Content uploaded by Sheila E Crowell
All content in this area was uploaded by Sheila E Crowell on Mar 31, 2018
Content may be subject to copyright.
Cardiovascular Reactivity During Marital Conflict in
Laboratory and Naturalistic Settings: Differential
Associations with Relationship and Individual
Functioning Across Contexts
BRIAN R. W. BAUCOM*
KATHERINE J. W. BAUCOM*
JASARA N. HOGAN*
ALEXANDER O. CRENSHAW*
STACIA V. BOURNE*
SHEILA E. CROWELL*
MATTHEW S. GOODWIN
Cardiovascular reactivity during spousal conflict is considered to be one of the main path-
ways for relationship distress to impact physical, mental, and relationship health. However,
the magnitude of association between cardiovascular reactivity during laboratory marital con-
flict and relationship functioning is small and inconsistent given the scope of its importance in
theoretical models of intimate relationships. This study tests the possibility that cardiovascu-
lar data collected in laboratory settings downwardly bias the magnitude of these associations
when compared to measures obtained in naturalistic settings. Ambulatory cardiovascular
reactivity data were collected from 20 couples during two relationship conflicts in a research
laboratory, two planned relationship conflicts at couples’ homes, and two spontaneous relation-
ship conflicts during couples’ daily lives. Associations between self-report measures of relation-
ship functioning, individual functioning, and cardiovascular reactivity across settings are
tested using multilevel models. Cardiovascular reactivity was significantly larger during
planned and spontaneous relationship conflicts in naturalistic settings than during planned
relationship conflicts in the laboratory. Similarly, associations with relationship and individ-
ual functioning variables were statistically significantly larger for cardiovascular data col-
lected in naturalistic settings than the same data collected in the laboratory. Our findings
suggest that cardiovascular reactivity during spousal conflict in naturalistic settings is statis-
tically significantly different from that elicited in laboratory settings both in magnitude and in
the pattern of associations with a wide range of inter- and intrapersonal variables. These dif-
ferences in findings across laboratory and naturalistic physiological responses highlight the
value of testing physiological phenomena across interaction contexts in romantic relationships.
Keywords: Heart Rate Reactivity; Romantic Relationships; Marital Conflict
Fam Proc x:1–17, 2018
*Department of Psychology, University of Utah, Salt Lake City, UT.
Department of Electrical Engineering, University of Southern California, Los Angeles, CA.
Department of Health Sciences, Northeastern University, Boston, MA.
Correspondence concerning this article should be addressed to Brian R. W. Baucom, Department of
Psychology, University of Utah, 380 South 1350 East, BEHS 502, Salt Lake City, UT 84112.
This manuscript was supported in part by start-up funding from the University of Utah and a Vice
President for Research Seed Grant from the University of Utah awarded to Brian Baucom.
Family Process, Vol. x, No. x, 2018 ©2018 Family Process Institute
Cardiovascular reactivity is one of the major pathways by which marital conflict is
thought to impact overall relationship functioning, mental health, and physical well-
being (e.g., Robles & Kiecolt-Glaser, 2003; Whisman & Uebelacker, 2003). For example,
faster heart rate (HR) during couple conflict is associated with lower levels of concurrent
relationship satisfaction and greater longitudinal decline in relationship satisfaction (e.g.,
Levenson & Gottman, 1985), higher levels of negative communication behaviors (e.g.,
Newton & Sanford, 2003), and increased risk for hypertension and cardiovascular disease
(e.g., Robles & Kiecolt-Glaser, 2003). These effects are generally understood to occur due
to distress that partners experience during stressful couple interactions (e.g., Robles &
Kiecolt-Glaser, 2003). However, empirical evidence has thus far failed to support the theo-
rized magnitude of association between conflict-related cardiovascular reactivity and rela-
tionship functioning. For example, Robles and colleagues (Robles, Slatcher, Trombello, &
McGinn, 2014) reported meta-analytic effect sizes of r=.10, .18, and .18 (ps<.001)
for associations between relationship satisfaction and HR, diastolic blood pressure (DBP),
and systolic blood pressure (SBP) reactivity, respectively.
Existing research on cardiovascular reactivity during marital conflict is primarily
based on physiological data collected in laboratory settings while spouses discuss an area
of disagreement in their relationship (Robles et al., 2014). This methodology grew out of
a behavior analytic tradition that emphasized the utility of directly observing spouses’
behavior during arguments recorded in laboratory settings (for a review see Heyman,
2001). However, empirical evidence suggests that while conflict behavior enacted in a
laboratory is generally representative of how spouses typically behave, it tends to be less
negative and more positive than conflict that occurs outside of a laboratory (e.g., at
home; Gottman & Krokoff, 1989). Romantic relationship researchers have frequently
explained this finding as reflecting a social desirability effect wherein spouses do not
want to behave badly when being recorded (e.g., Heyman, 2001). While researchers have
not viewed this behavioral discrepancy as an impediment to studying marital conflict in
the laboratory generally, it may have a more pernicious effect when one seeks to identify
representative psychophysiological processes associated with spousal conflict in natural-
Additional concern about the generalizability of laboratory-based psychophysiological
reactivity during marital conflict arises from the literature examining the consistency of
an individual’s cardiovascular reactivity during stress tasks across time and place. The
fields of health psychology and behavioral medicine have long been concerned with the
consistency of an individual’s physiological reactivity to stressors because stable individ-
ual differences in physiological reactivity are a key component of stress-related process
models of disease. Early research into the stability of physiological reactivity to stress,
and cardiovascular reactivity specifically, found that within-person correlations between
cardiovascular reactivity during repeated laboratory stressors varies considerably from
small (e.g., r=.15; Smith & O’Keeffe, 1988) to large (r=.68; Gerin et al., 1998). Larger
correlations typically emerge between cardiovascular reactivity during identical stress
tasks performed on different occasions than cardiovascular reactivity during different
tasks performed either on the same or different occasions.
Conversely, varying the context or location of a stress task has been associated with
lower within-person correlations in cardiovascular reactivity relative to correlations that
emerge for stress tasks that occur in the same place or similar settings. For example,
Gerin et al. (1998) measured three indices of cardiovascular reactivity—HR,
SBP—during a serial subtraction task performed in the laboratory, a classroom, and
Correlations between measures of HR reactivity are not reported because they were all very small, and
study authors suggest that they are unreliable because of measurement imprecision.
participants’ homes. Reactivity scores were larger, though not statistically significantly
so, in naturalistic settings than in laboratory settings. Furthermore, within-person corre-
lations between (1) laboratory and classroom and (2) laboratory and home blood pressure
reactivity (average r=.37) were smaller than those observed between classroom and
home (average r=.51), which were in turn smaller than those observed between repeated
laboratory stressors (average r=.65). Study authors interpreted these findings to indicate
that correlations between repeated measurements of cardiovascular reactivity are highest
when comparing measures obtained under similar circumstances (i.e., laboratory vs. natu-
ralistic) even if the actual location varies.
