Reciprocal associations between major depressive disorder and cognitive impairment as well as between depression or cognitive decline and impaired sleep have been widely recognized. Thus, in many cases, impaired sleep is the chief complaint of depression, and vice versa, several longitudinal studies have identified insomnia as an independent risk factor for the development of emerging or recurrent depression (Fang et al., Reference Fang, Tu, Sheng and Shao2019).
Likewise, a series of epidemiological studies indicate that multiple forms of sleep disturbances (particularly, decreased sleep duration, increased sleep fragmentation, and sleep-disordered breathing) are associated with increased risk of developing cognitive impairment and neurodegenerative diseases, such as Alzheimer’s and Parkinson’s diseases (Yaffe et al., Reference Yaffe, Falvey and Hoang2014). Notably, sleep impairment may even trigger disease onset at an earlier age (Mander, Reference Mander2020), while depressive symptoms may mediate at least some of the association between subjective sleep complaints and cognitive decline (Yaffe et al., Reference Yaffe, Falvey and Hoang2014), being a risk factor or marker for current poorer cognitive function and future decline (Almeida et al., Reference Almeida, Hankey, Yeap, Golledge and Flicker2017; Deckers et al., Reference Deckers2015; Opdebeeck et al., Reference Opdebeeck, Yates, Kudlicka and Martyr2019).
The associations between impaired mood, sleep, and cognition may be due to some shared mechanism (Byers and Yaffe, Reference Byers and Yaffe2011), including disrupted neurogenesis, synaptic plasticity, and neurotransmission (Yaffe et al., Reference Yaffe, Falvey and Hoang2014). Further, sleep deficiency contributes to increased levels of inflammatory cytokines (e.g., the interleukine-6, tumor necrosis factor) throughout the day. Neuroinflammation, in turn, is implicated in contributing to a variety of illnesses, including Alzheimer’s disease, Parkinson’s disease, and depression.
Thus far, most literature has focused on long-term associations between mood, sleep, and cognitive performance (in all possible combinations). Nevertheless, these three states may also be reciprocally associated with each other from one day to the next. Surprisingly, the short-term associations are rather underexplored, even though it is well known that depressive symptoms, cognitive performance, and sleep quality show high day-to-day variability (Meeusen, Reference Meeusen2014; Kim et al., Reference Kim, Nunes, Oliveira, Young and Lafer2016). The study by Zuidersma et al. (Reference Zuidersma, Lugtenburg, Van Zelst, Reesink, De Deyn, Strijkert and Oude Voshaar2021) in this issue of International Psychogeriatrics addresses this subject by exploring short-term temporal associations of depressive symptoms, cognitive performance, and sleep in eight older persons with both depression and cognitive impairment.
Whereas most previous studies on daily associations between depression, sleep, and cognitive performance focused on the group-averages analysis, Zuidersma et al. (Reference Zuidersma, Lugtenburg, Van Zelst, Reesink, De Deyn, Strijkert and Oude Voshaar2021) take into account individual differences. In the current work, the use of the single-subject design, an approach that analyzes each individual separately, was motivated by the fact that the group averages rarely can be applicable to individual persons, thus, limiting the value of the group-level analysis in clinical settings. Therefore, Zuidersma et al. (Reference Zuidersma, Lugtenburg, Van Zelst, Reesink, De Deyn, Strijkert and Oude Voshaar2021) evaluated sleep, depression severity, and working memory during 63 consecutive assessments for each participant.
Depression was defined according to Diagnostic Statistical Manual (DSM) criteria or a Geriatric Depression Scale (score > 4). Cognitive impairment was defined as a diagnosis of dementia (a clinical dementia rating scale < 2) or mild cognitive impairment according to the team of the Memory Clinic (a Mini-Mental State Examination score < 25, or a Montreal Cognitive Assessment score < 26). Eight participants were included in the final analyses, however it is not reported how many had depression and how many had MCI or dementia. Sleep was assessed with the total sleep time as recorded with actigraphy, cognition was evaluated based on working memory using the One Back Task, and depression severity was assessed with the Patient Health Questionnaire.
Even though the authors did not find any consistent relationships between the variables of interest, the study provides an important insight into the nature and direction of the temporal associations between depressive symptoms, cognitive performance, and sleep, demonstrating how they vary within individuals. Importantly, the findings relate to real-world sleep and mood disturbances, and, therefore, can be valuable for the development of personalized intervention strategies. Due to the bidirectional relationships between sleep and psychiatric or cognitive conditions (Fang et al., Reference Fang, Tu, Sheng and Shao2019; Yaffe et al., Reference Yaffe, Falvey and Hoang2014), which also have been shown in the current study, sleep interventions may have the potential to improve depressive symptomatology as well as cognitive performance. In addition, individual-level sleep data in a free-living environment may substantially advance the fields of wearable sensors and machine learning algorithms for sleep stage scoring.
