Subjective complaints in mild cognitive impairment make a difference

Abstract

Mild cognitive impairment (MCI) is an intermediate stage between normal cognitive ageing and dementia (1). Understanding the incidence of MCI has important implications for public health planning, development of interventions, and prevention of the condition. Several studies have reported the incidence of MCI primarily in population-based settings (2-5), but few have reported estimates of incidence in the clinical setting (6). In this issue of Acta Psychiatrica Scandinavica, Luck et al. (7) report estimates of MCI incidence in a clinical setting. Population-based studies have reported estimates that range from 21.5 to 76.8 per 1000 person-years (8, 9). The reported MCI incidence of 56.5 per 1000 person-years by Luck et al. is within this range, but may be lower than would be expected for the clinical setting (6), possibly because estimates were derived from family practice rather than from memory clinics where a more severe spectrum of cognitive impairment would be present. The paper by Luck et al. lends support for the importance of the subjective memory complaint as an essential element in the criteria for MCI. In their paper, subjective memory complaints predicted overall MCI. While some investigators do not consider subjective memory complaints to be important (10), this measure may enhance diagnosis in subjects with mild MCI or facilitate detection of those at increased risk of progression from normal cognition to MCI. The findings, therefore, suggest that subjective memory complaints should not be taken lightly by providers of health care for the elderly since it may indeed herald imminent MCI. Incidence studies are important for several reasons. They are a reliable estimate of disease frequency and, if the average duration of disease is relatively stable, may provide more accurate estimates of the prevalence of disease in a community. This is particularly relevant for MCI; since mild cases may revert to normal cognition, the estimates of prevalence may vary in the same population and across studies. Incidence studies identify causal risk factors for disease, provide insights into the pathophysiology of disease, and inform on the design of interventions and strategies to reduce disease risk and promote health. The study by Luck et al. identified older age, vascular conditions, and the ApoE ɛ4 allele as risk factors for MCI. It also suggested etiological differences in MCI subtypes. ApoE ɛ4 allele and older age predicted amnestic MCI, suggesting an underlying neurodegenerative aetiology that may be associated with an increased likelihood of progression to Alzheimer’s disease. In contrast, vascular disease, impairment in vision, and subjective memory complaints were predictive of non-amnestic MCI, and ApoE ɛ4 had a lesser impact. Thus, non-amnestic MCI may have a more heterogeneous aetiology, and may be more likely to progress to non-Alzheimer’s disease dementias such as vascular dementia. The study by Luck et al. reported a higher incidence of MCI in men than in women. Other studies have shown a higher prevalence of MCI in men than in women (3, 11-13). Thus, the study by Luck et al. suggests that the sex difference in MCI prevalence observed in cross-sectional studies may be valid; however, this remains to be confirmed in prospective population-based studies of longer duration. Despite efforts to standardize the criteria for the diagnosis of MCI, this issue remains problematic. Some studies have used a purely algorithmic approach for a diagnosis of MCI that is based on neuropsychological testing. Others use a combination that includes information from a clinical evaluation and psychometric testing. Luck et al. used the Structured Interview for the Diagnosis of Dementia of the Alzheimer Type, Multi-Infarct Dementia, and Dementias of Other Aetiology cognitive test battery to assess objective cognitive impairment (14). Although this tool is useful for assessing dementia, 30 of the 55 questions are the Mini-Mental State Exam questions; thus, the battery may have lacked sensitivity for MCI. Various cutpoints have also been used to assess impairment in cognitive domains. Use of a cutpoint of 1 standard deviation below the mean of age- and education-specific norms to determine impairment in specific cognitive domains could have resulted in a lower threshold for impairment. Limitations of the study include the relatively short duration of the study (3 years), which raises questions about the precision of the estimates of MCI incidence. Due to greater attrition in the early years of a study and the tendency of some mild cases of MCI to revert to normal, estimates of incidence are more stable over a longer duration of follow-up. The recruitment of participants from the clinical setting raises questions about the potential for selection bias among subjects seeking care; thus, the generalizability of the study findings may be limited to studies conducted in a similar setting. Subjective memory complaints predicted overall and non-amnestic MCI, but not amnestic MCI as would have been presumed. This suggests that subjective memory complaints may encompass or herald impairment in certain non-memory cognitive domains. Dr Rosebud O. Roberts is supported by the National Institutes of Health (NIH) under project number: U01 AG06786, and by the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer’s Disease Research Program. Dr Ronald C. Petersen is supported by the National Institutes of Health (NIH) under project numbers: P50 AG016574, AG006786, R01 AG011378, and U01 AG024904.