As may be expected from variability in the magnitude and rank ordering of cardiovascu-
lar reactivity across settings, the predictive utility of cardiovascular measures obtained in
different settings is also widely variable. A prime example of this phenomenon is research
examining the association between hypertension and risk for cardiovascular outcomes.
Individuals who exhibit hypertension (i.e., elevated SBP) in both a clinic setting and dur-
ing daily life are at significantly greater risk for cerebrovascular and coronary events rela-
tive to individuals who exhibit hypertension only in a clinic setting (e.g., Khattar, Senior,
& Lahiri, 1998). Such “white coat” hypertension findings have been interpreted as evi-
dence that SBP can be artificially elevated due to the stress of being in a clinic setting,
rather than a true index of overall health. Extending this phenomenon to physiological
assessments of marital conflict in the laboratory invites the possibility that cardiovascular
reactivity in the laboratory may also be less strongly related to individual and relationship
functioning than would be observed in the real world.
One way spouses can behave less negatively in the laboratory than in more naturalistic
environments is to inhibit impulses to engage in negative behaviors. Such a process could
be described as “editing” from a relationship science perspective (Gottman, Notarius,
Gonso, & Markman, 1976) or as “suppression” from an emotion regulation perspective
(e.g., Gross & Levenson, 1993). Editing refers to refraining from engaging in negative
behaviors and is viewed as a relationship enhancing process (Gottman et al., 1976), while
suppression refers to inhibiting emotional expression and is linked to poorer interpersonal
functioning (e.g., Gross & John, 2003). Of relevance to this study, editing should lead to
lower cardiovascular reactivity during relationship conflict to the extent that it reduces
overall level of negative behavior (e.g., Gottman & Krokoff, 1989), whereas suppression
typically (e.g., Butler et al., 2003), but not always (e.g., Butler, Gross, & Barnard, 2014),
leads to higher levels of cardiovascular reactivity.
Building on the theory and findings presented above, the aims of this study were two-
fold. First, we sought to test mean level differences in cardiovascular reactivity during
repeated marital conflict that occurs in and outside of a laboratory setting. Second, we
sought to evaluate the magnitude of associations between individual and relationship
functioning variables and cardiovascular reactivity during marital conflicts across labora-
tory and naturalistic settings. Consistent with the general pattern of mean level differ-
ences in cardiovascular reactivity observed in stress tasks completed in laboratory and
naturalistic settings (e.g., Gerin et al., 1998), we hypothesized that cardiovascular reactiv-
ity occurring in the real-world environment would be larger in magnitude than cardiovas-
cular reactivity elicited in the laboratory. We also hypothesized that cardiovascular
reactivity in naturalistic settings would be more strongly related to individual and rela-
tionship functioning variables relative to cardiovascular reactivity recorded in the labora-
tory. Finally, exploratory analyses were conducted to examine differences in the
magnitude and strength of associations involving cardiovascular reactivity during
planned and spontaneous marital conflicts that took place in naturalistic settings (i.e., the
Fam. Proc., Vol. x, xxxx, 2018
BAUCOM ET AL.
Participants were 20 married, heterosexual, community couples (N=40 individuals)
recruited for a study of couple communication. Inclusion criteria included being legally
married for at least 1 year, living within 20 miles of the University, both spouses being
between 18 and 65 years of age, reliable access to the Internet throughout the day, and
ability to speak fluent English. Exclusion criteria included self-reports of moderate to sev-
ere physical aggression assessed using the Conflict Tactics Scale-Revised (Straus, Hamby,
Boney-McCoy, & Sugarman, 1996), history of medical conditions that affect the cardiovas-
cular system (e.g., Coronary Artery Disease), current or previous medication use associ-
ated with cardiovascular disease (e.g., angiotensin-converting enzyme inhibitors) or
current medication use that affects cardiovascular functioning (e.g., Beta-blockers), cur-
rent use of tobacco products, consuming an average of more than 10 alcoholic beverages
per week, the wife being pregnant, and the presence of other family members (e.g., chil-
dren or parents) or individuals (e.g., tenants or friends) living in the couple’s home and
sharing the same living space. Finally, one spouse was required to report a score of 18 or
less on the Couple Satisfaction Inventory, 4-item version (CSI4; Funk & Rogge, 2007) dur-
ing the screening assessment to increase the likelihood of spontaneous conflict occurring
during the week of daily data collection (described below). A score of 18 on the CSI4 is
approximately 0.5 SD below the maximum possible score of 21 based on established norms
for the scale (SD =4.7, Funk & Rogge, 2007).
Participants ranged from 22 to 64 years old, with a mean age for men of 29.3 (SD =7.8)
years and a mean age for women of 28.1 (SD =9.0) years. They were, on average, college
educated and earned a median annual household income of $32,400. Spouses largely self-
identified as Caucasian (80%), with 12.5% Asian or Pacific Islander, and 7.5% “Other.”
Self-reported religious identity was 40% Church of Jesus Christ of Latter-Day Saints
(LDS), 20% Atheist, 17.5% Spiritual/Agnostic, 5% Non-LDS Christian, and 17.5% chose
not to answer.
Couples completed a 3- to 4-hour laboratory assessment that included self-report ques-
tionnaires, physiological baselines, and four videotaped discussions. After consent,
spouses were outfitted with physiological recording equipment that continuously recorded
HR (described below) and then completed two 5-minute resting physiological baselines;
one resting baseline was collected while spouses were in the same room and the other was
collected while spouses were in separate rooms. The order of the two resting baselines was
randomized and counterbalanced. While continuing to wear physiological recording equip-
ment, spouses then completed a battery of self-report questionnaires that took approxi-
mately 45–90 minutes followed by four videotaped conversations consisting of a: (1) 5-
minute events of the day conversation; (2) 7-minutes relationship history conversation; and
(3–4) two 10-minutes relationship conflict discussions. Each spouse determined the topic
for one of the two relationship conflict discussions; the order of which spouse’s topic was
discussed first was randomized and counterbalanced. In addition, all spouses completed
the events of the day conversation first and the order of remaining types of discussions
(i.e., relationship history vs. relationship conflict) were randomized and counterbalanced.
