Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Apr 1.
Published in final edited form as: Stress Health. 2023 Aug 1;40(2):e3294. doi: 10.1002/smi.3294

The relationship between loneliness and the experiences of everyday stress and stressor-related emotion

Jee eun Kang 1, Jennifer E Graham-Engeland 2, Stacey Scott 3, Joshua M Smyth 2, Martin J Sliwinski 1,*
PMCID: PMC10830881  NIHMSID: NIHMS1926303  PMID: 37526522

Abstract

Loneliness influences how people experience and respond to stressors, which may account for its role as a risk factor for morbidity and mortality. The present study was motivated by emerging evidence that affective responses to minor daily events have long-term implications for health and well-being. Specifically, we evaluated how individual differences in loneliness relate to the frequency of everyday stressors and stressor-related negative emotions. A diverse community sample of 255 adults (age 25–65 years) completed ecological momentary assessments (EMA), during which they reported recent stressors and current negative affect (NA) five times a day for 14 days. Multilevel logistic analyses indicated that there was a quadratic association between loneliness and likelihood of reporting stressors, controlling for demographics, social isolation, depressive symptoms, and context (current activities, current location). Multilevel regression indicated that loneliness was unrelated to the concurrent effect of stressors on NA but significantly larger lagged stressor effects were observed among individuals in the low and high ranges of loneliness. These findings suggest that individuals with high levels of loneliness are more likely to experience everyday stressors and have prolonged emotional responses following stressors.

Keywords: loneliness, subjective social isolation, daily stress, emotional stress response, negative affect

1. Introduction

Loneliness refers to the distress, discomfort, or dissatisfaction that results when one perceives a gap between one’s desires for social connection and their actual social experiences (Peplau & Perlman, 1982). Loneliness refers to the distress, discomfort, or dissatisfaction that results when one perceives a gap between one’s desires for social connection and their actual social experiences (Peplau & Perlman, 1982). Loneliness represents the subjective feelings or perception of social isolation. Importantly, loneliness is distinct from social isolation itself, which refers to objective aspects of social deficiencies (Holt-Lunstad et al., 2015) and is often measured by marital status (not being married), living arrangement (living alone), social network size (small network), or social contact (less frequent). Accumulating evidence points to the negative association of loneliness with a broad range of health outcomes. For instance, loneliness was associated with poor physical (Christiansen et al., 2016) and mental health (Alpass & Neville, 2003; Stephens et al., 2011), cognitive decline (Boss et al., 2015), and all-cause mortality (Holt-Lunstad et al. 2015).

Stress is one possible mechanism through which loneliness may affect health (Hawkley & Cacioppo, 2010; Christiansen et al., 2016). Loneliness is associated with feelings of stress and anxiety (Cacioppo et al., 2006), and may induce cognitive biases (Cacioppo & Hawkley, 2009; Hawkley & Cacioppo, 2010) that lead people to expect, attend to, and remember more negative social information. Social support mediates impacts of stress-provoking conditions and provides a buffer against the effects of stressors (Pearlin et al., 1981). Loneliness, which has been associated with lower social support (Segrin & Passalacqua, 2010), potentially contributes to amplified and prolonged stress responses. Therefore, lonely individuals may be more likely to experience stress and slower to recover from its effects.

Negative emotional responses to everyday stressors are associated with negative health outcomes, including chronic physical health conditions (Piazza et al., 2013), affective disorders (Charles et al., 2013), depressive symptoms (Parrish et al., 2011), and all-cause mortality (Mroczek et al., 2015). Examining the role that loneliness plays in emotional responses to everyday stress may provide insight into its effect on long-term mental and physical health. Accordingly, the goal of the present study is to evaluate whether individual differences in loneliness relate to the likelihood of reporting stressors, and the magnitude of concurrent or prolonged negative emotional states associated with stressors in the context of everyday life.

1.1. What is the evidence that loneliness is related to stress?

The evidence linking loneliness and stress comes primarily from three paradigms: cross-sectional survey studies, experimental (stress induction) studies, and experience sampling studies. In cross-sectional surveys, people who report greater levels of loneliness report more stressful life events (Gaudin et al., 1993) and more exposure to chronic stressors (Hawkley et al., 2008). Cacioppo and colleagues (2000) found that loneliness was related to reports of increased frequency of daily hassles, increased fear of stressful tasks, and a higher level of perceived stress in college students. Additionally, a recent longitudinal survey study (Laustsen et al., 2023) revealed that loneliness and perceived stress mutually predicted each other over time and that high loneliness predicted greater perceived stress cross-sectionally.

Experimental work has provided fairly consistent evidence linking loneliness and experimentally induced physiological reactivity. Lonely people tend to exhibit amplified cardiovascular, inflammatory, and immunosuppressive responses immediately following experimental stressors, as well as slower recovery of those responses (Balter et al., 2019; Brown et al., 2018; Hackett et al., 2012; Jaremka et al., 2013; Kiecolt-Glaser et al., 1984; Ong et al., 2012; Steptoe et al., 2004), despite a few inconsistent results regarding cardiovascular reactivity (e.g., Brown et al., 2019; Brown et al., 2022). Although these findings overall suggest that loneliness moderates physiological responses to stressors, results focused on subjective stress responses in laboratory settings are not as consistent. Three experimental studies found that individual differences in loneliness were not related to subjective stress (Hackett et al., 2012; Steptoe et al., 2004) and affective responses (Ong et al., 2012) immediately following the performance of a stressful task. However, Hackett and colleagues (2012) observed that higher levels of loneliness corresponded with elevated subjective stress ratings during the post-stressor recovery period. Thus, evidence from experimental stress paradigms raises the possibility that loneliness may not moderate immediate subjective stress responses (reactivity), but that it might contribute to prolonged responses (recovery).

