Stronger right hemisphere functional connectivity supports executive aspects of language in older adults
Introduction
Healthy older adults commonly exhibit age-related decline in a variety of cognitive functions, including memory, attention, executive function, and language production (Buckner, 2004, Burke and Light, 1981, Burke and Shafto, 2004, Burke and Shafto, 2008, Craik, 1994, Park et al., 2002, Rogers, 2000). Additionally, both brain structure and function decline with age, including declines in white matter integrity (Stamatakis et al., 2011, Troutman and Diaz, 2019), as well as age-related differences in activation and functional connectivity. For example, the brain’s functional networks during rest (e.g., default mode network, executive control network) become weaker and less segregated with age, and this has been associated with cognitive decline in episodic memory, working memory, and attention (Andrews-Hanna et al., 2007, Chan et al., 2014, Tian et al., 2012, Tomasi and Volkow, 2012, Zou et al., 2013). However, the extent to which these brain differences are observed in regions involved in language processing and their relationship with age-related differences in language abilities remains unknown.
In addition to age-related neural and network differences, several aspects of language production suggest there is age-related decline. For instance, older adults generally experience more tip-of-the tongue states (Burke, MacKay, Worthley, & Wade, 1991), produce slower (Burke and Shafto, 2004, Burke and Shafto, 2008), more disfluent, and syntactically simpler speech (Kemper et al., 2003, Kemper et al., 2001), and have more slips of the tongue (MacKay & James, 2004). In addition to older adults citing these declines in production as being highly frustrating (Ossher, Flegal, & Lustig, 2013), they have the potential to cause older adults to withdraw from social interactions (Hummert, Garstka, Ryan, & Bonnesen, 2004). Although language production abilities decline with age, language comprehension abilities remain largely intact in older adults as semantic knowledge generally increases with age (Burke and Shafto, 2008, Park and Reuter-Lorenz, 2009, Salthouse, 2010). While older adults often have more semantic knowledge than younger adults, older adults perform worse on tasks requiring semantic control consistent with declines in executive aspects of language (Hoffman, 2018, Krieger-Redwood et al., 2019). One possibility, as described by the Inhibition Deficit Hypothesis (Hasher and Zacks, 1988, Lustig et al., 2007) suggests that with increased age, inhibiting information becomes more difficult and the excess information can become distracting to older adults. Since older adults have more semantic knowledge, this may lead to increased competition among lexical items, particularly when semantic selection demands are high. Others have noted phonological aspects of age-related production failures (i.e., tip-of-the-tongue incidences) and have suggested that production declines arise from phonological deficits (Burke et al., 2000, Burke et al., 1991). It may be the case that both inhibitory and phonological deficits contribute to language production declines, with the impact of each varying according to the demands of the situation.
While age-related deficits in language production are commonly reported (e.g., Burke and Shafto, 2008, Hummert et al., 2004, Kemper et al., 2001), how this maps on to patterns of functional activation is less clear. Although age-related increases in task-based functional activation are commonly observed, there is considerable debate about whether such increases are compensatory or reflect neural decline (Dennis and Cabeza, 2011, Park and Bischof, 2013). According to the hemispheric asymmetry reduction in older adults (HAROLD) hypothesis, increased bilateral frontal activation in older adults may serve a compensatory function if this increased activation is associated with improved or maintained behavioral performance (Cabeza, 2002); however, if this overactivation is associated with behavioral declines, it is often interpreted as evidence for dedifferentiation, in which desegregation of brain networks results in performance declines (Li & Lindenberger, 1999). In looking specifically at functional activation associated with language, there have been variable reports about the nature and extent of age-related differences in the engagement of the core left-hemisphere language network (e.g., Shafto & Tyler, 2014). For example, some have suggested that age-related differences in activation may reflect the increased cognitive demands of the task as opposed to age-related changes in language (Davis, Zhuang, Wright, & Tyler, 2014). Davis et al. (2014) observed minimal age-related differences during naturalistic speech comprehension. In contrast, during speech comprehension with an explicit task, there were large age-related increases in activation in bilateral prefrontal cortex, suggesting that age-related differences in fMRI activation may be due to external task demands. Other studies examining language production have found that task difficulty modulated functional activation during language production, and that older adults were more sensitive to task difficulty, and less neurally responsive to increases in task demands (Zhang et al., 2018, Zhang et al., 2019). Moreover, Zhang et al. (2019) noted that brain-behavior relationships changed as a function of task difficulty, becoming weaker with increased task difficulty. Critically, task modulation of neural activity may confound interpretations of age-related differences in functional activation because these observed age-related differences in fMRI activation may be due to the task demands themselves.
