Abstract
Objectives: Off-topic verbosity (OTV) refers to prolonged speech that derails from the initial conversational topic by including more loosely associated speech and becoming increasingly more unfocused and distant from the initial topic. Previous research has suggested that, among older adults, loneliness may be associated with greater OTV. The purpose of this study was to investigate the nature of the relationship between loneliness and OTV among young adults (n = 62) and older adults (n = 80). Methods: Participants completed a measure of loneliness and provided speech samples, which were transcribed and rated for OTV. Results: Results indicated some relationship between loneliness and tangentiality of speech, particularly among older adults. Discussion: Overall, loneliness may be associated with greater OTV in older adults, which could further explain the connection between increased loneliness and worse health outcomes in older adulthood.
Loneliness has been described as an epidemic in modern society (Killeen, 1998). Loneliness is associated with several adverse outcomes, including worse psychological health, cardiovascular health, and cognitive decline (Courtin & Knapp, 2017; Gardiner et al., 2018; Luo et al., 2012; Medvene et al., 2016). Older adults may be particularly vulnerable to loneliness and social isolation due to common problems with deterioration in health, lack of physical activity, and restricted living situations (Victor & Bowling, 2012). Impairments in communication in particular have been shown to predict higher levels of loneliness in older adults (Palmer at al., 2019). Previous research has indirectly connected loneliness with unfocused speech qualities, such as off-topic verbosity (OTV), among older adults (Arbuckle et al., 2004). The purpose of this study was to investigate the direct relationship between loneliness and off-topic verbosity in the speech among both young and older adult age cohorts.
Loneliness
Loneliness affects individuals of all ages and life stages (Green et al., 2001). Loneliness may be salient for college students because they often find themselves in a new environment away from established relationships (Stewart-Brown et al., 2000). Among college students, loneliness has been linked with poor physical health (Segrin & Passalacqua, 2010), poor sleep quality (Zawadzki et al., 2013), life dissatisfaction (Diener & Seligman, 2002), and suicide and suicidal ideation (Stravynski & Boyer, 2001). The effects of loneliness in young adults may be long-lasting; longitudinal research suggests that young adults who reported higher loneliness are at increased risk of disability and obtaining a lower income once they reach midlife (von Soest et al., 2020a).
Loneliness may also be a concern for older adults (Theeke, 2009). Similar to college students, older adults may find themselves moving to new environments (e.g., a retirement community or relocating to be near family members) and away from familiar individuals (Erickson et al., 2006). Aging itself is often associated with disruption in individuals’ social networks later in life, as individuals move away or die (Wrzus et al., 2013). Loss of loved ones and friends represents common and major negative changes in individuals’ social networks that have been associated with loneliness in older adults (Böger & Huxhold, 2018; Pettigrew & Roberts, 2008; von Soest et al., 2020b). Loneliness is associated with accelerated cognitive decline (Cacioppo & Hawkley, 2009; Boss et al., 2015; Donovan et al., 2017). Increased loneliness in older adults has been linked to a steeper decline in general cognitive ability (Tilvis et al., 2004; O’Luanaigh et al., 2012; Gow et al., 2013), dementia/Alzheimer’s Disease (Holwerda et al., 2014; Wilson et al., 2007), memory (Lara et al., 2019; O’Luanaigh et al., 2012; Schnittger et al., 2012; Shankar et al., 2013; Wilson et al., 2007; J. Yin et al., 2019), executive functioning (Schnittger et al., 2012), verbal fluency (Lara et al., 2019; J. Yin et al., 2019), and processing speed (Gow et al., 2013; O’Luanaigh et al., 2012). A systematic review indicated that these findings are inconsistent regarding whether other demographic and psychosocial factors impact the relationship between loneliness and cognitive function (Boss et al., 2015).