Several daily diary and ecological momentary assessment (EMA) studies have examined the relationships between loneliness and stress-related processes in everyday experiences. Hawkley and colleagues (2003) showed that individual differences in loneliness was associated with higher negative ratings of daily interactions, stress appraisals, and threat ratings. Results from several daily diary (Russell et al., 2012; Wolf & Davis, 2014) and EMA (Hawkley et al., 2007) studies have demonstrated that individual differences in loneliness were associated with more frequent negative social interactions, but that loneliness did not moderate immediate changes in affect relating to negative social interactions. Nowland and colleagues (2018), focusing on a naturally occurring social challenge (3-day university preparation program for prospective undergraduates), revealed that loneliness did not moderate the level of HPA stress reactivity, although the high loneliness group had higher levels of perceived stress and social threat sensitivity towards the event. In sum, daily diary and EMA studies have provided some evidence that individual differences in loneliness are associated with greater typical/mean levels of reported emotional distress, higher frequency of negative social interactions, and more negative appraisals of those interactions, but not that loneliness is related to changes in affect relevant to negative social interactions. However, other types of everyday occurrences can be a source of stress in addition to negative interactions, such as work-related stressors and network stressors (i.e., events that occur to close others) (Almeida et al., 2002). If loneliness moderates the magnitude and duration of negative emotional states following multiple types of daily stressors, then prior research may have underestimated its effects by examining only one type of stressful occurrence (i.e., negative social interactions).

1.2. The Present Study

The aim of the paper was to evaluate whether individual differences in loneliness was associated with experiences of everyday stressors and immediate and prolonged emotional responses to stressors. Specifically, based on the theoretical arguments connecting loneliness to hypervigilant tendencies and lack of social resources, as well as the empirical studies linking loneliness and stress discussed above, we hypothesized that loneliness would be associated with greater likelihood of reporting everyday stressors, and greater immediate and more prolonged responses to stressors. Our approach used an EMA design in which participants completed mobile surveys up to 5 times daily for 14 days. On each assessment, participants reported current negative affect (NA) and whether anything stressful happened since the last assessment. Following prior work (e.g., Charles et al., 2013), we operationally defined concurrent emotional responses as the difference in NA on assessment occasions when a stressor was reported compared to non-stressor occasions. We should note that this approach to defining concurrent stress responses assumes that stressors precede shifts in negative affect but would also be sensitive to the effect of negative emotional states on the likelihood that a stressful event is reported. Prolonged emotional responses were operationalized by the difference in NA on assessment occasions when a stressor was reported on the prior occasions compared to non-stressor occasions (i.e., the lagged stressor effect).

Cacioppo and colleagues (2006) questioned the assumption that effects of loneliness are linear across the severity continuum. Consistent with this view, previous studies have found evidence that loneliness has a curvilinear (i.e., quadratic) associations with hypervigilance to social threat (e.g., Qualter et al., 2013; Bangee et al., 2014). Specifically, people in the highest range of loneliness exhibited visual vigilance against threating situations compared to others. Accordingly, the current study investigated both linear and quadratic associations of loneliness and likelihood of stressors and stress responses.

Previous studies showed that loneliness and depression are two distinct constructs but are highly associated with each other (e.g., Cacioppo et al., 2006). Moreover, some studies showed that controlling for depression attenuates effects of loneliness (e.g., Patterson & Veenstra, 2010). Thus, we examined whether the effect of loneliness on stress reports and stress responses were robust to controlling for depressive symptoms.

2. Method

2.1. Sample

Data for the present research were from the Effects of Stress on Cognitive Aging, Physiology and Emotion (ESCAPE) project (for details, see Scott et al., 2015). A racially and economically diverse sample of adults was systematically recruited using voter registration lists from Co-Op City, Bronx, New York. Eligible participants were 25–65 years of age, ambulatory, fluent in English, and free of visual impairment. The average age of the sample (N=255) was 46.5 (SD = 11.14). 65.1% were women, 43.9% had a college degree or greater education. In terms of race, 69.8% were Black, 26.3% were White, and 3.9% were others. Married participants comprised 31.4% of the sample, and participants who were living alone represented 24.3% of the sample.

2.2. Procedures

The ESCAPE study was approved by the Albert Einstein College of Medicine of Yeshiva University ethical review board. Eligible participants completed baseline surveys to assess demographic and psychosocial characteristics at their home. Following the baseline surveys (2–4 weeks later), participants completed a 14-day EMA protocol, during which they answered surveys 5 times a day when signaled by a “beep” from a study-provided smartphone. As such, each participant received beep signals to complete 70 EMA surveys during the entire EMA period. Of those 70 surveys, 83.9% (median [IQR], 77.1% [91.4%−98.6%]; SD= 20.7%) were completed on average, and 20% of the participants completed all the scheduled assessment. EMA surveys included questions about their experiences of recent stressors, activities, and current emotions. Beeps were scheduled to sample the entire waking day, calibrated by each participant’s normal wakeup time, with quasi-random timing to ensure that participants could not anticipate the beeps; the average time between scheduled beeps was 2 hours and 33 minutes. In addition to these beeped EMA surveys, participants completed a brief survey each morning and at the end of each day during the EMA period; however, these morning and bedtime surveys were not included in the present research, as they did not measure recent stressors. A custom EMA software platform was developed by the Penn State Survey Research Center, for this study.

2.3. Measures

Loneliness.

Loneliness was measured in the baseline survey with a short version of the Patient-Reported Outcomes Measurement Information System (PROMIS) Social Isolation scale (Cella et al., 2010). Six items asked perceptions of being left out, isolated, avoided, disconnected from, or unknown by others. Responses were on a five-point scale (from 1 [Never] to 5 [Very often]). Following PROMIS scoring guidelines, the sum of the six items was transformed to a standardized score with a mean 50 and a standard deviation of 10 (PROMIS Health Organization, 2021). These standardized scores were grand-mean centered; Both linear and quadratic terms of loneliness were included in the models.

Momentary Negative Affect (NA).

In each momentary assessment, participants were asked to report their negative emotions using a slider scale anchored by “Not at all” to “Extremely” (coded as 0 to 100). The average of five items (tense, depressed, angry, unhappy, and frustrated) was used as an outcome variable to indicate momentary NA (between-person reliability intraclass correlation [ICC]= 0.99, within-person reliability ICC= 0.84).

Current and lagged stressor.

