Issues about the use of subjective cognitive decline in Alzheimer's disease research

Recently, Jessen et al. 1 have described the conceptual framework for “subjective cognitive decline” (SCD) in Alzheimer's disease (AD) research. SCD has been indicated as a possible risk factor for incident cognitive impairment and dementia 2, and pathophysiological commonalities have been reported between subjects with SCD and AD 3. The hypothesis that SCD might represent a preliminary stage of the dementia cascade has already brought to the development of specific preventive trials targeting individuals with subjective cognitive complaints 4. Nevertheless, to date, the study of SCD is hampered by the lack of a shared and agreed terminology. Thus, SCD has so far been differently defined, screened, and measured 5. By proposing a new lexicon for SCD, Jessen et al. 1 address an important gap in the field. The authors proposed the core criteria for defining SCD and recommendations for its adoption as a preclinical model of AD in research. The article indeed poses itself as a possible cornerstone for future research in the field. However, we would like to point out some issues that we believe should have been better considered and discussed by the authors. The course of SCD over time is quite unstable (and often unpredictable), so that it can progress toward overt cognitive impairment and revert to a subjective experience of normal mental efficiency 6, 7. Interestingly, despite being commonly described, the reversion to normal cognition has also been frequently neglected in other predementia conditions, including mild cognitive impairment (MCI) 8. Moreover, because all the assessment tools we use in clinics and research for objectively defining cognitive abnormalities present methodological and accuracy limitations, the risk of misclassification across these conditions is present and enhanced by the limited knowledge about them. The possible inversion of the SCD trajectory can be easily explained when the condition is sustained/influenced by temporary reversible clinical conditions (e.g., pain, depressive symptoms, fatigue). To operationalize the novel construct, Jessen et al. proposed to exclude any “psychiatric or neurologic disease, medical disorder, medication, or substance use” when detecting SCD (or, at least, carefully considering them when reporting the results). Nevertheless, it might be argued that disentangling the relationship between SCD and such potential confounders is likely unfeasible, especially among elders (frequently characterized by polypharmacy and multiple interacting comorbidities). In other words, how many elders with subjective cognitive complaints do not simultaneously present a psychiatric or neurologic disease, a medical disorder, or assume a medication potentially explaining the SCD condition? On the other hand, if we accept certain heterogeneity in the definition of the target population, the exploration of underlying pathophysiological mechanisms of AD will become complicated (although the SCD condition may get closer to the clinical implementation). Although the authors clearly proposed the SCD as a preclinical and research condition, it should also not be underestimated how many times other research conditions have almost automatically acquired clinical value. For example, the definition of MCI was similarly motivated at the beginning by the need of anticipating the detection of pathophysiological mechanisms responsible for dementia. Nevertheless, MCI is today widely adopted as a specific nosographic entity in clinics 9. In a scenario where the clinical relevance of the SCD is ambiguous, the potential negative consequences of “labeling” an individual as affected by it (even if only in the research practice) should not be minimized. The health care professionals should be spared by the risk of the SCD condition as the next subject for potential “overdiagnosis” 10. Persons identified as presenting SCD may develop unnecessary levels of anxiety or stress, reactions that may induce requests for unexpected/unjustified examinations and care. Moreover, the risk exists that the generation of a new category of “patients” may lead to forms of discrimination or stigmatization. The foreseen research targeting SCD must be balanced against the potential harms associated with its use and misuse. This is important if SCD will be considered as a biomarker and (even more) if it will turn out to become a diagnostic tool (for instance in the suggested form of “SCD plus”). In this respect, we draw attention to the fact that a net distinction between risk factors, biomarkers, and diseases is not always easy to establish and maintain 10. In conclusion, the proposal of SCD as a new paradigm for preclinical AD cannot take shape independently of such crucial considerations. The adoption of a new condition in research should always be accompanied by the definition of the borders for its use. Such perimeter is dictated by the clinical relevance of the condition, a concept implying conceptual, ethical, and societal evaluations.