The authors list some limitations to their study, including the use of the single-subject design, which does not allow one to generalize the findings, whereas generalizability is one of the main aims of science. On the other hand, the single-subject design may lead to the development of daily intervention strategies to maintain health, quality of life, functional outcomes, and independence (Zuidersma et al., Reference Zuidersma, Lugtenburg, Van Zelst, Reesink, De Deyn, Strijkert and Oude Voshaar2021). This is the strength of the current study, and it should be applauded because the single-subject design is a crucial method and certainly should be recognized and encouraged to be used when appropriate. The present study clearly shows what single-subject design can do and what it cannot do, suggesting that this approach should be considered by more researchers.
Furthermore, the current report gives an impression that combining the single-subject and group-level approaches would be a promising future direction, which will allow one to generalize the findings. For example, the single-subject design can be implemented prior to performing a randomized controlled trial to get a better handle on the magnitude and direction of effects. In addition, the repeated measures of the data collected with the single-subject design automatically allow one to assess several other important metrics that predict many health-related outcomes but are, nevertheless, rarely collected (Costa et al., Reference Costa, Dogan, Schulz and Reetz2019; Fischer et al., Reference Fischer, Klerman and Phillips2021). One such metric is sleep regularity, as its impairment has been associated with lower academic performance, impaired daytime function, poorer mood/well-being, depression, and insomnia (Watson et al., Reference Watson, McGlashan, Hosken, Anderson, Phillips and Cain2020; Fischer et al., Reference Fischer, Klerman and Phillips2021; Pye et al., Reference Pye, Phillips, Cain, Montazerolghaem, Mowszowski, Duffy, Hickie and Naismith2021). Additional examples include intra-individual variability, inter-daily stability, and social jet lag that assess deviations in sleep patterns from an individual’s average, or composite phase deviation, and sleep regularity index, which evaluate variability between consecutive days (Fischer et al., Reference Fischer, Klerman and Phillips2021). The idea is that the metrics collected using the single-subject design can be later analyzed on the group level, allowing one to generalize the findings. Adding to this, collecting the metrics reflecting day-to-day variability is important since these variables can be entered as confounding factors to be controlled for to a model that studies other parameters of interest, such as cognitive performance and depression severity, on a group level.
The current study shows that the single-subject design is a valuable method that should be extended to several lines of inquiry. For example, it can be used in the areas where it is hard to get the number of participants needed for the gold standard randomized controlled trial, such as the rare subtypes of Alzheimer’s or Parkinson’s diseases associated with genetic mutations.
In addition, the current study gives an impression that the single-subject design may be a valuable tool to study fluctuations in attention and alertness, seen in such neurodegenerative diseases as Parkinson’s disease dementia, and dementia with Lewy bodies (McKeith et al., Reference McKeith2017). Fluctuation in attention and alertness/sleepiness is a prominent symptom of these diseases, which are, however, less explored compared to other symptoms (e.g., cognitive, motor, and sleep impairments) due to methodological difficulties. The single-subject design may address this issue and further allow researchers and clinicians to create databases of within-individual fluctuations as well as other variables of interest (e.g., fluctuations in sleep quality, depression severity, and cognitive performance). This is of particular importance since one of the first steps in developing any treatment is understanding what an individual does.
While the current study by Zuidersma and colleagues highlights the importance of considering day-to-day relationships between mood, cognition, and sleep, it could not provide any information on the mechanism of the reported associations. Therefore, other future considerations for the present work may include the use of objective measures or interventions, for example, polysomnography, more comprehensive assessments of cognition, auditory closed-loop stimulation, and pharmacological interventions.
In summary, the current study by Zuidersma et al. (Reference Zuidersma, Lugtenburg, Van Zelst, Reesink, De Deyn, Strijkert and Oude Voshaar2021) is a valuable addition to the research, which clearly shows that the presence, nature, and direction of the associations between sleep, depressive symptoms, and cognitive performance differ from individual to individual. The study further provides an insight into the day-to-day variations in these associations. The authors suggest that the promotion of the single-subject design in the elderly with cognitive impairments and depressive symptoms may lead to the development of a personalized treatment and intervention strategy. Larger investigations of simultaneous short- and long-term associations between sleep, depression, and cognition patterns using a wider range of objectively assessed variables, which also consider racial and ethnic minority as well as the genetic profile of aging populations and patients with neurodegenerative diseases may shed greater light on this issue.
