Introduction
The prevalence of dementia and of subjective cognitive impairment (SCI) is increasing. Reference Verity, Kirk and O’Connell1 This indicates a laudable increase in awareness of dementia’s presentation and its impact on patients, but also an increased burden on the healthcare system and healthcare providers. Reference Verity, Kirk and O’Connell1 Alzheimer’s disease (AD) is predicted to increase in incidence in the coming decades due to more accurate diagnosis and aging of the population. Reference Cutler2 Contrastingly, some population-based studies demonstrate a trend of decreasing incidence of dementia. Reference Tom, Phadke and Hubbard3,Reference Wu, Beiser and Breteler4 Nonetheless, public awareness of AD along with its burdens and its symptoms is increasing. AD is quickly emerging as a particularly feared condition, creating anxiety and concern in the general population. Reference Takizawa, Thompson and van Walsem5 These concerns are especially common among middle-aged and older adults with first-degree relatives who have been diagnosed with AD. Reference Cutler and Hodgson6 In addition to the impact on the general population, this disease and its increasing prevalence has a significant impact on primary care providers and specialists. The number of referrals to specialists for memory complaints has become disproportionate to dementia diagnoses. Reference Verity, Kirk and O’Connell1 Additionally, epidemiological studies have shown that many people with dementia do not have a diagnosis. Reference Cagliarini, Price and Livemore7 This is known as the diagnosis gap, and indicates a need for improved detection of dementia. Reference Verity, Kirk and O’Connell1,Reference Cagliarini, Price and Livemore7 In the majority of cases, the diagnosis is largely based on clinical evaluation. Reference Verity, Kirk and O’Connell1 Due to this, factors increasing patients’ risk need to be identified to narrow the number of patients referred, and to ensure that those with dementia are diagnosed and receive prompt treatment.
Some factors that have already been associated with AD risk include older age and lower Mini-Mental State Examination (MMSE) score. Higher levels of education and working at least part time may be associated with an increased chance of presenting with cognitive complaints. This may be due to increased awareness of the impact of dementia and memory changes becoming more apparent due to consistent routines and tasks, respectively. Reference Verity, Kirk and O’Connell1 Depression and other psychiatric illnesses may also be risk factors for dementia. Depression can be both an early symptom of dementia, and a contributor to memory concerns. Reference Verity, Kirk and O’Connell1 There are also similarities between patients with dementia and patients with depression, as apathy often develops in both. Reference Verity, Kirk and O’Connell1 Depression is an important factor to consider when studying patients with suspected cognitive decline, as up to 50% of patients with AD also have depression. Reference Chi, Yu and Tan8 This comorbidity impacts the patient’s quality of life and clinical presentation, and worsens the caregiver burden. Reference Chi, Yu and Tan8 A study of cognitive decline in rural and remote populations found that a history of hypertension and decreased ability to perform daily activities predicted greater cognitive decline 1 year later. Reference Lacny, Kirk and Morgan9
SCI has increasingly been investigated as a potential risk factor for the development of dementia. The “worried well”, a term used to describe individuals with SCI, refers to people who are concerned that they may have dementia, but are neurologically normal relative to others in their demographic upon examination and testing. Reference Verity, Kirk and O’Connell1 Psychological and environmental factors may play a role in these concerns, as this group is more likely to have relatives with dementia, or other personal relationships with dementia patients. Reference Verity, Kirk and O’Connell1 SCI has been identified as a risk factor for mild cognitive impairment. Reference Verity, Kirk and O’Connell1 It is important to note that the pathologic changes that occur in AD patients often take place years before symptoms develop. Reference Godbolt, Waldman and MacManus10 Biomarker and imaging studies support the idea that pathological processes may begin more than a decade before a diagnosis of dementia is made. Reference Sperling, Rentz and Johnson11
Based on the information gathered to date, psychiatric illness and psychological factors may be predictors of SCI, or may independently lead to memory concerns. Reference Verity, Kirk and O’Connell1 Psychiatric symptoms may also overlap with behaviors present in the early stages of dementia. Reference Verity, Kirk and O’Connell1 It has been established that long-term, severe worries surrounding dementia are associated with decreased psychological well-being, as well as poorer physical health. Reference Cutler and Hodgson12 Patients with dementia also clearly benefit from early assessment and diagnosis. Reference Chi, Yu and Tan8 It is crucial that the association between SCI and objective cognitive impairment be established, to better understand SCI as a potential predictor of dementia diagnosis to both improve care for patients in need and to decrease the burden on the healthcare system. We aimed to explore measures of cognitive function, function in general, and depressive symptoms through comparing the MMSE, FAQ, and Center for Epidemiologic Studies Depression (CESD) scores, respectively, at clinic day and at 1-year follow-up. In addition, we aimed to explore the MMSE scores over time for those who were followed up long term, as well as the diagnoses given to some of these patients.
