A seed and soil model of loneliness in Alzheimer’s disease

Loneliness is classically defined as a result of perceiving a discrepancy between the desired quantity and quality of one's social life and actual social relationships (Perlman and Peplau, 1984). Lifespan research has indicated older adults are among the highest risk for experiencing loneliness because their social network size decreases more frequently than that of younger adults. In part, this is due to the growing risk of losing one's partner and peers through death, the loss of social roles, and limitations of mobility. Despite this risk, levels of loneliness are highly stable in old age and relatively low compared to other points in the lifespan, likely owing to the improved emotion regulation skills often found among healthy older adults (Mund et al., 2020). However, to the extent that increases in loneliness do occur among older adults, this subset of individuals might be at advanced risk for Alzheimer's disease (AD) and related neurocognitive disorders (Lara et al., 2019). In the present perspective paper, we use the seed and soil model of neurocognitive disorders to better understand how loneliness impacts the brain to potentially increase one's risk for AD prior to frank cognitive decline and an AD diagnosis (McDonough and Allen, 2019). Seed and soil model of neurocognitive disorders: The seed and soil model of neurocognitive disorders (McDonough and Allen, 2019) was created to explain why some older adults who are predisposed to develop neurocognitive disorders, like AD, do not develop them even when surviving into late life (Figure 1). Specifically, it describes how one's lived environment and experiences throughout the life course set the stage for a toxic brain environment and greater susceptibility to pathology linked to AD later in life. Based on this model, the "seed" represents a predisposition to a neurocognitive disorder (e.g., genetic profile) and the "soil" refers to factors that moderate the expression of that seed. For example, the soil could be identified as the sum of lifetime exposures in one's physical, social, emotional, and cultural environments that contribute to non-genetic drivers of health (e.g., the exposome). Because the brain is a malleable organ, the external circumstances that comprise one's exposome can contribute to the disposition of the brain's microenvironment, and how it adapts to the consequences of such circumstances. An accumulation of negative circumstances can leave the brain vulnerable to the development of pathology such as amyloid-beta and tau, leading to progressive cognitive impairment and a diagnosis of AD. Critically, both the seed and the soil together determine whether an older adult will develop the disease. In its original form, this model did not address loneliness specifically, but the model is sufficiently general to apply new contexts as research and ideas evolve. As such, it might be the case that loneliness is one critical factor that can cause or accelerate the formation of brain pathology and thus, neurocognitive disorders later in life.Figure 1: Seed and soil model of neurocognitive disorders in relation to loneliness.Many factors in one's exposome (the soil) interact with one's predisposition to developing a neurocognitive disorder (the seed) to change the balance of neuroprotection vs. toxicity of the brain's micro-environment. Cumulative or persistent loneliness might serve as one critical negative factor in one's exposome that contributes to a toxic brain microenvironment. The more toxic the brain's micro-environment, the more vulnerable the brain is to pathogenesis, thus increasing the likelihood of the onset of a neurocognitive disorder, including Alzheimer's disease. To the extent that critical brain hubs like the DMN or connected regions like the entorhinal cortex are impacted, the higher the risk of such a disorder. A non-linear pattern of brain activation is hypothesized to occur in these early stages but with noticeable hyperactivity as an early indication of abnormal functioning. These brain structural and functional insults result in observable changes in cognition (e.g., episodic memory) and emotion (e.g., depression). Prevention of surpassing the disorder threshold can occur by (1) reducing the toxic brain load, (2) increasing neuroprotection, or (3) implementing interventions that strengthen the at-risk brain hubs. The brain map on the right side of the figure displays resting-state functional connectivity using the medial prefrontal cortex as a seed region to display the default mode network. Created with Microsoft PowerPoint.In addition to influencing the development of brain pathology, the model also predicts that brain regions with high metabolic consumption such as those found within the default mode network (including the hippocampus) should be at the earliest risk in those with negative exposomes. The model states that such accumulated risk would first manifest itself as hyperactivity (i.e., elevated activity or connectivity) using functional brain measures that would eventually turn into hypoactivity (or hypoconnectivity) and neurodegeneration in the same regions. Initial evidence for this model was provided by McDonough et al. (2019). Specifically, they found that dementia risk, as measured by multiple self-reported factors (e.g., education, hypertension, memory concerns) was associated with elevated brain activity in the default mode network and brain regions connected to this network (like the entorhinal cortex and the hippocampus). However, this study did not include loneliness as a part of their dementia risk score. Impact of loneliness on the brain: According to the seed and soil model of neurocognitive disorders, factors across the life course may contribute to late-life brain alterations, either through an accumulation of insults or latent impacts. To better understand how loneliness might impact the brain, we briefly review studies across the adult life course. One of the first studies to investigate brain structural correlates of loneliness was in young adults. This study found that greater loneliness was associated with smaller white matter microstructure in brain regions involved in emotion processing (e.g., anterior insula) and parts of the default mode network (e.g., lateral parietal and temporal cortex; Nakagawa et al., 2015). Subsequent studies using functional magnetic resonance imaging focused on brain regions involved in emotional processing and also in younger adults (e.g., Inagaki et al., 2016). Inagaki et al. (2016) showed that greater loneliness was associated with greater brain activity in a combined region of interest of left and right caudate and putamen. These authors suggested that this hyperactivity might be due to a greater "hunger" for