You have
Access
Reciprocal associations between major depressive disorder and cognitive impairment as well as between depression or cognitive decline and impaired sleep have been widely recognized. Thus, in many cases, impaired sleep is the chief complaint of depression, and vice versa, several longitudinal studies have identified insomnia as an independent risk factor for the development of emerging or recurrent depression (Fang et al., Reference Fang, Tu, Sheng and Shao2019).
Likewise, a series of epidemiological studies indicate that multiple forms of sleep disturbances (particularly, decreased sleep duration, increased sleep fragmentation, and sleep-disordered breathing) are associated with increased risk of developing cognitive impairment and neurodegenerative diseases, such as Alzheimer’s and Parkinson’s diseases (Yaffe et al., Reference Yaffe, Falvey and Hoang2014). Notably, sleep impairment may even trigger disease onset at an earlier age (Mander, Reference Mander2020), while depressive symptoms may mediate at least some of the association between subjective sleep complaints and cognitive decline (Yaffe et al., Reference Yaffe, Falvey and Hoang2014), being a risk factor or marker for current poorer cognitive function and future decline (Almeida et al., Reference Almeida, Hankey, Yeap, Golledge and Flicker2017; Deckers et al., Reference Deckers2015; Opdebeeck et al., Reference Opdebeeck, Yates, Kudlicka and Martyr2019).
The associations between impaired mood, sleep, and cognition may be due to some shared mechanism (Byers and Yaffe, Reference Byers and Yaffe2011), including disrupted neurogenesis, synaptic plasticity, and neurotransmission (Yaffe et al., Reference Yaffe, Falvey and Hoang2014). Further, sleep deficiency contributes to increased levels of inflammatory cytokines (e.g., the interleukine-6, tumor necrosis factor) throughout the day. Neuroinflammation, in turn, is implicated in contributing to a variety of illnesses, including Alzheimer’s disease, Parkinson’s disease, and depression.
Thus far, most literature has focused on long-term associations between mood, sleep, and cognitive performance (in all possible combinations). Nevertheless, these three states may also be reciprocally associated with each other from one day to the next. Surprisingly, the short-term associations are rather underexplored, even though it is well known that depressive symptoms, cognitive performance, and sleep quality show high day-to-day variability (Meeusen, Reference Meeusen2014; Kim et al., Reference Kim, Nunes, Oliveira, Young and Lafer2016). The study by Zuidersma et al. (Reference Zuidersma, Lugtenburg, Van Zelst, Reesink, De Deyn, Strijkert and Oude Voshaar2021) in this issue of International Psychogeriatrics addresses this subject by exploring short-term temporal associations of depressive symptoms, cognitive performance, and sleep in eight older persons with both depression and cognitive impairment.
Whereas most previous studies on daily associations between depression, sleep, and cognitive performance focused on the group-averages analysis, Zuidersma et al. (Reference Zuidersma, Lugtenburg, Van Zelst, Reesink, De Deyn, Strijkert and Oude Voshaar2021) take into account individual differences. In the current work, the use of the single-subject design, an approach that analyzes each individual separately, was motivated by the fact that the group averages rarely can be applicable to individual persons, thus, limiting the value of the group-level analysis in clinical settings. Therefore, Zuidersma et al. (Reference Zuidersma, Lugtenburg, Van Zelst, Reesink, De Deyn, Strijkert and Oude Voshaar2021) evaluated sleep, depression severity, and working memory during 63 consecutive assessments for each participant.
Depression was defined according to Diagnostic Statistical Manual (DSM) criteria or a Geriatric Depression Scale (score > 4). Cognitive impairment was defined as a diagnosis of dementia (a clinical dementia rating scale < 2) or mild cognitive impairment according to the team of the Memory Clinic (a Mini-Mental State Examination score < 25, or a Montreal Cognitive Assessment score < 26). Eight participants were included in the final analyses, however it is not reported how many had depression and how many had MCI or dementia. Sleep was assessed with the total sleep time as recorded with actigraphy, cognition was evaluated based on working memory using the One Back Task, and depression severity was assessed with the Patient Health Questionnaire.
Even though the authors did not find any consistent relationships between the variables of interest, the study provides an important insight into the nature and direction of the temporal associations between depressive symptoms, cognitive performance, and sleep, demonstrating how they vary within individuals. Importantly, the findings relate to real-world sleep and mood disturbances, and, therefore, can be valuable for the development of personalized intervention strategies. Due to the bidirectional relationships between sleep and psychiatric or cognitive conditions (Fang et al., Reference Fang, Tu, Sheng and Shao2019; Yaffe et al., Reference Yaffe, Falvey and Hoang2014), which also have been shown in the current study, sleep interventions may have the potential to improve depressive symptomatology as well as cognitive performance. In addition, individual-level sleep data in a free-living environment may substantially advance the fields of wearable sensors and machine learning algorithms for sleep stage scoring.
The authors list some limitations to their study, including the use of the single-subject design, which does not allow one to generalize the findings, whereas generalizability is one of the main aims of science. On the other hand, the single-subject design may lead to the development of daily intervention strategies to maintain health, quality of life, functional outcomes, and independence (Zuidersma et al., Reference Zuidersma, Lugtenburg, Van Zelst, Reesink, De Deyn, Strijkert and Oude Voshaar2021). This is the strength of the current study, and it should be applauded because the single-subject design is a crucial method and certainly should be recognized and encouraged to be used when appropriate. The present study clearly shows what single-subject design can do and what it cannot do, suggesting that this approach should be considered by more researchers.
Furthermore, the current report gives an impression that combining the single-subject and group-level approaches would be a promising future direction, which will allow one to generalize the findings. For example, the single-subject design can be implemented prior to performing a randomized controlled trial to get a better handle on the magnitude and direction of effects. In addition, the repeated measures of the data collected with the single-subject design automatically allow one to assess several other important metrics that predict many health-related outcomes but are, nevertheless, rarely collected (Costa et al., Reference Costa, Dogan, Schulz and Reetz2019; Fischer et al., Reference Fischer, Klerman and Phillips2021). One such metric is sleep regularity, as its impairment has been associated with lower academic performance, impaired daytime function, poorer mood/well-being, depression, and insomnia (Watson et al., Reference Watson, McGlashan, Hosken, Anderson, Phillips and Cain2020; Fischer et al., Reference Fischer, Klerman and Phillips2021; Pye et al., Reference Pye, Phillips, Cain, Montazerolghaem, Mowszowski, Duffy, Hickie and Naismith2021). Additional examples include intra-individual variability, inter-daily stability, and social jet lag that assess deviations in sleep patterns from an individual’s average, or composite phase deviation, and sleep regularity index, which evaluate variability between consecutive days (Fischer et al., Reference Fischer, Klerman and Phillips2021). The idea is that the metrics collected using the single-subject design can be later analyzed on the group level, allowing one to generalize the findings. Adding to this, collecting the metrics reflecting day-to-day variability is important since these variables can be entered as confounding factors to be controlled for to a model that studies other parameters of interest, such as cognitive performance and depression severity, on a group level.
The current study shows that the single-subject design is a valuable method that should be extended to several lines of inquiry. For example, it can be used in the areas where it is hard to get the number of participants needed for the gold standard randomized controlled trial, such as the rare subtypes of Alzheimer’s or Parkinson’s diseases associated with genetic mutations.
In addition, the current study gives an impression that the single-subject design may be a valuable tool to study fluctuations in attention and alertness, seen in such neurodegenerative diseases as Parkinson’s disease dementia, and dementia with Lewy bodies (McKeith et al., Reference McKeith2017). Fluctuation in attention and alertness/sleepiness is a prominent symptom of these diseases, which are, however, less explored compared to other symptoms (e.g., cognitive, motor, and sleep impairments) due to methodological difficulties. The single-subject design may address this issue and further allow researchers and clinicians to create databases of within-individual fluctuations as well as other variables of interest (e.g., fluctuations in sleep quality, depression severity, and cognitive performance). This is of particular importance since one of the first steps in developing any treatment is understanding what an individual does.
While the current study by Zuidersma and colleagues highlights the importance of considering day-to-day relationships between mood, cognition, and sleep, it could not provide any information on the mechanism of the reported associations. Therefore, other future considerations for the present work may include the use of objective measures or interventions, for example, polysomnography, more comprehensive assessments of cognition, auditory closed-loop stimulation, and pharmacological interventions.
In summary, the current study by Zuidersma et al. (Reference Zuidersma, Lugtenburg, Van Zelst, Reesink, De Deyn, Strijkert and Oude Voshaar2021) is a valuable addition to the research, which clearly shows that the presence, nature, and direction of the associations between sleep, depressive symptoms, and cognitive performance differ from individual to individual. The study further provides an insight into the day-to-day variations in these associations. The authors suggest that the promotion of the single-subject design in the elderly with cognitive impairments and depressive symptoms may lead to the development of a personalized treatment and intervention strategy. Larger investigations of simultaneous short- and long-term associations between sleep, depression, and cognition patterns using a wider range of objectively assessed variables, which also consider racial and ethnic minority as well as the genetic profile of aging populations and patients with neurodegenerative diseases may shed greater light on this issue.