Methods
A total of 672 consecutive patients were seen for the initial assessment at the Rural and Remote Memory Clinic (RRMC) between March 2004 and June 2019, 166 of whom were neurologically and neuropsychologically normal at clinic day and, therefore, included in this analysis. 54.8% of patients returned for follow-up. Patients were followed for between 1 and 15.25 years, with a mean follow-up of 2.95 years (SD 2.87). Follow-up appointments occurred between May 2004 up to and including May 2020. Data collection occurred simultaneously. The eligibility criteria for participation in this study were patients presenting to the RRMC within the timeframe of data collection deemed to have SCI at initial assessment and no other neurological diagnosis. The University of Saskatchewan’s RRMC provides an interdisciplinary assessment for rural patients with memory concerns from across Saskatchewan. Reference Verity, Kirk and O’Connell1 The team involved at the initial visit consists of a neurologist, a physiotherapist, a dietician, a nurse, and a neuropsychology team. The patient undergoes neuropsychological testing and a standard workup for reversible causes of memory concerns. Reference Verity, Kirk and O’Connell1 This workup includes a complete blood count (CBC), electrolytes including calcium, thyroid-stimulating hormone (TSH), vitamin B12, neuroimaging, and other investigations when indicated. Reference Verity, Kirk and O’Connell1 The neuropsychological battery includes measures of premorbid ability, attention, speeded mental processing, receptive and expressive language, visuospatial abilities, executive function, and verbal and visual memory. Reference Verity, Kirk and O’Connell1 The RRMC began operating in 2004 and is located in the University of Saskatchewan in Saskatoon. The multidisciplinary staff at the RRMC assess and provide care to rural patients with cognitive impairments. The assessments take one full clinic day and follow-up assessments are conducted at 6 weeks, 12 weeks, 6 months, 1 year, and then annually, with added appointments in between as clinically needed. Clinical and evaluation data are centrally managed to enhance data quality and accessibility, and to facilitate interdisciplinary analyses. Reference Morgan, Crossley and Kirk13 The neuropsychological testing battery is approximately 2 h. More detailed information regarding the RRMC can be found in previous publications. Reference Morgan, Crossley and Kirk13–Reference Crossley, Morgan and Lanting16
Data collection at initial visit includes age, sex, years of formal education, MMSE score, CESD score, FAQ score, Self-Rating of Memory Scale, alcohol consumption, marital status, hours per week of work, past medical history, sleep concerns, possession of a driver’s license, and information regarding a family history of memory concerns. Reference Verity, Kirk and O’Connell1 The assessment team agrees on a diagnosis, which is delivered to the patient at the end of the clinic day. Reference Verity, Kirk and O’Connell1 A diagnosis of SCI is given if the following criteria are met: no clinical evidence of neurologic disease; normal neuroimaging; and normal performance on neuropsychological testing relative to normative comparison standards that are adjusted for demographic variables, which include age, sex, and education. Reference Verity, Kirk and O’Connell1 The patients with a diagnosis of SCI at initial assessment were included in this study, as they were cognitively normal at clinic day. We then compared MMSE, FAQ, and CESD scores at clinic day to those at the 1-year follow-up and examined all of the MMSE scores collected at follow-up appointments at various times for each patient to determine if there was a pattern present. Statistical analyses were conducted using SPSS version 27. 17
The MMSE has been shown to be a reasonable screening tool for assessing possible dementia in the community and primary care. Reference Mitchell18 The neurological examination conducted at clinic day also includes relevant features of a motor exam, which has been demonstrated to be relevant to SCI. Reference Morgan, Funk and Crossley19 Ability to perform daily activities can be measured by many scales, one of which is the Functional Activities Questionnaire (FAQ). Reference Olazarán, Mouronte and Bermejo20 The FAQ has been shown to have internal consistency, reliability, and diagnostic validity. Reference Olazarán, Mouronte and Bermejo20 Other scales, such as the Scale of Instrumental Activities of Daily Living have similar characteristics, but the FAQ is superior in terms of specificity and sensitivity. The FAQ is also less influenced by sex and age. Reference Olazarán, Mouronte and Bermejo20 Studies have shown that using the CESD is an accurate and valid way to screen for depression in the general population. Reference Van Dam and Earleywine21
The outcomes assessed include a significant change in MMSE, CESD, and FAQ scores or lack thereof, and subsequent neurological diagnosis at follow-up or continued SCI designation. Lifestyle and overall health-related variables are outlined in Table 1. Information on these variables was collected by surveying the patients on clinic day. Age may be a potential confounding variable. Potential bias was avoided by analyzing the information without any patient identifiers. Additionally, those that administered the testing of the patients did not analyze the data.