At each beep, participants received the prompt “Did anything stressful occur since the last survey? A stressful event is any event, even a minor one, which negatively affected you.” Participants answered no (0) or yes (1). Considering the average time intervals between each beep, this current stressor variable represents the endorsement of stressful events that occurred within approximately 2.5 hours. We used a lagged stressor variable to examine the relatively prolonged effects of stressors on NA. The lagged stressor variable reflected whether participants reported a stressor at the previous (lag 1) beep. In the few cases when a previous assessment was missing, the observations were set to missing. Additionally, to exclude overnight time lag, all of the first observations of each day were set to missing.

Person-level covariates.

Age was centered at the sample mean. Gender (female=0, male=1) was dummy coded. Race/ethnicity was coded as follows: Caucasian, Hispanic White, Hispanic Black, Asian, Black, or other.

As loneliness is often related to objective social isolation (Hughes et al., 2004), measures of objective isolation were included as covariates. The objective social isolation measures, both marital status (0=single, 1=married) and living arrangement (0=living with others, 1=living alone) were dummy coded.

We utilized a short version (8 items, Cronbach reliability α = .93) of the PROMIS Depression scale to measure depressive symptoms. The sum of 8 items was transformed to a standardized score following PROMIS guidelines (PROMIS Health Organization, 2021).

Momentary covariates.

Time-varying covariates reflecting current location and recent activities were included because prior work has shown these variables to be related to negative affect (e.g., Scott et al., 2017). Current location was coded as follows: at home (reference category, in another person’s home, outside, vehicle, work/school, or elsewhere. Current activity indicated recent engagement in chores (the reference category), eating, physical activity, recreation, resting, self-care, socializing, work-related activity, or other activities.

2.4. Analytic Approach

We used multilevel logistic regression (SAS PROC GLIMMIX) to examine whether loneliness was associated with a likelihood of reporting stressors. The outcome variable was coded 1 if a stressful event was reported at a given momentary assessment, and 0 otherwise. The main predictor variables were terms for linear and quadratic loneliness. In model 1, we controlled for gender, age, race/ethnicity, marital status, living arrangement, current location, and current activity; in model 2, depressive symptoms were included to evaluate whether effects of loneliness were independent from depressive symptoms.

We used multilevel regression to test whether loneliness was related to both immediate and prolonged emotional responses. Models were fit using PROC MIXED in SAS with a spatial power covariance structure to adjust for the potential bias due to the serial dependency (or autocorrelation) of NA. The outcome variable was momentary NA and two primary predictor variables were used to represent momentary stressors: A dummy variable of current stressor indicating whether a stressor was reported on occasion i, and a dummy variable of lagged stressor indicating whether a stressor was reported on occasion i-1. We tested the interactions of current and lagged stressor with linear and quadratic loneliness to determine whether loneliness moderated current and prolonged emotional responses. To disentangle individual (between-person) differences in stressor exposure from momentary (within-person) stressor effects following prior works (e.g., Scott et al. 2013), a variable representing each person’s “stressor frequency” was calculated the proportion of all observations on which stressors were reported (range 0–1) and entered into the multilevel linear models. In model 1, we controlled for gender, age, race/ethnicity, marital status, living arrangement, current location and current activity; in model 2, depressive symptoms and its interactions with both current and lagged stressors were additionally controlled.

3. Results

Descriptive statistics and bivariate correlations at the person-level are presented in Table 1. Loneliness was correlated with greater depressive symptoms (r=0.59, p<.001), higher person-mean of stressors frequency (r=0.16, p<.05), and higher person-mean of NA (r=0.28, p<.001).

Table 1.

Descriptive statistics on person-level covariates and bivariate correlations with loneliness and covariates

Variables Mean (SD) 1 2 3 4 5 6 7 8
1. Age 46.48 (11.14) 1.00
2. Gender
(1=women)		 0.65 (0.48) −0.02 1.00
3. Marital Status
(1= married)		 0.31 (0.46) 0.19 −0.14* 1.00
4. Living arrangement
(1= living alone)		 0.24 (0.43) 0.12 0.01 −0.30*** 1.00
5. Depressive symptoms 0.02 (9.37) −0.05 0.04 −0.09 0.03 1.00
6. Loneliness 0.04 (9.90) −0.05 −0.02 −0.11 0.00 0.59*** 1.00
7. Person-mean stressor frequency 0.17 (0.16) 0.13* 0.06 −0.01 −0.03 0.21*** 0.16* 1.00
8. Person-mean NA 22.80 (15.99) −0.05 0.05 0.01 0.10 0.37*** 0.28*** 0.30*** 1.00
Open in a new tab

Note. N of persons= 255.

*

p < .05.

**

p < .01.

***

p < .001.

3.1. Is loneliness related to higher probability of stressors?

Multilevel logistic analyses showed that there was a small but significant linear association of loneliness (Odd ratio=1.030, p<.05, 95% C.I. [1.013–1.045]) with probability of stressors, controlling for demographics, objective isolation measures, and momentary covariates (model 1 in Table 2.) Examination of these odds ratios indicated that for a 1SD increase in loneliness (i.e. 10-unit increase) there were 34% increases in the odds of reporting a stressor (C.I. [1.144–1.560]). The association with quadratic loneliness was marginally significant (Odd ratio=0.999, p=.06, 95% C.I. [0.997–1.000]). Figure 1 shows that the expected probability of reporting stressors increased between low and middle range of loneliness (70th percentile) and reached a plateau. After controlling for depressive symptoms, the linear effect of loneliness was no longer significant (b=0.01, p>0.5), but the quadratic association become significant ( Odd ratio=0.999, p<.05, 95% C.I. [0.997–1.000]).

Table 2.

Multilevel logistic analysis predicting the likelihood of reporting stressors

Model 1
 Model 2
Estimate Odds ratio 95% CI Estimate Odds ratio 95% CI
Intercept −2.185*** 0.112 0.064–0.196 −2.211*** 0.110 0.063–0.190
Loneliness 0.029*** 1.030 1.014–1.046 0.011 1.011 0.992–1.030
Loneliness2 −0.001* 0.999 0.997–1.000 −0.001* 0.999 0.997–1.000
Depressive symptoms 0.033** 1.033 1.012–1.054
−2 Log Likelihood 77181.50 77217.73
Open in a new tab

Note. N of persons= 255. Model 1 adjusted for demographic information, objective isolation measures, and situational information (current location and current activity). Model 2 additionally adjusted for the depressive symptom variables.

