Article

Social Psychological and

Personality Science
Perceived Partner Responsiveness 2017, Vol. 8(1) 83-92
ª The Author(s) 2016
Reprints and permission:
Predicts Better Sleep Quality sagepub.com/journalsPermissions.nav
DOI: 10.1177/1948550616662128
Through Lower Anxiety journals.sagepub.com/home/spp

Emre Selcuk1, Sarah C. E. Stanton2, Richard B. Slatcher2,

and Anthony D. Ong3

Abstract
The present study investigated whether perceived partner responsiveness—the extent to which individuals feel cared for,
understood, and validated by their partner—predicted subjective sleep problems and objective (actigraph-based) sleep efficiency
through lower anxiety and depression symptoms. A life span sample of 698 married or cohabiting adults (35–86 years old)
completed measures of perceived partner responsiveness and subjective sleep problems. A subset of the sample (N ¼ 219)
completed a weeklong sleep study where actigraph-based measures of sleep efficiency were obtained. Perceived partner
responsiveness predicted lower self-reported global sleep problems through lower anxiety and depression and greater actigraph-
assessed sleep efficiency through lower anxiety. All indirect associations held after controlling for emotional support provision to
the partner, agreeableness, and demographic and health covariates known to affect sleep quality. These findings are among the
first to demonstrate how perceived partner responsiveness, a core aspect of romantic relationships, is linked to sleep behavior.

Keywords
perceived partner responsiveness, sleep, marriage, well-being, anxiety

Sleep is a critical health behavior reducing the risk for morbid- et al., 2011) and physiological arousal (Slatcher, Selcuk, &
ity and mortality (e.g., Dew et al., 2003; Reid et al., 2006). Ong, 2015).
Given the well-established link between social relationships Given that adult sleep is typically a shared activity between
and health (e.g., Holt-Lunstad, Smith, & Layton, 2010), romantic partners (National Sleep Foundation, 2013) and
research has increasingly focused on the role of close relation- romantic relationships have a unique capacity to influence the
ships in sleep. Although both sleep quality (Dew et al., 2003) quality of human health and well-being (Loving & Slatcher,
and total sleep duration (Shen, Wu, & Zhang, 2016) have been 2013), the role of marital and cohabiting relationships in sleep
linked to health outcomes, social relationships or lack thereof quality has received increased research attention (Troxel,
have typically been found to be linked with sleep quality—for 2010). Although studies have established that individuals’
instance, subjective evaluations of how well individuals sleep sleep quality is closely linked to how happy (or unhappy) they
or how much daytime dysfunction they experience, or objec- are in their relationship, the psychological processes through
tive assessments of sleep efficiency (the ratio of time spent which relationships affect sleep are still not well understood
sleeping to the time spent in bed)—rather than total duration (Troxel, 2010). Growing work, primarily led by social psychol-
(Bordeleau, Bernier, & Carrier, 2012; Cacioppo et al., 2002). ogists, aiming at explaining the psychological pathways by
These findings suggest that social relationships are associated which long-term romantic relationships are linked to physical
with reduced nonrestorative sleep, which is defined as sleep health demonstrates that relationship processes (e.g.,
that is interrupted with frequent awakenings and not refreshing,
despite normal duration (Hawkley, Preacher, & Cacioppo,
2010). Restorative sleep depends on perceived absence of 1
Middle East Technical University, Ankara, Turkey
2
threat in the environment and downregulation of arousal. Per- Wayne State University, Detroit, MI, USA
3
sistent high arousal—a marker of anxiety—disrupts sleep by Cornell University, Ithaca, NY, USA
increasing nightly awakenings and resulting in poorer daytime
Corresponding Author:
functioning. Social relationships are thought to counteract this Emre Selcuk, Middle East Technical University, B45 Human Sciences Building,
process, as they are a potent source of safety and protection, Ankara 06800, Turkey.
and they downregulate perceptions of threat (Eisenberger Email: [email protected]
84 Social Psychological and Personality Science 8(1)