Following completion of the laboratory procedures, participants completed 6–7 days of
data collection during their daily lives. Data collection included continuous measurement
of HR during all waking hours collected using the same equipment as used to collect HR in
the laboratory as well as completion of daily diaries twice a day. Daily diaries assessed the
beginning and ending time of interpersonal conflict (including conflict with the spouse) and
self-reported relationship functioning, mood, and stress, as well as physiological covariates
(e.g., timing and duration of eating, drinking, and exercise); daily diaries were completed
between 11:00 a.m. and 1:00 p.m. and between 8:00 p.m. and 10:00 p.m. each day.
On the final day of data collection, participants were asked to spend the day with their
spouse and to generate a continuous audio recording as well as a videotaped session
repeating the two conflict discussions they had during the laboratory assessment. Partici-
pants were provided with a video-recorder, given a written reminder of the conflict topics
discussed during the laboratory assessment, and asked to discuss the same two areas of
conflict in reverse order. For example, if the wife’s topic was discussed first during the lab-
oratory assessment, the husband’s topic was discussed first at home. The order of which
spouse’s topic was discussed first was counterbalanced between laboratory and home set-
tings to minimize any potential order effects. Participants were also asked to sit in an
arrangement as similar as possible to the session in the laboratory, follow the same set of
instructions as the laboratory conflict (e.g., discuss the same topic for 10 minutes, do not
get up and move around, etc.), and video-record the entirety of their discussions. Finally,
participants were provided with a digital timer to help them know when 10 minutes had
elapsed. All procedures were approved by the Institutional Review Board at the Univer-
sity of Utah.
Heart rate data for this study were based on the average of the two resting baselines,
the two 10-minutes relationship conflict discussions from the laboratory assessment, the
two 10-minutes relationship conflict discussions participants video-recorded in their
homes, and the first 10 minutes of the first two spontaneous relationship conflicts partici-
pants reported in their daily diaries.
We selected these periods of measurement for analy-
sis because they maximize the similarity of the stressors by indexing the first 10 minutes
of relationship conflict in all instances and included up to two assessments of conflict in
each interaction context.
HR during the two resting baselines were averaged to improve
stability of the estimate by including a maximum amount of data; this decision was judged
to be reasonable given that HR values were highly correlated across the two baselines,
r=.93, p<.001. All self-report measures used in our analyses were collected during the
Heart rate reactivity
Continuous HR data during the laboratory assessment and daily life portions of the
study were collected using ambulatory Actiheart biosensors (CamNtech Ltd U.K., 2014).
The Actiheart is a two-lead, miniaturized biosensor validated against gold standard clini-
cal/laboratory (3-lead, medical grade electrocardiogram [ECG] recordings) and ambulatory
(e.g., Holter monitor) equipment at rest (Brage, Brage, Franks, Ekelund, & Wareham,
2005; Kristiansen et al., 2011), while exercising (Brage et al., 2005), and during activities
of daily living (Kristiansen et al., 2011). The biosensor connects to standard ECG elec-
trodes placed at midline and left ventral positions of V4. Waveform data were sampled at
As expected, not every couple reported having a spontaneous conflict. Twelve couples reported having
at least one spontaneous conflict during the week of daily data collection. HR data from the planned home
conflicts in two couples were missing because of equipment malfunction. In addition, one couple opted not
to complete the daily data collection procedures; data for planned and spontaneous naturalistic conflict
were not collected for this couple.
Including multiple measurements of physiological reactivity across contexts has been shown to increase
the stability of reactivity within and across contexts and is the currently recommended method for studies
comparing reactivity across contexts (e.g., Kamarck, Debski, & Manuck, 2000).
Fam. Proc., Vol. x, xxxx, 2018
BAUCOM ET AL.
128 Hz and processed in real-time to generate 15-seconds epochs of HR values, which were
manually inspected for outliers and poor signal quality prior to analysis. Standard regres-
sion diagnostics were used to test for outliers, and signal quality was assessed using the
percent of missing data for each epoch provided by the Actiheart software. Signal quality
was considered to be acceptable if 10% or fewer data points were missing in any epoch. No
HR values were identified as outliers or poor quality using these methods.
To account for baseline differences in cardiovascular functioning across settings, aver-
age HR during the two resting baselines was subtracted from the average HR during all
relationship conflicts prior to analysis to create change scores. We decided to use the labo-
ratory baseline to calculate heart rate reactivity (HRR) in all relationship conflicts because
there are no currently established methods for selecting baselines for naturally occurring
stressors, and cardiovascular baselines have been shown to be stable across laboratory
and naturalistic settings (e.g., average within-subject correlation: for SBP and DPB,
r=.73; Gerin et al., 1998; for HR, SBP, and DPB, r=.70; Cohen et al., 2000). While this
decision makes interpretation of cardiovascular reactivity scores during the naturalistic
stressors less precise, it allowed HRR values across settings to be maximally comparable
to one another because they were all calculated relative to the same resting baseline
Relationship functioning variables
Relationship satisfaction was assessed using the total score on the Couples Satisfaction
Index, 4-item version (CSI4; Funk & Rogge, 2007), where higher scores indicate greater
relationship satisfaction. Perceived emotional flooding during relationship conflict was
assessed using the total score on the 10-item Flooding questionnaire (Gottman, 1999),
where higher scores indicate feeling more overwhelmed by emotion during relationship
Individual functioning variables
Emotion dysregulation was assessed using the total score on the 36-item Difficulties in
Emotion Regulation Scale (DERS; Gratz & Roemer, 2004), where higher scores indicate
greater emotion dysregulation. Emotion regulation strategies were assessed using the
Reappraisal (six items) and Suppression (four items) scales of the Emotion Regulation
Questionnaire (Gross & John, 2003), where higher scores on both scales indicate a stron-
ger tendency to use that emotion regulation strategy. Psychological distress was assessed
using the total score on the 21-item version of the Depression, Anxiety, and Stress Scale
(DASS; Antony, Bieling, Cox, Enns, & Swinson, 1998), where higher scores indicate
greater psychological distress. Positive and negative mood were assessed using the Posi-
tive (10 items) and Negative (10 items) scales of the 20-item version of the Positive and
Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988), where higher scores
indicate a stronger general experience of the respective mood.
Body Mass Index
Body Mass Index was computed as BMI =703 *weight in lb./height in inches
was measured using a stadiometer, and weight was measured using a beam scale.