*

p<.05

**

p<.01

***

p<.00

Figure 1.

Figure 1

Open in a new tab

Probability of reporting stressors by loneliness

Note. N of persons= 255. PCTL=Percentile

3.2. Does loneliness moderate the association of current and lagged stressors and NA?

Results from multilevel regression analyses showed that the association of concurrent stressors with NA was significant; participants reported significantly higher NA (b= 16.55, p<.001) at times when recent stressors were reported compared to occasions when no recent stressor was reported (model 1 in Table 3). The lagged stressor effect on NA was also significant; NA was significantly higher on occasions following a stressor report (b=3.16, p<.001). There was no interaction between concurrent and lagged stressors (b= 0.45, p>.05).

Table 3.

Multilevel analysis predicting momentary negative affect

Model 1
 Model 2
Estimate SE Estimate SE
Fixed effects        
Intercept 23.90*** 3.43 24.40*** 3.35
Stressor frequency (BP) 26.10*** 6.47 21.84*** 6.47
Current stressor (WP) 16.55*** 1.26 16.55*** 1.27
Lagged stressor (WP) 3.16*** 0.69 3.13*** 0.69
Current stressor X Lagged stressor (WP) 0.45 1.23 0.44 1.24
Loneliness 0.38*** 0.10 0.13 0.10
Loneliness2 0.00 0.01 0.00 0.01
Current stressor X Loneliness 0.12 0.10 0.11 0.12
Current stressor X Loneliness2 0.00 0.01 0.00 0.01
Lagged stressor X Loneliness 0.01 0.06 −0.02 0.08
Lagged stressor X Loneliness2 0.01* 0.00 0.01* 0.00
Depressive symptoms     0.47*** 0.13
Current stressor X Depressive symptoms     0.01 0.13
Lagged stressor X Depressive symptoms     0.05 0.08
Random effects        
Var (Intercept) 210.93*** 19.32 199.45*** 18.31
Var (Current Stress) 149.17*** 18.90 149.11*** 18.89
Var (Lagged Stress) 28.41*** 5.96 28.37*** 5.96
Covariance (Intercept X Current Stress) −33.15* 14.05 −34.45* 13.72
Covariance (Intercept X Lagged Stress) −5.85 7.99 −6.85 7.77
Covariance (Current Stress X Lagged Stress) 39.51*** 7.99 39.57*** 8.07
Spatial power covariance 11.39*** 1.89 11.41*** 1.89
Residual 160.75*** 2.27 160.75*** 2.27
−2LL 86588.1 86572.8***
AIC 86672.1 86662.8***
Open in a new tab

Note. N of persons= 255. Model 1 adjusted for demographic information, objective isolation measures, and situational information (current location and current activity). Model 2 additionally adjusted for the depressive symptom variables.

Note:

*

p<.05

**

p<.01

***

p<.001.

WP: within-person effect, BP: between-person effect.

There was a significant main effect of loneliness on NA (b=0.38, p<.001), such that people with higher loneliness showed higher average NA. The quadratic effect of loneliness was not significant (b=0.00, p>.05). There was no moderating effect of linear (b=0.12, p>.05) or quadratic loneliness (b=0.00, p>.05) on the association of concurrently reported stressors with NA (i.e., this association did not vary across levels of loneliness). There was a significant interaction between quadratic loneliness and lagged stressors (b=0.01, p<.05). To interpret the quadratic association between lagged stressors and loneliness, we calculated the lagged stressor effect for individuals at different levels of loneliness. We observed that the effect of lagged stressors for individuals in mid-levels of loneliness (e.g., coefficient of lagged stressors for 50th percentile = 3.16) was the smallest and increased for lower (e.g., coefficient of lagged stressors for 25th percentile = 3.61) and higher (e.g., 75th percentile = 3.78) loneliness levels. The interaction between quadratic loneliness and lagged stressors remained significant after controlling for depressive symptoms (model 2 in Table 3.) We conducted a sensitivity analysis by fitting a model that included NA from the prior occasion in the fixed effects part of the model: the critical interaction between loneliness and lagged stressor, remained significant (b=0.01, p<.05) using this model specification.

4. Discussion

Previous studies have suggested that loneliness correlates with stress-related health outcomes. This study aimed to clarify this relationship by examining how loneliness relates to the likelihood of reporting stressors and emotional responses to stressors in the context of everyday life.

4.1. Loneliness and likelihood of reporting daily stressors

We found a significant association between loneliness and greater likelihood of reporting stressors, controlling for demographic information, objective social isolation indicators, and momentary covariates. This result is consistent with prior daily diary studies (Russell et al., 2012; Wolf & Davis, 2014), which have demonstrated positive associations between loneliness and negative events. This result is broadly consistent with the notion that loneliness biases people to perceive their environment as threatening (Cacioppo & Hawkley, 2009; Hawkley & Cacioppo, 2010), which could lead to a lower threshold for perceiving and reporting neutral or mildly aversive events as “stressful”.

Despite the above association, we found that the linear association between loneliness and increased likelihood of reporting stressors was not independent of depressive symptoms. This result is not entirely surprising, as both loneliness and depression seem to share a similar cognitive mechanism – negative cognitive biases – that can lead to more frequent reports of stressors. Similar to loneliness, depression has been associated with negative attention biases, which include a tendency to focus on negatively valenced information (e.g., Gilboa-Schechtman et al., 2002), which may lead depressed people to perceive neutral or mildly negative events as threatening and stressful. However, we cannot determine whether the association between loneliness and stressor frequency is a byproduct of being depressed (or some other factor), or whether depression mediates the effect of loneliness. Regardless, the present results suggest that loneliness does not uniquely impact stressor frequency, and instead mechanisms common to both loneliness and depression (e.g., negative cognitive bias) may be responsible.