responsiveness, support provision) predict psychological Of particular relevance to the present study, a central func-
symptoms (e.g., anxiety, depression) and well-being (e.g., life tion of perceived partner responsiveness involves downregulat-
satisfaction), which in turn predict physical health (for reviews, ing anxiety and arousal and instilling a sense of security and
see Slatcher, 2010; Slatcher & Selcuk, in press). In the few quiescence (Selcuk, Zayas, & Hazan, 2010). When individuals
studies taking this approach (e.g., the links between self- encounter threats and stressors, the primary coping strategy for
disclosure to one’s partner and sleep quality; Kane, Slatcher, most adults is to turn to their partners for safety and protection
Reynolds, Repetti, & Robles, 2014), the small sample sizes (Mikulincer & Shaver, 2007). Responsive partner support dur-
limited the ability to detect between-person differences and ing these times alleviates distress and downregulates anxious
sleep quality was measured only with self-reports. However, arousal. Indeed, when individuals were faced with an
a multimethod approach to measuring sleep is important anxiety-provoking experience in the laboratory (e.g., talking
because self-reported sleep quality is at best weakly correlated about a stressful problem or anticipating giving a public
with objective measures such as actigraphy (Grandner, Kripke, speech), their partner’s responsive support alleviated both
Yoon, & Youngstedt, 2006), suggesting that the two types of self-reported (Collins & Feeney, 2000) and observer-rated
measures tap different aspects of sleep quality. Whereas self- (Simpson, Rholes, & Nelligan, 1992) anxiety. Repeated
reports typically measure subjective (dis)satisfaction about responsive interactions with partners translate over time to a
sleep quality, actigraphy provides indices of objective sleep long-term decline in anxiety, both psychologically and physio-
disruptions during a night’s interval. The two measures are also logically (e.g., endocrine functioning; Feeney & Collins,
differentially related to health and well-being (e.g., Lemola, 2015). For instance, a recent daily experience study (Slatcher
Ledermann, & Friedman, 2013; Liu et al., 2013). It may be the et al., 2015) demonstrated that high partner responsiveness pre-
case that romantic relationships may be associated with one dicts a steeper decline in diurnal cortisol a decade later, sug-
type of sleep measure but not the other, or they may predict gesting that the effect of responsiveness goes beyond the
both subjective and objective sleep quality but through differ- immediate stressful context and may potentially be associated
ent psychological mechanisms. with lower chronic levels of anxiety over the long term. This
Perceived partner responsiveness (i.e., the extent to which finding is important in the present context also given that prior
individuals perceive their partner as caring, understanding, and work has linked steeper diurnal cortisol slopes to lower anxious
appreciative; Reis, 2007) may be one important process by arousal (Doane et al., 2013), which would be expected to pre-
which romantic relationships affect sleep quality. In this study, dict higher quality sleep.
we investigated the associations between perceived partner Responsiveness (or lack thereof) is also thought to be one
responsiveness and sleep quality in a large sample using both key process that explains how social relationships affect
self-report measures of sleep problems and an objective depression across the life span (Bowlby, 1980). More specif-
actigraph-based measure of sleep efficiency. The large sample ically, relationships are argued to reduce an individual’s risk
size provided us with sufficient statistical power to investigate for depression to the extent that a partner’s support (a) effec-
the potential psychological mechanisms through which partner tively meets the demands of the stressful life situations and (b)
responsiveness is associated with sleep. Specifically, we tested does not undermine the individual’s sense of autonomy
two potential mediators of the link between perceived partner (Ibarra-Rovillard & Kuiper, 2011). Those who are perceived
responsiveness and sleep quality—anxiety and depressed affect as responsive are more likely to engage in support behaviors
two classes of psychological symptoms that are among the that are appropriately contingent on the demands of the situ-
most common predictors of sleep disturbances (Koffel & Wat- ation (Collins, Guichard, Ford, & Feeney, 2006) and, more-
son, 2009)—hypothesizing that perceived partner responsive- over, their support also enhances the partner’s autonomy,
ness would positively predict sleep quality through lower self-efficacy, and independent goal pursuit (Feeney, 2007).
anxiety and depressive symptoms. Thus, it is not surprising that perceived partner responsiveness
is a strong predictor of lower symptoms of depression in daily
life (Fekete, Stephens, Mickelson, & Druley, 2007; Khan
et al., 2009).
Perceived Partner Responsiveness and
Prior studies have illuminated consistent associations
Well-Being between anxiety, depression, and sleep quality as well
Perceived partner responsiveness has been identified as a key (Alvaro, Roberts, & Harris, 2013; Koffel & Watson, 2009;
process that influences the extent to which romantic relation- Magee & Carmin, 2010; Revenson, Marı́n-Chollom, Rundle,
ships are satisfying and intimate. It focuses on partners’ posi- Wisnivesky, & Neugut, 2016). Increasing research indicates
tive responses to each other in contrast to negative responses that disruptions in nighttime sleep is associated with indica-
or indifference (Reis, 2007). When one perceives her or his tors of physiological hyperarousal such as increased meta-
partner as caring, understanding, and appreciative, one is more bolic rates, heart rate variability, and cortisol output (for a
likely to self-disclose and also to react responsively to the part- review, see Stepanski & Rybarczyk, 2006). Watson and
ner’s disclosures. When this process is enacted reciprocally and colleagues (1995) referred to these indicators as anxious
mutually, it reinforces the development and maintenance of arousal, and in extensive work with both clinical and noncli-
intimacy in the relationship (Reis & Patrick, 1996). nical samples of adults from different age-groups, they
Selcuk et al. 85