Table 1 presents descriptive statistics for and correlations between all study vari-
ables. Of particular note, HRR during laboratory conflict and during planned conflict at
Descriptive Statistics for and Correlations Between All Study Variables
123 4 5 6789101112
2 .70*** —
3.17 .34 —
4.05 .29 .07 —
5.04 .20 .66** .14 —
6.01 .06 .44* .55*** .38* —
7.22 .21 .31 .26 .38* .23 —
8 .22 .30 .05 .22 .22 .54* .02 —
9.08 .07 .45* .42** .05 .41** .09 .21 —
10 .12 .21 .40* .47** .08 .35* .39* .10 .19 —
11 .02 .02 .19 .57*** .00 .49* .14 .27 .58*** .31 —
12 .34* .03 .20 .05 .08 .08 .09 .19 .04 .20 .19
Mean 0.44 8.17 9.05 17.09 1.29 74.18 32.80 13.91 20.87 38.02 19.70 26.21
SD 6.24 9.87 8.16 3.19 9.33 16.94 4.96 5.49 13.57 5.67 6.39 5.25
Alpha ——— .92 .86 .89 .71 .78 .89 .83 .85 —
;4—CSI; 5—flooding; 6—DERS; 7—reappraisal; 8—suppression; 9—DASS; 10—PANAS
positive; 11—PANAS negative; 12—BMI.
*p<.05; **p<.01; ***p<.001.
Fam. Proc., Vol. x, xxxx, 2018
BAUCOM ET AL.
Heart Rate Reactivity in the This Study, Other Studies of Marital Conflict, and Studies of Other Social Stressors
This study Other studies of marital conflict Trier studies
Laboratory Home—planned Home—spontaneous 1 2 3 4 5 6 7 8
Mean 0.44 8.17 9.05 3.71 2.45 1.37 1.21 2.2 3.6 30.63 19.47
Minimum 11.16 11.54 14.64 20.85 —————— —
Maximum 16.43 33.86 31.16 23.05 —————— —
SD 6.43 9.65 8.98 4.23 ————0.18 ——
Note. 1—Gottman et al. (1995); 2—Levenson and Gottman (1985); 3—Levenson et al. (1994); 4—Denton, Burleson, Hobbs, Von Stein, and Rodriguez
(2001); 5—Nealey-Moore et al. (2007); 6—Smith et al. (2009); 7—Ditzen et al. (2007); 8—Larson et al. (2001).
home was statistically significantly and positively associated (r=.7, p<.001). How-
ever, correlations between HRR during laboratory conflict and spontaneous conflict at
home and between planned and spontaneous conflict at home were nonsignificant
To better contextualize mean levels of HRR presented in Tables 1 and 2 presents
descriptive statistics from other published studies of laboratory marital conflict that pro-
vide descriptive statistics for HRR and of the Trier Social Stress Task (TSST; Kirschbaum
et al., 1993), considered to be a gold standard laboratory social stress task that reliably
provokes a cardiovascular response (Larson, Ader, & Moynihan, 2001). Descriptive statis-
tics for the TSST are taken from a meta-analysis of the TSST (Larson et al., 2001) and
from the only existing study of the TSST that recruited couples and reported HRR (Ditzen
et al., 2007). The mean level of HRR during laboratory conflict observed in this study was
somewhat lower than, though within 1 SD of, HRR reported in other studies of marital
conflict. In addition, the minimum and maximum levels of HRR obtained in this study
were consistent with those in the only study that reported these values (Gottman, Jacob-
son, Rushe, & Shortt, 1995). Mean levels of HRR during planned and spontaneous conflict
in this study were also between mean HRR values reported in previous studies of labora-
tory marital conflict and laboratory studies of the TSST (Ditzen et al., 2007; Gottman
et al., 1995; Larson et al., 2001; Levenson, Carstensen, & Gottman, 1994; Levenson &
Gottman, 1985; Nealey-Moore, Smith, Uchino, Hawkins, & Olson-Cerny, 2007; Smith
et al., 2009). Finally, maximum HRR values during marital conflict in naturalistic set-
tings in this study were consistent with mean HRR values during TSST reported in Ditzen
et al. (2007). Taken together, the consistency of HRR values in this study and those
reported in previous work on social stress indicates that HRR during laboratory and natu-
ralistic conflicts is consistent with values that would be expected based on similar previous
Differences in the Magnitude of HRR Across Settings
A 3-level, multilevel model (MLM) was used to test differences in the magnitude of
HRR during conversations in the laboratory, planned conversations at home, and sponta-
neous conflicts outside of the laboratory, where HRR was regressed onto dummy-coded
variables indicating planned conversations at home (0 =laboratory conflict, 1 =planned
conversations at home) and spontaneous conflicts outside of the laboratory (0 =laboratory
conflict, 1 =spontaneous conflict outside of the laboratory), an effect-coded spouse vari-
able (0.5 =husband, 0.5 =wife), and grand-mean centered BMI as a covariate,
trated by the following series of equations:
Level-1: YðHRRÞijk ¼p0jk þp1jk ðPlannedijkÞþp2jk ðSpontaneousijk Þþeijk
Level-2:p0jk ¼b00kþb01kðSpousejkÞþb02kðBMIjk Þþr0jk
for i¼1;2:pijk ¼bi0kþbi1kðSpousejkÞþbi2kðBMIjk Þ
Level-3: b00k¼c000 þu00k
for i¼0to2;j¼0;1:bijk ¼cij0
where iindexes conversations, jindexes spouses, and kindexes couples.
BMI was included in models to statistically control for variability in baseline heart rate associated with
obesity in young adults (e.g., Rossi et al., 2015).
Fam. Proc., Vol. x, xxxx, 2018
BAUCOM ET AL.
Consistent with our hypotheses, HRR during planned conflicts and spontaneous con-
flicts outside of the laboratory were statistically significantly greater than HRR during
laboratory conflicts (B=8.18, p<.001; B=8.14, p<.001, respectively). HRR was not sta-
tistically significantly different during planned conflicts outside of the laboratory and
spontaneous conflicts outside of the laboratory (B=.04, p>.50). As such, we refer to con-
flict in the naturalistic setting as “naturalistic conflict” below.
Differences in the Strength of Association with HRR Across Settings
The 3-level MLM used to test differences in the magnitude of HRR was expanded by
adding predictors at levels 2 and 3 to test differences in the strength of association with
cardiovascular reactivity across settings (represented by the cross-level interactions
between predictor and planned/spontaneous dummy-coded variables
) as illustrated by
the following series of equations:
Level-1: YðHRRÞijk ¼p0jk þp1jk ðPlannedijkÞþp2jk ðSpontaneousijk Þþeijk
Level-2:p0jk ¼b00kþb01kðSpousejkÞþb02kðBMIjkÞþb03kðPredictorcouple centered;jkÞþr0jk
fori¼1;2:pijk ¼bi0kþbi1kðSpousejkÞþbi2kðBMIjkÞþbi3kðPredictorcouple centered;jkÞ
Level-3: b00k¼c000 þc001 ðPredictorgrand centered;jkÞþu00k
for i¼1;2;j¼0:bijk ¼cij0þci01 ðPredictorgrand centered;jkÞ
for i¼0to2;j¼1;2;3:bijk ¼cij0
where iindexes conversations, jindexes spouses, and kindexes couples. Separate models
were run for each variable resulting in eight total models.