Characteristics of normal SCI patients at baseline including those who remained normal and those who received a diagnosis
*Fisher’s Exact Test P-values were reported due to small cell counts.
Descriptive analyses were completed using frequencies, measures of central tendency, and variability. The two times (clinic day and 1-year follow-up) were compared using paired sample t-tests for continuous variables. Case summaries were used to build the profiles of the nine patients who developed a neurologic diagnosis. Ethics approval was obtained from the University of Saskatchewan Biomedical Research Ethics Board.
Results
The population of 166 patients with SCI had a higher proportion of females than males. The education level of the patient population was predominantly a high school diploma or more/postsecondary education. Most patients were married or common law, owned their house, had a valid driver’s license, and did not experience problems due to drug use. The number of times per week that the patients exercised varied, but most exercised one or more times per week. Most of the patients had been smokers at some point, but a small portion of the patients were current smokers. The number of patients with a family history of dementia was similar to those without.
The majority of the 166 SCI patients had been diagnosed with arthritis. Conversely, only a small portion of the population had been diagnosed with diabetes, heart disease, kidney trouble, or stroke. The proportion of the population diagnosed with hyperlipidemia, hypertension, poor hearing, poor vision, chronic respiratory problems, head injury, stomach problems, and psychiatric problems was relatively similar to the number of those who were not diagnosed with these conditions. The average number of drinks of alcohol per week for the population assessed was 3.8 (SD 3.7), which is well under the recommended maximum for both males and females. The average age at clinic day, years of formal education, and number of times engaged in exercise per week are all included in Table 1. All variables, aside from age at clinic day, have missing values due to the data being primarily self-reported in questionnaires.
Eighty-one of the 166 cases with SCI were followed up to year 1. The results of the analysis of the scores for MMSE, FAQ, and CESD at clinic day and at 1-year follow-up showed no significant difference, as seen in Table 2. Ninety-one of the 166 patients’ MMSE scores over time were examined in the second analysis, as these patients had initial clinic day MMSE scores as well as follow-up MMSE scores at various times after clinic day. No consistent pattern in MMSE scores over time was demonstrated, as seen in Figure 1.
Comparison of MMSE, CESD and FAQ Scores at Clinic Day and Year 1 for patients with SCI at initial assessment using paired t-test
MMSE scores over time of 91individual normal SCI patients separated into four groups based on length of follow-up.
Nine of the 166 patients developed a neurological diagnosis. These diagnoses were made between 0.5 and 8.5 years after the initial assessment. Of these, five were diagnosed with AD; two with MCI; one with vascular cognitive impairment; and one with frontotemporal dementia (FTD). One of the patients with MCI had mild vascular cognitive impairment specifically. Of the patients diagnosed with AD, three were previously diagnosed with MCI. One of the patients with AD had previously undergone chemotherapy prior to and following the clinic day. The patient with vascular cognitive impairment had a history of cardiovascular events leading up to the diagnosis that occurred after the initial clinic visit. The patient with FTD was very slow progressing, and was temporarily suspected to have FTD phenocopy syndrome due to this. These cases are detailed in Table 3.