showed that these symptoms are specific to anxiety (and not Method
to depression). In a similar vein, they found that loss of
interest and low positive affect, which are collectively
Sample and Procedure
referred to as anhedonic depression, are specific to depres- The data for the present study came from the National Survey
sion (and not anxiety; Watson et al., 1995). Focusing on of Midlife Development in the United States II (MIDUS II), a
these nonoverlapping aspects of anxiety and depression is study on health and aging conducted in 2004–2006 (N ¼ 4,963;
important to identify specifically through which symptoms age range ¼ 32–84). The MIDUS II survey consisted of a
partner responsiveness would be linked to different aspects phone interview and a self-administered questionnaire (Ryff
of sleep quality. et al., 2007). Upon completion of MIDUS II, a subset of
respondents (N ¼ 1,255) participated in the Biomarkers Study
(Dienberg Love, Seeman, Weinstein, & Ryff, 2010), which
included sleep assessments. Mean time lag between the
The Present Research MIDUS II self-administered questionnaire and the Biomarkers
The present study aimed to further extend our conceptual Study was 25 months (SD ¼ 14 months). The current sample
understanding of how partner responsiveness is linked to health consisted of 698 married or cohabiting adults (mean age ¼
and well-being by investigating the role of perceived partner 57 years, range ¼ 35–86 years) who completed the perceived
responsiveness in sleep. Importantly, we tested the associations partner responsiveness measure, all covariates, and at least one
of perceived partner responsiveness with both self-report mea- of the sleep measures (self-reported global quality or objective
sures of sleep problems and an objective actigraph-based sleep efficiency) and reported still being together with their
assessment of sleep efficiency—that is, the ratio of total sleep partner over the course of data collection (i.e., between the
time to the total time spent in bed. Based on prior theorizing MIDUS II phone interview and the Biomarkers Study). Of
and empirical work on social relationships and sleep (Borde- these participants, 50% were female and 50% male; 94% were
leau et al., 2012; Cacioppo et al., 2002; Hawkley et al., White and 6% were from other racial backgrounds; 24% grad-
2010), we prioritized testing the links between partner respon- uated from high school or less and 76% had some college edu-
siveness and sleep quality, although we included sleep duration cation or more. In the final sample, 479 participants completed
in analyses as well. only the self-reported global sleep quality measure, 16 com-
Moreover, we investigated the potential mechanisms by pleted only the objective sleep efficiency measure, and 203
which perceived partner responsiveness was linked to sleep. completed both measures. Thus, analyses testing the associa-
Specifically, we chose anxiety and depression, given their tions of perceived partner responsiveness with global sleep
well-established links to both perceived partner responsive- quality and objective sleep efficiency were based on 682 and
ness (Selcuk et al., 2010; Slatcher et al., 2015) and sleep 219 adults, respectively. Participants who had data for objec-
quality (Alvaro et al., 2013; Koffel & Watson, 2009; tive sleep efficiency were slightly younger (M ¼ 56.22, SD
Revenson et al., 2016). Given the theoretical function of ¼ 10.86 vs. M ¼ 58.03, SD ¼ 11.28, p ¼ .048, d ¼ .16, 95%
partner responsiveness in downregulating arousal and anxi- CI [0.02, 3.59]) and scored slightly lower on anxiety (M ¼
ety (Selcuk et al., 2010) and prior empirical work docu- 20.84, SD ¼ 4.20 vs. M ¼ 21.54, SD ¼ 4.41, p ¼ .047, d ¼
menting that perceived partner responsiveness predicts .16, 95% CI [0.01, 1.40]). The two groups, however, did not
lower depression and anxiety (e.g., Fekete et al., 2007; differ on other variables of interest including perceived partner
Simpson et al., 1992), we expected that perceived partner responsiveness, global sleep quality, depression, or any of the
responsiveness would be meaningfully linked to sleep qual- covariates (ps > .071).
ity via its associations with those two symptomologies, a
hypothesis hitherto unexplored in the relationships and
health literatures. Previous research indicated that individu- Measures
als project their own support provision to their partner, that Perceived partner responsiveness. Following prior work (Selcuk,
is, individuals who provide more support to their partner Gunaydin, Ong, & Almeida, 2016; Selcuk & Ong, 2013;
are more likely to perceive their partner as responsive Slatcher et al., 2015), perceived partner responsiveness was
(Lemay, Clark, & Feeney, 2007). Moreover, individuals measured with 3 items in the MIDUS II self-administered ques-
who perceive their partner as responsive may be more tionnaire. Participants indicated the extent to which their part-
agreeable people in general. Thus, following prior work ner or spouse cares about them, understands the way they feel
on partner responsiveness (Slatcher et al., 2015), we con- about things, and appreciates them (1 ¼ a lot to 4 ¼ not at all,
trolled for emotional support provision to the partner and a ¼ .82). Responses were reverse scored, so that higher scores
agreeableness in the analyses. In addition, our analysis con- reflected greater partner responsiveness.
trolled for demographic factors (age, gender, race, and edu-
cation) and physical health factors (perceived health, health Anxiety and depression symptoms. Anxiety and depression were
symptoms, and body mass index [BMI]) known to affect assessed in the Biomarker Study using the Anxious Arousal
sleep quality (Mezick, Wing, & McCaffery, 2014; Ong subscale of the Mood and Symptom Questionnaire (MASQ;
et al., 2013). Watson et al., 1995), measuring specific anxiety symptoms
86 Social Psychological and Personality Science 8(1)