Again consistent with our hypotheses, there were statistically significant interactions
between measurement context and most predictors in predicting HRR; differences in the
strength of associations between HRR during laboratory conflict and HRR during natural-
istic conflict emerged for relationship satisfaction, flooding, suppression, difficulties with
emotion regulation, positive affect, and negative affect (all ps<.034); trends for differ-
ences in the strength of associations emerged for psychological distress (see Table 3). In
all cases, associations between HRR and predictor variables were larger in magnitude for
HRR during naturalistic conflict than during laboratory conflict. In addition, decomposing
these interactions revealed that simple slopes for the effects of naturalistic conflict HRR
were statistically significant in eight of nine interactions (and 10 of 13 statistically signifi-
cant interactions and nonsignificant, trend-level interactions considered jointly), indicat-
ing that HRR measured in naturalistic contexts was significantly associated with most
expected predictors. None of the associations between any predictor variable and HRR
during laboratory conflict (i.e., c
) were statistically significant, indicating HRR
measured in the laboratory context was not significantly associated with expected
The coefficients for these cross-level interactions indicate the magnitude and significance of the fixed
effect regression coefficients in the planned/spontaneous at-home relative to the same association in the
laboratory conflict. For example, a significant cross-level interaction between a couple centered predictor
and the dummy code for planned at home conflict (b
) would indicate that the association between the
predictor and HRR was significantly different for HRR during planned laboratory conflict and planned at-
home conflict; the sign of this interaction coefficient indicates the relative magnitude of the association
across settings. If the interaction term was positive, it would indicate that the association between the pre-
dictor and HRR was more positive/less negative for planned conflict at home relative to planned conflict in
Parameter Estimates of Multilevel Models Regressing HRR onto Relationship Functioning, Emotion Regulation, Psychological Well-Being, and Mood Variables
Variable 0 1 2 3 4 5 6 7 8
Intercept 0.44 0.44 0.44 0.44 0.44 0.44 0.44 0.44 0.44
—0.15 0.03 0.25 0.53 0.02 0.05 0.21 0.14
—0.12 0.96 0.32 0.01 0.00 0.02 0.37 0.05
At-home planned 8.18*** 11.12*** 8.15*** 8.39*** 8.20*** 8.11*** 8.42*** 8.62*** 8.28***
At-home planned 9Predictor
—2.35** 0.2 0.37 0.67* 0.01 0.1 0.49
At-home planned 9Predictor
—0.16 1.02 0.26 0.44 0.23* 0.16 0.43 0.27
At-home spontaneous 8.14*** 8.6*** 8.27*** 7.97*** 8.37*** 7.07*** 9.9*** 7.37*** 6.63***
At-home spontaneous 9Predictor
—0.21 0.4*** 0.13 0.31 0.43*** 0.53
At-home spontaneous 9Predictor
—0.75 1.64* 0.24 0.39 0.04 0.19
At-home planned —2.50** —— 1.20** ——0.71* 0.34
At-home spontaneous —— 0.36** —— 0.41*** 0.56
At-home planned —————0.24* ———
At-home spontaneous ——0.68 ———0.21* 0.58* 0.49*
Level-1 33.29 32.00 30.09 32.48 31.18 27.07 36.03 28.78 29.88
Level-2 13.41*** 10.56*** 13.45*** 12.65*** 14.03*** 14.34*** 4.08** 15.86*** 12.83***
Level-3 13.17*** 11.93*** 8.85** 11.29** 6.00* 13.48*** 14.18*** 8.86* 14.55***
Notes. 0—Baseline model; 1—CSI; 2—Flooding; 3—Reappraisal; 4—Suppression; 5—DERS; 6—DASS; 7—PANAS positive; 8—PANAS negative.
Parameter estimates for simple slopes are unstandardized regression coefficients.
p<.10; *p<.05; **p<.01; ***p<.001.
Fam. Proc., Vol. x, xxxx, 2018
BAUCOM ET AL.
Taken together, our collection of findings indicates that HRR during marital conflict in
daily life is robustly associated in expected directions with a wide range of relationship
and individual functioning variables, that these associations are statistically significantly
greater than those involving HRR during laboratory conflict. Moreover, none of the associ-
ations involving HRR during laboratory conflict are predictive of individual or relationship
which may indicate that laboratory associations are downwardly
This study examined differences in the magnitude of HRR during relationship conflict
in- and outside of the laboratory and differences in the magnitude of associations of HRR,
individual functioning, and relationship functioning variables across laboratory and natu-
ralistic contexts. Consistent with our hypotheses, there were significant interactions
between measurement context and individual and relationship functioning on HRR such
that HRR during naturalistic conflict was larger in magnitude and more strongly associ-
ated with both individual and relationship functioning variables relative to HRR during
laboratory conflict. Moreover, when decomposing these interactions, statistically signifi-
cant associations emerged in expected directions between seven of eight variables and
HRR during naturalistic conflict, but no statistically significant associations emerged
between any variable and HRR during laboratory conflict.
This collection of findings has both conceptual and methodological implications for the
study of psychophysiology, behavior, emotion, and psychological distress in romantic rela-
tionships. In contrast to meta-analytic findings documenting weak and inconsistent asso-
ciations between HRR and relationship adjustment, this study found strong and
consistent associations between HRR during relationship conflict, relationship functioning
variables, and individual functioning variables—but only when HRR was measured in
naturalistic settings. These findings for HRR in naturalistic settings are consistent with
theoretical models that consider physiological reactivity during conflict to be a primary
mechanism by which romantic relationships impact overall well-being (e.g., Robles & Kie-
colt-Glaser, 2003) and support the notion that HRR during marital conflict is robustly
associated with a wide variety of individual and relationship functioning domains.
In addition to supporting the general premise of theoretical models (e.g., Robles & Kie-
colt-Glaser, 2003; Robles et al., 2014), the findings in this study suggest that such theoret-
ical models would benefit by incorporating context as a moderator of associations.
Statistically significant effects only emerged in HRR data collected during relationship
conflict in daily life. This pronounced difference in the strength of association involving a
wide range of individual and relationship variables suggests that HRR is not associated
with individual and relationship functioning variables in all interaction contexts.