social connection. These findings suggest that loneliness can start to have an impact on brain structure and function early in the life course. Since those initial studies, research on the impact of loneliness on both brain structure and function has increased at a rapid rate. In one notable study consisting of over 40,000 middle-aged and older participants from the UK Biobank, the association with loneliness and the brain was tested across multiple MRI modalities (Spreng et al., 2020). Brain regions within the default network and limbic system reliably showed the strongest associations with brain structure and function. Notably, greater loneliness was associated with larger brain structures and more brain activity, suggesting a greater reliance on or upregulation of these networks. From this perspective, these findings might be consistent with the seed and soil model of neurocognitive disorders such that seemingly healthy middle-aged and older adults with a negative exposome appear to be first characterized by greater reliance on key brain hubs. Alternatively, these relationships might be characteristic of only normative or healthy aging. Brain function and structure in people on the pathway towards AD or related dementia might exhibit different relationships, especially in the presence of pathology. Indeed, research investigating middle-aged and older adults on the AD spectrum shows different results. Two studies took a region of interest approach in the temporal lobe given the previous link between loneliness and memory-related brain structures. For example, in the Alabama brain study on risk for dementia, middle-aged and older adults with greater loneliness showed lower gray matter density (as measured by voxel-based morphometry) in the entorhinal cortex (Apostolou et al., 2023). Similar results in the entorhinal cortex were found using voxel-based morphometry in a sample with cognitively normal and amnestic mild cognitive impairment (Imai et al., 2022). The significant relationship between the entorhinal cortex and loneliness is notable because this region was found to be associated with risk for AD (e.g., McDonough et al., 2019). The entorhinal cortex is one of the earliest brain regions to develop tau pathology and projects to the hippocampus, creating a clear pathway to impact both the hippocampus and the default mode network. Consistent with this notion, in vivo proxies for tau pathology in the entorhinal cortex have been associated with loneliness among cognitively normal older adults (for a review, see Lam et al., 2021). Similarly, some studies also have shown a relationship between in vivo proxies for beta-amyloid and loneliness among cognitively normal older adults (Lam et al., 2021). These studies support the possibility that loneliness plays an early role in increasing one's risk for AD through brain alterations. Although the aforementioned studies provide support for altered default mode network and limbic regions (as well as reward regions), more research is needed to understand the reliability of these effects. For example, a recent meta-analysis failed to find any association between loneliness and brain structure (Wong et al., 2022). In the same meta-analysis, loneliness was related to brain activity largely in the ventral attention and frontoparietal networks, as well as the default mode network. Thus, the default mode network is clearly not the only network to be impacted and may not even be the most strongly impacted by loneliness. Interestingly, this meta-analysis also revealed a large heterogeneity of effects with some studies showing positive associations with loneliness, some negative, and some null effects. As mentioned above, one possible explanation for the opposing direction of effects could be related to age (younger vs. older) and propensity to develop neurocognitive disorders like AD (i.e., having the "seed"). Another explanation for some of the different associations across studies is that the samples might consist of different types of loneliness that often are not captured. For example, Tao et al. (2022) distinguished between different profiles of loneliness depending on their temporal consistency every five years across two to three time points. Middle-aged adults who expressed high levels of loneliness during at least two of the three time points were more likely to show a cognitive decline and lower temporal and hippocampal brain volume at their last time point relative to the non-lonely group. Similarly, an analysis showing the accumulation of loneliness across the three time points resulted in the strongest associations with brain structure. In contrast, middle-aged adults who showed transient/incidental loneliness at just one of the time points did not differ from non-lonely adults in cognition or brain structure. Thus, cross-sectional studies are unable to tease apart transient from persistent loneliness unless measured across multiple time points retrospectively, which is uncommon. The seed and soil model of neurocognitive disorders emphasizes the accumulation of life experiences across time that should give rise to the greatest brain vulnerabilities rather than single incidents. Concluding remarks: While previous research has demonstrated that loneliness can have deleterious effects on health, its relation to brain health and the development of neurocognitive disorders is still in its infancy. Although additional research is necessary to clarify the reason for heterogeneous patterns related to loneliness, the evidence of the science thus far suggests that multiple brain networks are impacted across the life course both in health and pathology. This information can provide further support to guide clinicians on how to educate their patients on the importance of maintaining social relationships throughout their lives. Recommendations to improve social connection include engaging in volunteer activities, recreational activities, scheduling regular activities with friends and family, engaging with your neighbors and community, and simply getting out of the house and into environments with other people present. While such recommendations can improve social engagement directly, other interventions aimed at targeting the brain regions impacted by loneliness might minimize brain atrophy or even restore brain functioning in downstream regions affected by loneliness, as proposed by McDonough and Allen (2019). Such attempts could be implemented via neurostimulation techniques such as transcranial magnetic stimulation or transcranial electric stimulation. Additional methods include implementing cognitive or behavioral interventions that have reliably been shown to impact brain regions within the temporal lobe and default mode network. C-Editors: Zhao M, Sun Y, Qiu Y; T-Editor: Jia Y