We found some limited evidence for non-linearity of the association between loneliness and likelihood of reporting stressors, the patterning of which was not entirely consistent with prior studies. Prior findings have suggested that severely lonely individuals (vs. not or moderately lonely) show a distinct behavioral tendency (e.g., greater attention to the socially threatening stimuli) (Qualter et al., 2013; Bangee et al., 2014); in contrast, our results showed that individuals with the lowest loneliness levels reported the fewest stressors, opposed to the mid- and high-level of loneliness. One possibility is that levels of loneliness in our sample were relatively low; thus, even individuals in the high range of loneliness in our sample might not have been “high enough” in loneliness to show distinctively different behavioral characteristics. As the quadratic association observed in the current study was relatively weak, this pattern should be verified further.

4.2. Loneliness and emotional responses to daily stressors

We found that loneliness was not related to concurrent effects of stressors on NA. This result is consistent with previous experimental (Ong et al., 2012), daily diary (Russell et al., 2012; Wolf & Davis, 2014), and EMA (Hawkley, et al., 2007) studies that have also found loneliness to be unrelated to immediate effects of stressors on affect. However, our results revealed an association of loneliness with emotional responses to lagged stressors. Individuals with high levels of loneliness evidenced a larger lagged stressor effect on their NA compared to moderately lonely individuals. One possible explanation is that loneliness leads to higher initial NA levels following stressors, which affects NA at the following assessment. This explanation is unlikely though, because we did not observe any difference in concurrent NA levels associated with stressors. Instead, the present results suggest that in the approximately 2.5 hours following report of a stressor, people high in loneliness may show less emotional recovery (i.e., they remain further from their emotional baseline) compared to people with average levels of loneliness. This result is consistent with those from laboratory studies that have demonstrated an association between loneliness and greater prolonged physiological stress responses during recovery (Hackett et al., 2012; Ong et al., 2012).

There are a few possible explanations for why loneliness moderated lagged but not concurrent stressor effects in the present work. One possible explanation is that an immediate increase in NA following stressors is normative (Scott et al., 2017; Stawski et al., 2008) and might more directly reflect objective features of the stressor (e.g., type and severity), such that loneliness might not amplify NA at or very near to the time that a stressor occurs (Hackett et al., 2012; Steptoe et al., 2004). At the same time, however, loneliness may impact the temporal extension of emotional stress responses by reducing the likelihood that positive reappraisals or social support come into play with the passage of time. Another possible explanation could be related to maladaptive cognitions, such as perseverative cognitions, which are manifested in phenomena such as worry, rumination, and anticipatory stress (Brosschot et al., 2005). Perseverative cognition can sustain the effect of stressors even in the absence of stressors (Smyth et al., 2013). In light of previous findings demonstrating that rumination is a mediator between loneliness and depressive symptoms (Vanhalst et al., 2012), lonely people’s perseverative cognitive activity (i.e., rumination) might also play a role in their prolonged emotional response to stressors.

Unexpectedly, prolonged negative emotional responses were also observed among individuals in the low range of loneliness. We did not have a priori expectation for this finding, but provide here some speculative possibilities that may help explain it. Among individuals in the low range of loneliness, there were some who endorsed “never” to all of the six items of the loneliness measure in this study (18% of the sample). It is possible that this may reflect under-reporting of loneliness, perhaps due to the social stigma attached to loneliness (Barreto et al., 2022), which may make people hesitant to admit to feeling isolated. Previous studies suggest that when individuals are asked to report their emotion with no clear timeframe (or too broad of a timeframe), they tend to report about their experiences of emotions based more on self-schema, or general beliefs about themselves, rather than based on their actual emotional experience (Robinson & Clore, 2002a, 2002b). Considering that we assessed loneliness using a retrospective survey that does not have a timeframe, our loneliness measure may be insensitive to recent status or feelings, as such some individuals might have endorsed that they never felt lonely at all not because they actually never felt lonely but because of their idea of themselves as a “not lonely person”. For example, some individuals who have been generally socially connected and satisfied with their social relationships but experienced recent critical changes that may increase loneliness, such as family loss, relocation, or unemployment, may experience but not report loneliness. We speculate that the prolonged responses shown among individuals in the low range of loneliness may be driven by these individuals who were uncharacteristically lonely. Future study may benefit from an EMA approach to measure more proximal, recent feelings of loneliness, to unpack the relationship between loneliness and prolonged stress responses.

4.3. Limitations & Conclusions

One limitation of the present study is that using the PROMIS Social isolation measure may not capture the full range of loneliness construct. Future studies that include more comprehensive and fine-grained assessments of the various aspects of loneliness (e.g., social and emotional loneliness), in conjunction with detailed measures of objective social isolation would be beneficial in providing a more complete understanding of the relationship between loneliness and stressAnother limitation is that we did not have the ability to measure intraindividual variability in loneliness, as we only measured loneliness once based on a retrospective survey. Even those with relatively high loneliness in this study may not always feel lonely and it might be times when people are feeling particularly lonely that the effect of loneliness is largest. Also, the current study included limited measures of social isolation, married and living alone, lacking consideration of the current social context that can be indicative of social isolation at the moment. By using an EMA approach to measure loneliness and social isolation (e.g., being alone vs. being with others; no social interaction vs. interacting with others), future studies would further examine how loneliness and social isolation, on a daily basis, distinctively or similarly affect stress experiences and processes. Another limitation is sample selectivity: Although we used probability sampling methods, socially isolated individuals or those most affected by loneliness may not have agreed to participate, and thus we could have underestimated or missed certain effects of loneliness.