Future research on HRR, and autonomic and endocrine reactivity more broadly, during
conjoint couple therapy is an area of particular need (e.g., Karvonen, Kykyri, Kaartinen,
Penttonen, & Seikkula, 2016). Spontaneous and planned (e.g., re-enactments of recent
arguments) couple conflict during couple therapy are both assumed to be representative of
couple conflict that occurs outside of the therapy room (e.g., Epstein & Baucom, 2002).
This assumed similarity includes form (i.e., how the couple behaves during and experi-
ences conflict) and functional implications (i.e., associations with other aspects of relation-
Because of the dummy coding scheme used to test differences in associations across interaction con-
texts, the conditional main effects in the MLMs represent the associations between each predictor and
HRR in planned laboratory conflicts.
Study of autonomic and endocrine reactivity during conjoint couple therapy would be
beneficial not only because it would provide a further test of contextually dependent physi-
ological activation during couple conflict but also because it could provide practitioners
with information that could be used to guide treatment decisions. For example, if assump-
tions that conflicts during couple therapy are representative of how the couple normally
argues during daily life are correct, it is possible that physiological reactivity during
psychotherapy sessions may be associated with domains of emotional functioning similar
to those that were associated with at home HRR in this study (e.g., flooding, emotional
suppression, difficulties with emotion regulation). High levels of emotional reactivity
during couple conflict have been linked to poor treatment outcomes in couple therapy
(Baucom, Atkins, Simpson, & Christensen, 2009; Baucom, Weusthoff, Atkins, & Hahlweg,
2012) and an improved understanding of the physiological underpinnings of these associa-
tions could help therapists decide when and how to heighten versus dampen emotion
during couple therapy sessions (e.g., Baucom & Crenshaw, in press).
It is important to note that the individual and relationship functioning variables
included in this study were global assessments of what typically occurs in a relationship,
enduring psychological states that persisted over at least the previous week, and stable
individual differences in affectivity. While these findings demonstrate that HRR during
marital conflict in naturalistic settings is more consistently associated with such vari-
ables, it is unknown whether these findings will extend to other more proximal measures,
such as behavior during (or perceptions of) an instance of marital conflict. Examination of
differential patterns of associations involving HRR and proximal measures during marital
conflict in laboratory and naturalistic settings should be explored in future research.
The observed differences in the statistical significance and magnitude of associations
involving HRR during planned naturalistic conflict and those involving laboratory conflict
is particularly noteworthy given that couples were videotaped discussing the same two
topics for the same amount of time in similar physical settings during both data collection
settings. Given these similarities, it is not surprising that HRR values during laboratory
and planned naturalistic conflict were strongly correlated, r=.7, but it is surprising that
the statistical significance and magnitude of associations with individual and relationship
functioning variables were so different across the two contexts. One possible explanation
for these differences is that, even though couples were aware they were being videotaped
at home, there were fewer cues to remind them that they were participating in a study
during their home conversations. It is also possible that being at home contributed to
spouses behaving more naturalistically because they were in a more familiar environ-
ment. While such explanations are plausible, future research would benefit from including
a self-report measure of the representativeness of interactions in both settings to directly
test these possibilities.
Planned and spontaneous naturalistic conflict produced similar levels of HRR when
averaged across participants as a group, but HRR values during the two types of conflicts
were only modestly correlated, r=.34, and associations involving individual and relation-
ship functioning variables were inconsistent across the two types of conflict. No hypothe-
ses were made regarding differences in the strength of association across the two types of
conflict, and it is not immediately clear why such differences emerged. One possibility is
that HRR during spontaneous conflict may be considerably more variable than HRR dur-
ing planned conflict, and may be differentially associated with variables as a result. For
example, the topics discussed during planned conflicts both in and outside of the labora-
tory are topics that one or both spouses identified as being among the most upsetting in
their relationship. In contrast, spontaneous conflicts could have involved topic areas of a
wide range of importance to spouses.
Fam. Proc., Vol. x, xxxx, 2018
BAUCOM ET AL.
The patterns of associations with HRR across interaction contexts have the potential to
create valuable new avenues for relationship scientists to study marital conflict. The
field’s original goal of studying behavior and physiological reactivity during marital con-
flict in laboratory settings was to improve the quality of such data through tight experi-
mental control of the topics being discussed and the environments in which they occurred.
The results of this study suggest that studying HRR during marital conflict as it naturally
occurs during daily life with few or no constraints placed on topics, timing, or location of
discussions offers valuable insights into well-being. This pattern of findings has implica-
tions for a wide range of important domains of functioning (e.g., relationship satisfaction,
emotion regulation, and mood) that cannot be similarly assessed using laboratory conflict
data. As passive data acquisition technologies (i.e., methods for collecting data from partic-
ipants that involve little or no direct effort from participants themselves; e.g., Atkins &
Baucom, 2016; Goodwin, Velicer, & Intille, 2008) continue to develop, relationship scien-
tists will have an increasingly large number of options for collecting high quality data
about marital conflict using designs that maximize ecological validity (e.g., Hogan & Bau-
com, 2016; Walls, H€
oppner, & Goodwin, 2007).
There are a number of limitations to keep in mind when considering this study’s
results. First, the sample size of couples in this study was small by current standards in
relationship science. The repeated measurement of HRR during conflict across contexts
reduces this concern because of the increase in precision of HRR estimates generated from
repeated measures. However, it is possible that this sample size of couples resulted in
Type II error in failing to find relationships between HRR in the laboratory and self-report
measures. Second, the timing of spontaneous naturalistic conflicts was based on self-
reports. The timing of the daily diaries was designed to assess potential relationship con-
flicts shortly after participants typically parted in the mornings and again after partici-
pants reunited in the evening, thereby minimizing reporting inaccuracies. Imprecision in
the reported timing of spontaneous conflicts could result in the inclusion of HR values out-
side of conflict and such data would likely reduce the magnitude of HRR and their
strength of associations with individual and relationship variables. Thus, while impreci-
sion in the timing of reported spontaneous conflicts may have occurred, concerns about its
impact on findings are reduced given the statistical significance and consistency of find-
ings across multiple individual and relationship variables. Third, the planned conflict dis-
cussions at home always occurred after the planned conflict discussions in the laboratory
and it is therefore not possible to disentangle order effects from locations effects in tests of
laboratory versus planned at home data. It is possible that HRR was higher during the
planned at home discussions simply because it was the second time that partners dis-
cussed the same topics within a week’s time. Significantly greater HRR during sponta-
neous conflicts relative to laboratory conflicts and nonsignificant differences in HRR
between planned at home and spontaneous conflicts mitigates this concern. Replication of
these findings with data from a fully factorial, order (1st vs. 2nd) by location (laboratory
vs. at home) design is an important direction for future research. Fourth, equipment mal-
function resulted in the unavailability of HRR data during planned conflicts for two cou-
ples. These malfunctions appeared to be random, but it is possible that these missing data
biased results. Finally, participants were largely Caucasian and did not have children. In
addition, the average age of men (M=29.3) in the sample was below that for first time
fathers in the United States (M=30.9; Khandwala, Zhang, Lu, & Eisenberg, 2017), but
the average age of women in the sample (M=28.1) was above that for first time mothers
in the United States (M=26.4; CIA, 2017). These demographic characteristics of the sam-
ple may limit its generalizability and highlight the importance of replicating study find-
ings in a larger and more diverse sample of couples.