MMSE scores and important events of patients with neurological diagnoses
Discussion
SCI often leads to referral to a neurologist or other specialist, and further testing to rule out a neurological diagnosis. This can put a strain on the healthcare system due to the influx of specialist consults and can cause a great degree of stress for these patients and their families. It is, therefore, important to determine whether or not these patients truly benefit from referral. Through our exploration of this topic, we were able to gain insight into the progression of the “worried well” following their initial visit to the RRMC, and whether or not they progress to having a neurological diagnosis.
Deficits related to attention that are associated with brain activation changes in early AD progression may be associated with SCI. One study showed that patients in the SCI group had increased activation in the anatomical areas in which decreased activation normally occurs during divided attention in AD patients. Reference Lacny, Kirk and Morgan9 This indicates that compensatory changes may occur in patients with SCI. Reference Lacny, Kirk and Morgan9 Patients with SCI have been shown to have a significant increase in the risk of conversion to AD compared to people without SCI. Reference Lacny, Kirk and Morgan9,Reference Rodda, Dannhauser and Cutinha22 One study found that those with SCI and those with MCI have the same risk of progression to AD. Reference Jessen, Wolfsgruber and Wiese23 This information collectively supports the suggestion that SCI may be an early stage of AD occurring prior to clinical detection or diagnosis. However, other studies suggest that SCI is associated with psychological factors including health anxiety and dementia worry rather than objective cognitive impairment. Reference Kinzer and Suhr24 These studies suggest that dementia worry acts as a confounding factor in the relationship between objective memory impairment and SCI. Reference Kinzer and Suhr24 It may well be that there are two groups of patients with SCI: those who are noticing subtle changes presaging dementia and those who are not at risk for dementia, but are concerned about normal changes with age.
Biomarkers such as serum neurofilament light (NfL) have also been shown to be altered in patients with mutations prior to symptom onset. Reference Weston, Poole and Ryan25 One study demonstrated that patients with a mutation associated with familial AD, but without symptoms, have metabolic changes years before the expected onset of the disease. Reference Godbolt, Waldman and MacManus10 The proteinopathies that have the most evidence supporting their role in AD biomarker changes are tau and amyloid beta. Reference Jack, Knopman and Jagust26
The characteristics of the patient population outlined in Table 1 likely have more to do with the age range of the patients, with a mean age of 61, rather than the SCI. However, the family history of dementia in 51.2% of patients, and the predominant high school or higher educated characteristic (70.1%) may be relevant. A family history of dementia often means that patients are more aware of the signs of cognitive decline, as well as its effects. Reference Verity, Kirk and O’Connell1 Higher levels of education are also associated with the “worried well”, as their education often gives them insight into medical conditions, and they are often well-versed in utilizing critical thinking. Reference Verity, Kirk and O’Connell1
The data we collected showed no significant overall change in MMSE, CESD, or FAQ over 1 year in patients with SCI at initial assessment, and showed no consistent pattern in MMSE scores throughout long-term follow-up. It is important to note that follow-up did not occur in all patients. Ninety-one out of the 166 patients returned for follow-up assessments, making the follow-up rate 54.8%. Patients usually informed us that they did not wish to attend follow-up because they did not have progressing or concerning symptoms, and presumably stayed neurologically normal.
In a study comparing cognitively normal patients to patients with SCI over time, a significantly larger proportion of the SCI group developed a neurological diagnosis. Reference Reisberg, Shulman and Torossian27 This supports the theory that patients with SCI, or the “worried well”, are more at risk for neurological decline than the general population. We observed a subsequent neurological diagnosis in 9/166, or 5.4%, of the SCI patients analyzed. This suggests that, for the most part, those with SCI could be reassured at their initial visit, not following up or being referred further. Though this would be sufficient for most of the patients in this analysis, as the vast majority did not receive a neurological diagnosis throughout follow-up, early detection and treatment of disease would be important for those few who did receive a later diagnosis. The trouble, of course, is how to tell the difference between patients likely to remain stable and those likely to show deterioration.