As discussed in Section 1 above, some research suggests that behavioral and fMRI age-related differences in language production are, at least in part, due to task modulation. One method that can help disentangle the influences of age and task on brain activity is resting state functional connectivity (RSFC). RSFC examines how activation correlates across brain regions while participants are not performing an explicit task (Biswal, Yetkin, Haughton, & Hyde, 1995). Interestingly, brain regions that function together during tasks, also demonstrate synchronous correlations while participants are at rest (Rosazza & Minati, 2011). Importantly, studies have demonstrated that brain activity elicited while not actively engaged during a task can be used to predict off-line behavioral task performance. For example, a study using resting-state fMRI by Tian et al. (2012) observed that positive correlations between bilateral inferior frontal cortex (IFC) and the Default Mode Network (DMN) and negative correlations between bilateral middle occipital cortex, left inferior temporal cortex (ITC), and posterior insula activity reliably predicted shorter stop signal reaction times. Others have shown that resting-state activity predicted performance on an N-back working memory task (Zou et al., 2013), in which increased resting-state brain connectivity between the middle frontal gyrus and parietal regions and decreased connectivity between the medial prefrontal cortex and posterior cingulate was correlated with improved working memory. Additionally, studies have shown that age-related alterations during RSFC exist in a variety of networks including the default mode, executive control, and dorsal attentional networks, and these age-related declines have been associated with cognitive deficits in older adults (Andrews-Hanna et al., 2007, Chan et al., 2014, Sala-Llonch et al., 2015). However, fewer studies have examined RSFC within language regions (Zhu et al., 2014) or how such connectivity relates to task performance (Ferre et al., 2019, Miró-Padilla et al., 2017). Additionally, to our knowledge, only two RSFC studies measured the extent to which RSFC in a language network related to behavioral performance (Ferre et al., 2019, Miró-Padilla et al., 2017). Miró-Padilla et al. (2017) found that among younger adults, improved performance on the verbal fluency task was associated with higher functional connectivity between the thalamus and cerebellum and lower functional connectivities between the left inferior frontal gyrus (IFG) and right insula and between the left supplementary motor area (SMA) and right insula (Miró-Padilla et al., 2017). In another study, Ferre and colleagues examined both task-based functional connectivity and resting-state functional connectivity in language and default mode regions (Ferre et al., 2019). Of particular relevance to the present study, they found that overall older adults had weaker resting-state connectivities compared to younger adults. They also observed age-related increases in connectivity during language tasks, particularly in the left inferior frontal, left occipital, and bilateral parietal regions. Although Ferre and colleagues examined the relation between functional connectivity, age, and vocabulary, there were no significant interactions, suggesting that although vocabulary increases with age, the functional networks that support this ability remain intact across the lifespan. These studies provide important insights into how functional connectivity relates to language production in younger adults and vocabulary in both younger and older adults; however, the effect of age and the role that domain general resources, such as cognitive control, may have on language-related differences in older adults remains less clear.
Few studies have looked at whole-brain functional connectivity with key language regions during resting-state in younger and older adults and its relationship with behavioral performance (Ferre et al., 2019, Miró-Padilla et al., 2017). Davis et al. (2014) highlighted the notion that age-related declines observed in language may not be due to declines in the language network itself, but to declines in executive function, and more specifically the relationship between language and executive function regions. Additionally, conducting resting-state functional connectivity analyses allowed us to further investigate the whole-brain RSFC in language processing regions while removing the potential confound of task demands. The present study examined whole-brain functional connectivity with two language regions—left posterior superior temporal gyrus (pSTG) and the left inferior frontal gyrus (IFG), pars triangularis division. The left pSTG was chosen for its role in lexical selection and phonological retrieval and encoding (Graves, Grabowski, Mehta, & Gordon, 2007). The left IFG was selected for its role in executive function processes, including semantic control, and for its involvement in lexical selection during language processing (Sowell et al., 2001, Thompson-Schill et al., 1997). Consistent with prior reports (e.g., Ferre et al., 2019), we predicted that compared to younger adults, older adults would show weaker functional connectivity overall.
We were also interested in the relationship between RSFC and behavior. To examine this, we correlated whole-brain functional connectivity using the language seed regions with a task involving both language processing and executive function—the Stroop task—to better understand the relationship between language and executive function regions, and to examine how RSFC may predict age-related differences in behavioral performance. Because prior research has found that increased task-based functional activation in right hemisphere regions was associated with improved word retrieval performance in older adults (e.g., Wierenga et al., 2008), and others have observed increased age-related functional connectivity between left frontal and bilateral parietal regions, we predicted that in older adults, stronger functional connectivity between our seed language regions and right hemisphere regions involved in executive function would correlate with better Stroop performance. Moreover, given well-documented age-related decline in the frontal cortex, age-related differences in RSFC and RSFC-behavior relations may be greatest in frontal regions.