Despite these limitations, the results of this study suggest that assessing HRR—and
potentially physiological reactivity more broadly—during relationship conflict outside of
the laboratory may improve relationship scientists’ ability to understand various ways
that physiological reactivity during conflict can differentially associate with individual
and relational functioning. Assessing physiological reactivity outside of the laboratory can
involve loss of experimental control, decrease measurement precision, and add noise to
data. However, these drawbacks appear to be outweighed by the unique associations with
multiple variables provided by data acquired outside of laboratory settings. Studying asso-
ciations involving physiological reactivity across multiple contexts appears to be a fruitful
direction for future research, and relationship scientists interested in physiological mech-
anisms are encouraged to consider incorporating ambulatory methods to complement labo-
Antony, M. M., Bieling, P. J., Cox, B. J., Enns, M. W., & Swinson, R. P. (1998). Psychometric properties of the 42-
item and 21-item versions of the Depression Anxiety Stress Scales in clinical groups and a community sample.
Psychological Assessment,10, 176–181. https://doi.org/10.1037/1040-3518.104.22.168.
Atkins, D. A., & Baucom, B. R. (2016). Emerging methodological and statistical techniques in couple research. In
E. Lawrence & K. Sullivan (Eds.), Oxford handbook of relationship science and couple interventions (pp. 148–
163). New York: Oxford University Press.
Baucom, B. R., Atkins, D., Simpson, L., & Christensen, A. (2009). Prediction of response to treatment in a ran-
domized clinical trial of couple therapy: A 2-year follow-up. Journal of Consulting and Clinical Psychology,
77, 160–173. https://doi.org/10.1037/a0014405.
Baucom, B. R. W., & Crenshaw, A. O. (in press). Evaluating the efficacy of couple and family therapy. In B. Fiese,
M. Whisman, M. Celano, K. Deater-Deckard, & E. Jouriles (Eds.), APA handbook of contemporary family psy-
chology. Washington, DC: American Psychological Association.
Baucom, B. R., Weusthoff, S., Atkins, D., & Hahlweg, K. (2012). Greater emotional arousal predicts poorer long-
term memory of communication skills in couples. Behaviour Research and Therapy,50, 442–447. https://doi.
Brage, S., Brage, N., Franks, P., Ekelund, U., & Wareham, N. (2005). Reliability and validity of the combined
heart rate and movement sensor Actiheart. European Journal of Clinical Nutrition,59, 561–570. https://doi.
Butler, E. A., Egloff, B., Wilhelm, F. H., Smith, N. C., Erickson, E. A., & Gross, J. J. (2003). The social conse-
quences of expressive suppression. Emotion,3,48–67. https://doi.org/10.1037/1528-3522.214.171.124.
Butler, E. A., Gross, J. J., & Barnard, K. (2014). Testing the effects of suppression and reappraisal on emotional
concordance using a multivariate multilevel model. Biological Psychology,98,6–18. https://doi.org/10.1016/j.
CamNTech Inc. (2014). Actiheart [equipment]. Boerne, TX: CamNTech Inc.
Central Intelligence Agency (2017). The World Factbook. Retrieved from https://www.cia.gov/library/publi
Cohen, S., Hamrick, N. M., Rodriguez, M. S., Feldman, P. J., Rabin, B. S., & Manuck, S. B. (2000). The stability of
and intercorrelations among cardiovascular, immune, endocrine, and psychological reactivity. Annals of
Behavioral Medicine,22, 171–179. https://doi.org/10.1007/BF02895111
Denton, W. H., Burleson, B. R., Hobbs, B. V., Von Stein, M., & Rodriguez, C. P. (2001). Cardiovascular reactivity
and initiate/avoid patterns of marital communication: A test of Gottman’s psychophysiologic model of marital
interaction. Journal of Behavioral Medicine,24, 401–421. https://doi.org/10.1023/a:1012278209577.
Ditzen, B., Neumann, I. D., Bodenmann, G., von Dawans, B., Turner, R. A., Ehlert, U. et al. (2007). Effects of dif-
ferent kinds of couple interaction on cortisol and heart rate responses to stress in women. Psychoneuroen-
docrinology,32, 565–574. https://doi.org/10.1016/j.psyneuen.2007.03.011.
Epstein, N. B., & Baucom, D. H. (2002). Enhanced cognitive-behavioral therapy for couples: A contextual
approach. Washington, DC: American Psychological Association. https://doi.org/10.1037/10481-000
Funk, J. L., & Rogge, R. D. (2007). Testing the ruler with item response theory: Increasing precision of measure-
ment for relationship satisfaction with the Couples Satisfaction Index. Journal of Family Psychology,21, 572–
Gerin, W., Christenfeld, N., Pieper, C., Derafael, D. A., Su, O., Stroessner, S. J. et al. (1998). The generalizability
of cardiovascular responses across settings. Journal of Psychosomatic Research,44, 209–218. https://doi.org/
BAUCOM ET AL.
Fam. Proc., Vol. x, xxxx, 2018
Goodwin, M. S., Velicer, W. F., & Intille, S. S. (2008). Telemetric monitoring in the behavior sciences. Behavioral
Research Methods,40, 328–341. https://doi.org/10.3758/brm.40.1.328.
Gottman, J. M. (1999). The marriage clinic: A scientifically-based marital therapy. New York: WW Norton &
Gottman, J. M., Jacobson, N. S., Rushe, R. H., & Shortt, J. W. (1995). The relationship between heart rate reactiv-
ity, emotionally aggressive behavior, and general violence in batterers. Journal of Family Psychology,9, 227–
Gottman, J. M., & Krokoff, L. J. (1989). Marital interaction and satisfaction: A longitudinal view. Journal of Con-
sulting and Clinical Psychology,57,47–52. https://doi.org/10.1037/0022-006x.57.1.47.