Despite these limitations, the current study revealed that high levels of loneliness were not associated with emotional states shortly after reporting stressors, yet loneliness was associated with NA 3–5 hours following everyday stressors. A somewhat lower capacity to quickly recover from minor stressful events along with experiencing frequent stressors might play a role in understanding how loneliness contributes to adverse health outcomes associated with chronic stress. Without proper recovery, the effects of frequent daily stressors may pile up, with the potential to result in chronic stress and poorer health and well-being. Importantly, stress reactivity and stress recovery have been identified as two different ways in which stress can impact health (McEwen, 2000; Smyth et al., 2018), such that it is important to distinguish them. Although there have been some studies to examine loneliness and emotional reactivity in daily life, prior work has not distinguished between concurrent and prolonged responses. By exploring the shifts in emotional states following a stressor, the present work revealed that examining the inability to quickly recover from minor stressful events (rather than being over-reactive) along with experiencing frequent stressors might play a role in understanding how loneliness contributes to adverse health outcomes associated with chronic stress. Future research into such patterns is warranted and may help inform interventions as to what aspects of the stress response to target (e.g., recovery) to improve health and well-being.

Acknowledgments

This work was supported by the National Institute of Health (NIH) grants R01 AG039409 and P01-AG0394. Kang was partially supported by NIA T32AG049676 to The Pennsylvania State University.

Footnotes

Conflict of Interest

The authors have no conflict of interests.

Data accessibility

The data, analytic methods, and study materials on which the manuscript is based will be made available per request. This study is not preregistered.

