Physical activity also differed significantly across diagnoses (H(3) = 15.29, p = 0.002), driven by significantly more activity reported by healthy volunteers compared to patients with a dementia-related diagnosis. Patients with a diagnosis of MCI had a significantly greater BMI (mean = 26.9) than patients with a dementia-related diagnosis (mean = 24.9; H(3) = 10.09, p = 0.01), and healthy volunteers and patients with no memory-related/no formal diagnosis reported significantly less sleep than patients with a dementia diagnosis (H(3) = 20.91, p < 0.001). Long-term conditions (LTCQ-8) total score also differed significantly (H(3) = 35.85, p < 0.001); healthy volunteers (mean = 85.7) scored significantly higher than patients with no memory-related/no formal diagnosis (mean = 77.9), MCI diagnosis (mean = 73.1), or a dementia-related diagnosis (mean = 69.6), and patients with no memory-related/no formal diagnosis scored significantly higher than patients with a dementia-related diagnosis. There was no significant difference in the prevalence of high-risk factors across diagnoses for education (H(3) = 5.15, p = 0.16) or alcohol consumption (H(3) = 7.43, p = 0.06), likely due to the small prevalence of these risk factors in all groups. There was no significant difference across diagnoses for the presence of an Ɛ4 allele in patients (H(2) = 1.10, p = 0.58) or family history of dementia (H(3) = 1.61, p = 0.66). Figure 3 summarises the potentially modifiable risk for each diagnostic group: DRD, MCI, no DRD and healthy volunteers.

Total number of potentially modifiable risks

67.4% of OBHC patients have one or more potentially modifiable dementia risk factors compared to 37% of healthy volunteers. This difference in total risk prevalence between groups was statistically significant (W = 17,350, p < 0.001, effect size = − 0.32), with OBHC patients having a greater number of risks (mean = 1.0 ± 0.6, range: 0–4) than healthy volunteers (mean = 0.4 ± 0.9, range: 0–3). There was also a significant difference in the total number of potentially modifiable dementia risk factors across diagnoses (H(3) = 32.64, p < 0.001), with healthy volunteers having significantly fewer risks (mean = 0.4) than patients with no dementia-related diagnosis (mean = 0.9), MCI (mean = 1.1) or dementia-related diagnosis (mean = 1.1). There was also a significant difference in cognition when the total number of potentially modifiable risks were grouped into 0, 1 and 2 + risks, with a significantly higher ACE-III total score in individuals with no risks (n = 147, mean = 84.5 ± 14.2, range: 23–100), than in those with one (n = 150, mean = 76.8 ± 18.4, range: 9–99) or 2 or more risks (n = 84, mean = 73.0 ± 19.3, range: 20–99). Figure 4 summarises the total number of potentially modifiable risk factors (education, physical activity, depression, BMI, sleep, alcohol consumption) for patients with a dementia-related diagnosis (a), MCI and no dementia-related diagnosis combined (b) and healthy volunteers (c) and the total ACE-III score across risk groups.