(somatic tension and hyperarousal) that are critical for sleep diary). Sleep efficiency was computed as the percentage ratio
quality (Stepanski & Rybarczyk, 2006), and the Anhedonic of total sleep time to the total time spent in bed. Sleep effi-
Depression subscale measuring specific depression symptoms ciency may suffer due to two reasons: difficulty to fall asleep
(low positive affect and loss of interest). Prior work showed or difficulty to stay asleep. Therefore, sleep onset (a measure
that these subscales are less correlated with each other (r ¼ of difficulty falling asleep) and wake after sleep onset (a mea-
.369, p < .001 in the current sample) and show higher discrimi- sure of difficulty staying asleep) were also included in the anal-
nant validity compared to other measures of anxiety and yses to figure out which aspects of sleep efficiency were linked
depression while maintaining convergent validity, in both clin- with partner responsiveness. Sleep onset corresponded to the
ical and nonclinical samples (Watson et al., 1995). Participants time required, in minutes, for the onset of sleep after attempting
indicated how much they experienced each symptom during to get to sleep. Finally, wake after sleep onset corresponded to
the past week (1 ¼ not at all to 5 ¼ extremely). The Anxious the total time of awakenings during the night’s interval after
Arousal subscale consisted of 17 items (e.g., ‘‘heart was racing falling asleep.
or pounding’’) and the Anhedonic Depression subscale con-
sisted of 22 items (e.g., ‘‘felt nothing was very enjoyable’’). Covariates
Anxiety and depression scores were computed by summing
across all items for participants who had no missing value (for Demographic covariates. Demographic covariates included age
participants who had a missing value for only 1 item, mean sub- at completion of the Biomarker study, and gender (0 ¼ male,
stitution was used for the item), a ¼ .73 for anxiety and .93 for 1 ¼ female), race (0 ¼ White, 1 ¼ Non-White), and education
depression. (1 ¼ no school/some grade school to 12 ¼ doctoral degree)
assessed at MIDUS II.
Sleep outcomes. All sleep assessments were obtained in the Bio-
marker Study. Global sleep quality was measured with the Physical health covariates. We controlled for three physical health
widely used Pittsburgh Sleep Quality Index (PSQI; Buysse, predictors of sleep quality: perceived physical health, health
Reynolds, Monk, Berman, & Kupfer, 1989). The PSQI symptoms, and BMI. Perceived physical health was measured
includes subjective assessments of seven sleep components: in the MIDUS II phone interview via a single item asking par-
sleep quality (overall assessment of sleep quality), sleep ticipants to evaluate their physical health (1 ¼ excellent to 5 ¼
latency (time and difficulty to fall asleep at night), sleep dura- poor). Participants also completed a health symptoms checklist
tion (hours of sleep gotten at night), habitual sleep efficiency (e.g., ‘‘ever had heart disease?’’ ‘‘ever had cancer?’’) in the
(the ratio of actual sleep to the time spent in bed), sleep distur- Biomarker Study. The total number of health symptoms ever
bance (trouble staying asleep), use of sleeping medication, and experienced was included in the analyses. Finally, the BMI was
daytime dysfunction (trouble staying awake during daytime). computed by dividing weight in kilograms by height squared in
Each category receives a score between 0 and 3, with higher meters. These measurements were obtained by clinical staff
scores reflecting worse sleep quality. Although the components during a physical exam as part of the Biomarker Study.
measure different aspects of sleep, the PSQI is typically ana-
lyzed using a global score, especially given that all components Relationship covariates. Emotional support provision to the part-
reflect an underlying subjective (dis)satisfaction with sleep ner was measured by a single item asking how many hours per
(e.g., Grandner et al., 2006). Thus, in line with prior work using month participants give emotional support to their partner (e.g.,
the PSQI, a global sleep problems index was computed by sum- comforting, lending a listening ear, giving advice; Rossi,
ming the seven sleep components for each participant with 2001). Given the open-ended nature of the item, there were
complete data (a ¼ .69; see Online Supplemental Material for some outliers with very high values on this variable. Responses
supplemental analyses using the component scores separately). higher than 2.5 standard deviations of the mean were recoded
Objective sleep outcomes were measured by collecting acti- to the highest value below 2.5 standard deviations to reduce the
graphy data. Participants wore a Mini Mitter Actiwatch1-64 influence of the outliers on the results.
activity monitor on their nondominant wrist for 7 consecutive Agreeableness was measured by asking participants the
days and nights starting on a Tuesday morning at 7:00 a.m. and extent to which each of five adjectives (helpful, warm, caring,
ending the next Tuesday morning. Using a built-in sensor, the softhearted, sympathetic) described them (1 ¼ a lot to 4 ¼ not
monitor detects the number of movements made by the wearer. at all; Rossi, 2001). Responses were reverse coded, so that
The start and end times of actigraphic records were determined higher scores indicated greater agreeableness (a ¼ .81).
using diary logs in which participants entered their bedtime and
risetime. Activity counts within 30-s epochs were used to esti-
mate sleep statistics. Whether participants were asleep or Results
awake was estimated by comparing activity counts in each Table 1 provides the correlations among variables of interest.
epoch and the epochs surrounding it to a predetermined thresh- As in prior work (Grandner et al., 2006), the PSQI global score
old value. Sleep duration was computed by summing the showed weak correlations with the actigraph assessments (all
epochs, in minutes, marked as sleep during a night’s interval rs < .24), suggesting that the two measures tap different aspects
(the difference between the start and end times logged in the of sleep quality. Looking at the specific PSQI components that
 Table 1. Descriptive Statistics and Correlations Among Variables.

Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1. Partner –
responsiveness
2. Sleep problems .144*** –
(PSQI)
3. Sleep efficiency .015 .159* –
(actigraph)
4. WASO (actigraph) .118 .242*** .608*** –
5. Sleep onset .039 .089 .728*** .197** –
(actigraph)
6. Sleep duration .036 .068 .549*** .032 .358*** –
(actigraph)
7. Anxiety .148*** .359*** .186** .294*** .053 .008 –
8. Depression .212*** .401*** .070 .054 .042 .042 .369*** –
9. Agreeableness .110** .005 .097 .026 .168* .035 .071 .212*** –
10. Provision of .142*** .046 .092 .041 .067 .134* .002 .045 .106** –
emotional support
11. Poor perceived .049 .256*** .195** .218** .153* .009 .340*** .346*** .044 .039 –
health
12. Health symptoms .001 .254*** .131 .193** .175** .070 .308*** .164*** .076* .056 .345*** –
13. BMI .069 .062 .176** .114 .151* .147* .104** .092* .014 .048 .240*** .151*** –
14. Age .137*** .073 .138* .150* .086 .002 .021 .172*** .086* .069 .022 .355*** .004 –
15. Education .028 .033 .011 .071 .012 .077 .124*** .066 .094* .110** .119** .006 .059 .059 –
16. Racea .046 .015 .031 .035 .035 .068 .038 .067 .031 .029 .162*** .014 .080* .108** .031 –
17. Genderb .151*** .162*** .313*** .126 .230*** .316*** .096* .012 .221*** .101** .034 .068 .100** .167*** .049 .014