Loneliness has been deemed a concern for both young and older adults (Stewart-Brown et al., 2000; Theeke, 2009), and therefore included for investigation in the current study. In fact, research often indicates that young and older adults have the highest levels of loneliness across the lifespan (Luhmann & Hawkley, 2016; Victor & Yang, 2012). Studies are mixed regarding the direction of the relationship between loneliness and age. A recent, cross-cultural study found that loneliness decreases with age (Barreto et al., 2021). However, multiple factors, such as loss of social connections and increasing health concerns, support the possibility of higher levels of loneliness in older adults than in young adults (Pinquart & Sörensen, 2001; von Soest et al., 2020b). Additionally, older adults have been found to have higher levels of loneliness than younger adults on an indirect measure of loneliness (Nicolaisen & Thorsen, 2014). In the current study, we investigated age cohort differences in loneliness to further examine the complex relationship between loneliness and age as well as explore the role quality of speech may play.
Speech Quality
Speech is humans’ primary form of communication, and the quality of speech has significant impacts on individuals’ social functioning (Bowie et al., 2011; Heine & Browning, 2002; Palmer at al., 2019). A small body of research has investigated tangential, self-focused, and verbose speech among older adults (i.e., OTV; D. Gold et al., 1988; D. P. Gold & Arbuckle, 1995). Early studies of OTV focused on speech that was defined as loosely associated and increasingly extraneous to the initial conversational stimulus (D. Gold et al., 1988). Subsequent studies (Arbuckle & Gold, 1993; D. P. Gold & Arbuckle 1995) continued to utilize D. Gold et al. (1988) definition of OTV, with the focus remaining on tangential speech. However, later studies (Blake, 2006) expanded on constructs of speech quality by focusing on not only tangentiality but also egocentrism and quantity of speech. Egocentric speech is a type of verbal discourse that is often overly personal in nature and includes inserting oneself into a story (Tompkins et al., 1993; Blake, 2006). D. Gold et al. (1988) postulated that egocentrism may lead to increased OTV. Although few studies have researched this direct relationship between egocentric speech and aging, older adults often perform higher on measures of egocentrism when compared to younger adults (Looft & Charles, 1971; McDonald & Stuart-Hamilton, 2003). Quantity of speech refers to loquacity, or the degree of verbosity and paucity of speech (Juncos-Rabadán et al., 2005; Blake, 2006). Numerous studies have established the relationship between higher quantity of speech and age (Gould & Dixon, 1993; Horton et al., 2010; James et al., 1998; Juncos-Rabadán et al., 2005; Obler, 1980). Research indicates that increased age is associated with greater OTV, throughout the lifespan as well as throughout older adulthood (Arbuckle et al., 2000; Pereira et al., 2019). Age effects in OTV may reflect changes in cognition (Arbuckle & Gold, 1993; Hoffman et al., 2018), including declining everyday competence and executive functioning (Arbuckle et al., 2004; Hoffman, 2019; S. Yin & Peng, 2016).
The Current Study
Previous research has linked increased OTV in older adults with greater loneliness through everyday competence (Arbuckle et al., 2004). The direction of this relationship is supported by research connecting increased loneliness with decreased cognitive abilities and greater OTV with declining cognition (Goverover, 2004; Lipskaya-Velikovsky et al., 2018; Shankar et al., 2017). Conversely, evidence from Arbuckle et al. (2004) also suggests that higher levels of OTV predicted lower levels of loneliness when controlling for everyday competence. In this study, we sought to elucidate the nature of the relationship between OTV and loneliness in both young and older adult age cohorts.
The purpose of this study was to investigate the relationship between loneliness and off-topic verbosity among young adult and older adult age cohorts. In line with Nicolaisen & Thorsen (2014) findings that older adults were significantly more lonely than younger adults, we hypothesized (H1) older adults would be lonelier than young adults. Consistent with previous research (Arbuckle & Gold, 1993; Pereira et al., 2019), we also hypothesized that (H2) older adults would display greater off-topic verbosity than young adults. We further hypothesized (H3) that the relationship between loneliness and off-topic verbosity would be stronger among older adults than among young adults.