References

  1. Almeida DM, Wethington E, & Kessler RC (2002). The daily inventory of stressful events: An interview-based approach for measuring daily stressors. Assessment, 9(1), 41–55. 10.1177/1073191102091006 [DOI] [PubMed] [Google Scholar]
  2. Alpass FM, & Neville S (2003). Loneliness, health and depression in older males. Aging & Mental Health, 7(3), 212–216. 10.1080/1360786031000101193 [DOI] [PubMed] [Google Scholar]
  3. Balter LJT, Raymond JE, Aldred S, Drayson MT, Veldhuijzen van Zanten JJCS, Higgs S, & Bosch JA (2019). Loneliness in healthy young adults predicts inflammatory responsiveness to a mild immune challenge in vivo. Brain, Behavior, and Immunity, 82, 298–301. 10.1016/j.bbi.2019.08.196 [DOI] [PubMed] [Google Scholar]
  4. Bangee M, Harris RA, Bridges N, Rotenberg KJ, & Qualter P (2014). Loneliness and attention to social threat in young adults: Findings from an eye tracker study. Personality and Individual Differences, 63, 16–23. 10.1016/j.paid.2014.01.039 [DOI] [Google Scholar]
  5. Barreto M, van Breen J, Victor C, Hammond C, Eccles A, Richins MT, & Qualter P (2022). Exploring the nature and variation of the stigma associated with loneliness. Journal of Social and Personal Relationships, 39(9), 2658–2679. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brosschot JF, Pieper S, & Thayer JF (2005). Expanding stress theory: Prolonged activation and perseverative cognition. Psychoneuroendocrinology, 30(10), 1043–1049. 10.1016/j.psyneuen.2005.04.008 [DOI] [PubMed] [Google Scholar]
  7. Brown EG, Gallagher S, & Creaven A (2018). Loneliness and acute stress reactivity: A systematic review of psychophysiological studies. Psychophysiology 55(5), e13031. 10.1111/psyp.13031 [DOI] [PubMed] [Google Scholar]
  8. Brown EG, Creaven A, & Gallagher S (2019). Loneliness and cardiovascular reactivity to acute stress in younger adults. International Journal of Psychophysiology, 135, 121–125. 10.1016/j.ijpsycho.2018.07.471 [DOI] [PubMed] [Google Scholar]
  9. Brown EG, Creaven AM, & Gallagher S (2022). Loneliness and cardiovascular reactivity to acute stress in older adults. Psychophysiology, 59(7), e14012. 10.1111/psyp.14012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Boss L, Kang DH, & Branson S (2015). Loneliness and cognitive function in the older adult: a systematic review. International Psychogeriatrics, 27(4), 541–553. 10.1017/S1041610214002749 [DOI] [PubMed] [Google Scholar]
  11. Cacioppo JT, Ernst JM, Burleson MH, McClintock MK, Malarkey WB, Hawkley LC, Kowalewski RB, Paulsen A, Hobson J. Allan, Hugdahl K, Spiegel D, & Berntson GG (2000). Lonely traits and concomitant physiological processes: The MacArthur social neuroscience studies. International Journal of Psychophysiology, 35(2–3), 143–154. 10.1016/S0167-8760(99)00049-5 [DOI] [PubMed] [Google Scholar]
  12. Cacioppo JT, Hawkley LC, Ernst JM, Burleson M, Berntson GG, Nouriani B, & Spiegel D (2006). Loneliness within a nomological net: An evolutionary perspective. Journal of research in Personality, 40(6), 1054–1085. 10.1016/j.jrp.2005.11.007 [DOI] [Google Scholar]
  13. Cacioppo JT, Hughes ME, Waite LJ, Hawkley LC, & Thisted RA (2006). Loneliness as a specific risk factor for depressive symptoms: Cross-sectional and longitudinal analyses. Psychology and Aging, 21(1), 140–151. 10.1037/0882-7974.21.1.140 [DOI] [PubMed] [Google Scholar]
  14. Cacioppo JT, & Hawkley LC (2009). Perceived social isolation and cognition. Trends in Cognitive Sciences, 13(10), 447–454. 10.1016/j.tics.2009.06.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Cella D, Riley W, Stone A, Rothrock N, Reeve B, Yount S, Amtmann D, Bode R, Buysse D, Choi S, Cook K, DeVellis R, DeWalt D, Fries JF, Gershon R, Hahn E, Lai J, Pilkonis P, Revicki D, … & PROMIS Cooperative Group (2010). The Patient-Reported Outcomes Measurement Information System (PROMIS) developed and tested its first wave of adult self-reported health outcome item banks: 2005–2008. Journal of Clinical Epidemiology, 63(11), 1179–1194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Charles ST, Piazza JR, Mogle J, Sliwinski MJ, & Almeida DM (2013). The Wear and Tear of Daily Stressors on Mental Health. Psychological Science, 24(5), 733–741. 10.1177/0956797612462222 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Christiansen J, Larsen FB, & Lasgaard M (2016). Do stress, health behavior, and sleep mediate the association between loneliness and adverse health conditions among older people? Social Science & Medicine, 152, 80–86. 10.1016/j.socscimed.2016.01.020 [DOI] [PubMed] [Google Scholar]
  18. Gaudin JM, Polansky NA, Kilpatrick AC, & Shilton P (1993). Loneliness, depression, stress, and social supports in neglectful families. American Journal of Orthopsychiatry, 63(4), 597–605. 10.1037/h0079475 [DOI] [PubMed] [Google Scholar]
  19. Gilboa-Schechtman E, Erhard-Weiss D, & Jeczemien P (2002). Interpersonal deficits meet cognitive biases: Memory for facial expressions in depressed and anxious men and women. Psychiatry Research, 113(3), 279–293. 10.1016/S0165-1781(02)00266-4 [DOI] [PubMed] [Google Scholar]
  20. Hackett RA, Hamer M, Endrighi R, Brydon L, & Steptoe A (2012). Loneliness and stress-related inflammatory and neuroendocrine responses in older men and women. Psychoneuroendocrinology, 37(11), 1801–1809. 10.1016/j.psyneuen.2012.03.016 [DOI] [PubMed] [Google Scholar]
  21. Hawkley LC, Burleson MH, Berntson GG, & Cacioppo JT (2003). Loneliness in everyday life: cardiovascular activity, psychosocial context, and health behaviors. Journal of Personality and Social Psychology, 85(1), 105. 10.1037/0022-3514.85.1.105 [DOI] [PubMed] [Google Scholar]
  22. Hawkley LC, Preacher KJ, & Cacioppo JT (2007). Multilevel modeling of social interactions and mood in lonely and socially connected individuals: The MacArthur social neuroscience studies. In Ong AD, & van Dulmen M (Eds.), Oxford handbook of methods in positive psychology (pp. 559–575). New York: Oxford University Press. [Google Scholar]
  23. Hawkley LC, Hughes ME, Waite LJ, Masi CM, Thisted RA, & Cacioppo JT (2008). From Social Structural Factors to Perceptions of Relationship Quality and Loneliness: The Chicago Health, Aging, and Social Relations Study. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 63(6), S375–S384. 10.1093/geronb/63.6.S375 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hawkley LC, & Cacioppo JT (2010). Loneliness Matters: A Theoretical and Empirical Review of Consequences and Mechanisms. Annals of Behavioral Medicine, 40(2), 218–227. 10.1007/s12160-010-9210-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Holt-Lunstad J, Smith TB, Baker M, Harris T, & Stephenson D (2015). Loneliness and Social Isolation as Risk Factors for Mortality: A Meta-Analytic Review. Perspectives on Psychological Science, 10(2), 227–237. 10.1177/1745691614568352 [DOI] [PubMed] [Google Scholar]
  26. Jaremka LM, Fagundes CP, Peng J, Bennett JM, Glaser R, Malarkey WB, & Kiecolt-Glaser JK (2013). Loneliness Promotes Inflammation During Acute Stress. Psychological Science, 24(7), 1089–1097. 10.1177/0956797612464059 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Laustsen LM, Christiansen J, Maindal HT, Plana-Ripoll O, Lasgaard M (2023). The longitudinal relation between loneliness and perceived stress: A structural equation modelling analysis of 10,159 individuals. Scandinavian Journal of Public Health, 0(0). 10.1177/14034948231151716 [DOI] [PubMed] [Google Scholar]
  28. Kiecolt-Glaser JK, Garner W, Speicher C, Penn GM, Holliday J, & Glaser R (1984). Psychosocial Modifiers of Immunocompetence in Medical Students: Psychosomatic Medicine, 46, 7–14. 10.1097/00006842-198401000-00003 [DOI] [PubMed] [Google Scholar]