M 3.592 5.640 82.975 42.522 24.001 384.913 21.321 50.534 3.420 27.546 2.280 3.930 29.092 57.460 7.830 –
SD 0.538 3.357 7.887 18.391 20.921 62.514 4.356 12.338 0.496 38.758 0.909 2.807 5.565 11.174 2.449 –

Note. PSQI ¼ Pittsburgh Sleep Quality Index; WASO ¼ wake after sleep onset; BMI ¼ body mass index. The sample size was 682 for estimates including the PSQI, 219 for estimates including actigraph assessments, and 698
for the remaining estimates.
a
0 ¼ White, 1 ¼ non-White. b0 ¼ male, 1 ¼ female.
*p < .05. **p < .01. ***p < .001.

87
88 Social Psychological and Personality Science 8(1)

Figure 1. The indirect associations of perceived partner responsiveness with global sleep problems (Panel A) and actigraph-assessed sleep
efficiency (Panel B) through anxiety and depression. Numbers outside the parentheses are unstandardized regression coefficients and numbers
inside the parentheses are standard errors. The sample size was 682 in analyses predicting global sleep problems and 219 in analyses predicting
sleep efficiency. PSQI ¼ Pittsburgh Sleep Quality Index. **p <.01. ***p < .001.

map onto the actigraph-based measures, PSQI sleep efficiency (although the effect size was similar to that of the IA through
was unrelated to the actigraph sleep efficiency (r ¼ .104, anxiety, B ¼ .272, SE ¼ .218, p ¼ .213). The indirect asso-
p ¼ .140) and the PSQI sleep duration was moderately related ciations between perceived partner responsiveness and global
to actigraph sleep duration (r ¼ .379, p < .001; see Table S1 in sleep problems held even when the analyses were repeated
Online Supplemental Materials for all pairs of correlations by controlling for emotional support provision to the partner,
between the actigraphy indices and the PSQI components). agreeableness, demographic factors, and physical health fac-
tors (IA ¼ .129, 95% CI [0.297, 0.027] for anxiety and
IA ¼ .280, 95% CI [0.465, 0.143] for depression; see
Perceived Partner Responsiveness and Self-Reported Table S2 in Online Supplemental Materials for all direct and
Sleep Problems indirect associations between partner responsiveness and the
Participants who perceived their partner as responsive reported PSQI subcomponents).
lower sleep problems as measured by the global PSQI score
(B ¼ .901, SE ¼ .237, p < .001). Partner responsiveness also
Perceived Partner Responsiveness and Actigraph-
indirectly predicted lower global sleep problems through lower
anxiety (indirect association [IA] ¼ .223, 95% CI: [0.426,
Assessed Sleep Efficiency
0.081]) and depression (IA ¼ .406, 95% CI: [0.620, Partner responsiveness was not directly associated with
0.251]; Figure 1A; see Online Supplemental Materials for actigraph-assessed sleep efficiency or sleep duration (Table
complete details on the data analytic approach for testing indi- 1). However, partner responsiveness indirectly predicted
rect associations). Once anxiety and depression were included greater objective sleep efficiency via lower anxiety (IA ¼
in the model, the direct association between partner responsive- .678, 95% CI [0.100, 2.014]) but not depression (IA ¼ .052,
ness and self-reported sleep problems was not significant 95% CI [0.398, 0.504]; Figure 1B). Improved sleep
Selcuk et al. 89