Methods
Participants
Young adult participants (n = 62; age 18–28, M = 20.69, SD = 2.44) consisted of undergraduate students enrolled in a psychology course at a large public university in the southern U.S. Participants were recruited through the department’s website, where students can volunteer to participate in research studies in order to earn course credit. Regarding gender, 62.9% of young adult participants identified as female and 37.1% identified as male. Regarding ethnicity, 37.1% of young adult participants identified as white/Caucasian, 17.7% as black/African American, 19.4% as Latinx, 19.4% as multiracial, 3.2% as Asian/Pacific Islander, and 3.2% as Native American.
Older adult participants (n = 80; age 60–98, M = 76.54, SD = 8.32) consisted of community-dwelling individuals from a major metropolitan area and a small city in the southern U.S. Older adult participants were recruited through announcements in semi-public settings such as retirement communities, senior centers, civic organizations, and religious organizations. Regarding gender, 65.0% of older adult participants identified as female, 27.5% identified as male, and 7.5% as other or unspecified. Regarding ethnicity, 88.8% of the older adult participants identified as white/Caucasian, 5.0% as multiracial, 2.5% as Latinx, 2.5% as black/African American, and 1.3% as Asian/Pacific Islander.
While the young adult and older adult samples were not matched on education level or cognitive ability, each sample consisted of people living independently within the community, and the two samples did not differ significantly on an estimate of premorbid intellectual functioning (t [140] = 1.35, p = .18).
Measures
Loneliness
The Three-Item Loneliness Scale (TILS; Hughes et al., 2004) is a brief self-report measure of loneliness. The TILS consists of three items (e.g., “How often do you feel left out?”) to which participants responded using a scale ranging from 1 = hardly ever to 3 = often. Item scores were averaged for a total loneliness score, so that higher scores indicate a higher level of loneliness (in this study, Cronbach’s α = .71).
Speech Quality
All participants provided two samples of speech for analysis of speech quality. These speech samples were prompted by examiners with a set script for two verbal description tasks adapted from Trunk and Abrams (2009), in order to best measure participants’ everyday speech in different communication situations. One verbal description task was of an episodic memory: “Tell me about a trip or vacation that you have taken; you can provide as many details as necessary.” The other verbal description task was of a procedural memory: “I would like you to please explain how you would make the following meal step-by-step: a cup of coffee, scrambled eggs, and a bowl of fruit. Pretend I do not know how to make the meal and provide details.”
Estimated Premorbid Intellectual Functioning
To obtain an estimate of premorbid intellectual functioning, participants were administered the Wechsler Test of Adult Reading (WTAR; Wechsler, 2001). The WTAR is composed of 50 irregularly spelled words that participants are asked to pronounce aloud. Given that the WTAR norms do not cover the entire age range of the present study, raw scores were used to estimate premorbid intellectual functioning.
Procedures
This study was approved by the committee for the protection of human subjects. Informed consent was obtained from all participants, and all participants were interviewed in-person. Young adult participants were given course credit. Older adult participants were entered into a drawing for one of several gift cards.
All participants were provided up to 5 min to respond to each verbal description task. The speech samples for both verbal description tasks were recorded and transcribed in order to blind raters to the demographic characteristics of each participant. Three graduate students (ages 22–25) acted as raters and independently evaluated the speech quality of all participant transcriptions on 5-point Likert scales adapted from Blake (2006). Each speech sample was rated separately for extent of verbal tangentiality, quantity, and egocentrism. The rating scale for tangentiality ranged from a score of 1 for very tangential or excessively off topic responses to a score of 5 for not tangential responses where all information was on topic. Raters were instructed to score egocentrism on a scale from 1 for extremely egocentric or excessively over-personalized responses to a score of 5 for not egocentric responses with no references to overly personal experiences. The rating scale for quantity of speech ranged from a score of 1 for verbose or excessively long responses to a score of 5 for a paucity of speech or very short responses, with a score of 3 representing appropriate length responses. Interrater reliability was calculated between all three raters for speech quality variables on both verbal description tasks with intraclass correlation coefficients—which ranged from .51 to .92, with all but one coefficient falling above .70 and a total rater agreement of .89 across all rated variables. For more manageable score interpretation, the rating scale variables on speech quality were reversed so that a higher score represented a higher level of that speech quality (i.e., a tangentiality score of 5 represented a high level of tangentiality).
