Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Genuine Smiles by Patients During Marital Interactions are Associated with Better Caregiver Mental Health

Genuine Smiles by Patients During Marital Interactions are Associated with Better Caregiver... Objective: Providing care for a spouse with dementia is associated with an increased risk for poor mental health. To deter- mine whether this vulnerability in caregivers is related to the expression of positive emotion, we examined 57 patients with Alzheimer’s disease and behavioral variant frontotemporal dementia and their spouses as they discussed a marital conflict. Method: Facial behavior during the discussion was objectively coded to identify Duchenne (i.e., genuine) smiles and non- Duchenne (i.e., polite) smiles. Caregiver mental health was measured using the Medical Outcomes Survey. Results: Greater expression of Duchenne smiles by patients was associated with better caregiver mental health, even when accounting for covariates (i.e., diagnosis, patient cognitive functioning, and caregiver marital satisfaction). Greater expression of non-Duchenne smiles by patients was associated with worse caregiver health, but only when covariates were entered in the model. Expression of Duchenne and non-Duchenne smiles by caregivers was not associated with caregiver mental health. Discussion: Patients’ expression of Duchenne and non-Duchenne smiles may reveal important aspects of the emotional quality of the patient–caregiver relationship that influence caregiver burden and mental health. Keywords: Caregiving, Dementia, Mental health, Positive emotion, Smiling. Familial caregivers of patients with dementia are faced focused on an important aspect of patient behavior, the with the enormous challenge of caring for a loved one expression of positive emotion that occurs during patient– whose level of functioning progressively deteriorates. As caregiver interactions, and its association with caregivers’ a group, dementia caregivers are more prone to mental mental health. health problems than non-caregiving older adults (Schulz & Eden, 2016). However, individual caregivers differ Patient Emotional Behavior, Patient– greatly in the extent to which they experience these nega- Caregiver Relationship Quality, and Caregiver tive outcomes. Prior research indicates that behavioral and Mental Health psychological symptoms in patients can be particularly dif- ficult for caregivers, even more so than cognitive and func- Patients with Alzheimer’s disease (AD) and behavioral tional symptoms (Schulz, O’Brien, Bookwala, & Fleissner, variant frontotemporal dementia (bvFTD) undergo a 1995). These behavioral symptoms in patients can be par- number of changes in emotional behavior that can have a ticularly challenging when manifest in interactions with negative impact on caregivers. For example, as AD patients’ caregivers (Ascher et  al., 2010). In the present study, we dementia progresses, problematic behaviors such as anger Published by Oxford University Press on behalf of The Gerontological Society of America 2018. This work is written by (a) US 975 Government employee(s) and is in the public domain in the US. Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 976 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 and aggression become increasingly difficult to manage, (Harker & Keltner, 2001). Greater expressions of love and leading to worse mental health in caregivers (Schulz & other positive emotions during a conflict are associated Sherwood, 2008). Similarly, as bvFTD progresses, patients with higher levels of relationship satisfaction and a greater undergo dramatic changes in personality and express likelihood of engaging in constructive behaviors (e.g., af- greater levels of apathy and social inappropriateness, which firmation, soothing contact, expressing concern; Gonzaga, leads to greater levels of burden, depression, and distress in Keltner, Londahl, & Smith, 2001). caregivers (Merrilees et al., 2013). Despite the established links between patient behav- Positive Emotional Behavior and Health ioral symptoms and worse caregiver mental health, stud- ies of caregiver health that have focused on qualities of In the literature on emotion and health, individuals who the patient–caregiver relationship have been rare. This experience greater positive emotions also experience bet- is unfortunate, because spouses and relationship part- ter health outcomes (e.g., Wichers et  al., 2007). Positive ners often are called on to serve as primary caregivers emotions are thought to help build personal resources for patients with dementia (Schulz & Eden, 2016), and (Fredrickson, 2004) which, in turn, have been associated dementia can profoundly affect these relationships. For with individuals’ ability to utilize coping strategies and be example, patients with bvFTD may become more emo- resilient (Fredrickson, Tugade, Waugh, & Larkin, 2003), tionally blunted and less empathic (Rascovsky et  al., both of which are important for mental health. Although 2011), leading to a weakening of the emotional con- rarely examined, the link between positive emotion and nection with their partner. In previous studies of dyadic health is also found in close relationships. For example, interactions, patients with FTD and their spousal car- individuals with happier spouses report better health, fewer egivers showed less mutual gaze during conversations physical impairments, and less chronic disease (Chopik about a relationship problem and reported lower marital & O’Brien, 2016). Exhibiting and eliciting positive emo- satisfaction than healthy controls (Sturm et  al., 2011). tions fosters supportive environments that promote cop- Alzheimer’s disease can also affect relationships. For ex- ing (Fredrickson, 2004), and smiling has been found to be ample, a longitudinal study found that caregivers’ social particularly effective for eliciting greater cooperation from intimacy with the patient declined over time (Blieszner others (Johnston, Miles, & Macrae, 2010). In the realm & Shifflett, 1990). Loss of closeness between the patient of caregiving where one spouse has dementia, patients’ and caregiver can be devastating, especially in later life positive communication behaviors (e.g., humor) have been when social networks shrink and close relationships be- linked to lower depression in the caregiving spouse (Braun, come increasingly more important (Cornwell, Laumann, Mura, Peter-Wight, Hornung, & Scholz, 2010). & Schumm, 2008). Thus, not surprisingly, poor rela- tionship quality has been strongly associated with worse Measuring Positive Emotion caregiver mental health (Mahoney, Regan, Katona, & Livingston, 2005). Emotions can be measured via self-reported subjective experience, expressive behavior, and peripheral and cen- tral nervous system physiology (Levenson et  al., 2017). Positive Emotional Behavior and Intimate Emotional facial expressions, with their high signal value Relationships for conspecifics, may be particularly important indicators Positive emotional behaviors are particularly important of the emotional life of couples (Levenson, Haase, Bloch, indicators of the state of intimate relationships. Consistent Holley, & Seider, 2013). Although there are a number with this, a primary function of positive emotions is to of different positive emotions (e.g., pride, amusement, broaden and build social relationships (Fredrickson, contentment), most positive emotions share the smile as 2004). Experiencing positive emotions has been linked a common expressive element (Campos, Shiota, Keltner, to stronger social bonds and social connection (Losada Gonzaga, & Goetz, 2013). Moreover, particular morpho- & Heaphy, 2004), and encourages individuals to explore logical features of the smile are thought to convey whether and engage in new experiences, which increase social inte- the emotion is genuinely felt (i.e., “Duchenne smiles,” gration (Fredrickson, 2004). For example, when mothers which involve the raising of the lip corners and the rais- look at photographs of their infants smiling, dopaminergic ing of the cheeks) or not genuinely felt (“non-Duchenne” reward-related brain areas show greater activation than or “polite” smiles, which only involve the raising of lip when infants’ neutral or sad expressions are viewed; this re- corners; Ekman & Friesen, 1982). In dementia research, inforcement associated with smiles may play an important assessing emotion via facial expressions has the additional role in building mother–infant attachments (Strathearn, advantage of reducing the problems associated with retro- Li, Fonagy, & Montague, 2008). Positive emotions are spective self-reports of emotion in patients who may have also important in romantic relationships; people who ex- deficits in language, memory, and self-awareness. Despite press greater happiness in college yearbook photographs these advantages, we are aware of no prior studies of the report higher levels of marital satisfaction 30  years later associations between patients’ emotional behaviors and Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 977 caregivers’ mental health that have utilized objective cod- through the Memory and Aging Center at the University ing of emotional facial expressions during patient–car- of California, San Francisco. Patients were evaluated by egiver interactions. a multidisciplinary team and diagnosed based on neu- rological, neuropsychological, and neuroimaging data using consensus criteria for AD (McKhann et al., 1984) The Present Study and FTD (Rascovsky et al., 2011). Patients were gener- ally in the early stage of their disease (see Mini-Mental We studied positive emotional facial expressions (Duchenne State Examination [MMSE] scores in Table 1) and thus and non-Duchenne smiles) that occurred during the first were able to understand and follow task instructions (as 30 s of a 10-min discussion of an area of marital conflict confirmed by verbal checks with session facilitators). Both in a sample of patients with dementia and their spousal patients and caregivers needed to be sufficiently healthy caregivers. Expressions were coded using the Facial Action to travel to the University of California, Berkeley to com- Coding System (FACS, Ekman & Friesen, 1977), a precise plete the day-long laboratory session. Patients who met anatomically based system that can decompose observable criteria for Mild Cognitive Impairment (i.e., patients with facial behavior in terms of its underlying muscular con- cognitive impairments that were not significant enough tractions. The first 30 s of the interaction were chosen be- to interfere with daily activities) were excluded. Table cause they capture a period of reconnection after couples 1 presents demographic characteristics for patients and had sat for 5  min of enforced silence. In healthy couples, caregivers. All couples were paid $30 in addition to any these “reunion” periods are highly diagnostic of the quality transportation costs they incurred. of the relationship. For example, after a similar period of enforced silence, greater positive emotion expressed during the first 3 min of a discussion of an area of marital conflict Procedure predicted lower likelihood of divorce (Carrère & Gottman, 1999). Focusing on only 30 s of facial behavior has practical A week before their laboratory visit, caregivers completed advantages as well. FACS coding is highly time-consuming; a questionnaire packet including measures of mental when applied thoroughly, it typically takes 100 min to code health and marital satisfaction. Patients and caregivers 1 min of behavior, and it is a slow process even when only then came to the Berkeley Psychophysiology Laboratory coding for Duchenne and non-Duchenne smiles. Moreover, for a day-long comprehensive assessment of emotional studying “thin slices” of behavior (typically ranging from and social functioning. Upon arrival, participants were 10 to 30  s) is a well-established procedure for capturing informed that their physiological, behavioral, and self- important qualities of individuals and dyads (Ambady, reported responses would be recorded and videotaped. Bernieri, & Richeson, 2000). Prior to the start of the laboratory session, participants Because different forms of dementia affect different brain had sensors attached for monitoring autonomic and regions and have different effects on emotional functioning somatic physiological responses (these data were not used (Seeley et al., 2007), we included patients with two common for the present study). Throughout the session, partici- forms of dementia: AD—a dementia that affects the tempo- pants’ upper body and face were filmed using a partially ral and parietal lobes and primarily impairs memory and concealed video camera. cognition, and bvFTD—a dementia that affects the frontal The present study focused on a laboratory task in which and temporal lobes and primarily produces changes in emo- couples sat quietly during a 5-min baseline period and then tion, personality, and behavior. Among the various subtypes had an unrehearsed discussion about an area of marital con- of frontotemporal dementia, we focused on bvFTD because flict (i.e., conflict conversation) for 10 min. This procedure language problems associated with other subtypes (i.e., was originally developed for studying marital interactions semantic variant primary progressive aphasia, non-fluent in healthy couples (Levenson & Gottman, 1983), but has variant primary progressive aphasia) could interfere with also been used with dementia patients and their caregivers patients’ ability to engage in the conflict conversation. (Ascher et al., 2010; Sturm et al., 2011). The discussion Our primary hypothesis was that more Duchenne smiles occurred for each couple immediately after an hour-long (thought to indicate genuine positive emotion; Ekman & break for lunch; anecdotally, many couples reported having Friesen, 1982) expressed by patients and their spousal car- good energy levels at this point in the day. egivers would be associated with better caregiver mental health. We did not expect this relationship to be found for Measures non-Duchenne smiles. Positive emotional expressions Using the Facial Action Coding System (FACS; Ekman & Method Friesen, 1977), trained coders blind to diagnosis and care- Participants giver outcomes measured smiling behaviors expressed by Twenty-nine patients with AD and 28 patients with both spouses during the first 30  s of the 10-min conflict bvFTD and their spousal caregivers were recruited conversation. Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 978 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 Table 1. Sociodemographic Characteristics of Patients with Behavioral Variant Frontotemporal Dementia, Alzheimer’s Disease, and Patients’ Spousal Caregivers Spousal caregivers of Spousal caregivers of Patients with AD patients with AD Patients with bvFTD patients with bvFTD n 29 29 28 28 Age (M [SD]) 61.92 (8.89) 61.08 (8.77) 61.78 (8.22) 60.21 (8.04) Sex 15 M, 14 F 13 M, 16 F 20 M, 8 F 9 M, 19 F Smiles Duchenne Frequency 1.03 (1.30) 0.97 (1.13) 0.97 (1.11) 1.04 (1.05) Intensity 1.39 (1.67) 1.51 (1.70) 0.98 (1.45) 1.01 (1.44) Duration 4.72 (5.93) 3.97 (5.41) 4.69 (6.41) 3.74 (4.60) Composite 0.04 (0.99) 0.03 (0.93) −0.05 (0.90) −0.03 (0.91) Non-Duchenne Frequency 1.26 (1.48) 1.19 (1.29) 0.88 (1.17) 1.19 (1.15) Intensity 1.16 (1.14) 1.43 (1.41) 0.99 (1.04) 1.35 (1.05) Duration 3.88 (4.63) 5.52 (7.58) 3.04 (4.17) 4.79 (5.97) Composite 0.08 (1.03) −0.02 (0.99) −0.08 (0.89) 0.02 (0.88) Marital satisfaction — 111.65 (30.71) — 96.40 (28.25) MMSE (M [SD]) 21.97 (5.12) — 25.07 (4.12) — Mental Health (M [SD]) — 0.35 (0.91) — −0.36 (1.12) Note. For smiles, three raw subscores are presented: frequency (number of Duchenne smiles), intensity (average intensity of the Duchenne smile), and duration (number of seconds during the film during which there is a Duchenne smile). Logarithmic transformations were applied to the subscores to reduce skewness. Subscores were then normalized (using the means and standard deviations for the entire sample) and averaged to derive the score for the two kinds of smiles (see Keltner & Bonanno, 1997). Higher scores on marital satisfaction indicate greater satisfaction (range: 2–158); lower scores on MMSE indicate lower cognitive functioning (range: 0–30). A dash (—) indicates that the given variable was not measured. AD = Alzheimer’s disease; bvFTD = behavioral variant frontotemporal dementia; MMSE = Mini-Mental State Exam. To develop reliability, all coders completed practice cod- with negative emotions generally are not deemed to be ing assignments of older adults’ emotional behavior and signs of genuine positive emotion. Thus, smiles accom- met weekly to discuss discrepancies before coding the reli- panied by action units typically associated with negative ability sample of 18 dyads (30% of the sample). High reli- emotions (e.g., disgust, sadness, anger) such as AU9 (nose ability was required for the practice coding assignments wrinkling), AU15 (frowning), or AU4 (brow furrowing) (Cronbach’s α > 0.70) before coders completed their work were not included in the analyses. These exclusionary for the remaining dyads. criteria have been used in previous studies of Duchenne In order to ensure that individual differences in facial smiles (Haase et