As aforementioned, the pathology of AD is present long before a diagnosis can be made. Reference Godbolt, Waldman and MacManus10 The early detection and treatment of AD are associated with better outcomes for patients. Reference Chi, Yu and Tan8 With this being taken into account, it can be said that the optimal course of action for the “worried well” is to reassure and educate them on the low probability of their progression to a neurological diagnosis, but to nonetheless follow-up with them to ensure that they receive adequate and early care and intervention if they require it. MCI has been described as being a precursor to AD in some cases. Reference Reisberg and Gauthier28 It may, therefore, be worthwhile to increase the frequency of follow-ups following an MCI diagnosis, so as to ensure early detection of AD. Though our sample size of patients who developed AD was too small to be significant, the majority of them did have a previous diagnosis of MCI, supporting this idea. The only variable found to have a significant difference between the population of patients who remained cognitively normal and those who developed a neurological diagnosis is age, as seen in Table 1. Most of the patients who received a diagnosis were older than those who did not. Additional studies surrounding the variables summarized in Table 1 should also be conducted to investigate who is at greater risk with respect to potentially modifiable risk factors such as exercise per week, general health, education level, living situation, and degree of hearing loss. Previous studies suggest that moderate exercise, balanced diet, and educational activities may decrease the likelihood of one developing dementia, but further inspection is required. Reference Rockwood, Andrew and Aubertin-Leheudre29
Limitations
As previously discussed, not all of the patients attended follow-ups. We were usually informed that their lack of attendance was due to improvement or stability of their condition, but we cannot know for certain in all cases. Another feature of the cohort that should be noted is the average age of 60.9 (SD 12.0). This relatively young age for a study on cognitive decline and dementia, which are strongly associated with older age, does create limitations. The data in Tables 1 and 3 highlighting features of those who received a neurological diagnosis may also suggest some possible risks for developing a neurological diagnosis, but the sample size is too small to be significant. Follow-up studies exploring this with a larger patient population may be useful.
Additionally, the setting of data collection needs to be considered. In a study comparing SCI in the patient presenting to memory clinics and those in the general population, assessed through a population study, it was found that SCI is more likely to progress to MCI in the group that attended the memory clinic. Reference Snitz, Wang and Cloonan30 As we do not have a comparison in this study, we cannot know for certain that the attendance at the RRMC was not a factor in the presence of later diagnoses of MCI in the population we studied. Reference Chi, Yu and Tan8 Other techniques, such as neuroimaging to assess biomarkers or CSF analysis are also used for diagnosing AD. Reference Frisoni, Boccardi and Barkhof31,Reference Olsson, Lautner and Andreasson32 Amyloid-beta imaging techniques are also useful in early diagnosis of AD, and potential monitoring of AD progression. Reference Nordberg33 We did not have access to these modalities, which could be useful for prognostication, during the conduct of our study.
Conclusion
Overall, we did not observe a significant change in MMSE, FAQ, or CESD score at 1-year follow-up compared to clinic day, nor did we observe a consistent pattern of MMSE score decline over long-term follow-up. This supports the course of action of being optimistic regarding the “worried well”, as measurable neurological decline is unlikely. A small proportion of patients analyzed developed a diagnosis. The pathology of AD occurs before diagnosis, meaning that the disease process was likely underway in these patients at initial assessment, despite normal neurological test results. Reference Chi, Yu and Tan8 This, along with the fact that AD has better outcomes with early recognition, demonstrates the need for follow-up, despite the low probability of progression. Reference Godbolt, Waldman and MacManus10
Acknowledgments
The authors would like to acknowledge the Rural and Remote Memory Clinic at the University of Saskatchewan for the support of this research.
Funding
Support for this research was received from CIHR Foundation Grant #FDN-148444.
Conflicts of Interest
There are no conflicts of interest.
Statement of Authorship
All authors contributed in the preparation and editing of the final manuscript. MS wrote the manuscript under the guidance of AK. AK and MS developed the study design and oversaw the data collection. CK conducted the statistical analysis. DM was a member of the team that conceived the clinic model and managed data entry. MEO and AK oversaw the clinical assessments and performed the diagnosis. MEO administered and oversaw neuropsychological testing, depression testing, and function testing.