Results
The young adult and older adult samples did not differ on estimated intellectual functioning, t (140) = 1.35, p = .18.
A multivariate analysis of variance (MANOVA) indicated a significant multivariate effect of age cohort (i.e., young vs. older adults) on loneliness and speech quality variables, F (4, 137) = 2.58, p = .04, Wilks’ λ = .93, partial ηp2 = .07. Univariate analyses showed that older adults had more tangential speech, F (1, 140) = 7.32, p = .008, ηp2 = .05 and more egocentrism, F (1, 140) = 4.00, p = .05, ηp2 = .03, than young adults. Young adults and older adults did not differ in loneliness or quantity of speech.
An interesting pattern was observed in participant responses to the two verbal description tasks, so further MANOVAs were run as post-hoc analyses to investigate the difference in speech quality by age cohort separately for the episodic and procedural prompts. No significant effect of age cohort was found in episodic speech quality. Age cohort had a significant multivariate effect on procedural speech quality (F [3, 138] = 3.38, p = .02, Wilks’ λ = .93, ηp2 = .07), and univariate analyses suggest that older adults had more tangential (F [1, 140] = 6.01, p = .02, ηp2 = .04) and more egocentric (F [1, 140] = 5.24, p = .02, ηp2 = .04) speech than young adults.
Correlations and Descriptive Statistics of Loneliness and Speech Quality Variables.
*p < .05, **p < .001
Regression Analyses for Speech Quality Predicting Loneliness (N = 142).
*p < .05
Discussion
This study investigated the loneliness and speech quality of young adult and older adult age cohorts and the relationships between these factors. Young adults and older adults did not differ with regards to loneliness, meaning that H1 was not supported. Results suggest that both young adults and older adults suffer from loneliness at similar rates in their everyday lives. While these results are inconsistent with some previous results (Nicolaisen & Thorsen, 2014), they are consistent with other known studies (Pinquart & Sörensen, 2001; von Soest et al., 2020b) that suggest that both young adults and older adults have the highest levels of loneliness across the lifespan without much measurable differences between the two age cohorts.
Older adults demonstrated greater tangentiality and egocentrism of speech than young adults, supporting H2. These results suggest that older adults’ everyday speech may include more off-topic and overly self-referential information than young adults’ speech. This is consistent with research on OTV and age-related changes in conversational speech (Arbuckle & Gold, 1993; Pereira et al., 2019). When separated into procedural and episodic speech, older adults were more tangential and egocentric in their procedural descriptions but not in their episodic descriptions. This is consistent with previous research investigating shorter instances of speech (Trunk & Abrams, 2009), and suggests that these differences measured in young and older adults’ speech may only reflect everyday speech in instances of structured or instructional speech.
Loneliness was associated with greater tangentiality of speech among all participants (i.e., young adults and older adults considered together). This suggests that as people’s everyday speech increases in off-topic information, their loneliness increases as well—which is partially consistent with the results of Arbuckle et al. (2004). Although these variables have only been studied together in older adults, this relationship may apply to other age cohorts as well. It may be that as loneliness increases from middle adulthood to older adulthood (Victor & Yang, 2012), so too does tangential speech increase—which would suggest a similar age-related change in both variables. However, while the relationship between loneliness and tangentiality of speech was significant among older adults, it was not significant among young adults. Additionally, OTV was found to predict a significant portion of the variance in the loneliness of the older adult age cohort but not of the young adult age cohort. These results suggest that loneliness is related to the tendency for speech tangents in older adults but not in young adults. While it is unknown whether any relationship exists between loneliness and tangentiality in any unstudied age cohorts (e.g., middle adulthood), there appears to be some evidence that the relationship between loneliness and tangentiality varies by age cohort—offering partial support for H3.
Multiple hypotheses could potentially explain the relationship between speech quality and loneliness in older adults but not in young adults. First, in keeping with the pragmatic change hypothesis (James et al., 1998; Trunk & Abrams, 2009), it is possible that loneliness affects the communication goals of older adults. Older adults who are lonelier may go on more tangents in order to express themselves. However, that would not explain why lonelier older adults would have lower speech quantity. Second, it is possible that older adults with high OTV speech quality tend to alienate the people around them who might otherwise be social connections, leading to higher levels of loneliness in high OTV older adult populations. This possibility would be consistent with previous findings that other limitations to verbal communication (e.g., linguistic barriers) are associated with higher loneliness in older adults and that older adults without family social systems tend to rely on casual socialization in public settings with other older adults for building social connection and support (Nyqvist et al., 2021; Torres, 2019). Finally, in keeping with the inhibitory deficit hypothesis (Arbuckle & Gold, 1993; D. P. Gold & Arbuckle, 1995), loneliness may be associated with higher tangentiality and lower quantity of speech due to age-related declines in cognitive abilities. Executive functions, which are impacted by age (Arbuckle & Gold, 1993; Hoffman et al., 2018; Zelazo et al., 2004), include inhibition of irrelevant details and generativity, which would explain both the tangentiality and quantity findings. It is possible that loneliness and declines in cognition exacerbate each other given previous literature linking them (Arbuckle et al., 2004).
Results of post-hoc analyses considering relationships between loneliness and speech qualities but broken down by episodic and procedural prompts indicated that older adults’ tangentiality predicted higher loneliness for episodic but not procedural speech. This is consistent with the flexible and descriptive nature of older adults’ speech in Arbuckle et al. (2004), but inconsistent with the direction of the direct association found between loneliness and OTV. These conflicting results suggest a complicated but true relationship between loneliness and OTV among older adults, which is consistent with past literature linking loneliness and communication/social contact in older adulthood (Palmer at al., 2019; von Soest et al., 2020b).
This study further elucidates age cohort differences and the relationship between loneliness and OTV; however, several limitations occurred. Samples sizes were small and might not be representative of either age cohort’s U.S. population, as both cohorts were primarily female. Additionally, demographic variables were not matched among age cohort samples. For example, 37.1% of the young adult cohort and 88.8% of the older adult cohort identified as white/Caucasian. This was also apparent in education, as the young adult cohort sampled undergraduate college students, while the older adult cohort varied among education levels. The small sample sizes might also have resulted in analyses being underpowered.
Future studies should aim to match cohort samples as well as represent national demographics in order to increase the generalizability of the results. Additionally, larger samples may allow future researchers to further explore the nature of the relationship between loneliness and speech quality as well as determine whether a true age cohort difference in this relationship occurs. A longitudinal study could further investigate the difference between cohort effects, the direction of these effects, and age-related changes in the relationship between loneliness and OTV. Future studies specifically aimed at differentiating between the speech differences in describing episodic and procedural memories may also be able to further illuminate the relationship between communication and loneliness across the lifespan.
Overall, the results of this study are consistent with the notion that increased loneliness is associated with greater tangentiality in speech. This may be especially true among older adults, who showed evidence of a relationship between loneliness and tangentiality as well as quantity of speech. Greater loneliness appears to be linked with lower quantity of speech and greater tangentiality of speech, which may be connected to age-related cognitive decline. While this is not consistent with some of the previous findings by Arbuckle et al. (2004), it is consistent with the direction of their theoretically based hypotheses. Additionally, older adults showed a stronger relationship between loneliness and tangentiality when recounting episodic memories. Loneliness may increase tangentiality for episodic speech because lonelier people get less practice organizing their thoughts into a narrative. The open-ended nature of conveying episodic memories may therefore make them particularly prone to tangentiality. In everyday life, lonelier older adults may have lower quantity of speech and greater tangentiality when interacting with health care providers, which could be part of the link between loneliness and worse health outcomes in older adults (Courtin & Knapp, 2017; Gardiner et al., 2018). This study highlights the need for future research to focus on further investigating the complicated combination of factors that affect loneliness in both young and older adulthood, as well as the differences and similarities of these relationships across the lifespan.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