al., 2015). features (e.g., elasticity) were taken into account when For each participant, one score was derived to charac- coding, all coders were provided with neutral stills of each terize Duchenne smiling and a second score was derived patient and caregiver, and instructed to only code facial to characterize non-Duchenne smiling. To produce these expressions that clearly resulted from changes in specific scores, we first computed three subscores for each kind facial muscle movements. FACS is a well-validated and of smile: (a) the number of smiles that occurred during widely used measure (Ekman & Rosenberg, 2005), and has the conversation; (b) the average intensity of all smiles been successfully utilized in prior studies examining facial at their apex or most intense level of facial action (based expressions in older adults and adults with neurodegenera- on 1–5 intensity of AU12), and (c) the average duration tive diseases (Lints-Martindale, Hadjistavropoulos, Barber, of all smiles. Our measures of Duchenne and non-Duch- & Gibson, 2007). enne smiles were moderately skewed (skew ≥1). In order Based on prior research (Ekman & Friesen, 1982), to reduce skewness, a constant of one was added to each smiles were classified as genuine Duchenne smiles if they subscore (i.e., frequency, duration, intensity) before loga- included contraction of both the orbicularis oculi (cheek rithmic transformations were applied. This reduced the raise, action unit [AU]6 and/or AU7) and zygomatic major skew of each variable to ≤1, providing a more normal (lip corner raise, AU12) or as “polite” non-Duchenne distribution for analyses. Subscores were then normalized smiles if they consisted only of the contraction of the (using the means and standard deviations for the entire zygomatic major (AU12). Duchenne and non-Duchenne sample) and averaged to derive the score for the two kinds smiles with an opened mouth (AU25 or AU26) were also of smiles (see Keltner & Bonanno, 1997). To determine included. Smiles that are accompanied by AUs associated inter-rater reliability, two FACS-certified coders and two Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 979 trained FACS coders scored 18 conversations. Inter-rater whether caregivers differed in age, sex, marital satisfaction, reliability for the Duchenne and non-Duchenne scores was Duchenne smiles, and non-Duchenne smiles. high (Cronbach’s α = 0.84). For our primary hypothesis that Duchenne smiles by patients and caregivers would be associated with better Caregiver mental health caregiver mental health, we conducted a linear regres- Caregiver mental health was assessed using the Medical sion analysis in which the four kinds of smiles (patient Outcomes Study (MOS SF-36; Tarlov et  al., 1989). The and caregiver Duchenne and non-Duchenne) were entered MOS SF-36 is a 36-item self-report measure designed to together on the same step as predictors and caregiver men- assess eight health domains: (a) physical functioning, (b) tal health was the dependent variable. This analysis was role limitations due to physical health, (c) energy/fatigue, then re-run with diagnosis included as a covariate, reflect- (d) pain, (e) general health problems, (f) role limitations ing prior findings that indicate that symptoms associated due to emotional problems, (g) emotional well-being, and with bvFTD and AD create different kinds of burdens (h) social functioning. Scores for each of these domains and challenges for caregivers (Nunnemann, Kurz, Leucht, ranged from 0 (worst) to 100 (best). According to con- & Diehl-Schmid, 2012). When associations were found vention (Ware, 2000), all subscales were z-scored and between a particular kind of smile and caregiver mental weighted to create a composite score of mental health health, we conducted regression analyses that included the such that mental health subscales are weighted more heav- interaction of that kind of smile with diagnosis. Finally, to ily than other subscales; this weighting reduces but does evaluate the robustness of our findings, we repeated the not eliminate the influence of physical health on caregiv- main regression analysis while accounting for variables ers’ mental health scores. This scoring scheme has been that differed across diagnostic groups (i.e., marital satis- used extensively in the literature (Bourke-Taylor, Pallant, faction, cognitive functioning; see Preliminary Analyses). Law, & Howie, 2012; Hawthorne, Osborne, Taylor, & This analysis revealed a potential suppressor effect, thus, Sansoni, 2007) and its reliability and validity for predict- additional post hoc analyses were conducted to under- ing mental health is well-established (McHorney, Ware, stand this effect. An additional exploratory analysis was & Raczek, 1993). conducted to examine whether marital satisfaction medi- ated the relationship found between patient Duchenne smiles and caregiver mental health. Marital satisfaction Caregivers completed the Locke–Wallace Marital Adjustment Test (Locke & Wallace, 1959), a well-validated Preliminary Analyses 15-item scale (e.g., “To what extent do you and your mate agree or disagree on handling family finances” [0 = Always Age and sex differences disagree, 5 = Always agree]) that we have used in prior re- Group differences in AD and bvFTD patients’ age were search with dementia patients and caregivers (Ascher et al., analyzed using an independent t-test; sex differences were 2010). Scores can range between 2 and 158, with higher analyzed using a Chi-square test. Results revealed no scores indicating greater marital satisfaction. 2 patient differences in age, t(55) = 0.07, p = .948 or sex, χ (1, N=57) = 2.33, p = .127. Parallel analyses were conducted to Cognitive functioning examine group differences for caregivers. Results revealed Patients’ cognitive functioning was assessed using the no caregiver differences in age, t(55) = 0.39, p = .699 or sex, MMSE (Folstein, Folstein, & McHugh, 1975), a well- χ (1, N = 57) = 0.97, p = .325. validated 30-item test that measures memory, orienta- tion, attention, and language. Items were summed, with Duchenne and non-duchenne smiles higher scores indicating higher cognitive functioning. Group differences in bvFTD and AD Duchenne and non- When interpreting scores, sums between 25 and 30 indi- Duchenne smiles were examined using independent t-tests. cate questionable cognitive impairment, 20–25 indicate No group differences were found in patient Duchenne mild cognitive impairment, 10–20 indicate moderate t(55) = 0.36, p = .721 and non-Duchenne smiles t(55) = 0.63, cognitive impairment, and 0–10 indicate severe cognitive p = .530, or in caregiver Duchenne t(55) = 0.24, p = .809 impairment. and non-Duchenne smiles t(55) = −0.15, p = .881. To examine correlations between patient and caregiver Duchenne smiles, Pearson’s r and Spearman’s rho were Results both used. Spearman’s rho was included because behav- Data Analysis ioral data often shows non-normal distributions, and Preliminary analyses were conducted to examine whether thus may benefit from a non-parametric analysis. Results patients with bvFTD and AD differed in age, sex, cogni- indicated a significant association between patient and tive functioning, Duchenne smiles, and non-Duchenne caregiver Duchenne smiles (r  =  0.38, p  =  .004; r   =  0.38, smiles. Similar analyses were conducted to examine p = .003), but no significant association between patient Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 980 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 and caregiver non-Duchenne smiles (r  =  0.21, p  =  .120; also significantly associated with caregiver mental health, r  = .20, p = .131. B = −0.71, SE(B) = 0.26, β = −0.34, p = .009, CI = [−1.24, −0.18]; caregivers of patients with AD reported having bet- ter mental health than caregivers of patients with bvFTD. Marital satisfaction Patient and caregiver non-Duchenne smiles were not asso- Group differences in caregivers’ self-reported marital ciated with caregiver mental health (ps > 0.159). Model fit satisfaction were analyzed using an independent t-test. was significant (R  = 0.144, F(5, 51) = 2.89, p = .023). Differences in caregiver marital satisfaction approached adjusted Because we observed an association between patient significance across groups, t(55)  =  1.95, p  =  .056, with Duchenne smiles and caregiver mental health, we deter- caregivers of patients with bvFTD reporting lower mari- mined whether this relationship varied by diagnosis by fit- tal satisfaction scores (M  =  96.40, SD  =  28.25) than car- ting another regression model that included the interaction egivers of patients with AD (M  =  111.65, SD  =  30.71). of patient Duchenne smiles and diagnosis. As seen in Table 2 Although group differences only approached significance, (model 3), analyses revealed that patient Duchenne smiles caregiver marital satisfaction was entered as a covariate in (B  =  0.38, SE(B)  =  0.16, β  =  0.33, p  =  .027, CI  =  [0.05, our analyses given prior findings of links between marital 0.70]) and patient diagnosis remained significantly asso- satisfaction and caregiver mental health (Kouros, Papp, & ciated with better caregiver mental health (B  =  −0.71, Cummings, 2008). SE(B)  =  0.27, β  =  −0.34, p  =  .010, CI  =  [−1.25, −0.18]), but no association was found between the interaction term Cognitive functioning (patient Duchenne smiles × diagnosis) and caregiver men- Group differences in AD and bvFTD patients’ Mini- tal health (B  =  −0.03, SE(B)  =  0.14, β  =  −0.03, p  =  .850, Mental State Exam (MMSE) scores were analyzed using an CI = [−0.31, 0.26]). The associations between patient non- independent t-test. Patient groups differed significantly in Duchenne smiles and both caregiver smiles (Duchenne and MMSE, t(55) = −2.52, p = .015, with patients with bvFTD non-Duchenne) and caregiver mental health also remained having higher scores (M = 25.07, SD = 4.12) than patients non-significant (ps > .166). Model fit remained significant with AD (M  =  21.97, SD  =  5.12). Consequently, patient (R  = 0.128, F(6, 50) = 2.37, p = .043). MMSE scores were entered as a covariate in our analyses. adjusted To examine the robustness of our findings, we conducted the main analysis while also accounting for patient diag- Duchenne smiles and caregiver mental health nosis and two additional variables that differed between As depicted in Table  2 (model 1), the regression ana- diagnostic groups—cognitive functioning (i.e., MMSE) and lysis revealed that patient Duchenne smiles were signifi- caregiver marital satisfaction. As depicted in Table 2 (model cantly associated with caregiver mental health, B  =  0.39, 4), results indicated that patient Duchenne smiles remained SE(B) = 0.17, β = 0.34, p = .029, CI = [0.05, 0.73] whereas significantly associated with better caregiver mental health caregiver Duchenne smiles were not, B = −0.08, SE(B) = 0.18, (B  =  0.42, SE(B)  =  0.15, β  =  0.37, p  =  .008, CI  =  [0.12, β = −0.07, p = .669, CI = [−0.43, 0.28]. When this analysis 0.72]). In addition, patient non-Duchenne smiles became was repeated with diagnosis included as a covariate (see associated with worse caregiver mental health (B = −0.30, Table  2, model 2), these results remained significant for SE(B)  =  0.14, β  =  −0.27, p  =  .036, CI  =  [−0.57, −0.02]). patient Duchenne smiles (B = 0.37, SE(B) = 0.16, β = 0.32, Marital satisfaction was also significantly associated with p  =  .025, CI  =  [0.05, 0.69]) while caregiver Duchenne caregiver mental health (B = 0.01, SE(B) = 0.00, β = 0.38, smiles remained non-significant (B =  −0.09, SE(B) = 0.17, 2,3 p = .003, CI = [0.005, 0.022]. while diagnosis showed an β  =  −0.07, p  =  .609, CI  =  [−0.42, 0.25]). Diagnosis was 1. The moderate correlation between patient and care- 2. Studies have also computed the MOS-SF 36 mental giver Duchenne smiles suggests that patients who expressed health composite scale by summing the following four more genuine smiles have caregivers who do the same. To subscales: vitality, social functioning, limitations due examine whether patients and caregivers who expressed to emotional problems, and mental health (Zhu et  al., more genuine smiles may also report greater caregiver 2016). When analyses were conducted using this alter- marital satisfaction and mental health, a median split native approach, results indicated that patient Duchenne was computed in order to group patients who expressed smiles were associated with caregiver mental health at fewer Duchenne smiles and patients who expressed greater near significant levels B =25.65, SE( B) = 13.06, β = 0.29, Duchenne smiles. Group differences in caregiver marital p  =  .055, CI  =  [−0.56, 51.86], but caregiver Duchenne satisfaction and mental health were then examined using smiles, and both patient and caregiver non-Duchenne independent t-tests. Analyses revealed a non-significant dif- smiles were not (ps > .203). When accounting for patient ference in marital satisfaction t(55)  =  0.14, p  =  .892 and cognitive functioning and caregiver marital satisfac- a difference that trended towards significance in caregiver tion, patient Duchenne smiles remained positively asso- mental health t(55) = −1.94, p = .058, such that caregivers ciated with better caregiver mental health (B  =  29.57, of patients who expressed more Duchenne smiles reported SE(B)  =  12.22, β  =  0.33, p  =  .019, CI  =  [5.01, 54.12]) higher mental health scores (M = 0.27, SD = 0.77) than car- while patient non-Duchenne smiles remained associ- egivers of patients who expressed fewer Duchenne smiles ated with worse caregiver mental health (B  =  23.33, (M = −0.26, SD = 1.26). SE(B) = 11.12, β = −0.27, p = .041, CI = [−45.67, −0.99]). Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 981 Table 2. Duchenne and non-Duchenne Smiles as Predictors of Caregiver Mental Health Caregiver mental health Model 1 Model 2 Model 3 Model 4 Model 5 B (SE[B]) β B (SE[B]) β B (SE[B]) β B (SE[B]) β B (SE[B]) β Patient Duchenne 0.39 (0.17) 0.34* 0.37 (0.16) 0.32* 0.38 (0.16) 0.33* 0.42 (0.15) 0.37** 0.37 (0.16) 0.33* Caregiver −0.08 (0.18) −0.07 −0.09 (0.17) −0.07 −0.08 (0.17) −0.07 −0.11 (0.16) −0.09 −0.10 (0.17) −0.09 Duchenne Patient −0.17 (0.15) −0.16 −0.21 (0.14) −0.18 −0.20 (0.15) −0.18 −0.30 (0.14) −0.27* −0.20 (0.15) −0.18 non-Duchenne Caregiver −0.12 (0.18) −0.11 −0.09 (0.17) −0.08 −0.10 (0.18) −0.09 −0.12 (0.16) −0.11 −0.09 (0.17) −0.08 non-Duchenne Diagnosis — — −0.71 (0.26) −0.34** −0.71 (0.27) −0.34** −0.53 (0.27) 0.25 — — Patient Duchenne — — — — −0.03 (0.14) −0.03 — — — — × diagnosis Patient non- — — — — — — — — −0.05 (0.15) −0.05 Duchenne × diagnosis Cognitive — — — — — — 0.004 (0.03) 0.02 — — functioning Marital — — — — — — 0.01 (0.00) 0.38** — — satisfaction Note. Covariates include patient diagnosis and cognitive functioning, and caregiver marital satisfaction. Diagnosis was dummy-coded; patients with AD were set as the reference group. A dash (—) indicates that the given variable was not included within the model. *p < .05, **p < .01. association that trended towards significance (B  =  −0.53, fit remained significant (R   =  0.255, F(7, 49)  =  3.74, adjusted SE(B)  =  0.27, β  =  −0.25, p  =  .052, CI  =  [−1.06, 0.004]). p = .003). Cognitive functioning was not associated with caregiver Because we observed an association between patient non- mental health (B = 0.004, SE(B) = 0.03, β = 0.02, p = .871, Duchenne smiles and caregiver mental health, we determined CI  =  [−0.05, 0.06]). The associations between caregiver whether this relationship varied by diagnosis by fitting another Duchenne and non-Duchenne smiles and caregiver men- regression model that included the four predictor variables tal health remained non-significant (ps  >  .439). Model (patient and caregiver Duchenne and non-Duchenne smiles) The associations between caregiver Duchenne and non- included both patient and caregiver Duchenne and non- Duchenne smiles and caregiver mental health remained Duchenne smiles). Patient Duchenne smiles remained sig- non-significant (ps > .628). nificantly associated with caregiver mental health, B = 0.39, 3. Same-sex spouses may reveal unique differences due to SE(B) = 0.17, β = 0.34, p = .028, CI = [0.04, 0.73] while care- gender, as women tend to smile more than men (LaFrance, giver Duchenne smiles remained non-significant, B = 0.05, Hecht, & Paluck, 2003), or additional stressors due to soci- SE(B) = 0.20, β = 0.04, p = .813, CI = [−0.36, 0.45]. Patient etal stigma (Lewis, Derlega, Griffin, & Krowinski, 2003). and caregiver non-Duchenne smiles likewise remained non- To examine whether same-sex couples impacted our find- significant (ps  >  .299). When analyses included length of ings, we re-ran analyses without these two couples. Results union as an additional covariate in Model 4 (where pre- remained significant; patient Duchenne smiles remained dictors included both patient and caregiver Duchenne and significantly associated with caregiver mental health, non-Duchenne smiles, as well as the covariates of patient B = 0.38, SE(B) = 0.16, β = 0.33, p = .023, CI = [0.05, 0.71] diagnosis and cognitive functioning, and caregiver mari- while caregiver Duchenne smiles were not, B  =  −0.11, tal satisfaction), patient Duchenne smiles remained signifi- SE(B) = 0.17, β = −0.10, p = .512, CI = [−0.46, 0.23]. The cantly associated with caregiver mental health, B  =  0.43, relationship between patient non-Duchenne smiles and SE(B)  =  0.16, β  =  0.37, p = .009, CI = [0.11, 0.74] while caregiver non-Duchenne smiles and caregiver mental health caregiver Duchenne smiles remained non-significant, remained non-significant (ps  >  .096) Results may also be B = −0.01, SE(B) = 0.18, β = −0.01, p = .974, CI = [−0.38, affected by the length of time spouses have been married, as 0.36]. Patient diagnosis B = −0.62, SE(B) = 0.28, β = −0.29, shorter times indicate newer spouses or second marriages, p = .030, CI = [−1.18, −0.06] and caregiver marital satisfac- which have been found to affect marital quality and sat- tion B = 0.01, SE(B) = 0.01, β = 0.34, p = .013, CI = [0.003, isfaction (Bograd & Spilka, 1996; Coleman, Ganong, & 0.02] remained significantly associated with caregiver men- Fine, 2000). To examine whether length of union impacted tal health, while patient and caregiver non-Duchenne smiles our findings, results were examined with length of union and patient cognitive functioning remained non-significant included as a covariate in Model 1 (where predictors (ps > .072). Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 982 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 and the interaction of patient non-Duchenne smiles and diag- cognitive functioning, and caregiver marital satisfaction), nosis. As seen in Table  2 (model 5), the analysis revealed a more patient Duchenne smiles remained a significant pre- non-significant interaction term (B  =  −0.04, SE(B)  =  0.15, dictor of better caregiver mental health, and fewer patient β = −0.05, p = .746, CI = [−0.35, 0.25]), which remained non- non-Duchenne smiles became a significant predictor of significant when the analysis was repeated with additional worse caregiver mental health. Further analysis revealed covariates included (i.e., diagnosis, martial satisfaction, and that marital satisfaction, but not patient cognitive function- cognitive functioning; B  =  −0.01, SE(B)  =  0.14, β  =  −0.01, ing, provided a likely suppressor effect in the association p = .945, CI = [−0.28, 0.26]). between greater patient non-Duchenne smiles and worse To further understand the potential suppressor effect in caregiver mental health. Finally, an exploratory analysis the association between patient non-Duchenne smiles and revealed that the association between patient Duchenne caregiver mental health, two additional post hoc regressions smiles and caregiver health was not mediated by caregiver were conducted. In the first, the four kinds of smiles (patient marital satisfaction. and caregiver Duchenne and non-Duchenne) and patient Although analyses revealed that marital satisfaction was cognitive functioning were entered as predictors, and car- not a significant mediator, both marital satisfaction and egiver mental health was the dependent variable. Patient patient diagnosis were significant predictors of caregiver non-Duchenne smiles were not significantly associated well-being. These findings support studies that have found with caregiver mental health in this regression (B = −0.19, marital satisfaction and marital quality to reduce negative SE(B)  =  0.15, β  =  −0.17, p  = .225, CI = [−0.49, 0.12]). In affect (Carr, Cornman, & Freedman, 2016) and serve as the second, the four kinds of smiles (patient and caregiver key protective factors against the adverse effects of care- Duchenne and non-Duchenne) and marital satisfaction giver burden. For example, caregivers who report higher were entered as predictors, and caregiver mental health levels of marital satisfaction have been found to be less was the dependent variable. In this regression, patient non- reactive to memory and behavioral changes in dementia Duchenne smiles were significantly associated with car - patients and engage in better problem solving skills than egiver mental health (B  =  −0.29, SE(B)  =  0.14, β  =  −0.26, caregivers who reported lower levels of marital satisfaction p  = .042, CI = [−0.57, −0.01]), indicating that marital sat- (Steadman, Tremont, & Duncan Davis, 2007). Similarly, isfaction may be acting as a suppressor variable. To further several studies have found that patient diagnosis can also examine this effect, a median split was computed in order to impact caregiver outcomes. Behavioral symptoms, such as group patients who expressed fewer non-Duchenne smiles apathy and disinhibition, which tend to have the greatest and patients who expressed greater non-Duchenne smiles. negative impact on caregiver outcomes (Merrilees et  al., Further analyses revealed marital satisfaction to be signifi- 2013; Mioshi et al., 2013), are more commonly observed in cantly associated with caregiver mental health (r  =  0.55, frontotemporal dementia than in AD. Thus, although our p = .002) for patients who expressed more non-Duchenne study revealed that patient smiles were positively associ- smiles; this association was not significant for patients who ated with caregiver mental health even when accounting expressed fewer non-Duchenne smiles (r = 0.33, p = .10). for patient diagnosis and caregiver marital satisfaction, it Finally, having established an association between illuminates just one piece of the complex processes that in- patient Duchenne smiles and caregiver mental health, fluence caregivers’ mental health. we conducted an exploratory mediation analysis to test whether this association was mediated by marital satisfac- Implications of Patient Duchenne Smiles and tion. Using the PROCESS macro (Hayes, 2008) with 50,000 Caregiver Mental Health bias-corrected bootstrapped samples, results revealed a non-significant mediation, B  =  −0.008, SE  =  (0.065), Our findings indicate that the mental health of caregiv- CI  = [−0.15, 0.12]. ers of patients with bvFTD and AD is associated with signs of genuine positive emotion expressed by the person in their care. However, the particular mechanisms driv- Discussion ing this association are still unknown. Smiles serve many We examined whether smiles expressed by patients with interpersonal functions. For example, Duchenne smiles bvFTD and AD and their spousal caregivers during an convey affiliation, warmth, and intimacy (Hess, Beaupré, unrehearsed, semi-naturalistic conversation about a rela- & Cheung, 2002), and elicit more positive judgments and tionship conflict were associated with caregiver men- more cooperation from others than do non-Duchenne tal health. Results revealed that more patient Duchenne smiles (Johnston et  al., 2010). These functions may be smiles were associated with better caregiver mental health. particularly important in late life, because older adults These findings generalized across diagnosis, and were spe- have been found to appraise social rejection more nega- cific to patient Duchenne smiles; no association was found tively than younger adults (Cheng & Grühn, 2015), and between caregiver smiles (Duchenne or non-Duchenne) and a lack of affiliation has been linked with greater cognitive caregiver mental health. When examining this association decline and dementia in late life (Rafnsson, Orrell, d’Orsi, while accounting for covariates (i.e., diagnosis, patient Hogervorst, & Steptoe, 2017). Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 983 Extrapolating from these findings, for married cou- emotional behavior, due to the substantial impact that ples where one spouse has dementia, Duchenne smiles patient behaviors have on caregiver mental health. Many expressed by the spouse with dementia may convey affili- studies have found that behavioral symptoms in patients ation, which could have a soothing effect on the caregiver. with dementia are most closely linked with adverse care- Previous research has shown that affiliation can reduce giver health outcomes (Schulz & Eden, 2016). stress and HPA activity (DeVries, Glasper, & Detillion, Our findings have a number of practical implications. 2003), and that holding a spouse’s hand reduces activity Caregivers of dementia patients experience significant lev- in brain regions associated with anxiety (Coan, Schaefer, els of burden and are vulnerable to declining mental health & Davidson, 2006). Duchenne smiles may be another way (Schulz & Eden, 2016). Interventions designed to prevent or in which partners affiliate and thereby soothe each other, treat these problems in caregivers could benefit from target- which may be especially helpful for caregivers who are ing specific factors, such as altering patient behaviors (e.g., likely to be experiencing high levels of stress. Consistent reinforcing the expression of positive emotion) or compen- with this, in previous studies with individuals and couples, sating for related losses (e.g., support groups and friends we have found that the expression of positive emotions is becoming a source of emotionally positive interactions for associated with a reduction of autonomic nervous system caregivers). Given the effects positive emotions have on arousal (Fredrickson & Levenson, 1998; Yuan, McCarthy, reducing autonomic arousal (Yuan et al., 2010), it may also Holley, & Levenson, 2010). be useful for caregiver interventions to incorporate com- Alternately, Duchenne smiles may serve to elicit greater ponents that focus on self-soothing, such as mindfulness cooperation from others (Johnston et al., 2010). Patients’ exercises (Raes, Bruyneel, Loeys, Moerkerke, & De Raedt, Duchenne smiles may make spousal caregivers more will- 2015) and other individualized activities that can increase ing to help the patient, thus helping caregivers feel less self-care. Finally, providing psychoeducation for caregivers resentful of or trapped in their caregiving role, which could concerning the likely reduction of positive emotion that will foster better mental health. Experiencing warmth and posi- accompany disease progression can help caregivers under- tive emotions has also been linked to caregiver satisfaction stand that this reduction is not a deliberate behavior on (Carruth, Tate, Moffett, & Hill, 1997), which in turn can the part of the patient. This understanding can help buffer increase caregiver mental health. caregivers from the frustration and sense of loss associated In the present study, we also found some evidence that with diminished expression of positive emotion on the part more patient non-Duchenne smiles were associated with of a loved one with dementia. worse caregiver mental health. This finding only emerged when we accounted for differences in patient diagnosis, pa- Strengths and Limitations tient cognitive functioning, and caregiver marital satisfaction. Examining these covariates separately revealed that it was the Strengths of this research include studying a community inclusion of marital satisfaction that revealed the association sample of participants with two different kinds of demen- between more non-Duchenne smiles by patients and worse tias, the use of objective behavioral measurement of positive caregiver mental health. These analyses suggest that higher emotions, and studying positive emotion in an ecologically marital satisfaction may serve as a particularly important valid semi-naturalistic interpersonal context. To our know- buffer for maintaining caregiver mental health in relation- ledge, this is the first study to use an objective behavioral ships where the patient expresses greater smiles that are measure to assess positive emotion in the interactions of merely “polite” when communicating with their caregiver. couples where one spouse has dementia, and the first to The negative association between more non-Duchenne smiles link positive emotion measured in this manner with care- and worse caregiver mental health is consistent with previous giver mental health. research that has found non-Duchenne smiles to be associ- The research also has several limitations. Although our ated with more negative outcomes. For example, in healthy sample size of 57 patient–caregiver dyads is relatively large married couples, non-Duchenne smiles during a conversation compared to some laboratory studies that have examined were associated with more time separated over the following dementia patients, sample size remains a limitation. A four years (Gottman, Levenson, & Woodin, 2001). priori power analyses indicated that for power = 0.80 and One striking non-finding in the present study was that four predictors, a sample size of 85 dyads would have been caregiver smiles (Duchenne and non-Duchenne) were not required to detect a small effect size. Thus, some of the asso- associated with caregivers’ mental health. We expected car- ciations that were non-significant within our sample may egivers’ positive emotional behavior to be related to their reach significance in larger samples with additional power. own mental health, consistent with research that has gener- Other methodological limitations include our exclusive ally found positive emotion to serve as a protective buffer focus on positive emotion (we did not consider negative against negative outcomes (Fredrickson & Joiner, 2002). [e.g., anger, sadness] or self-conscious [e.g., shame, pride] The association between positive emotional behavior and emotional expressions), examining a brief, “thin slice” be- caregiver mental health may have been significant for pa- havior (Ambady et al., 2000), using a cross-sectional rather tient positive emotional behavior, but not caregiver positive than a longitudinal design, and not including a sample of Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 984 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 healthy control couples (which would help determine the Duchenne smiles will be important factors to examine. extent to which findings are unique to couples dealing with Additionally, data on HPA activity might reveal a medi- dementia). ating effect at the physiological level. Further research An additional limitation is that we cannot know with examining these mechanisms would help to clarify the certainty whether observable Duchenne smiles are spon- association between patient Duchenne smiles and care- taneous, veridical indicators of participants’ underlying giver mental health. positive emotional states, or are being produced strategic- In research with individuals with dementia, it is com- ally and deliberately (Gunnery, Hall, & Ruben, 2013). pelling to consider the disease as the primary cause of Although impairments in patients’ cognitive, motor, and changes in patients’ emotional behavior and the behav- social functioning make this level of self-monitoring, vol- ioral changes as the primary cause of changes in car- untary control of facial behavior, and strategic impression egivers’ mental health. Thus, we suspect that dementia management less likely (Gregory et  al., 2002), it is still is causing declining levels of patient Duchenne smiles, possible that patients are not actually experiencing posi- which cause declines in caregiver mental health. However, tive emotion, but rather have learned to produce Duchenne elements of this causal chain may function in the reverse smiles as a way of eliciting cooperation and reducing dis- direction (i.e., caregivers with worse mental health may pleasure in caregivers. cause patients to produce lower levels of Duchenne smiles). Evaluating these possibilities will require longi- tudinal designs that measure potential mediating mecha- Future Directions nisms at multiple time-points. In future studies, it will be important to extend this research to include other indicators of emotion (e.g., words, ges- Conclusion tures), negative and self-conscious emotions, and other contexts (e.g., other kinds of conversations and shared The human face has an exquisitely tuned facial muscula- activities). Similarly, it will be useful to examine these asso- ture that plays a crucial role in communicating our emo- ciations in patients with other dementias as well as psychi- tions. A  Duchenne, or “genuine,” smile, moves only two atric disorders. facial muscles—the orbicularis oculi (cheek raise) and Several timing issues are also important directions for zygomatic major (lip corner raise). Nonetheless, this kind future studies. Patients with dementia and other neuro- of smile has been found to be associated with a number logical diseases may react to social stimuli at a slower of positive outcomes including, in the present study, better pace than adults without neurological disorders. No lit- mental health in spousal caregivers of dementia patients. erature currently addresses this question, but it would Given the increasing prevalence of dementia and the rap- be useful to determine whether patients show delayed idly growing aging population worldwide, understanding emotional responses to stimuli, and whether those delays the role that positive emotional behaviors play in caregiver impact caregiver mental health. In addition, the present mental health has important implications for helping pre- study focused on the initial segment of the conversation, serve the well-being of caregivers and for advancing our which has previously been found to be related to rela- understanding of emotional expression and its effects on tionship quality (Carrère & Gottman, 1999). However, others. future studies would benefit from examining emotional behavior during other parts of the conversation (e.g., the Funding ending) as well as the trajectory of emotional changes over time. This research was supported by National Institute on Aging grants It will also be important to consider possible mecha- 1R01AG041762-01A1 and 2P01AG019724-11 to R.W.L. and B.L.M. nisms of the effects found in this study. Exploratory analy- ses in the present study suggest that marital satisfaction is not a likely mechanism undergirding the relationship Acknowledgements between patient Duchenne smiles and caregiver mental S.J.L., J.J.C., J.M., and R.W.L. developed the study concept. S.J.L., health. Previous research has shown that Duchenne smiles J.J.C., A.V., and D.E.C.  performed the facial coding. S.J.L.  pro- elicit in others a greater willingness to cooperate (Johnston cessed and analyzed the data. S.J.L. and J.J.C. wrote the first draft et al., 2010). When patients show more Duchenne smiles, of the manuscript; and all authors contributed to revisions of the caregivers may feel more willing to help the patient with manuscript. R.W.L.  supervised all phases of the project. S.J.L.  and J.J.C. contributed equally to the study. activities of daily living instead of feeling obligated and trapped in the caregiving role, and may experience bet- ter mental health as a result. Thus, mechanisms such as Conflict of Interest caregivers’ subjective experience of decreased burden, or increased positive emotion as the result of receiving There were no identified conflicts of interest. Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 985 Coleman, M., Ganong, L., & Fine, M. (2000). Reinvestigating References remarriage: Another decade of progress. Journal of Ambady, N., Bernieri, F. J., & Richeson, J. A. (2000). Toward a hist- Marriage and Family, 62, 1288–1307. doi:10.1111/ ology of social behavior: Judgmental accuracy from thin slices j.1741-3737.2000.01288.x of the behavioral stream. In M. P. Zanna (Eds.), Advances in Cornwell, B., Laumann, E. O., & Schumm, L. P. (2008). experimental social psychology (Vol. 32, pp. 201–271). San The social connectedness of older adults: a national pro- Francisco, CA: Academic Press. file*. American Sociological Review, 73, 185–203. Ascher, E. A., Sturm, V. E., Seider, B. H., Holley, S. R., Miller, B. L., & doi:10.1177/000312240807300201 Levenson, R. W. (2010). Relationship satisfaction and emotional DeVries, A. C., Glasper, E. R., & Detillion, C. E. (2003). Social language in frontotemporal dementia and Alzheimer disease modulation of stress responses. Physiology & Behavior, 79, patients and spousal caregivers. Alzheimer Disease and Associated 399–407. doi:10.1016/S0031-9384(03)00152-5 Disorders, 24, 49–55. doi:10.1097/WAD.0b013e3181bd66a3 Ekman, P., & Friesen, W. V. (1977). Facial action coding system. Palo Blieszner, R., & Shifflett, P. A. (1990). The effects of Alzheimer’s Alto, CA: Consulting Psychologists Press. disease on close relationships between patients and caregivers. Ekman, P., & Friesen, W. V. (1982). Felt, false, and miserable smiles. Family Relations, 39, 57–62. doi:10.2307/584949 Journal of Nonverbal Behavior, 6, 238–252. doi:10.1007/ Bograd, R., & Spilka, B. (1996). Self-disclosure and marital satisfac- BF00987191 tion in mid-life and late-life remarriages. International Journal Ekman, P., & Rosenberg, E. L. (2005). What the face reveals: Basic of Aging & Human Development, 42, 161–172. doi:10.2190/ and applied studies of spontaneous expression using the facial W87M-WCK7-MHTT-N34F action coding system (FACS) (2nd ed.). New York, NY: Oxford Bourke-Taylor, H., Pallant, J. F., Law, M., & Howie, L. (2012). University Press. Predicting mental health among mothers of school-aged chil- Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). “Mini- dren with developmental disabilities: The relative contribu- mental state”. A practical method for grading the cognitive state tion of child, maternal and environmental factors. Research of patients for the clinician. Journal of Psychiatric Research, 12, in Developmental Disabilities, 33, 1732–1740. doi:10.1016/j. 189–198. doi:10.1016/0022-3956(75)90026-6 ridd.2012.04.011 Fredrickson, B. L. (2004). The broaden-and-build theory of posi- Braun, M., Mura, K., Peter-Wight, M., Hornung, R., & Scholz, U. tive emotions. Philosophical Transactions of the Royal Society (2010). Toward a better understanding of psychological well- of London Series B: Biological Sciences, 359, 1367–1378. being in dementia caregivers: The link between marital com- doi:10.1098/rstb.2004.1512 munication and depression. Family Process, 49, 185–203. Fredrickson, B. L., & Joiner, T. (2002). Positive emotions trigger doi:10.1111/j.1545-5300.2010.01317.x upward spirals toward emotional well-being. Psychological Campos, B., Shiota, M. N., Keltner, D., Gonzaga, G. C., & Goetz, J. L. Science, 13, 172–175. doi:10.1111/1467-9280.00431 (2013). What is shared, what is different? Core relational themes Fredrickson, B. L., & Levenson, R. W. (1998). Positive emo- and expressive displays of eight positive emotions. Cognition & tions speed recovery from the cardiovascular sequelae of Emotion, 27, 37–52. doi:10.1080/02699931.2012.683852 negative emotions. Cognition & Emotion, 12, 191–220. Carr, D., Cornman, J. C., & Freedman, V. A. (2016). Marital qual- doi:10.1080/026999398379718 ity and negative experienced well-being: An assessment of actor Fredrickson, B. L., Tugade, M. M., Waugh, C. E., & Larkin, G. and partner effects among older married persons. The Journals R. (2003). What good are positive emotions in crises? A  pro- of Gerontology Series B: Psychological Sciences and Social spective study of resilience and emotions following the ter- Sciences, 71, 177–187. doi:10.1093/geronb/gbv073 th rorist attacks on the United States on September 11 , 2001. Carrère, S., & Gottman, J. M. (1999). Predicting divorce Journal of Personality and Social Psychology, 84, 365–376. among newlyweds from the first three minutes of doi:10.1037/0022-3514.84.2.365 a marital conflict discussion. Family Process, 38, Gonzaga, G. C., Keltner, D., Londahl, E. A., & Smith, M. D. (2001). 293–301. doi:10.1111/j.1545-5300.1999.00293.x Love and the commitment problem in romantic relations and Carruth, A. K., Tate, U. S., Moffett, B. S., & Hill, K. (1997). friendship. Journal of Personality and Social Psychology, 81, Reciprocity, emotional well-being, and family functioning as 247–262. doi:10.1037/0022-3514.81.2.247 determinants of family satisfaction in caregivers of elderly par- Gottman, J., Levenson, R., & Woodin, E. (2001). Facial expressions ents. Nursing Research, 46, 93–100. doi:10.1097/00006199- during marital conflict. Journal of Family Communication, 1, 199703000-00006 37–57. doi:10.1207/S15327698JFC0101_06 Cheng, Y., & Grühn, D. (2015). Age differences in reactions to social Gregory, C., Lough, S., Stone, V., Erzinclioglu, S., Martin, L., rejection: The role of cognitive resources and appraisals. The Baron-Cohen, S., & Hodges, J. R. (2002). Theory of mind in Journals of Gerontology Series B: Psychological Sciences and patients with frontal variant frontotemporal dementia and Social Sciences, 70, 830–839. doi:10.1093/geronb/gbu054 Alzheimer’s disease: Theoretical and practical implications. Chopik, W. J., & O’Brien, E. (2017). Happy you, healthy me? Having Brain: A  Journal of Neurology, 125, 752–764. doi:10.1093/ a happy partner is independently associated with better health in brain/awf079 oneself. Health Psychology, 36, 21–30. doi:10.1037/hea0000432 Gunnery, S. D., Hall, J. A., & Ruben, M. A. (2013). The deliber- Coan, J. A., Schaefer, H. S., & Davidson, R. J. (2006). Lending a hand: ate Duchenne smile: Individual differences in expressive con- Social regulation of the neural response to threat. Psychological trol. Journal of Nonverbal Behavior, 37, 29–41. doi:10.1007/ Science, 17, 1032–1039. doi:10.1111/j.1467-9280.2006.01832.x s10919-012-0139-4 Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 986 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 Haase, C. M., Beermann, U., Saslow, L. R., et  al. (2015). Short Locke, H. J., & Wallace, K. M. (1959). Short marital-adjustment and prediction tests: Their reliability and validity. Marriage and alleles, bigger smiles? The effect of 5-HTTLPR on positive emo- tional expressions. Emotion (Washington, D.C.), 15, 438–448. Family Living, 21, 251–255. doi:10.2307/348022 Losada, M., & Heaphy, E. (2004). The role of positivity and con- doi:10.1037/emo0000074 Harker, L., & Keltner, D. (2001). Expressions of positive emo- nectivity in the performance of business teams: A  nonlinear dynamics model. American Behavioral Scientist, 47, 740–765. tion in women’s college yearbook pictures and their rela- tionship to personality and life outcomes across adulthood. doi:10.1177/0002764203260208 Mahoney, R., Regan, C., Katona, C., & Livingston, G. (2005). Anxiety Journal of Personality and Social Psychology, 80, 112–124. doi:10.1037/0022-3514.80.1.112 and depression in family caregivers of people with Alzheimer disease: The LASER-AD study. The American Journal of Geriatric Psychiatry, Hawthorne, G., Osborne, R. H., Taylor, A., & Sansoni, J. (2007). The SF36 Version 2: Critical analyses of population weights, scoring 13, 795–801. doi:10.1097/00019442-200509000-00008 McHorney, C. A., Ware, J. E. Jr, & Raczek, A. E. (1993). algorithms and population norms. Quality of Life Research, 16, 661–673. doi:10.1007/s11136-006-9154-4 The MOS 36-item short-form health survey (SF-36): II. Psychometric and clinical tests of validity in measuring phys- Hayes, A. F. (2008). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New ical and mental health constructs. Medical Care, 31, 247–263. doi:10.1097/00005650-199303000-00006 York, NY: Guilford Press. Hess, U., Beaupré, M. G., & Cheung, N. (2002). Who to whom McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., & Stadlan, E. M. (1984). Clinical diagnosis of Alzheimer’s dis- and why–cultural differences and similarities in the function of smiles. In M. Abel (Ed.), An empirical reflection on the smile (pp. ease: Report of the NINCDS-ADRDA work group under the auspices of department of health and human services task force 187–216). New York: Edwin Mellen Press. Johnston, L., Miles, L., & Macrae, C. N. (2010). Why are you smil- on Alzheimer’s disease. Neurology, 34, 939–944. doi:10.1212/ WNL.34.7.939 ing at me? Social functions of enjoyment and non-enjoyment smiles. The British Journal of Social Psychology, 49, 107–127. Merrilees, J., Dowling, G. A., Hubbard, E., Mastick, J., Ketelle, R., & Miller, B. L. (2013). Characterization of apathy in persons doi:10.1348/014466609X412476 Keltner, D., & Bonanno, G. A. (1997). A study of laughter and dis- with frontotemporal dementia and the impact on family caregiv- ers. Alzheimer Disease and Associated Disorders, 27, 62–67. sociation: Distinct correlates of laughter and smiling during bereavement. Journal of Personality and Social Psychology, 73, doi:10.1097/WAD.0b013e3182471c54 Mioshi, E., Foxe, D., Leslie, F., Savage, S., Hsieh, S., Miller, L., … 687–702. doi:10.1037/0022-3514.73.4.687 Kouros, C. D., Papp, L. M., & Cummings, E. M. (2008). Interrelations Piquet, O. (2013). The impact of dementia severity on care- giver burden in frontotemporal dementia and Alzheimer dis- and moderators of longitudinal links between marital satisfac- tion and depressive symptoms among couples in established ease. Alzheimer Disease and Associated Disorders, 27, 68–73. doi:10.1097/WAD.0b013e318247a0bc relationships. Journal of Family Psychology, 22, 667–677. doi:10.1037/0893-3200.22.5.667 Nunnemann, S., Kurz, A., Leucht, S., & Diehl-Schmid, J. (2012). Caregivers of patients with frontotemporal lobar degener- LaFrance, M., Hecht, M. A., & Paluck, E. L. (2003). The contingent smile: A meta-analysis of sex differences in smiling. Psychological ation: A review of burden, problems, needs, and interventions. International Psychogeriatrics, 24, 1368–1386. doi:10.1017/ Bulletin, 129, 305–334. doi:10.1037/0033-2909.129.2.305 Levenson, R. W., & Gottman, J. M. (1983). Marital interaction: S104161021200035X Raes, A. K., Bruyneel, L., Loeys, T., Moerkerke, B., & De Raedt, R. Physiological linkage and affective exchange. Journal of P ersonality and Social Psychology, 45, 587–597. doi:10.1037/0022- (2015). Mindful attention and awareness mediate the association between age and negative affect. The Journals of Gerontology 3514.45.3.587 Levenson, R. W., Haase, C., Bloch, L., Holley, S., & Seider, B. (2013). Series B: Psychological Sciences and Social Sciences, 70, 181– Emotion regulation in couples. In J. J. Gross (Ed.), Handbook of 190. doi:10.1093/geronb/gbt074 emotion regulation (2nd ed., pp. 267–283). New York: Guilford Rafnsson, S. B., Orrell, M., d’Orsi, E., Hogervorst, E., & Steptoe, A. Press. (2017). Loneliness, social Integration, and incident dementia over Levenson, R. W., Lwi, S. J., Brown, C. L., Ford, B. Q., Otero, M. C., & 6 years: Prospective findings from the English Longitudinal Study Verstaen, A. (2017). Emotion. In J. T. Cacioppo, L. G. Tassinary of Ageing. The Journals of Gerontology Series B: Psychological & G. G. Berntson (Eds.), Handbook of psychophysiology (4th Sciences and Social Sciences. 0,  1–11.  doi:10.1093/geronb/ ed.) (pp. 444–464). Cambridge: Cambridge University Press. gbx087 Lewis, R. J., Derlega, V. J., Griffin, J. L., & Krowinski, A. C. (2003). Rascovsky, K., Hodges, J. R., Knopman, D., Mendes, M. F., Kramer, Stressors for gay men and lesbians: Life stress, gay-related J. H., Neuhaus, J., … Miller, B. L. (2011). Sensitivity of revised stress, stigma consciousness, and depressive symptoms. Journal diagnostic criteria for the behavioural variant of frontotemporal of Social and Clinical Psychology, 22, 716–729. doi:10.1521/ dementia. Brain, 134, 2456–2477. doi:10.1093/brain/awr179 jscp.22.6.716.22932 Schulz, R., & Eden, J. (2016). Families caring for an aging America. Lints-Martindale, A. C., Hadjistavropoulos, T., Barber, B., & Gibson, Washington, DC: The National Academies Press. S. J. (2007). A psychophysical investigation of the facial action Schulz, R., O’Brien, A. T., Bookwala, J., & Fleissner, K. (1995). coding system as an index of pain variability among older adults Psychiatric and physical morbidity effects of dementia caregiv- with and without Alzheimer’s disease. Pain Medicine (Malden, ing: Prevalence, correlates, and causes. The Gerontologist, 35, Mass.), 8, 678–689. doi:10.1111/j.1526-4637.2007.00358.x 771–791. doi:10.1093/geront/35.6.771 Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 987 Schulz, R., & Sherwood, P. R. (2008). Physical and mental health medical care. JAMA, 262, 925–930. doi:10.1001/jama. effects of family caregiving. The American Journal of Nursing, 1989.03430070073033 108, 23–7; quiz 27. doi:10.1097/01.NAJ.0000336406.45248.4c Ware, J. (2000). SF-36 health survey update. Spine, 25, 3130–3139. Seeley, W. W., Allman, J. M., Carlin, D. A., Crawford, R. K., Macedo, doi:10.1097/00007632-200012150-00008 M. N., Greicius, M. D., Dearmond, S. J., … Miller, B. L. (2007). Ware, J. E., Kosinski, M., & Keller, S. (1994). SF-36 physical Divergent social functioning in behavioral variant frontotem- and mental health summary scales: A  user’s manual. Health poral dementia and Alzheimer disease: reciprocal networks Assessment Lab.  Boston, MA: Health Institute, New England and neuronal evolution. Alzheimer Disease and Associated Medical Center. Disorders, 21, S50–S57. doi:10.1097/WAD.0b013e31815c0f14 Wichers, M. C., Myin-Germeys, I., Jacobs, N., Peeters, F., Steadman, P. L., Tremont, G., & Davis, J. D. (2007). Premorbid rela- Kenis, G., Derom, C., … van Os, J. (2007). Evidence that tionship satisfaction and caregiver burden in dementia caregiv- moment-to-moment variation in positive emotions buf- ers. Journal of Geriatric Psychiatry and Neurology, 20, 115–119. fer genetic risk for depression: A  momentary assessment doi:10.1177/0891988706298624 twin study. Acta Psychiatrica Scandinavica, 115, 451–457. Strathearn, L., Li, J., Fonagy, P., & Montague, P. R. (2008). What’s doi:10.1111/j.1600-0447.2006.00924.x in a smile? Maternal brain responses to infant facial cues. Yuan, J. W., McCarthy, M., Holley, S. R., & Levenson, R. W. (2010). Pediatrics, 122, 40–51. doi:10.1542/peds.2007-1566 Physiological down-regulation and positive emotion in mari- Sturm, V. E., McCarthy, M. E., Yun, I., Madan, A., Yuan, J. W., Holley, S. tal interaction. Emotion (Washington, D.C.), 10, 467–474. R., … Levenson, R. W. (2011). Mutual gaze in Alzheimer’s disease, doi:10.1037/a0018699 frontotemporal and semantic dementia couples. Social Cognitive Zhu, Y. X., Li, T., Fan, S. R., Liu, X. P., Liang, Y. H., & Liu, P. (2016). Health-related quality of life as measured with the Short-Form and Affective Neuroscience, 6, 359–367. doi:10.1093/scan/nsq055 Tarlov, A. R., Ware, J. E. Jr, Greenfield, S., Nelson, E. C., Perrin, 36 (SF-36) Questionnaire in patients with recurrent vulvovagi- nal candidiasis. Health and Quality of Life Outcomes, 14, 65. E., & Zubkoff, M. (1989). The medical outcomes study. An application of methods for monitoring the results of doi:10.1186/s12955-016-0470-2 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png "The Journals of Gerontology - Series B: Psychological Sciences and Social Sciences" Oxford University Press

Genuine Smiles by Patients During Marital Interactions are Associated with Better Caregiver Mental Health

Download PDF
13 pages

Loading next page...