efficiency could be due to faster sleep onset or lower wake after a direct effect. The present findings are in line with this theo-
sleep onset. Our findings supported the latter possibility. rizing by showing that partner responsiveness is mainly linked
Whereas perceived partner responsiveness indirectly predicted to sleep through psychological symptoms—particularly anx-
lower wake after sleep onset through lower anxiety (IA ¼ ious arousal but also anhedonic depression.
2.495, 95% CI [5.650, 0.728]), it was not associated with Although partner responsiveness predicted better actigraph-
sleep onset (IA ¼ .529, 95% CI [03.052, 0.630]). The indi- assessed sleep efficiency through lower anxiety, there was no
rect associations between perceived partner responsiveness and such indirect association between partner responsiveness and
objective sleep quality through anxiety held, even after adjust- actigraph-assessed sleep duration. A prior study on parental
ing for emotional support provision to the partner, agreeable- responsiveness and child sleep reached a similar conclusion,
ness, demographic factors, and physical health covariates (IA with parental responsiveness predicting parent reports of child
¼ .566, 95% CI [0.013, 1.856] for sleep efficiency and IA ¼ sleep quality but not duration, although that study only focused
1.796, 95% CI [4.656, 0.140] for wake after sleep onset). on the direct links (Bordeleau et al., 2012). Taken together with
After adjusting for covariates, partner responsiveness was not the present findings, it seems that sleep duration is unrelated
associated with actigraph-assessed sleep duration through anxi- to responsiveness of partners or caregivers, but individuals
ety (IA ¼ 2.195, 95% CI [.719, 10.389]) or depression (IA ¼ with less responsive close others experience more disrupted
1.148, 95% CI [1.113, 5.691]). sleep. This qualitative difference between individuals who
have responsive versus unresponsive partners is important,
as chronic disruptions—that is, inefficient sleep—predicts
Discussion important physical health outcomes, including mortality
These findings are the first to demonstrate how perceived part- (Dew et al., 2003).
ner responsiveness is linked to subjective and objective sleep Depression also mediated the partner responsiveness–sleep
quality. Perceived partner responsiveness predicted lower glo- association but only for subjective sleep problems. This finding
bal sleep problems through lower anxiety and depression. replicates prior work on the association between depression
Importantly, perceived partner responsiveness was also associ- and the PSQI and extends a recent finding that depression med-
ated with actigraph-assessed sleep efficiency through lower iates the association between quality of general social ties and
anxiety (but not depression). These indirect associations the global PSQI score (Kent, Uchino, Cribbet, Bowen, &
remained significant after we statistically controlled for emo- Smith, 2015). Given our finding that subjective evaluations
tional support provision to the partner, agreeableness, and of sleep quality do not reflect actual sleep efficiency, depressed
demographic (age, gender, race, and education) and health cov- individuals may be negatively biased in their perceptions of
ariates (perceived health, health symptoms, and BMI) that their psychological and physiological states, which may extend
could have potentially accounted for the findings. to sleep (Grandner et al., 2006). This is not to say that the PSQI
An important strength of the present study was using a com- assessments are irrelevant to sleep quality; on the contrary, sub-
bination of subjective (the PSQI) and objective (actigraph) jective sleep quality does predict important health and well-
sleep measures. Past work showed that the PSQI, the most being outcomes (Lemola et al., 2013; Martin et al., 2011).
widely used subjective sleep quality measure, and actigraph Rather, the present findings show the importance of using mul-
assessments are not substitutes for each other but rather mea- tiple measures to study the links between close relationships
sure distinct aspects of sleep quality (Grandner et al., 2006; and sleep, as the nature of the associations and the mediating
Landry, Best, & Liu-Ambrose, 2015). The low correlations psychological mechanisms may be different across measures.
between the PSQI and actigraphy assessments (also replicated The present findings also dovetail with and extend past work
in the present work) have led researchers to suggest that both investigating the role of romantic attachment orientations in sleep
measures should be included in sleep studies whenever possi- quality. Insecure (i.e., anxious or avoidant) attachment, which is
ble (Landry et al., 2015). Using the two measures in the same thought to result from close others’ failure to behave responsively,
study enabled us to document the distinct pathways by which has been linked to poor self-reported sleep quality and sleep dis-
perceived partner responsiveness is associated with sleep. turbances such as difficulty falling asleep and staying asleep
We found a direct association between partner responsive- (e.g., Adams & McWilliams, 2015; Carmichael & Reis, 2005;
ness and sleep only for the PSQI but not for the actigraph- see also Adams, Stoops, & Skomro, 2014, for a review). There
assessed sleep quality. Indirect associations were much more was a significant direct association between perceived partner
pronounced across both subjective and objective sleep mea- responsiveness and subjective sleep quality in the present sample
sures. The more consistent pattern with indirect (vs. direct) as well. In addition, the present study extended and complemen-
associations is in line with theoretical models explaining how ted prior findings by documenting psychological mechanisms
romantic relationships are associated with physical health through which partner responsiveness is linked to both self-
(Burman & Margolin; 1992; Kiecolt-Glaser & Newton, 2001; reported sleep problems and objectively assessed sleep efficiency.
Slatcher, 2010; Slatcher & Selcuk, in press). These models sug- Recent studies have shown that perceived partner respon-
gest that romantic relationship processes are more likely to be siveness has relevance for health outcomes including
linked to physical health through psychological mechanisms all-cause mortality (Selcuk & Ong, 2013). There is increasing
(e.g., psychological symptoms, well-being) rather than having evidence that partner responsiveness predicts potential
90 Social Psychological and Personality Science 8(1)