Gottman, J., Notarius, C., Gonso, J., & Markman, H. (1976). A couple’s guide to communication. Champaign, IL:
Gratz, K. L., & Roemer, L. (2004). Multidimensional assessment of emotion regulation and dysregulation: Devel-
opment, factor structure, and initial validation of the difficulties in emotion regulation scale. Journal of Psy-
chopathology and Behavioral Assessment,26,41–54. https://doi.org/10.1037/0893-3126.96.36.199.
Gross, J. J., & John, O. P. (2003). Individual differences in two emotion regulation processes: Implications for
affect, relationships, and well-being. Journal of Personality and Social Psychology,85, 348–362. https://doi.
Gross, J. J., & Levenson, R. W. (1993). Emotional suppression: Physiological, self-report, and expressive behavior.
Journal of Personality and Social Psychology,64, 970–986. https://doi.org/10.1037/0022-35188.8.131.520.
Heyman, R. E. (2001). Observation of couple conflicts: Clinical assessment applications, stubborn truths, and
shaky foundations. Psychological Assessment,13,5–35. https://doi.org/10.1037/1040-35184.108.40.206.
Hogan, J. N., & Baucom, B. R. (2016). Behavioral, affective, and physiological monitoring. In J. K. Luiselli & A. J.
Fischer (Eds.), Computer-assisted and web-based innovations in psychology, special education, and health (pp.
3–32). New York: Academic Press. https://doi.org/10.1016/B978-0-12-802075-3.00001-2
Kamarck, T. W., Debski, T. T., & Manuck, S. B. (2000). Enhancing the laboratory-to-life generalizability of cardio-
vascular reactivity using multiple occasions of measurement. Psychophysiology,37, 533–542. https://doi.org/
Karvonen, A., Kykyri, V. L., Kaartinen, J., Penttonen, M., & Seikkula, J. (2016). Sympathetic nervous system
synchrony in couple therapy. Journal of Marital and Family Therapy,42, 383–395. https://doi.org/10.1111/jmf
Khandwala, Y. S., Zhang, C. A., Lu, Y., & Eisenberg, M. L. (2017). The age of fathers in the USA is rising: An
analysis of 168 867 480 births from 1972 to 2015. Human Reproduction,32, 2110–2116. https://doi.org/10.
Khattar, R. S., Senior, R., & Lahiri, A. (1998). Cardiovascular outcome in white-coat versus sustained mild hyper-
tension: A 10-year follow-up study. Circulation,98, 1892–1897. https://doi.org/10.1161/01.cir.98.18.1892.
Kirschbaum, C., Pirke, K. M., & Hellhammer, D. H. (1993). The ‘Trier Social Stress Test’: A tool for investigating
psychobiological stress responses in a laboratory setting. Neuropsychobiology,28, 76–81.
Kristiansen, J., Korshøj, M., Skotte, J. H., Jespersen, T., Søgaard, K., Mortensen, O. S. et al. (2011). Comparison
of two systems for long-term heart rate variability monitoring in free-living conditions-a pilot study. Biomedi-
cal Engineering Online,10, 27. https://doi.org/10.1186/1475-925x-10-27.
Larson, M. R., Ader, R., & Moynihan, J. A. (2001). Heart rate, neuroendocrine, and immunological reactivity in
response to an acute laboratory stressor. Psychosomatic Medicine,63, 493–501. https://doi.org/10.1097/
Levenson, R. W., Carstensen, L. L., & Gottman, J. M. (1994). Influence of age and gender on affect, physiology,
and their interrelations: A study of long-term marriages. Journal of Personality and Social Psychology,67,
Levenson, R. W., & Gottman, J. M. (1985). Physiological and affective predictors of change in relationship satis-
faction. Journal of Personality and Social Psychology,49,85–94. https://doi.org/10.1037/0022-35220.127.116.11.
Nealey-Moore, J. B., Smith, T. W., Uchino, B. N., Hawkins, M. W., & Olson-Cerny, C. (2007). Cardiovascular reac-
tivity during positive and negative marital interactions. Journal of Behavioral Medicine,30, 505–519.
Newton, T. L., & Sanford, J. M. (2003). Conflict structure moderates associations between cardiovascular reactiv-
ity and negative marital interaction. Health Psychology,22, 270–278. https://doi.org/10.1037/0278-6133.22.3.
Robles, T. F., & Kiecolt-Glaser, J. K. (2003). The physiology of marriage: Pathways to health. Physiology &
Behavior,79, 409–416. https://doi.org/10.1016/s0031-9384(03)00160-4.
Robles, T. F., Slatcher, R. B., Trombello, J. M., & McGinn, M. M. (2014). Marital quality and health: A meta-ana-
lytic review. Psychological Bulletin,140, 140–187. https://doi.org/10.1037/a0031859.
Rossi, R. C., Vanderlei, L. C. M., Gonc
ßalves, A. C. C. R., Vanderlei, F. M., Bernardo, A. F. B., Yamada, K. M. H.
et al. (2015). Impact of obesity on autonomic modulation, heart rate and blood pressure in obese young people.
Autonomic Neuroscience,193, 138–141. https://doi.org/10.1016/j.autneu.2015.07.424
Smith, T. W., & O’Keeffe, J. L. (1988). Cross-situational consistency of cardiovascular reactivity. Biological Psy-
chology,27, 237–243. https://doi.org/10.1016/0301-0511(88)90033-6
Smith, T. W., Uchino, B. N., Berg, C. A., Florsheim, P., Pearce, G., Hawkins, M. et al. (2009). Conflict and collabo-
ration in middle-aged and older couples: II. Cardiovascular reactivity during marital interaction. Psychology
and Aging,24, 274–286. https://doi.org/10.1037/a0016067
Straus, M. A., Hamby, S. L., Boney-McCoy, S., & Sugarman, D. B. (1996). The revised conflict tactics scales:
Development and preliminary psychometric data. Journal of Family Issues,17, 283–316. https://doi.org/10.
Walls, T. A., H€
oppner, B. B., & Goodwin, M. S. (2007). Statistical issues in intensive longitudinal data analysis.
In A. Stone, S. Shiffman, A. Atienza, & L. Nebelling (Eds.), The science of real-time data capture (pp. 338–
360). New York: Oxford University Press.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of Positive and
Negative Affect: The PANAS Scales. Journal of Personality and Social Psychology,54, 1063–1070. https://doi.
Whisman, M. A., & Uebelacker, L. A. (2003). Comorbidity of relationship distress and mental and physical health
problems. In D. K. Snyder & M. A. Whisman (Eds.), Treating difficult couples: Helping clients with coexisting
mental and relationship disorders (pp. 3–26). New York: The Guilford Press.
Fam. Proc., Vol. x, xxxx, 2018
BAUCOM ET AL.