 
/lp/ou_press/genuine-smiles-by-patients-during-marital-interactions-are-associated-9guig01zu9

References (69)

Publisher
Oxford University Press
Copyright
Copyright © 2022 The Gerontological Society of America
ISSN
1079-5014
eISSN
1758-5368
DOI
10.1093/geronb/gbx157
Publisher site
See Article on Publisher Site

Abstract

Objective: Providing care for a spouse with dementia is associated with an increased risk for poor mental health. To deter- mine whether this vulnerability in caregivers is related to the expression of positive emotion, we examined 57 patients with Alzheimer’s disease and behavioral variant frontotemporal dementia and their spouses as they discussed a marital conflict. Method: Facial behavior during the discussion was objectively coded to identify Duchenne (i.e., genuine) smiles and non- Duchenne (i.e., polite) smiles. Caregiver mental health was measured using the Medical Outcomes Survey. Results: Greater expression of Duchenne smiles by patients was associated with better caregiver mental health, even when accounting for covariates (i.e., diagnosis, patient cognitive functioning, and caregiver marital satisfaction). Greater expression of non-Duchenne smiles by patients was associated with worse caregiver health, but only when covariates were entered in the model. Expression of Duchenne and non-Duchenne smiles by caregivers was not associated with caregiver mental health. Discussion: Patients’ expression of Duchenne and non-Duchenne smiles may reveal important aspects of the emotional quality of the patient–caregiver relationship that influence caregiver burden and mental health. Keywords: Caregiving, Dementia, Mental health, Positive emotion, Smiling. Familial caregivers of patients with dementia are faced focused on an important aspect of patient behavior, the with the enormous challenge of caring for a loved one expression of positive emotion that occurs during patient– whose level of functioning progressively deteriorates. As caregiver interactions, and its association with caregivers’ a group, dementia caregivers are more prone to mental mental health. health problems than non-caregiving older adults (Schulz & Eden, 2016). However, individual caregivers differ Patient Emotional Behavior, Patient– greatly in the extent to which they experience these nega- Caregiver Relationship Quality, and Caregiver tive outcomes. Prior research indicates that behavioral and Mental Health psychological symptoms in patients can be particularly dif- ficult for caregivers, even more so than cognitive and func- Patients with Alzheimer’s disease (AD) and behavioral tional symptoms (Schulz, O’Brien, Bookwala, & Fleissner, variant frontotemporal dementia (bvFTD) undergo a 1995). These behavioral symptoms in patients can be par- number of changes in emotional behavior that can have a ticularly challenging when manifest in interactions with negative impact on caregivers. For example, as AD patients’ caregivers (Ascher et  al., 2010). In the present study, we dementia progresses, problematic behaviors such as anger Published by Oxford University Press on behalf of The Gerontological Society of America 2018. This work is written by (a) US 975 Government employee(s) and is in the public domain in the US. Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 976 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 and aggression become increasingly difficult to manage, (Harker & Keltner, 2001). Greater expressions of love and leading to worse mental health in caregivers (Schulz & other positive emotions during a conflict are associated Sherwood, 2008). Similarly, as bvFTD progresses, patients with higher levels of relationship satisfaction and a greater undergo dramatic changes in personality and express likelihood of engaging in constructive behaviors (e.g., af- greater levels of apathy and social inappropriateness, which firmation, soothing contact, expressing concern; Gonzaga, leads to greater levels of burden, depression, and distress in Keltner, Londahl, & Smith, 2001). caregivers (Merrilees et al., 2013). Despite the established links between patient behav- Positive Emotional Behavior and Health ioral symptoms and worse caregiver mental health, stud- ies of caregiver health that have focused on qualities of In the literature on emotion and health, individuals who the patient–caregiver relationship have been rare. This experience greater positive emotions also experience bet- is unfortunate, because spouses and relationship part- ter health outcomes (e.g., Wichers et  al., 2007). Positive ners often are called on to serve as primary caregivers emotions are thought to help build personal resources for patients with dementia (Schulz & Eden, 2016), and (Fredrickson, 2004) which, in turn, have been associated dementia can profoundly affect these relationships. For with individuals’ ability to utilize coping strategies and be example, patients with bvFTD may become more emo- resilient (Fredrickson, Tugade, Waugh, & Larkin, 2003), tionally blunted and less empathic (Rascovsky et  al., both of which are important for mental health. Although 2011), leading to a weakening of the emotional con- rarely examined, the link between positive emotion and nection with their partner. In previous studies of dyadic health is also found in close relationships. For example, interactions, patients with FTD and their spousal car- individuals with happier spouses report better health, fewer egivers showed less mutual gaze during conversations physical impairments, and less chronic disease (Chopik about a relationship problem and reported lower marital & O’Brien, 2016). Exhibiting and eliciting positive emo- satisfaction than healthy controls (Sturm et  al., 2011). tions fosters supportive environments that promote cop- Alzheimer’s disease can also affect relationships. For ex- ing (Fredrickson, 2004), and smiling has been found to be ample, a longitudinal study found that caregivers’ social particularly effective for eliciting greater cooperation from intimacy with the patient declined over time (Blieszner others (Johnston, Miles, & Macrae, 2010). In the realm & Shifflett, 1990). Loss of closeness between the patient of caregiving where one spouse has dementia, patients’ and caregiver can be devastating, especially in later life positive communication behaviors (e.g., humor) have been when social networks shrink and close relationships be- linked to lower depression in the caregiving spouse (Braun, come increasingly more important (Cornwell, Laumann, Mura, Peter-Wight, Hornung, & Scholz, 2010). & Schumm, 2008). Thus, not surprisingly, poor rela- tionship quality has been strongly associated with worse Measuring Positive Emotion caregiver mental health (Mahoney, Regan, Katona, & Livingston, 2005). Emotions can be measured via self-reported subjective experience, expressive behavior, and peripheral and cen- tral nervous system physiology (Levenson et  al., 2017). Positive Emotional Behavior and Intimate Emotional facial expressions, with their high signal value Relationships for conspecifics, may be particularly important indicators Positive emotional behaviors are particularly important of the emotional life of couples (Levenson, Haase, Bloch, indicators of the state of intimate relationships. Consistent Holley, & Seider, 2013). Although there are a number with this, a primary function of positive emotions is to of different positive emotions (e.g., pride, amusement, broaden and build social relationships (Fredrickson, contentment), most positive emotions share the smile as 2004). Experiencing positive emotions has been linked a common expressive element (Campos, Shiota, Keltner, to stronger social bonds and social connection (Losada Gonzaga, & Goetz, 2013). Moreover, particular morpho- & Heaphy, 2004), and encourages individuals to explore logical features of the smile are thought to convey whether and engage in new experiences, which increase social inte- the emotion is genuinely felt (i.e., “Duchenne smiles,” gration (Fredrickson, 2004). For example, when mothers which involve the raising of the lip corners and the rais- look at photographs of their infants smiling, dopaminergic ing of the cheeks) or not genuinely felt (“non-Duchenne” reward-related brain areas show greater activation than or “polite” smiles, which only involve the raising of lip when infants’ neutral or sad expressions are viewed; this re- corners; Ekman & Friesen, 1982). In dementia research, inforcement associated with smiles may play an important assessing emotion via facial expressions has the additional role in building mother–infant attachments (Strathearn, advantage of reducing the problems associated with retro- Li, Fonagy, & Montague, 2008). Positive emotions are spective self-reports of emotion in patients who may have also important in romantic relationships; people who ex- deficits in language, memory, and self-awareness. Despite press greater happiness in college yearbook photographs these advantages, we are aware of no prior studies of the report higher levels of marital satisfaction 30  years later associations between patients’ emotional behaviors and Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 977 caregivers’ mental health that have utilized objective cod- through the Memory and Aging Center at the University ing of emotional facial expressions during patient–car- of California, San Francisco. Patients were evaluated by egiver interactions. a multidisciplinary team and diagnosed based on neu- rological, neuropsychological, and neuroimaging data using consensus criteria for AD (McKhann et al., 1984) The Present Study and FTD (Rascovsky et al., 2011). Patients were gener- ally in the early stage of their disease (see Mini-Mental We studied positive emotional facial expressions (Duchenne State Examination [MMSE] scores in Table 1) and thus and non-Duchenne smiles) that occurred during the first were able to understand and follow task instructions (as 30 s of a 10-min discussion of an area of marital conflict confirmed by verbal checks with session facilitators). Both in a sample of patients with dementia and their spousal patients and caregivers needed to be sufficiently healthy caregivers. Expressions were coded using the Facial Action to travel to the University of California, Berkeley to com- Coding System (FACS, Ekman & Friesen, 1977), a precise plete the day-long laboratory session. Patients who met anatomically based system that can decompose observable criteria for Mild Cognitive Impairment (i.e., patients with facial behavior in terms of its underlying muscular con- cognitive impairments that were not significant enough tractions. The first 30 s of the interaction were chosen be- to interfere with daily activities) were excluded. Table cause they capture a period of reconnection after couples 1 presents demographic characteristics for patients and had sat for 5  min of enforced silence. In healthy couples, caregivers. All couples were paid $30 in addition to any these “reunion” periods are highly diagnostic of the quality transportation costs they incurred. of the relationship. For example, after a similar period of enforced silence, greater positive emotion expressed during the first 3 min of a discussion of an area of marital conflict Procedure predicted lower likelihood of divorce (Carrère & Gottman, 1999). Focusing on only 30 s of facial behavior has practical A week before their laboratory visit, caregivers completed advantages as well. FACS coding is highly time-consuming; a questionnaire packet including measures of mental when applied thoroughly, it typically takes 100 min to code health and marital satisfaction. Patients and caregivers 1 min of behavior, and it is a slow process even when only then came to the Berkeley Psychophysiology Laboratory coding for Duchenne and non-Duchenne smiles. Moreover, for a day-long comprehensive assessment of emotional studying “thin slices” of behavior (typically ranging from and social functioning. Upon arrival, participants were 10 to 30  s) is a well-established procedure for capturing informed that their physiological, behavioral, and self- important qualities of individuals and dyads (Ambady, reported responses would be recorded and videotaped. Bernieri, & Richeson, 2000). Prior to the start of the laboratory session, participants Because different forms of dementia affect different brain had sensors attached for monitoring autonomic and regions and have different effects on emotional functioning somatic physiological responses (these data were not used (Seeley et al., 2007), we included patients with two common for the present study). Throughout the session, partici- forms of dementia: AD—a dementia that affects the tempo- pants’ upper body and face were filmed using a partially ral and parietal lobes and primarily impairs memory and concealed video camera. cognition, and bvFTD—a dementia that affects the frontal The present study focused on a laboratory task in which and temporal lobes and primarily produces changes in emo- couples sat quietly during a 5-min baseline period and then tion, personality, and behavior. Among the various subtypes had an unrehearsed discussion about an area of marital con- of frontotemporal dementia, we focused on bvFTD because flict (i.e., conflict conversation) for 10 min. This procedure language problems associated with other subtypes (i.e., was originally developed for studying marital interactions semantic variant primary progressive aphasia, non-fluent in healthy couples (Levenson & Gottman, 1983), but has variant primary progressive aphasia) could interfere with also been used with dementia patients and their caregivers patients’ ability to engage in the conflict conversation. (Ascher et al., 2010; Sturm et al., 2011). The discussion Our primary hypothesis was that more Duchenne smiles occurred for each couple immediately after an hour-long (thought to indicate genuine positive emotion; Ekman & break for lunch; anecdotally, many couples reported having Friesen, 1982) expressed by patients and their spousal car- good energy levels at this point in the day. egivers would be associated with better caregiver mental health. We did not expect this relationship to be found for Measures non-Duchenne smiles. Positive emotional expressions Using the Facial Action Coding System (FACS; Ekman & Method Friesen, 1977), trained coders blind to diagnosis and care- Participants giver outcomes measured smiling behaviors expressed by Twenty-nine patients with AD and 28 patients with both spouses during the first 30  s of the 10-min conflict bvFTD and their spousal caregivers were recruited conversation. Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 978 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 Table 1. Sociodemographic Characteristics of Patients with Behavioral Variant Frontotemporal Dementia, Alzheimer’s Disease, and Patients’ Spousal Caregivers Spousal caregivers of Spousal caregivers of Patients with AD patients with AD Patients with bvFTD patients with bvFTD n 29 29 28 28 Age (M [SD]) 61.92 (8.89) 61.08 (8.77) 61.78 (8.22) 60.21 (8.04) Sex 15 M, 14 F 13 M, 16 F 20 M, 8 F 9 M, 19 F Smiles Duchenne Frequency 1.03 (1.30) 0.97 (1.13) 0.97 (1.11) 1.04 (1.05) Intensity 1.39 (1.67) 1.51 (1.70) 0.98 (1.45) 1.01 (1.44) Duration 4.72 (5.93) 3.97 (5.41) 4.69 (6.41) 3.74 (4.60) Composite 0.04 (0.99) 0.03 (0.93) −0.05 (0.90) −0.03 (0.91) Non-Duchenne Frequency 1.26 (1.48) 1.19 (1.29) 0.88 (1.17) 1.19 (1.15) Intensity 1.16 (1.14) 1.43 (1.41) 0.99 (1.04) 1.35 (1.05) Duration 3.88 (4.63) 5.52 (7.58) 3.04 (4.17) 4.79 (5.97) Composite 0.08 (1.03) −0.02 (0.99) −0.08 (0.89) 0.02 (0.88) Marital satisfaction — 111.65 (30.71) — 96.40 (28.25) MMSE (M [SD]) 21.97 (5.12) — 25.07 (4.12) — Mental Health (M [SD]) — 0.35 (0.91) — −0.36 (1.12) Note. For smiles, three raw subscores are presented: frequency (number of Duchenne smiles), intensity (average intensity of the Duchenne smile), and duration (number of seconds during the film during which there is a Duchenne smile). Logarithmic transformations were applied to the subscores to reduce skewness. Subscores were then normalized (using the means and standard deviations for the entire sample) and averaged to derive the score for the two kinds of smiles (see Keltner & Bonanno, 1997). Higher scores on marital satisfaction indicate greater satisfaction (range: 2–158); lower scores on MMSE indicate lower cognitive functioning (range: 0–30). A dash (—) indicates that the given variable was not measured. AD = Alzheimer’s disease; bvFTD = behavioral variant frontotemporal dementia; MMSE = Mini-Mental State Exam. To develop reliability, all coders completed practice cod- with negative emotions generally are not deemed to be ing assignments of older adults’ emotional behavior and signs of genuine positive emotion. Thus, smiles accom- met weekly to discuss discrepancies before coding the reli- panied by action units typically associated with negative ability sample of 18 dyads (30% of the sample). High reli- emotions (e.g., disgust, sadness, anger) such as AU9 (nose ability was required for the practice coding assignments wrinkling), AU15 (frowning), or AU4 (brow furrowing) (Cronbach’s α > 0.70) before coders completed their work were not included in the analyses. These exclusionary for the remaining dyads. criteria have been used in previous studies of Duchenne In order to ensure that individual differences in facial smiles (Haase et