mechanisms, including affective reactivity to stressors, trait the findings. Future studies would benefit from replicating
negative affect, depression, psychological well-being, and diur- these findings in a more heterogeneous sample.
nal cortisol (Fekete et al., 2007; Selcuk et al., 2016; Selcuk, In sum, the present study demonstrated the role of perceived
Zayas, Gunaydin, Hazan, & Kross, 2012; Slatcher et al., partner responsiveness in subjective and objective assessments
2015), that may ultimately affect adult morbidity and mortality. of sleep quality through lower anxiety and depression. Future
By showing that perceived partner responsiveness predicts self- research should further elucidate the mechanisms by which
reported sleep problems through lower anxiety and depression, higher partner responsiveness exerts a salutary influence on
and objective sleep efficiency through lower anxiety, the pres- health and well-being.
ent study extends the set of processes by which perceived part-
ner responsiveness potentially affects physical health. Declaration of Conflicting Interests
The findings also have implications for therapy and inter- The author(s) declared no potential conflicts of interest with respect to
vention design. The inherently interdependent nature of adult the research, authorship, and/or publication of this article.
romantic relationships means that romantic partners, as well
as perceptions of one’s romantic partner, play a meaningful Funding
role in promoting better health and well-being. Our findings
The author(s) disclosed receipt of the following financial support for the
suggest that enhancing perceived partner responsiveness has
research, authorship, and/or publication of this article: The MIDUS II
the potential to increase the effectiveness of interventions
research was supported by a grant from the National Institute on Aging
designed to reduce sleep disturbances in particular and improve
(P01-AG020166). The Biomarker Project was further supported by the
individual well-being in general.
following grants: M01-RR023942 (Georgetown), M01-RR00865
Before concluding, we acknowledge some limitations of the
(UCLA) from the General Clinical Research Centers Program and
present research. These data are correlational, meaning that we
UL1TR000427 (UW) from the National Center for Advancing Transla-
are unable to make claims about the causal direction of the asso-
tional Sciences (NCATS), National Institutes of Health.
ciations between partner responsiveness, anxiety and depression,
and sleep. For example, it is possible that partner responsiveness
Supplemental Material
may be linked to anxiety, depression, and sleep simultaneously,
or poor sleep could have affected scores on the MASQ as well as The online data supplements are available at http://journals.sagepub
individuals’ perceptions of their partner’s responsiveness. The .com/doi/suppl/10.1177/1948550616662128.
existing literature, however, makes a stronger theoretical case
for individuals who experience higher partner responsiveness References
to have better sleep outcomes, rather than the other way around Adams, G. C., & McWilliams, L. A. (2015). Relationships between
(cf. Carmichael & Reis, 2005; Selcuk et al., 2015; Troxel, adult attachment style and sleep disturbances in a nationally repre-
Buysse, Hall, & Matthews, 2009). Moreover, the fact that part- sentative sample. Journal of Psychosomatic Research, 79, 37–42.
ner responsiveness was assessed on average 25 months before Retrieved from http://doi.org/10.1016/j.jpsychores.2014.12.017
both the mediating and outcome variables supports this possi- Adams, G. C., Stoops, M. A., & Skomro, R. P. (2014). Sleep tight:
bility, although we should note that we were not able to model Exploring the relationship between sleep and attachment style
change in sleep behavior. Related to this point, it may be pos- across the life span. Sleep Medicine Reviews, 18, 495–507.
sible to observe stronger associations between partner respon- Retrieved from http://doi.org/10.1016/j.smrv.2014.03.002
siveness and sleep in a design with a smaller time interval Alvaro, P. K., Roberts, R. M., & Harris, J. K. (2013). A systematic
between the measurements. For instance, future daily experi- review assessing bidirectionality between sleep disturbances, anxi-
ence studies may investigate whether daily perceptions of ety, and depression. Sleep, 36, 1059–1068. Retrieved from http://
responsiveness predicts sleep, especially on days participants doi.org/10.5665/sleep.2810
experience stressors and exhibit greater anxiety. Bordeleau, S., Bernier, A., & Carrier, J. (2012). Longitudinal associa-
A second limitation of the current study involves the diver- tions between the quality of parent–child interactions and chil-
sity of the sample. The MIDUS sample is not racially diverse, dren’s sleep at preschool age. Journal of Family Psychology, 26,
limiting our ability to generalize our findings to non-White 254–262. Retrieved from http://doi.org/10.1037/a0027366
individuals. Furthermore, although the PSQI was administered Bowlby, J. (1980). Attachment and loss: Loss, sadness and depression.
at three different study sites (University of California Los New York, NY: Basic Books.
Angeles, University of Wisconsin, and Georgetown Univer- Burman, B., & Margolin, G. (1992). Analysis of the association
sity), the subset of participants who provided actigraph data between marital relationships and health problems: An interac-
of sleep quality completed the study at only one site (University tional perspective. Psychological Bulletin, 112, 39–63. doi:10.
of Wisconsin). Finally, participants were in their middle and 1037/0033-2909.112.1.39
late adulthood, leaving the question of whether the results Buysse, D. J., Reynolds, C. F., Monk, T. H., Berman, S. R., & Kupfer,
would look similar for younger individuals open for future D. J. (1989). The Pittsburgh Sleep Quality Index: A new instru-
research. Regardless, the emergence of the partner responsive- ment for psychiatric practice and research. Psychiatry Research,
ness–sleep link through lower anxiety in analyses that included 28, 193–213. Retrieved from http://doi.org/10.1016/0165-
potentially meaningful covariates bolsters our confidence in 1781(89)90047-4
Selcuk et al. 91