Section snippets
Participants
Forty younger adults between the ages of 18 and 32 (mean age = 23.18, 20 females) and 40 older adults participated in the study. Two older adults were excluded from further data analysis—one for an incidental finding and one for excessive head motion, resulting in the number of TRs censored preventing the calculation of a reliable correlation (Van Dijk et al., 2010)—resulting in 38 older adults between the ages 60 and 79 (mean age = 67.26, 28 females), remaining in the study. All participants
Stroop task behavioral results
We conducted a linear regression on reaction time (RT) to assess overall performance where Condition (incongruent, congruent, neutral) and Age Group (younger, older) were examined. There was a significant effect of Age Group—older adults performed significantly slower across conditions compared to the younger adults, β = 154.00, F(5, 228) = 17.72, p < 0.001. Follow up t-tests showed that there were significant Age Group differences across all three conditions: Incongruent, t(73.9) = 3.77, p
Discussion
Although older adults often report experiencing increased language production difficulties, how this corresponds to age-related differences in the language network is not well understood. For example, some have argued that age-related task differences in language may be due to age-related declines in executive function (Shafto & Tyler, 2014) because experimental tasks place additional demands on cognitive control (Zhang et al., 2018, Zhang et al., 2019). Moreover, age-related performance
Acknowledgments
This publication was supported by NIH R01 AG034138 (mtd) and funding from the Social Sciences Research Institute and the Department of Psychology at Penn State. The content is solely the responsibility of the authors and does not necessarily represent the official views of the funding agencies. Preliminary results were presented at the 2019 Dallas Aging and Cognition Conference. We thank the staff and scientists at the Social, Life, & Engineering Sciences Imaging Center and the Center for
References (72)
- et al.
Lateralization of resting state networks and relationship to age and gender
Neuroimage
(2015) - et al.
Disruption of large-scale brain systems in advanced aging
Neuron
(2007) - et al.
Prefrontal regions play a predominant role in imposing an attentional ‘set’: Evidence from fMRI
Cognitive Brain Research
(2000) Memory and executive function in aging and AD: Multiple factors that cause decline and reserve factors that compensate
Neuron
(2004)- et al.
On the tip of the tongue: What causes word finding failures in young and older adults?
Journal of Memory and Language
(1991) - et al.
Filtering induces correlation in fMRI resting state data
Neuroimage
(2013) - et al.
Age-related sensitivity to task-related modulation of language-processing networks
Neuropsychologia
(2014) - et al.
An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest
Neuroimage
(2006) - et al.
“Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician
Journal of Psychiatric Research
(1975) - et al.
Thresholding of statistical maps in functional neuroimaging using the false discovery rate
Neuroimage
(2002)
The lateral occipital complex and its role in object recognition
Vision Research
The nuisance of nuisance regression: Spectral misspecification in a common approach to resting-state fMRI preprocessing reintroduces noise and obscures functional connectivity
Neuroimage
Working memory, comprehension, and aging: A review and a new view
Functional connectivity in resting state as a phonemic fluency ability measure
Neuropsychologia
Regional homogeneity of resting state fMRI signals predicts Stop signal task performance
Neuroimage
Normative data stratified by age and education for two measures of verbal fluency: FAS and animal naming
Archives of Clinical Neuropsychology
Age-related changes in word retrieval: Role of bilateral frontal and subcortical networks
Neurobiology of Aging
Development and validation of a geriatric depression screening scale: A preliminary report
Journal of Psychiatric Research
Task difficulty modulates age-related differences in the behavioral and neural bases of language production
Neuropsychologia
New and updated tests of print exposure and reading abilities in college students
Behavior Research Methods
California verbal learning test: Performance by patients with focal frontal and non-frontal lesions
Brain
Executive function and the frontal lobes: A meta-analytic review
Neuropsychology Review
The Freiburg visual acuity test-automatic measurement of visual acuity
Optometry and Vision Science
Controlling the false discovery rate: A practical and powerful approach to multiple testing
Journal of the Royal Statistical Society: Series B (Methodological)
Functional connectivity in the motor cortex of resting human brain using echo-planar MRI
Magnetic Resonance in Medicine
Memory and aging: The role of retrieval processes
Psychological Bulletin
Aging and language production
Current Directions in Psychological Science
Language and aging
The counting Stroop: An interference task specialized for functional neuroimaging—validation study with functional MRI
Human Brain Mapping
Hemispheric asymmetry reduction in older adults: The HAROLD model
Psychology and Aging
Maintenance, reserve and compensation: The cognitive neuroscience of healthy ageing
Nature Reviews Neuroscience
Decreased segregation of brain systems across the healthy adult lifespan
Proceedings of the National Academy of Sciences
The Ishihara test for color blindness
American Journal of Physiological Optics
Memory changes in normal aging
Current Directions in Psychological Science
Neuroimaging of healthy cognitive aging
Cited by (12)
The multifaceted nature of language across adulthood
2022, Brain and LanguageThe intensity and connectivity of spontaneous brain activity in a language network relate to aging and language
2021, NeuropsychologiaCitation Excerpt :These age-related differences in network characteristics have been associated with worse behavioral performance across several different cognitive domains (King et al., 2018; Onoda et al., 2012; Sala-Llonch et al., 2015; Varangis et al., 2019; L. Wang et al., 2010). To date, only a few studies have investigated age-related differences in a language resting-state network and its relationship with language ability (Ferré et al., 2019; Gertel et al., 2020). For instance, Ferré et al. (2019) reported that increased age was associated with increased connectivity in a language network during language tasks, but decreased connectivity during resting-state.
Dependency network approach to the oral production of English- and Chinese-speaking healthy older adults
2023, Speech, Language and HearingResting State Network Segregation Modulates Age-Related Differences in Language Production
2023, Neurobiology of Language