al., 2015). features (e.g., elasticity) were taken into account when For each participant, one score was derived to charac- coding, all coders were provided with neutral stills of each terize Duchenne smiling and a second score was derived patient and caregiver, and instructed to only code facial to characterize non-Duchenne smiling. To produce these expressions that clearly resulted from changes in specific scores, we first computed three subscores for each kind facial muscle movements. FACS is a well-validated and of smile: (a) the number of smiles that occurred during widely used measure (Ekman & Rosenberg, 2005), and has the conversation; (b) the average intensity of all smiles been successfully utilized in prior studies examining facial at their apex or most intense level of facial action (based expressions in older adults and adults with neurodegenera- on 1–5 intensity of AU12), and (c) the average duration tive diseases (Lints-Martindale, Hadjistavropoulos, Barber, of all smiles. Our measures of Duchenne and non-Duch- & Gibson, 2007). enne smiles were moderately skewed (skew ≥1). In order Based on prior research (Ekman & Friesen, 1982), to reduce skewness, a constant of one was added to each smiles were classified as genuine Duchenne smiles if they subscore (i.e., frequency, duration, intensity) before loga- included contraction of both the orbicularis oculi (cheek rithmic transformations were applied. This reduced the raise, action unit [AU]6 and/or AU7) and zygomatic major skew of each variable to ≤1, providing a more normal (lip corner raise, AU12) or as “polite” non-Duchenne distribution for analyses. Subscores were then normalized smiles if they consisted only of the contraction of the (using the means and standard deviations for the entire zygomatic major (AU12). Duchenne and non-Duchenne sample) and averaged to derive the score for the two kinds smiles with an opened mouth (AU25 or AU26) were also of smiles (see Keltner & Bonanno, 1997). To determine included. Smiles that are accompanied by AUs associated inter-rater reliability, two FACS-certified coders and two Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 979 trained FACS coders scored 18 conversations. Inter-rater whether caregivers differed in age, sex, marital satisfaction, reliability for the Duchenne and non-Duchenne scores was Duchenne smiles, and non-Duchenne smiles. high (Cronbach’s α = 0.84). For our primary hypothesis that Duchenne smiles by patients and caregivers would be associated with better Caregiver mental health caregiver mental health, we conducted a linear regres- Caregiver mental health was assessed using the Medical sion analysis in which the four kinds of smiles (patient Outcomes Study (MOS SF-36; Tarlov et  al., 1989). The and caregiver Duchenne and non-Duchenne) were entered MOS SF-36 is a 36-item self-report measure designed to together on the same step as predictors and caregiver men- assess eight health domains: (a) physical functioning, (b) tal health was the dependent variable. This analysis was role limitations due to physical health, (c) energy/fatigue, then re-run with diagnosis included as a covariate, reflect- (d) pain, (e) general health problems, (f) role limitations ing prior findings that indicate that symptoms associated due to emotional problems, (g) emotional well-being, and with bvFTD and AD create different kinds of burdens (h) social functioning. Scores for each of these domains and challenges for caregivers (Nunnemann, Kurz, Leucht, ranged from 0 (worst) to 100 (best). According to con- & Diehl-Schmid, 2012). When associations were found vention (Ware, 2000), all subscales were z-scored and between a particular kind of smile and caregiver mental weighted to create a composite score of mental health health, we conducted regression analyses that included the such that mental health subscales are weighted more heav- interaction of that kind of smile with diagnosis. Finally, to ily than other subscales; this weighting reduces but does evaluate the robustness of our findings, we repeated the not eliminate the influence of physical health on caregiv- main regression analysis while accounting for variables ers’ mental health scores. This scoring scheme has been that differed across diagnostic groups (i.e., marital satis- used extensively in the literature (Bourke-Taylor, Pallant, faction, cognitive functioning; see Preliminary Analyses). Law, & Howie, 2012; Hawthorne, Osborne, Taylor, & This analysis revealed a potential suppressor effect, thus, Sansoni, 2007) and its reliability and validity for predict- additional post hoc analyses were conducted to under- ing mental health is well-established (McHorney, Ware, stand this effect. An additional exploratory analysis was & Raczek, 1993). conducted to examine whether marital satisfaction medi- ated the relationship found between patient Duchenne smiles and caregiver mental health. Marital satisfaction Caregivers completed the Locke–Wallace Marital Adjustment Test (Locke & Wallace, 1959), a well-validated Preliminary Analyses 15-item scale (e.g., “To what extent do you and your mate agree or disagree on handling family finances” [0 = Always Age and sex differences disagree, 5 = Always agree]) that we have used in prior re- Group differences in AD and bvFTD patients’ age were search with dementia patients and caregivers (Ascher et al., analyzed using an independent t-test; sex differences were 2010). Scores can range between 2 and 158, with higher analyzed using a Chi-square test. Results revealed no scores indicating greater marital satisfaction. 2 patient differences in age, t(55) = 0.07, p = .948 or sex, χ (1, N=57) = 2.33, p = .127. Parallel analyses were conducted to Cognitive functioning examine group differences for caregivers. Results revealed Patients’ cognitive functioning was assessed using the no caregiver differences in age, t(55) = 0.39, p = .699 or sex, MMSE (Folstein, Folstein, & McHugh, 1975), a well- χ (1, N = 57) = 0.97, p = .325. validated 30-item test that measures memory, orienta- tion, attention, and language. Items were summed, with Duchenne and non-duchenne smiles higher scores indicating higher cognitive functioning. Group differences in bvFTD and AD Duchenne and non- When interpreting scores, sums between 25 and 30 indi- Duchenne smiles were examined using independent t-tests. cate questionable cognitive impairment, 20–25 indicate No group differences were found in patient Duchenne mild cognitive impairment, 10–20 indicate moderate t(55) = 0.36, p = .721 and non-Duchenne smiles t(55) = 0.63, cognitive impairment, and 0–10 indicate severe cognitive p = .530, or in caregiver Duchenne t(55) = 0.24, p = .809 impairment. and non-Duchenne smiles t(55) = −0.15, p = .881. To examine correlations between patient and caregiver Duchenne smiles, Pearson’s r and Spearman’s rho were Results both used. Spearman’s rho was included because behav- Data Analysis ioral data often shows non-normal distributions, and Preliminary analyses were conducted to examine whether thus may benefit from a non-parametric analysis. Results patients with bvFTD and AD differed in age, sex, cogni- indicated a significant association between patient and tive functioning, Duchenne smiles, and non-Duchenne caregiver Duchenne smiles (r  =  0.38, p  =  .004; r   =  0.38, smiles. Similar analyses were conducted to examine p = .003), but no significant association between patient Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 980 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 and caregiver non-Duchenne smiles (r  =  0.21, p  =  .120; also significantly associated with caregiver mental health, r  = .20, p = .131. B = −0.71, SE(B) = 0.26, β = −0.34, p = .009, CI = [−1.24, −0.18]; caregivers of patients with AD reported having bet- ter mental health than caregivers of patients with bvFTD. Marital satisfaction Patient and caregiver non-Duchenne smiles were not asso- Group differences in caregivers’ self-reported marital ciated with caregiver mental health (ps > 0.159). Model fit satisfaction were analyzed using an independent t-test. was significant (R  = 0.144, F(5, 51) = 2.89, p = .023). Differences in caregiver marital satisfaction approached adjusted Because we observed an association between patient significance across groups, t(55)  =  1.95, p  =  .056, with Duchenne smiles and caregiver mental health, we deter- caregivers of patients with bvFTD reporting lower mari- mined whether this relationship varied by diagnosis by fit- tal satisfaction scores (M  =  96.40, SD  =  28.25) than car- ting another regression model that included the interaction egivers of patients with AD (M  =  111.65, SD  =  30.71). of patient Duchenne smiles and diagnosis. As seen in Table 2 Although group differences only approached significance, (model 3), analyses revealed that patient Duchenne smiles caregiver marital satisfaction was entered as a covariate in (B  =  0.38, SE(B)  =  0.16, β  =  0.33, p  =  .027, CI  =  [0.05, our analyses given prior findings of links between marital 0.70]) and patient diagnosis remained significantly asso- satisfaction and caregiver mental health (Kouros, Papp, & ciated with better caregiver mental health (B  =  −0.71, Cummings, 2008). SE(B)  =  0.27, β  =  −0.34, p  =  .010, CI  =  [−1.25, −0.18]), but no association was found between the interaction term Cognitive functioning (patient Duchenne smiles × diagnosis) and caregiver men- Group differences in AD and bvFTD patients’ Mini- tal health (B  =  −0.03, SE(B)  =  0.14, β  =  −0.03, p  =  .850, Mental State Exam (MMSE) scores were analyzed using an CI = [−0.31, 0.26]). The associations between patient non- independent t-test. Patient groups differed significantly in Duchenne smiles and both caregiver smiles (Duchenne and MMSE, t(55) = −2.52, p = .015, with patients with bvFTD non-Duchenne) and caregiver mental health also remained having higher scores (M = 25.07, SD = 4.12) than patients non-significant (ps > .166). Model fit remained significant with AD (M  =  21.97, SD  =  5.12). Consequently, patient (R  = 0.128, F(6, 50) = 2.37, p = .043). MMSE scores were entered as a covariate in our analyses. adjusted To examine the robustness of our findings, we conducted the main analysis while also accounting for patient diag- Duchenne smiles and caregiver mental health nosis and two additional variables that differed between As depicted in Table  2 (model 1), the regression ana- diagnostic groups—cognitive functioning (i.e., MMSE) and lysis revealed that patient Duchenne smiles were signifi- caregiver marital satisfaction. As depicted in Table 2 (model cantly associated with caregiver mental health, B  =  0.39, 4), results indicated that patient Duchenne smiles remained SE(B) = 0.17, β = 0.34, p = .029, CI = [0.05, 0.73] whereas significantly associated with better caregiver mental health caregiver Duchenne smiles were not, B = −0.08, SE(B) = 0.18, (B  =  0.42, SE(B)  =  0.15, β  =  0.37, p  =  .008, CI  =  [0.12, β = −0.07, p = .669, CI = [−0.43, 0.28]. When this analysis 0.72]). In addition, patient non-Duchenne smiles became was repeated with diagnosis included as a covariate (see associated with worse caregiver mental health (B = −0.30, Table  2, model 2), these results remained significant for SE(B)  =  0.14, β  =  −0.27, p  =  .036, CI  =  [−0.57, −0.02]). patient Duchenne smiles (B = 0.37, SE(B) = 0.16, β = 0.32, Marital satisfaction was also significantly associated with p  =  .025, CI  =  [0.05, 0.69]) while caregiver Duchenne caregiver mental health (B = 0.01, SE(B) = 0.00, β = 0.38, smiles remained non-significant (B =  −0.09, SE(B) = 0.17, 2,3 p = .003, CI = [0.005, 0.022]. while diagnosis showed an β  =  −0.07, p  =  .609, CI  =  [−0.42, 0.25]). Diagnosis was 1. The moderate correlation between patient and care- 2. Studies have also computed the MOS-SF 36 mental giver Duchenne smiles suggests that patients who expressed health composite scale by summing the following four more genuine smiles have caregivers who do the same. To subscales: vitality, social functioning, limitations due examine whether patients and caregivers who expressed to emotional problems, and mental health (Zhu et  al., more genuine smiles may also report greater caregiver 2016). When analyses were conducted using this alter- marital satisfaction and mental health, a median split native approach, results indicated that patient Duchenne was computed in order to group patients who expressed smiles were associated with caregiver mental health at fewer Duchenne smiles and patients who expressed greater near significant levels B =25.65, SE( B) = 13.06, β = 0.29, Duchenne smiles. Group differences in caregiver marital p  =  .055, CI  =  [−0.56, 51.86], but caregiver Duchenne satisfaction and mental health were then examined using smiles, and both patient and caregiver non-Duchenne independent t-tests. Analyses revealed a non-significant dif- smiles were not (ps > .203). When accounting for patient ference in marital satisfaction t(55)  =  0.14, p  =  .892 and cognitive functioning and caregiver marital satisfac- a difference that trended towards significance in caregiver tion, patient Duchenne smiles remained positively asso- mental health t(55) = −1.94, p = .058, such that caregivers ciated with better caregiver mental health (B  =  29.57, of patients who expressed more Duchenne smiles reported SE(B)  =  12.22, β  =  0.33, p  =  .019, CI  =  [5.01, 54.12]) higher mental health scores (M = 0.27, SD = 0.77) than car- while patient non-Duchenne smiles remained associ- egivers of patients who expressed fewer Duchenne smiles ated with worse caregiver mental health (B  =  23.33, (M = −0.26, SD = 1.26). SE(B) = 11.12, β = −0.27, p = .041, CI = [−45.67, −0.99]). Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 981 Table 2. Duchenne and non-Duchenne Smiles as Predictors of Caregiver Mental Health Caregiver mental health Model 1 Model 2 Model 3 Model 4 Model 5 B (SE[B]) β B (SE[B]) β B (SE[B]) β B (SE[B]) β B (SE[B]) β Patient Duchenne 0.39 (0.17) 0.34* 0.37 (0.16) 0.32* 0.38 (0.16) 0.33* 0.42 (0.15) 0.37** 0.37 (0.16) 0.33* Caregiver −0.08 (0.18) −0.07 −0.09 (0.17) −0.07 −0.08 (0.17) −0.07 −0.11 (0.16) −0.09 −0.10 (0.17) −0.09 Duchenne Patient −0.17 (0.15) −0.16 −0.21 (0.14) −0.18 −0.20 (0.15) −0.18 −0.30 (0.14) −0.27* −0.20 (0.15) −0.18 non-Duchenne Caregiver −0.12 (0.18) −0.11 −0.09 (0.17) −0.08 −0.10 (0.18) −0.09 −0.12 (0.16) −0.11 −0.09 (0.17) −0.08 non-Duchenne Diagnosis — — −0.71 (0.26) −0.34** −0.71 (0.27) −0.34** −0.53 (0.27) 0.25 — — Patient Duchenne — — — — −0.03 (0.14) −0.03 — — — — × diagnosis Patient non- — — — — — — — — −0.05 (0.15) −0.05 Duchenne × diagnosis Cognitive — — — — — — 0.004 (0.03) 0.02 — — functioning Marital — — — — — — 0.01 (0.00) 0.38** — — satisfaction Note. Covariates include patient diagnosis and cognitive functioning, and caregiver marital satisfaction. Diagnosis was dummy-coded; patients with AD were set as the reference group. A dash (—) indicates that the given variable was not included within the model. *p < .05, **p < .01. association that trended towards significance (B  =  −0.53, fit remained significant (R   =  0.255, F(7, 49)  =  3.74, adjusted SE(B)  =  0.27, β  =  −0.25, p  =  .052, CI  =  [−1.06, 0.004]). p = .003). Cognitive functioning was not associated with caregiver Because we observed an association between patient non- mental health (B = 0.004, SE(B) = 0.03, β = 0.02, p = .871, Duchenne smiles and caregiver mental health, we determined CI  =  [−0.05, 0.06]). The associations between caregiver whether this relationship varied by diagnosis by fitting another Duchenne and non-Duchenne smiles and caregiver men- regression model that included the four predictor variables tal health remained non-significant (ps  >  .439). Model (patient and caregiver Duchenne and non-Duchenne smiles) The associations between caregiver Duchenne and non- included both patient and caregiver Duchenne and non- Duchenne smiles and caregiver mental health remained Duchenne smiles). Patient Duchenne smiles remained sig- non-significant (ps > .628). nificantly associated with caregiver mental health, B = 0.39, 3. Same-sex spouses may reveal unique differences due to SE(B) = 0.17, β = 0.34, p = .028, CI = [0.04, 0.73] while care- gender, as women tend to smile more than men (LaFrance, giver Duchenne smiles remained non-significant, B = 0.05, Hecht, & Paluck, 2003), or additional stressors due to soci- SE(B) = 0.20, β = 0.04, p = .813, CI = [−0.36, 0.45]. Patient etal stigma (Lewis, Derlega, Griffin, & Krowinski, 2003). and caregiver non-Duchenne smiles likewise remained non- To examine whether same-sex couples impacted our find- significant (ps  >  .299). When analyses included length of ings, we re-ran analyses without these two couples. Results union as an additional covariate in Model 4 (where pre- remained significant; patient Duchenne smiles remained dictors included both patient and caregiver Duchenne and significantly associated with caregiver mental health, non-Duchenne smiles, as well as the covariates of patient B = 0.38, SE(B) = 0.16, β = 0.33, p = .023, CI = [0.05, 0.71] diagnosis and cognitive functioning, and caregiver mari- while caregiver Duchenne smiles were not, B  =  −0.11, tal satisfaction), patient Duchenne smiles remained signifi- SE(B) = 0.17, β = −0.10, p = .512, CI = [−0.46, 0.23]. The cantly associated with caregiver mental health, B  =  0.43, relationship between patient non-Duchenne smiles and SE(B)  =  0.16, β  =  0.37, p = .009, CI = [0.11, 0.74] while caregiver non-Duchenne smiles and caregiver mental health caregiver Duchenne smiles remained non-significant, remained non-significant (ps  >  .096) Results may also be B = −0.01, SE(B) = 0.18, β = −0.01, p = .974, CI = [−0.38, affected by the length of time spouses have been married, as 0.36]. Patient diagnosis B = −0.62, SE(B) = 0.28, β = −0.29, shorter times indicate newer spouses or second marriages, p = .030, CI = [−1.18, −0.06] and caregiver marital satisfac- which have been found to affect marital quality and sat- tion B = 0.01, SE(B) = 0.01, β = 0.34, p = .013, CI = [0.003, isfaction (Bograd & Spilka, 1996; Coleman, Ganong, & 0.02] remained significantly associated with caregiver men- Fine, 2000). To examine whether length of union impacted tal health, while patient and caregiver non-Duchenne smiles our findings, results were examined with length of union and patient cognitive functioning remained non-significant included as a covariate in Model 1 (where predictors (ps > .072). Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 982 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 and the interaction of patient non-Duchenne smiles and diag- cognitive functioning, and caregiver marital satisfaction), nosis. As seen in Table  2 (model 5), the analysis revealed a more patient Duchenne smiles remained a significant pre- non-significant interaction term (B  =  −0.04, SE(B)  =  0.15, dictor of better caregiver mental health, and fewer patient β = −0.05, p = .746, CI = [−0.35, 0.25]), which remained non- non-Duchenne smiles became a significant predictor of significant when the analysis was repeated with additional worse caregiver mental health. Further analysis revealed covariates included (i.e., diagnosis, martial satisfaction, and that marital satisfaction, but not patient cognitive function- cognitive functioning; B  =  −0.01, SE(B)  =  0.14, β  =  −0.01, ing, provided a likely suppressor effect in the association p = .945, CI = [−0.28, 0.26]). between greater patient non-Duchenne smiles and worse To further understand the potential suppressor effect in caregiver mental health. Finally, an exploratory analysis the association between patient non-Duchenne smiles and revealed that the association between patient Duchenne caregiver mental health, two additional post hoc regressions smiles and caregiver health was not mediated by caregiver were conducted. In the first, the four kinds of smiles (patient marital satisfaction. and caregiver Duchenne and non-Duchenne) and patient Although analyses revealed that marital satisfaction was cognitive functioning were entered as predictors, and car- not a significant mediator, both marital satisfaction and egiver mental health was the dependent variable. Patient patient diagnosis were significant predictors of caregiver non-Duchenne smiles were not significantly associated well-being. These findings support studies that have found with caregiver mental health in this regression (B = −0.19, marital satisfaction and marital quality to reduce negative SE(B)  =  0.15, β  =  −0.17, p  = .225, CI = [−0.49, 0.12]). In affect (Carr, Cornman, & Freedman, 2016) and serve as the second, the four kinds of smiles (patient and caregiver key protective factors against the adverse effects of care- Duchenne and non-Duchenne) and marital satisfaction giver burden. For example, caregivers who report higher were entered as predictors, and caregiver mental health levels of marital satisfaction have been found to be less was the dependent variable. In this regression, patient non- reactive to memory and behavioral changes in dementia Duchenne smiles were significantly associated with car - patients and engage in better problem solving skills than egiver mental health (B  =  −0.29, SE(B)  =  0.14, β  =  −0.26, caregivers who reported lower levels of marital satisfaction p  = .042, CI = [−0.57, −0.01]), indicating that marital sat- (Steadman, Tremont, & Duncan Davis, 2007). Similarly, isfaction may be acting as a suppressor variable. To further several studies have found that patient diagnosis can also examine this effect, a median split was computed in order to impact caregiver outcomes. Behavioral symptoms, such as group patients who expressed fewer non-Duchenne smiles apathy and disinhibition, which tend to have the greatest and patients who expressed greater non-Duchenne smiles. negative impact on caregiver outcomes (Merrilees et  al., Further analyses revealed marital satisfaction to be signifi- 2013; Mioshi et al., 2013), are more commonly observed in cantly associated with caregiver mental health (r  =  0.55, frontotemporal dementia than in AD. Thus, although our p = .002) for patients who expressed more non-Duchenne study revealed that patient smiles were positively associ- smiles; this association was not significant for patients who ated with caregiver mental health even when accounting expressed fewer non-Duchenne smiles (r = 0.33, p = .10). for patient diagnosis and caregiver marital satisfaction, it Finally, having established an association between illuminates just one piece of the complex processes that in- patient Duchenne smiles and caregiver mental health, fluence caregivers’ mental health. we conducted an exploratory mediation analysis to test whether this association was mediated by marital satisfac- Implications of Patient Duchenne Smiles and tion. Using the PROCESS macro (Hayes, 2008) with 50,000 Caregiver Mental Health bias-corrected bootstrapped samples, results revealed a non-significant mediation, B  =  −0.008, SE  =  (0.065), Our findings indicate that the mental health of caregiv- CI  = [−0.15, 0.12]. ers of patients with bvFTD and AD is associated with signs of genuine positive emotion expressed by the person in their care. However, the particular mechanisms driv- Discussion ing this association are still unknown. Smiles serve many We examined whether smiles expressed by patients with interpersonal functions. For example, Duchenne smiles bvFTD and AD and their spousal caregivers during an convey affiliation, warmth, and intimacy (Hess, Beaupré, unrehearsed, semi-naturalistic conversation about a rela- & Cheung, 2002), and elicit more positive judgments and tionship conflict were associated with caregiver men- more cooperation from others than do non-Duchenne tal health. Results revealed that more patient Duchenne smiles (Johnston et  al., 2010). These functions may be smiles were associated with better caregiver mental health. particularly important in late life, because older adults These findings generalized across diagnosis, and were spe- have been found to appraise social rejection more nega- cific to patient Duchenne smiles; no association was found tively than younger adults (Cheng & Grühn, 2015), and between caregiver smiles (Duchenne or non-Duchenne) and a lack of affiliation has been linked with greater cognitive caregiver mental health. When examining this association decline and dementia in late life (Rafnsson, Orrell, d’Orsi, while accounting for covariates (i.e., diagnosis, patient Hogervorst, & Steptoe, 2017). Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 983 Extrapolating from these findings, for married cou- emotional behavior, due to the substantial impact that ples where one spouse has dementia, Duchenne smiles patient behaviors have on caregiver mental health. Many expressed by the spouse with dementia may convey affili- studies have found that behavioral symptoms in patients ation, which could have a soothing effect on the caregiver. with dementia are most closely linked with adverse care- Previous research has shown that affiliation can reduce giver health outcomes (Schulz & Eden, 2016). stress and HPA activity (DeVries, Glasper, & Detillion, Our findings have a number of practical implications. 2003), and that holding a spouse’s hand reduces activity Caregivers of dementia patients experience significant lev- in brain regions associated with anxiety (Coan, Schaefer, els of burden and are vulnerable to declining mental health & Davidson, 2006). Duchenne smiles may be another way (Schulz & Eden, 2016). Interventions designed to prevent or in which partners affiliate and thereby soothe each other, treat these problems in caregivers could benefit from target- which may be especially helpful for caregivers who are ing specific factors, such as altering patient behaviors (e.g., likely to be experiencing high levels of stress. Consistent reinforcing the expression of positive emotion) or compen- with this, in previous studies with individuals and couples, sating for related losses (e.g., support groups and friends we have found that the expression of positive emotions is becoming a source of emotionally positive interactions for associated with a reduction of autonomic nervous system caregivers). Given the effects positive emotions have on arousal (Fredrickson & Levenson, 1998; Yuan, McCarthy, reducing autonomic arousal (Yuan et al., 2010), it may also Holley, & Levenson, 2010). be useful for caregiver interventions to incorporate com- Alternately, Duchenne smiles may serve to elicit greater ponents that focus on self-soothing, such as mindfulness cooperation from others (Johnston et al., 2010). Patients’ exercises (Raes, Bruyneel, Loeys, Moerkerke, & De Raedt, Duchenne smiles may make spousal caregivers more will- 2015) and other individualized activities that can increase ing to help the patient, thus helping caregivers feel less self-care. Finally, providing psychoeducation for caregivers resentful of or trapped in their caregiving role, which could concerning the likely reduction of positive emotion that will foster better mental health. Experiencing warmth and posi- accompany disease progression can help caregivers under- tive emotions has also been linked to caregiver satisfaction stand that this reduction is not a deliberate behavior on (Carruth, Tate, Moffett, & Hill, 1997), which in turn can the part of the patient. This understanding can help buffer increase caregiver mental health. caregivers from the frustration and sense of loss associated In the present study, we also found some evidence that with diminished expression of positive emotion on the part more patient non-Duchenne smiles were associated with of a loved one with dementia. worse caregiver mental health. This finding only emerged when we accounted for differences in patient diagnosis, pa- Strengths and Limitations tient cognitive functioning, and caregiver marital satisfaction. Examining these covariates separately revealed that it was the Strengths of this research include studying a community inclusion of marital satisfaction that revealed the association sample of participants with two different kinds of demen- between more non-Duchenne smiles by patients and worse tias, the use of objective behavioral measurement of positive caregiver mental health. These analyses suggest that higher emotions, and studying positive emotion in an ecologically marital satisfaction may serve as a particularly important valid semi-naturalistic interpersonal context. To our know- buffer for maintaining caregiver mental health in relation- ledge, this is the first study to use an objective behavioral ships where the patient expresses greater smiles that are measure to assess positive emotion in the interactions of merely “polite” when communicating with their caregiver. couples where one spouse has dementia, and the first to The negative association between more non-Duchenne smiles link positive emotion measured in this manner with care- and worse caregiver mental health is consistent with previous giver mental health. research that has found non-Duchenne smiles to be associ- The research also has several limitations. Although our ated with more negative outcomes. For example, in healthy sample size of 57 patient–caregiver dyads is relatively large married couples, non-Duchenne smiles during a conversation compared to some laboratory studies that have examined were associated with more time separated over the following dementia patients, sample size remains a limitation. A four years (Gottman, Levenson, & Woodin, 2001). priori power analyses indicated that for power = 0.80 and One striking non-finding in the present study was that four predictors, a sample size of 85 dyads would have been caregiver smiles (Duchenne and non-Duchenne) were not required to detect a small effect size. Thus, some of the asso- associated with caregivers’ mental health. We expected car- ciations that were non-significant within our sample may egivers’ positive emotional behavior to be related to their reach significance in larger samples with additional power. own mental health, consistent with research that has gener- Other methodological limitations include our exclusive ally found positive emotion to serve as a protective buffer focus on positive emotion (we did not consider negative against negative outcomes (Fredrickson & Joiner, 2002). [e.g., anger, sadness] or self-conscious [e.g., shame, pride] The association between positive emotional behavior and emotional expressions), examining a brief, “thin slice” be- caregiver mental health may have been significant for pa- havior (Ambady et al., 2000), using a cross-sectional rather tient positive emotional behavior, but not caregiver positive than a longitudinal design, and not including a sample of Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 984 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 healthy control couples (which would help determine the Duchenne smiles will be important factors to examine. extent to which findings are unique to couples dealing with Additionally, data on HPA activity might reveal a medi- dementia). ating effect at the physiological level. Further research An additional limitation is that we cannot know with examining these mechanisms would help to clarify the certainty whether observable Duchenne smiles are spon- association between patient Duchenne smiles and care- taneous, veridical indicators of participants’ underlying giver mental health. positive emotional states, or are being produced strategic- In research with individuals with dementia, it is com- ally and deliberately (Gunnery, Hall, & Ruben, 2013). pelling to consider the disease as the primary cause of Although impairments in patients’ cognitive, motor, and changes in patients’ emotional behavior and the behav- social functioning make this level of self-monitoring, vol- ioral changes as the primary cause of changes in car- untary control of facial behavior, and strategic impression egivers’ mental health. Thus, we suspect that dementia management less likely (Gregory et  al., 2002), it is still is causing declining levels of patient Duchenne smiles, possible that patients are not actually experiencing posi- which cause declines in caregiver mental health. However, tive emotion, but rather have learned to produce Duchenne elements of this causal chain may function in the reverse smiles as a way of eliciting cooperation and reducing dis- direction (i.e., caregivers with worse mental health may pleasure in caregivers. cause patients to produce lower levels of Duchenne smiles). Evaluating these possibilities will require longi- tudinal designs that measure potential mediating mecha- Future Directions nisms at multiple time-points. In future studies, it will be important to extend this research to include other indicators of emotion (e.g., words, ges- Conclusion tures), negative and self-conscious emotions, and other contexts (e.g., other kinds of conversations and shared The human face has an exquisitely tuned facial muscula- activities). Similarly, it will be useful to examine these asso- ture that plays a crucial role in communicating our emo- ciations in patients with other dementias as well as psychi- tions. A  Duchenne, or “genuine,” smile, moves only two atric disorders. facial muscles—the orbicularis oculi (cheek raise) and Several timing issues are also important directions for zygomatic major (lip corner raise). Nonetheless, this kind future studies. Patients with dementia and other neuro- of smile has been found to be associated with a number logical diseases may react to social stimuli at a slower of positive outcomes including, in the present study, better pace than adults without neurological disorders. No lit- mental health in spousal caregivers of dementia patients. erature currently addresses this question, but it would Given the increasing prevalence of dementia and the rap- be useful to determine whether patients show delayed idly growing aging population worldwide, understanding emotional responses to stimuli, and whether those delays the role that positive emotional behaviors play in caregiver impact caregiver mental health. In addition, the present mental health has important implications for helping pre- study focused on the initial segment of the conversation, serve the well-being of caregivers and for advancing our which has previously been found to be related to rela- understanding of emotional expression and its effects on tionship quality (Carrère & Gottman, 1999). However, others. future studies would benefit from examining emotional behavior during other parts of the conversation (e.g., the Funding ending) as well as the trajectory of emotional changes over time. This research was supported by National Institute on Aging grants It will also be important to consider possible mecha- 1R01AG041762-01A1 and 2P01AG019724-11 to R.W.L. and B.L.M. nisms of the effects found in this study. Exploratory analy- ses in the present study suggest that marital satisfaction is not a likely mechanism undergirding the relationship Acknowledgements between patient Duchenne smiles and caregiver mental S.J.L., J.J.C., J.M., and R.W.L. developed the study concept. S.J.L., health. Previous research has shown that Duchenne smiles J.J.C., A.V., and D.E.C.  performed the facial coding. S.J.L.  pro- elicit in others a greater willingness to cooperate (Johnston cessed and analyzed the data. S.J.L. and J.J.C. wrote the first draft et al., 2010). When patients show more Duchenne smiles, of the manuscript; and all authors contributed to revisions of the caregivers may feel more willing to help the patient with manuscript. R.W.L.  supervised all phases of the project. S.J.L.  and J.J.C. contributed equally to the study. activities of daily living instead of feeling obligated and trapped in the caregiving role, and may experience bet- ter mental health as a result. Thus, mechanisms such as Conflict of Interest caregivers’ subjective experience of decreased burden, or increased positive emotion as the result of receiving There were no identified conflicts of interest. Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 985 Coleman, M., Ganong, L., & Fine, M. (2000). Reinvestigating References remarriage: Another decade of progress. Journal of Ambady, N., Bernieri, F. J., & Richeson, J. A. (2000). Toward a hist- Marriage and Family, 62, 1288–1307. doi:10.1111/ ology of social behavior: Judgmental accuracy from thin slices j.1741-3737.2000.01288.x of the behavioral stream. In M. P. Zanna (Eds.), Advances in Cornwell, B., Laumann, E. O., & Schumm, L. P. (2008). experimental social psychology (Vol. 32, pp. 201–271). San The social connectedness of older adults: a national pro- Francisco, CA: Academic Press. file*. American Sociological Review, 73, 185–203. Ascher, E. A., Sturm, V. E., Seider, B. H., Holley, S. R., Miller, B. L., & doi:10.1177/000312240807300201 Levenson, R. W. (2010). Relationship satisfaction and emotional DeVries, A. C., Glasper, E. R., & Detillion, C. E. (2003). Social language in frontotemporal dementia and Alzheimer disease modulation of stress responses. Physiology & Behavior, 79, patients and spousal caregivers. Alzheimer Disease and Associated 399–407. doi:10.1016/S0031-9384(03)00152-5 Disorders, 24, 49–55. doi:10.1097/WAD.0b013e3181bd66a3 Ekman, P., & Friesen, W. V. (1977). Facial action coding system. Palo Blieszner, R., & Shifflett, P. A. (1990). The effects of Alzheimer’s Alto, CA: Consulting Psychologists Press. disease on close relationships between patients and caregivers. Ekman, P., & Friesen, W. V. (1982). Felt, false, and miserable smiles. Family Relations, 39, 57–62. doi:10.2307/584949 Journal of Nonverbal Behavior, 6, 238–252. doi:10.1007/ Bograd, R., & Spilka, B. (1996). Self-disclosure and marital satisfac- BF00987191 tion in mid-life and late-life remarriages. International Journal Ekman, P., & Rosenberg, E. L. (2005). What the face reveals: Basic of Aging & Human Development, 42, 161–172. doi:10.2190/ and applied studies of spontaneous expression using the facial W87M-WCK7-MHTT-N34F action coding system (FACS) (2nd ed.). New York, NY: Oxford Bourke-Taylor, H., Pallant, J. F., Law, M., & Howie, L. (2012). University Press. Predicting mental health among mothers of school-aged chil- Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). “Mini- dren with developmental disabilities: The relative contribu- mental state”. A practical method for grading the cognitive state tion of child, maternal and environmental factors. Research of patients for the clinician. Journal of Psychiatric Research, 12, in Developmental Disabilities, 33, 1732–1740. doi:10.1016/j. 189–198. doi:10.1016/0022-3956(75)90026-6 ridd.2012.04.011 Fredrickson, B. L. (2004). The broaden-and-build theory of posi- Braun, M., Mura, K., Peter-Wight, M., Hornung, R., & Scholz, U. tive emotions. Philosophical Transactions of the Royal Society (2010). Toward a better understanding of psychological well- of London Series B: Biological Sciences, 359, 1367–1378. being in dementia caregivers: The link between marital com- doi:10.1098/rstb.2004.1512 munication and depression. Family Process, 49, 185–203. Fredrickson, B. L., & Joiner, T. (2002). Positive emotions trigger doi:10.1111/j.1545-5300.2010.01317.x upward spirals toward emotional well-being. Psychological Campos, B., Shiota, M. N., Keltner, D., Gonzaga, G. C., & Goetz, J. L. Science, 13, 172–175. doi:10.1111/1467-9280.00431 (2013). What is shared, what is different? Core relational themes Fredrickson, B. L., & Levenson, R. W. (1998). Positive emo- and expressive displays of eight positive emotions. Cognition & tions speed recovery from the cardiovascular sequelae of Emotion, 27, 37–52. doi:10.1080/02699931.2012.683852 negative emotions. Cognition & Emotion, 12, 191–220. Carr, D., Cornman, J. C., & Freedman, V. A. (2016). Marital qual- doi:10.1080/026999398379718 ity and negative experienced well-being: An assessment of actor Fredrickson, B. L., Tugade, M. M., Waugh, C. E., & Larkin, G. and partner effects among older married persons. The Journals R. (2003). What good are positive emotions in crises? A  pro- of Gerontology Series B: Psychological Sciences and Social spective study of resilience and emotions following the ter- Sciences, 71, 177–187. doi:10.1093/geronb/gbv073 th rorist attacks on the United States on September 11 , 2001. Carrère, S., & Gottman, J. M. (1999). Predicting divorce Journal of Personality and Social Psychology, 84, 365–376. among newlyweds from the first three minutes of doi:10.1037/0022-3514.84.2.365 a marital conflict discussion. Family Process, 38, Gonzaga, G. C., Keltner, D., Londahl, E. A., & Smith, M. D. (2001). 293–301. doi:10.1111/j.1545-5300.1999.00293.x Love and the commitment problem in romantic relations and Carruth, A. K., Tate, U. S., Moffett, B. S., & Hill, K. (1997). friendship. Journal of Personality and Social Psychology, 81, Reciprocity, emotional well-being, and family functioning as 247–262. doi:10.1037/0022-3514.81.2.247 determinants of family satisfaction in caregivers of elderly par- Gottman, J., Levenson, R., & Woodin, E. (2001). Facial expressions ents. Nursing Research, 46, 93–100. doi:10.1097/00006199- during marital conflict. Journal of Family Communication, 1, 199703000-00006 37–57. doi:10.1207/S15327698JFC0101_06 Cheng, Y., & Grühn, D. (2015). Age differences in reactions to social Gregory, C., Lough, S., Stone, V., Erzinclioglu, S., Martin, L., rejection: The role of cognitive resources and appraisals. The Baron-Cohen, S., & Hodges, J. R. (2002). Theory of mind in Journals of Gerontology Series B: Psychological Sciences and patients with frontal variant frontotemporal dementia and Social Sciences, 70, 830–839. doi:10.1093/geronb/gbu054 Alzheimer’s disease: Theoretical and practical implications. Chopik, W. J., & O’Brien, E. (2017). Happy you, healthy me? Having Brain: A  Journal of Neurology, 125, 752–764. doi:10.1093/ a happy partner is independently associated with better health in brain/awf079 oneself. Health Psychology, 36, 21–30. doi:10.1037/hea0000432 Gunnery, S. D., Hall, J. A., & Ruben, M. A. (2013). The deliber- Coan, J. A., Schaefer, H. S., & Davidson, R. J. (2006). Lending a hand: ate Duchenne smile: Individual differences in expressive con- Social regulation of the neural response to threat. Psychological trol. Journal of Nonverbal Behavior, 37, 29–41. doi:10.1007/ Science, 17, 1032–1039. doi:10.1111/j.1467-9280.2006.01832.x s10919-012-0139-4 Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV 986 Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 Haase, C. M., Beermann, U., Saslow, L. R., et  al. (2015). Short Locke, H. J., & Wallace, K. M. (1959). Short marital-adjustment and prediction tests: Their reliability and validity. Marriage and alleles, bigger smiles? The effect of 5-HTTLPR on positive emo- tional expressions. Emotion (Washington, D.C.), 15, 438–448. Family Living, 21, 251–255. doi:10.2307/348022 Losada, M., & Heaphy, E. (2004). The role of positivity and con- doi:10.1037/emo0000074 Harker, L., & Keltner, D. (2001). Expressions of positive emo- nectivity in the performance of business teams: A  nonlinear dynamics model. American Behavioral Scientist, 47, 740–765. tion in women’s college yearbook pictures and their rela- tionship to personality and life outcomes across adulthood. doi:10.1177/0002764203260208 Mahoney, R., Regan, C., Katona, C., & Livingston, G. (2005). Anxiety Journal of Personality and Social Psychology, 80, 112–124. doi:10.1037/0022-3514.80.1.112 and depression in family caregivers of people with Alzheimer disease: The LASER-AD study. The American Journal of Geriatric Psychiatry, Hawthorne, G., Osborne, R. H., Taylor, A., & Sansoni, J. (2007). The SF36 Version 2: Critical analyses of population weights, scoring 13, 795–801. doi:10.1097/00019442-200509000-00008 McHorney, C. A., Ware, J. E. Jr, & Raczek, A. E. (1993). algorithms and population norms. Quality of Life Research, 16, 661–673. doi:10.1007/s11136-006-9154-4 The MOS 36-item short-form health survey (SF-36): II. Psychometric and clinical tests of validity in measuring phys- Hayes, A. F. (2008). Introduction to mediation, moderation, and conditional process analysis: A regression-based approach. New ical and mental health constructs. Medical Care, 31, 247–263. doi:10.1097/00005650-199303000-00006 York, NY: Guilford Press. Hess, U., Beaupré, M. G., & Cheung, N. (2002). Who to whom McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., & Stadlan, E. M. (1984). Clinical diagnosis of Alzheimer’s dis- and why–cultural differences and similarities in the function of smiles. In M. Abel (Ed.), An empirical reflection on the smile (pp. ease: Report of the NINCDS-ADRDA work group under the auspices of department of health and human services task force 187–216). New York: Edwin Mellen Press. Johnston, L., Miles, L., & Macrae, C. N. (2010). Why are you smil- on Alzheimer’s disease. Neurology, 34, 939–944. doi:10.1212/ WNL.34.7.939 ing at me? Social functions of enjoyment and non-enjoyment smiles. The British Journal of Social Psychology, 49, 107–127. Merrilees, J., Dowling, G. A., Hubbard, E., Mastick, J., Ketelle, R., & Miller, B. L. (2013). Characterization of apathy in persons doi:10.1348/014466609X412476 Keltner, D., & Bonanno, G. A. (1997). A study of laughter and dis- with frontotemporal dementia and the impact on family caregiv- ers. Alzheimer Disease and Associated Disorders, 27, 62–67. sociation: Distinct correlates of laughter and smiling during bereavement. Journal of Personality and Social Psychology, 73, doi:10.1097/WAD.0b013e3182471c54 Mioshi, E., Foxe, D., Leslie, F., Savage, S., Hsieh, S., Miller, L., … 687–702. doi:10.1037/0022-3514.73.4.687 Kouros, C. D., Papp, L. M., & Cummings, E. M. (2008). Interrelations Piquet, O. (2013). The impact of dementia severity on care- giver burden in frontotemporal dementia and Alzheimer dis- and moderators of longitudinal links between marital satisfac- tion and depressive symptoms among couples in established ease. Alzheimer Disease and Associated Disorders, 27, 68–73. doi:10.1097/WAD.0b013e318247a0bc relationships. Journal of Family Psychology, 22, 667–677. doi:10.1037/0893-3200.22.5.667 Nunnemann, S., Kurz, A., Leucht, S., & Diehl-Schmid, J. (2012). Caregivers of patients with frontotemporal lobar degener- LaFrance, M., Hecht, M. A., & Paluck, E. L. (2003). The contingent smile: A meta-analysis of sex differences in smiling. Psychological ation: A review of burden, problems, needs, and interventions. International Psychogeriatrics, 24, 1368–1386. doi:10.1017/ Bulletin, 129, 305–334. doi:10.1037/0033-2909.129.2.305 Levenson, R. W., & Gottman, J. M. (1983). Marital interaction: S104161021200035X Raes, A. K., Bruyneel, L., Loeys, T., Moerkerke, B., & De Raedt, R. Physiological linkage and affective exchange. Journal of P ersonality and Social Psychology, 45, 587–597. doi:10.1037/0022- (2015). Mindful attention and awareness mediate the association between age and negative affect. The Journals of Gerontology 3514.45.3.587 Levenson, R. W., Haase, C., Bloch, L., Holley, S., & Seider, B. (2013). Series B: Psychological Sciences and Social Sciences, 70, 181– Emotion regulation in couples. In J. J. Gross (Ed.), Handbook of 190. doi:10.1093/geronb/gbt074 emotion regulation (2nd ed., pp. 267–283). New York: Guilford Rafnsson, S. B., Orrell, M., d’Orsi, E., Hogervorst, E., & Steptoe, A. Press. (2017). Loneliness, social Integration, and incident dementia over Levenson, R. W., Lwi, S. J., Brown, C. L., Ford, B. Q., Otero, M. C., & 6 years: Prospective findings from the English Longitudinal Study Verstaen, A. (2017). Emotion. In J. T. Cacioppo, L. G. Tassinary of Ageing. The Journals of Gerontology Series B: Psychological & G. G. Berntson (Eds.), Handbook of psychophysiology (4th Sciences and Social Sciences. 0,  1–11.  doi:10.1093/geronb/ ed.) (pp. 444–464). Cambridge: Cambridge University Press. gbx087 Lewis, R. J., Derlega, V. J., Griffin, J. L., & Krowinski, A. C. (2003). Rascovsky, K., Hodges, J. R., Knopman, D., Mendes, M. F., Kramer, Stressors for gay men and lesbians: Life stress, gay-related J. H., Neuhaus, J., … Miller, B. L. (2011). Sensitivity of revised stress, stigma consciousness, and depressive symptoms. Journal diagnostic criteria for the behavioural variant of frontotemporal of Social and Clinical Psychology, 22, 716–729. doi:10.1521/ dementia. Brain, 134, 2456–2477. doi:10.1093/brain/awr179 jscp.22.6.716.22932 Schulz, R., & Eden, J. (2016). Families caring for an aging America. Lints-Martindale, A. C., Hadjistavropoulos, T., Barber, B., & Gibson, Washington, DC: The National Academies Press. S. J. (2007). A psychophysical investigation of the facial action Schulz, R., O’Brien, A. T., Bookwala, J., & Fleissner, K. (1995). coding system as an index of pain variability among older adults Psychiatric and physical morbidity effects of dementia caregiv- with and without Alzheimer’s disease. Pain Medicine (Malden, ing: Prevalence, correlates, and causes. The Gerontologist, 35, Mass.), 8, 678–689. doi:10.1111/j.1526-4637.2007.00358.x 771–791. doi:10.1093/geront/35.6.771 Downloaded from https://academic.oup.com/psychsocgerontology/article/74/6/975/4827599 by DeepDyve user on 19 July 2022 Copyedited by: SV Journals of Gerontology: PSYCHOLOGICAL SCIENCES, 2019, Vol. 74, No. 6 987 Schulz, R., & Sherwood, P. R. (2008). Physical and mental health medical care. JAMA, 262, 925–930. doi:10.1001/jama. effects of family caregiving. The American Journal of Nursing, 1989.03430070073033 108, 23–7; quiz 27. doi:10.1097/01.NAJ.0000336406.45248.4c Ware, J. (2000). SF-36 health survey update. Spine, 25, 3130–3139. Seeley, W. W., Allman, J. M., Carlin, D. A., Crawford, R. K., Macedo, doi:10.1097/00007632-200012150-00008 M. N., Greicius, M. D., Dearmond, S. J., … Miller, B. L. (2007). Ware, J. E., Kosinski, M., & Keller, S. (1994). SF-36 physical Divergent social functioning in behavioral variant frontotem- and mental health summary scales: A  user’s manual. Health poral dementia and Alzheimer disease: reciprocal networks Assessment Lab.  Boston, MA: Health Institute, New England and neuronal evolution. Alzheimer Disease and Associated Medical Center. Disorders, 21, S50–S57. doi:10.1097/WAD.0b013e31815c0f14 Wichers, M. C., Myin-Germeys, I., Jacobs, N., Peeters, F., Steadman, P. L., Tremont, G., & Davis, J. D. (2007). Premorbid rela- Kenis, G., Derom, C., … van Os, J. (2007). Evidence that tionship satisfaction and caregiver burden in dementia caregiv- moment-to-moment variation in positive emotions buf- ers. Journal of Geriatric Psychiatry and Neurology, 20, 115–119. fer genetic risk for depression: A  momentary assessment doi:10.1177/0891988706298624 twin study. Acta Psychiatrica Scandinavica, 115, 451–457. Strathearn, L., Li, J., Fonagy, P., & Montague, P. R. (2008). What’s doi:10.1111/j.1600-0447.2006.00924.x in a smile? Maternal brain responses to infant facial cues. Yuan, J. W., McCarthy, M., Holley, S. R., & Levenson, R. W. (2010). Pediatrics, 122, 40–51. doi:10.1542/peds.2007-1566 Physiological down-regulation and positive emotion in mari- Sturm, V. E., McCarthy, M. E., Yun, I., Madan, A., Yuan, J. W., Holley, S. tal interaction. Emotion (Washington, D.C.), 10, 467–474. R., … Levenson, R. W. (2011). Mutual gaze in Alzheimer’s disease, doi:10.1037/a0018699 frontotemporal and semantic dementia couples. Social Cognitive Zhu, Y. X., Li, T., Fan, S. R., Liu, X. P., Liang, Y. H., & Liu, P. (2016). Health-related quality of life as measured with the Short-Form and Affective Neuroscience, 6, 359–367. doi:10.1093/scan/nsq055 Tarlov, A. R., Ware, J. E. Jr, Greenfield, S., Nelson, E. C., Perrin, 36 (SF-36) Questionnaire in patients with recurrent vulvovagi- nal candidiasis. Health and Quality of Life Outcomes, 14, 65. E., & Zubkoff, M. (1989). The medical outcomes study. An application of methods for monitoring the results of doi:10.1186/s12955-016-0470-2

Journal

"The Journals of Gerontology - Series B: Psychological Sciences and Social Sciences" – Oxford University Press

Published: Aug 21, 2019

Keywords: dementia; emotions; caregivers; marriage, life event; mental health; positive emotionality; mental processes; cognitive ability

There are no references for this article.