  29. McEwen B (2000). Allostasis and Allostatic Load Implications for Neuropsychopharmacology. Neuropsychopharmacology, 22(2), 108–124. 10.1016/S0893-133X(99)00129-3 [DOI] [PubMed] [Google Scholar]
  30. Mroczek DK, Stawski RS, Turiano NA, Chan W, Almeida DM, Neupert SD, & Spiro A (2015). Emotional Reactivity and Mortality: Longitudinal Findings From the VA Normative Aging Study. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 70(3), 398–406. 10.1093/geronb/gbt107 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Nowland R, Robinson SJ, Bradley BF, Summers V, & Qualter P (2018). Loneliness, HPA stress reactivity and social threat sensitivity: Analyzing naturalistic social challenges. Scandinavian Journal of Psychology, 59(5), 540–546. [DOI] [PubMed] [Google Scholar]
  32. Ong AD, Rothstein JD, & Uchino BN (2012). Loneliness accentuates age differences in cardiovascular responses to social evaluative threat. Psychology and Aging, 27(1), 190–198. 10.1037/a0025570 [DOI] [PubMed] [Google Scholar]
  33. Parrish BP, Cohen LH, & Laurenceau J-P (2011). Prospective Relationship between Negative Affective Reactivity to Daily Stress and Depressive Symptoms. Journal of Social and Clinical Psychology, 30(3), 270–296. 10.1521/jscp.2011.30.3.270 [DOI] [Google Scholar]
  34. Patterson AC, & Veenstra G (2010). Loneliness and risk of mortality: A longitudinal investigation in Alameda County, California. Social Science & Medicine, 71(1), 181–186. 10.1016/j.socscimed.2010.03.024 [DOI] [PubMed] [Google Scholar]
  35. Pearlin LI, Menaghan EG, Lieberman MA, & Mullan JT (1981). The stress process. Journal of Health and Social behavior, 22(4), 337–356. 10.2307/2136676 [DOI] [PubMed] [Google Scholar]
  36. Peplau LA, & Perlman D (1982). Perspectives on loneliness. In Peplau LA & Perlman D (Eds.), Loneliness: A sourcebook of current theory, research and therapy, (pp. 1–18). New York: Wiley. [Google Scholar]
  37. Piazza JR, Charles ST, Sliwinski MJ, Mogle J, & Almeida DM (2013). Affective Reactivity to Daily Stressors and Long-Term Risk of Reporting a Chronic Physical Health Condition. Annals of Behavioral Medicine, 45(1), 110–120. 10.1007/s12160-012-9423-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. PROMIS Health Organization. (2012). PROMIS Scoring Manuals. Unpublished instrument Retrieved from https://www.healthmeasures.net/index.php?option=com_content&view=article&id=180&Itemid=994
  39. Qualter P, Brown SL, Rotenberg KJ, Vanhalst J, Harris RA, Goossens L, Bangee M, & Munn P (2013). Trajectories of loneliness during childhood and adolescence: Predictors and health outcomes. Journal of Adolescence, 36(6), 1283–1293. 10.1016/j.adolescence.2013.01.005 [DOI] [PubMed] [Google Scholar]
  40. Robinson MD, & Clore GL (2002) Episodic and semantic knowledge in emotional self-report: Evidence for two judgment processes. J Pers Soc Psychol 83(1). 198–215. 10.1037/0022-3514.83.1.198 [DOI] [PubMed] [Google Scholar]
  41. Robinson MD, & Clore GL (2002) Belief and feeling: Evidence for an accessibility model of emotional self-report. Psychol Bull 128(6). 934–960. 10.1037/0033-2909.128.6.934 [DOI] [PubMed] [Google Scholar]
  42. Russell A, Bergeman CS, & Scott SB (2012). Daily Social Exchanges and Affect in Middle and Later Adulthood: The Impact of Loneliness and Age. The International Journal of Aging and Human Development, 74(4), 299–329. 10.2190/AG.71076.4.4.c [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Scott SB, Sliwinski MJ, & Blanchard-Fields F (2013). Age differences in emotional responses to daily stress: the role of timing, severity, and global perceived stress. Psychology and aging, 28(4), 1076. 10.1037/a0034000 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Scott SB, Graham-Engeland JE, Engeland CG, Smyth JM, Almeida DM, Katz MJ, Lipton RB, Mogle JA, Munoz E, Ram N, & Sliwinski MJ (2015). The Effects of Stress on Cognitive Aging, Physiology and Emotion (ESCAPE) Project. BMC Psychiatry, 15(1), 146. 10.1186/s12888-015-0497-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Scott SB, Ram N, Smyth JM, Almeida DM, & Sliwinski MJ (2017). Age differences in negative emotional responses to daily stressors depend on time since event. Developmental Psychology, 53(1), 177–190. 10.1037/dev0000257 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Segrin C, & Passalacqua SA (2010). Functions of Loneliness, Social Support, Health Behaviors, and Stress in Association With Poor Health. Health Communication, 25(4), 312–322. 10.1080/10410231003773334 [DOI] [PubMed] [Google Scholar]
  47. Smyth J, Zawadzki M, & Gerin W (2013). Stress and disease: A structural and functional analysis. Social and Personality Psychology Compass, 7(4), 217–227. 10.1111/spc3.12020 [DOI] [Google Scholar]
  48. Smyth JM, Sliwinski MJ, Zawadzki MJ, Scott SB, Conroy DE, Lanza ST, Marcusson-Clavertz D, Kim J, Stawski RS, Stoney CM, Buxton OM, Sciamanna CN, Green PM, & Almeida DM (2018). Everyday stress response targets in the science of behavior change. Behaviour Research and Therapy, 101, 20–29. 10.1016/j.brat.2017.09.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Smyth JM, Zawadzki MJ, Marcusson-Clavertz D, Scott SB, Johnson JA, Kim J, Toledo MJ, Stawski RS, Sliwinski MJ, & Almeida DM (2022). Computing Components of Everyday Stress Responses: Exploring Conceptual Challenges and New Opportunities. Perspectives on Psychological Science, 17456916221082108. 10.1177/17456916221082108 [DOI]
  50. Stawski RS, Sliwinski MJ, Almeida DM, & Smyth JM (2008). Reported exposure and emotional reactivity to daily stressors: The roles of adult age and global perceived stress. Psychology and Aging, 23(1), 52–61. 10.1037/0882-7974.23.1.52 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Stephens C, Alpass F, Towers A, & Stevenson B (2011). The Effects of Types of Social Networks, Perceived Social Support, and Loneliness on the Health of Older People: Accounting for the Social Context. Journal of Aging and Health, 23(6), 887–911. 10.1177/0898264311400189 [DOI] [PubMed] [Google Scholar]
  52. Steptoe A, Owen N, Kunz-Ebrecht SR, & Brydon L (2004). Loneliness and neuroendocrine, cardiovascular, and inflammatory stress responses in middle-aged men and women. Psychoneuroendocrinology, 29(5), 593–611. 10.1016/S0306-4530(03)00086-6 [DOI] [PubMed] [Google Scholar]
  53. Steptoe A, Leigh ES, & Kumari M (2011). Positive affect and distressed affect over the day in older people. Psychology and Aging, 26(4), 956. 10.1037/a0023303 [DOI] [PubMed] [Google Scholar]
  54. Vanhalst J, Luyckx K, Raes F, & Goossens L (2012). Loneliness and Depressive Symptoms: The Mediating and Moderating Role of Uncontrollable Ruminative Thoughts. The Journal of Psychology, 146(1–2), 259–276. 10.1080/00223980.2011.555433 [DOI] [PubMed] [Google Scholar]
  55. Wolf LD, & Davis MC (2014). Loneliness, daily pain, and perceptions of interpersonal events in adults with fibromyalgia. Health Psychology, 33(9), 929–937. 10.1037/hea0000059 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Yu B, Steptoe A, Chen Y, & Jia X (2021). Social isolation, rather than loneliness, is associated with cognitive decline in older adults: the China Health and Retirement Longitudinal Study. Psychological medicine, 51(14), 2414–2421. [DOI] [PubMed] [Google Scholar]
  57. Zhang W, Liu L, Tang F, & Dong X (2018). Social engagement and sense of loneliness and hopelessness: Findings from the PINE study. Gerontology and Geriatric Medicine, 4, 2333721418778189. 10.1177/2333721418778189 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES