Since there are no guidelines for PD-SCD, researchers have classified PD-SCD differently, and some studies have classified PD-SCD as coexisting with objective impairment. For example, among PD patients in the Dujardin cross-sectional study, 32.22% matched a classification of PD-SCD, of these patients with SCD, 44.83% of PD-SCD met the dementia criteria, of these patients without SCD, just 25.41% met the dementia criteria, suggesting that PD-SCD has a higher probability of manifesting cognitive dysfunction than PD without SCD (Dujardin et al., 2010). Erro reported that approximately 25% of patients with PD underwent SCD, and PD-SCD at baseline may be a risk factor for PD-MCI (Erro et al., 2014). Lehrner studied PD-SCD with FAI and found that 31% of the PD patients reported SCD. Pan showed 30.3% and 12.1% of SCD in PD-MCI, PD-NC, respectively, he found SCD in PD-NC exhibited declined attention/working memory and thought there might be different pathogenicity in SCD with different cognitive status (Pan et al., 2021). Xiao classified patients into early-onset PD (EOPD) and late-onset PD (LOPD), there were 18.66% of EOPD and 24.74% of LOPD reported SCD (Xiao et al., 2021). Siciliano focused on PD with fatigue, and he found the prevalence of SCD was higher in fatigued PD patients when comparing with those with no fatigued (35 vs. 9%) (Siciliano et al., 2020). The PD-SCD evaluated by other studies demonstrated proportions of 32.6% (Baschi et al., 2018), 44.74% (Hogue et al., 2018), 85% (Barbosa et al., 2019), 29.7% (Lee et al., 2020), 23% (Jones et al., 2021), 32.3% (Siciliano et al., 2021), 28.6%, and 30.2% (Galtier et al., 2019, 2021).

Some researchers defined SCD as subjective cognitive impairment without objective cognitive decline. Purri described a 53% incidence of PD-SCD among PD patients with normal cognition and thought PD-SCD may be an indicator of subsequent cognitive impairment (Purri et al., 2020). Hong showed a proportion of 54.3% and found that PD-SCD can predict cognitive decline later (Hong et al., 2014a). Yang found the prevalence of PD-SCD was 28.1% (Yang et al., 2022). Galtier exhibited 30.5% were diagnosed with PD-SCD according to subjective cognitive decline semi-structured interview, and the longitudinal study result showed poor performance in verb naming test in PD-SCD, which suggested the role of linguistic impairment in PD-SCD (Galtier et al., 2022). Other studies demonstrated rates of 16.3% (Lehrner et al., 2014), 22.36% (Hogue et al., 2018), and 27.2% (Baschi et al., 2018).

Above all, there were some variations in the prevalence of PD-SCD among those studies, and the following reasons may be responsible for this. First, different demographic characteristics may contribute to the discrepancy; for example, an older cohort is prone to SCD since cognition declines with age. Second, the tool chosen to assess SCD may also be a factor, and a complete questionnaire may be more accurate than a single item. Third, the definition of SCD used in studies also influenced the results greatly, since PD may have normal memory but impaired planning ability, leading to some people being missed if the focus is only on memory. Finally, a state of anxiety, depression, and apathy in the PD group also accounted for the higher existence of SCD.

Although studies on neuroimaging in PD-SCD have been scarce, the existing results are encouraging. In a study of FDG-PET in PD with normal cognition (n = 18 as PD-SCD, n = 12 as PD control), results revealed reduced FDG metabolism in the middle frontal, middle temporal, occipital areas, and angular gyrus of the cortex. These regions may be neural correlates of PD-SCD, and could demonstrate an early pathological change in PD-SCD (Ophey et al., 2022), consistent with the finding of hypometabolism in the middle frontal gyrus and inferior parietal lobule in PD-MCI (Huang et al., 2008). Yoo classified PD patients had both CCI score >3 and intact cognition as underestimation of cognitive function, and the group here means to PD-SCD. MRI suggested a close association between the frontal lobes and PD-SCD and thought it important to measure frontal structural connectivity in the early stages of PD (Yoo et al., 2021). Song and colleagues found reduced perfusion in the frontal and inferior temporal cortical regions as well as the anterior cingulate gyrus and thalamus in PD-SCD patients compared to PD patients without SCD, which may provide a potential biomarker of predementia. Other studies by Hong found a decreased gray matter density and cortical thinning in the anterior cingulate gyrus, right inferior parietal lobule and parahippocampal cortices in PD-SCD (Hong et al., 2012, 2014b), suggesting an aberrant PD-related pathology.

The frontal lobe is responsible for executive functions, the temporal area is associated with semantic memory, the anterior cingulate gyrus seems to be relevant to verbal fluency, attention is dominated by the parietal lobe as well as the anterior cingulate gyrus, and the occipital lobe plays a significant role in visuospatial ability (Mohanty et al., 2007; Hong et al., 2012, 2014b; Song et al., 2014; Ophey et al., 2022). The neuroimaging studies above revealed potential neural correlates that underlie PD-SCD, making PD-SCD a promising group to be emphasized in clinician. However, samples from neuroimaging studies to date are relatively small, and larger sample sizes and longitudinal studies are needed for further validation.

Because SCD was mainly defined through several subjective questions, the mood features in patients with PD should be taken into account, as we mentioned before. Currently, the relationship between PD-SCD and emotional symptoms was debated, as some studies have described a marked correlation between anxiety, depression or apathy and PD-SCD, while other studies have not shown this. For example, Siciliano et al. (2021) defined underestimators as patients with subjective cognitive complaints but no objective cognitive impairment, they regarded PD-SCD as an underestimator actually and found a positive correlation between underestimator scores and fatigue, depression, and anxiety, suggesting a greater focus on mood features in patients with early PD in the clinical setting. In Yang et al.’s (2022) study, they defined PD-SCD according to NMSs-5 ≥ 1, subsequently, they found PD-SCD was associated with worse Hamilton Depression Scale (HAMD) and Hamilton Anxiety Scale (HAMA) scores. Rosenblum et al. (2022a) regarded PD patients with sMCD as PD-SCD, and he found PD-sMCD showed higher depression when compared to those with no suspected. Chua also obtained the association between PD-SCD and emotional factors (depression, anxiety, apathy) (Chua et al., 2021), and other studies also showed association between PD-SCD and emotional factors (Dujardin et al., 2010; Lehrner et al., 2014; Santangelo et al., 2014; Koster et al., 2015; Castro et al., 2016; Baschi et al., 2018; Barbosa et al., 2019; Purri et al., 2020; Han et al., 2021). At the same time, there still remains unclear between PD-SCD and emotions, since some studies showed no relation between these factors (Hong et al., 2012; Dupouy et al., 2018). According to Baschi, emotional disorders may be both a cause and a consequence of PD-SCD; for example, anxiety in individuals with PD-SCD may be caused by their awareness of cognitive function loss (Baschi et al., 2018). This data reinforced the point that when we encounter PD patients with emotional disorders, it is necessary to incorporate this information to further assess objective cognitive functions. Of course, the discrepancy of the results may be related to the methodology, as age, education, duration and severity of PD influenced moods and should be adjusted.

Studies have chosen complete neuropsychological tests for patients with PD, such as the MoCA/Mini Mental State Examination (MMSE) for global cognition, the semantic fluency/trail making test (TMT) for executive function, the word list verbal learning test for memory, the digit span test (DST)/symbol digits modalities test (SDMT) for attention, the naming test for language, and the Rey-Osterrieth complex figure test (ROCFT) for visuospatial ability. Table 2 listed detailed information about neuropsychological tests.

Lehrner was interested in the relationship between cognitive complaints and cognitive impairment, he found significant correlations between PD-SCD and worse MMSE performance (Lehrner et al., 2014). Hong et al. (2014b) focused PD-SCD on memory impairment, he investigated the cognitive performance and cortical thickness (n = 49 PD-SCD, n = 23 controls), results showed PD-SCD have poorer performance in visual memory and executive functions, which was consistent with cortical thinning in frontal and posterior cortical areas, since frontal and posterior regions responsible for executive function and visual function, respectively. In Song et al.’s (2014) study, PD-SCD group (n = 30) differed PD-without SCD (n = 47) in global cognition measured by MMSE and suggested PD-SCD may be a predictive biomarker on predementia. Nakhla et al. (2021) took informant-based responses into account and measured them by IQCODE and ultimately found that higher scale scores were negatively correlated with objective performance, including attention, executive function, memory and global cognition. Hong et al. (2018) assessed PD-SCD by CCI and suggested that an increasing score was strongly correlated with poorer objective cognitive functions (global cognition and all five cognitive domains) after adjusting for a depressive score. Baschi found a lower MoCA score in PD-SCD patients than in PD patients without SCD (Baschi et al., 2018). Regression models analyzed by Mills et al. (2016) suggested lower scores on memory, executive and visuospatial functions in PD-SCD. Koster et al. (2015) showed that patients with PD have higher proportion of self-reported difficulties in attention and executive functions but not memory. Dujardin reported more objective cognitive dysfunction among PD patients with cognitive complaints than among PD patients without SCD (Dujardin et al., 2010). In a word, studies referred above showed a relationship between SCD and decreased cognitive manifestation, while the remaining reported mixed results (Castro et al., 2016; Copeland et al., 2016; Dupouy et al., 2018). Castro obtained a contradictory conclusion with the above results. In this study, PD with cognitive complaints performed at a higher level, suggesting better cognitive status (Castro et al., 2016). Additionally, Dupouy and Copeland found no association between neither the patient’s nor the caregiver’s complaints and the patient’s objective cognitive manifestation (Copeland et al., 2016; Dupouy et al., 2018).

Attention and executive functions, which have neural correlates with the parietal and frontal lobes, were the most reported to have a strong correlation with the existence of PD-SCD, and these findings were consistent with the pathological changes in the early stage of PD. Some studies also found that PD-SCD was related to memory, and we think the results were affected by the assessed tools used as well as the targeted population. On the one hand, studies defined SCD as self-reported memory complaints, and the association investigated may be memory rather than other domains. On the other hand, studies comparing PD-SCD and PD-MCI concurrently may increase the correlation with one another. Dupouy and Copeland found no association between SCD and objective cognitive performance, which may be due to the assessment used or the psychiatric symptoms among their cohort. Overall, the association between SCD and objective cognitive performance remains unclear now. SCD are subjective and may be influenced by countless factors and the conflicting results obtained can be associated with measurements of cognitive status, definition of SCD, sample size, study design and so on. Thus, it is difficult to show a clear relationship between the two, especially when the methods applied are not precise or extensive enough to evaluate SCD or objective cognitive functions.

In Jones et al.’s (2021) study (n = 483 PD patients), a single item was used to assess PD-SCD, and the results revealed no relation between PD-SCD at baseline and PD-MCI/PDD 5 years later. A longitudinal study (n = 139 non-demented PD patients) adopted 4 methods to elicit PD-SCD, it explored the link between PD-SCD and cognitive performance subsequently, results here showed there was no correlation between PD-SCD and cognitive decline and no predictive value of PD-SCD over time (1–2 years) (AlDakheel et al., 2019). Conversely, a follow-up study by Galtier suggested that PD-SCD patients (36.4%) are at higher risk of converting to dementia than PD without SCD (14.3%) 7.5 years later, and the PD-SCD group showed impairments in visuospatial and visuoperceptual functions, which was linked to thinning of posterior cortices (Galtier et al., 2021), at the same time, PD-SCD showed poor performance in verb naming test in the same longitudinal study (n = 46 PD, n = 20 controls), which suggested the possible linguistic dysfunction in PD-SCD (Galtier et al., 2022). Rosenblum followed up PD-sMCD (n = 25) and found self-reported cognitive decline may be a marker for predicting cognitive ability decline after 1 year (Rosenblum et al., 2022b). Han and colleagues regarded PD-SCD as an intermediate status of PD-NC and PD-MCI, they divided people into three groups: PD-NC (n = 189), PD-SCD (n = 59), PD-MCI (n = 135), the longitudinal study (1–7.5 years) found PD-SCD showed memory impairment and greater reductions in attention and executive function compared with PD-NC, which suggested PD-SCD may be high risk individuals progression to cognitive decline later (Han et al., 2021), and the results were similar to Hong’s and Purri’s study (Hong et al., 2014a; Purri et al., 2020). In Purri’s study (n = 153 PD patients with normal cognition), they defined PS-SCD as subjective cognitive complaints with objective cognitive normality and found that the PD-SCD group performed worse on global cognition, executive function and processing speed than the PD without SCD group. Additionally, PD-SCD more easily converted cognitive impairment longitudinally, indicating that PD-SCD may be a warning of subtle cognitive decline (Purri et al., 2020). A larger longitudinal study (n = 336 patients with PD) led by Mills et al. (2020) showed that PD-SCD patients were more likely to develop PD-MCI over 3 years of follow-up and can be used to predict future cognitive impairment, which was similar to other studies (Hong et al., 2014a; Galtier et al., 2019). In Hong et al.’s (2014a) study, the findings were consistent with the conclusion that PD-SCD may be a risk factor for future cognitive dysfunction, since their results showed that PD-SCD could predict progression to decreased cognition 1–4 years later. Erro et al. (2014) also focused on the power of PD-SCD to predict further cognitive decline, he found patients with subjective memory decline at baseline (n = 76) were independent predictor, which can predict the progression to MCI with 2 years longitudinal study, Hogue et al.’s (2018) conclusion was similar to Erro, he created a regression model included SCD in baseline, and the model can predict future cognitive decline, which showed encouraging predictive value of PD-SCD.

In summary, the majority of the above studies considered PD-SCD to be risk factor for impaired objective cognition, and PD-SCD may be a potential way to identify early cognitive decline. However, some problems should be noted. Most studies evaluated PD-SCD with a single item partly concerned with memory, which could eventually influence the results, thus, assessing PD-SCD according to more comprehensive cognitive domains may contribute to the consistency of subjects selected. Secondly, emotional factors should be emphasized and adjusted when exploring the correlation between cognitive performance and PD-SCD, because emotional factors influenced cognitive function and may be linked with PD-SCD, which interfered the results. Thirdly, it may be potential way to associate longitudinal study with more neuroimaging researches in PD-SCD.