Cacioppo, J. T., Hawkley, L. C., Berntson, G. G., Ernst, J. M., Gibbs, A. Holt-Lunstad, J., Smith, T. B., & Layton, J. B. (2010). Social relation-
C., Stickgold, R., & Hobson, J. A. (2002). Do lonely days invade the ships and mortality risk: A meta-analytic review. PLoS Medicine,
nights? Potential social modulation of sleep efficiency. Psychologi- 7, e1000316. doi:10.1371/journal.pmed.1000316
cal Science, 13, 384–387. doi:10.1111/1467-9280.00469 Ibarra-Rovillard, M. S., & Kuiper, N. A. (2011). Social support and
Carmichael, C. L., & Reis, H. T. (2005). Attachment, sleep quality, social negativity findings in depression: Perceived responsiveness
and depressed affect. Health Psychology, 24, 526–531. Retrieved to basic psychological needs. Clinical Psychology Review, 31,
from http://doi.org/10.1037/0278-6133.24.5.526 342–352. Retrieved from http://doi.org/10.1016/j.cpr.2011.01.005
Collins, N. L., & Feeney, B. C. (2000). A safe haven: An attachment Kane, H. S., Slatcher, R. B., Reynolds, B. M., Repetti, R. L., & Robles,
theory perspective on support seeking and caregiving in intimate T. F. (2014). Daily self-disclosure and sleep in couples. Health
relationships. Journal of Personality and Social Psychology, 78, Psychology, 33, 813–22. Retrieved from http://doi.org/10.1037/
1053–1073. Retrieved from http://doi.org/10.1037/0022-3514.78. hea0000077
6.1053 Kent, R. G., Uchino, B. N., Cribbet, M. R., Bowen, K., & Smith, T. W.
Collins, N. L., Guichard, A. C, Ford, M. B., & Feeney, B. C. (2006). (2015). Social relationships and sleep quality. Annals of Beha-
Responding to need in intimate relationships: Normative processes vioral Medicine, 49, 912–917. doi:10.1007/s12160-015-9711-6
and individual differences. In M. Mikulincer & G. S. Goodman Khan, C. M., Iida, M., Stephens, M. A. P., Fekete, E. M., Druley, J. A.,
(Eds.), Dynamics of romantic love: Attachment, caregiving, and & Greene, K. A. (2009). Spousal support following knee surgery:
sex (pp. 149–189). New York, NY: Guilford Press. Roles of self-efficacy and perceived emotional responsiveness.
Dew, M. A., Hoch, C. C., Buysse, D. J., Monk, T. H., Begley, A. E., Rehabilitation Psychology, 54, 28–32. Retrieved from http://doi.
Houck, P. R., . . . Reynolds, C. F. (2003). Healthy older adults’ org/10.1037/a0014753
sleep predicts all-cause mortality at 4 to 19 years of follow-up. Psy- Kiecolt-Glaser, J. K., & Newton, T. L. (2001). Marriage and health:
chosomatic Medicine, 65, 63–73. His and hers. Psychological Bulletin, 127, 472–503. doi:10.1037/
Dienberg Love, G., Seeman, T. E., Weinstein, M., & Ryff, C. D. 0033-2909.127.4.472
(2010). Bioindicators in the MIDUS national study: protocol, mea- Koffel, E., & Watson, D. (2009). The two-factor structure of sleep
sures, sample, and comparative context. Journal of Aging and complaints and its relation to depression and anxiety. Journal of
Health, 22, 1059–1080. Retrieved from http://doi.org/10.1177/ Abnormal Psychology, 118, 183–194. Retrieved from http://doi.
0898264310374355 org/10.1037/a0013945
Doane, L. D., Mineka, S., Zinbarg, R. E., Craske, M., Griffith, J. W., & Landry, G. J., Best, J. R., & Liu-Ambrose, T. (2015). Measuring sleep
Adam, E. K. (2013). Are flatter diurnal cortisol rhythms associated quality in older adults: A comparison using subjective and objec-
with major depression and anxiety disorders in late adolescence? tive methods. Frontiers in Aging Neuroscience, 7, 166.
The role of life stress and daily negative emotion. Development Lemay, E. P., Clark, M. S., & Feeney, B. C. (2007). Projection of
and Psychopathology, 25, 629–642. responsiveness to needs and the construction of satisfying commu-
Eisenberger, N. I., Master, S. L., Inagaki, T. K., Taylor, S. E., Shirin- nal relationships. Journal of Personality and Social Psychology,
yan, D., Lieberman, M. D., & Naliboff, B. D. (2011). Attachment 92, 834–853. doi:10.1037/0022-3514.92.5.834
figures activate a safety signal-related neural region and reduce Lemola, S., Ledermann, T., & Friedman, E. M. (2013). Variability of
pain experience. Proceedings of the National Academy of Sciences sleep duration is related to subjective sleep quality and subjective
of the United States of America, 108, 11721–11726. doi:10.1073/ well-being: An actigraphy study. PLoS ONE, 8, e71292.
pnas.1108239108 Liu, L., Fiorentino, L., Rissling, M., Natarajan, L., Parker, B. A.,
Feeney, B. C. (2007). The dependency paradox in close relationships: Dimsdale, J. E., . . . Ancoli-Israel, S. (2013). Decreased health-
Accepting dependence promotes independence. Journal of Person- related quality of life in women with breast cancer is associated
ality and Social Psychology, 92, 268–285. Retrieved from http:// with poor sleep. Behavioral Sleep Medicine, 11, 189–206. doi:
doi.org/10.1037/0022-3514.92.2.268 10.1080/15402002.2012.660589
Feeney, B. C., & Collins, N. L. (2015). Thriving through relationships. Loving, T. J., & Slatcher, R. B. (2013). Romantic relationships and
Current Opinion in Psychology, 1, 22–28. Retrieved from http:// health. In J. A. Simpson & L. Campbell (Eds.), The Oxford hand-
dx.doi.org/10.1016/j.copsyc.2014.11.001 book of close relationships (pp. 617–637). New York, NY: Oxford
Fekete, E. M., Stephens, M. A. P., Mickelson, K. D., & Druley, J. A. University Press.
(2007). Couples’ support provision during illness: The role of Magee, J. C., & Carmin, C. N. (2010). The relationship between sleep
perceived emotional responsiveness. Families, Systems, & and anxiety in older adults. Current Psychiatry Reports, 12, 13–19.
Health, 25, 204–217. Retrieved from http://doi.org/10.1037/ Retrieved from http://doi.org/10.1007/s11920-009-0087-9
1091-7527.25.2.204 Martin, J. L., Fiorentino, L., Jouldjian, S., Mitchell, M., Josephson,
Grandner, M. A., Kripke, D. F., Yoon, I. Y., & Youngstedt, S. D. K. R., & Alessi, C. A. (2011). Poor self-reported sleep quality
(2006). Criterion validity of the Pittsburgh Sleep Quality Index: predicts mortality within one year of inpatient post-acute rehabi-
Investigation in a non-clinical sample. Sleep and Biological litation among older adults. Sleep, 34, 1715–1721. doi:10.5665/
Rhythms, 4, 129–136. doi:10.1111/j.1479-8425.2006.00207.x sleep.1444
Hawkley, L. C., Preacher, K. J., & Cacioppo, J. T. (2010). Loneliness Mezick, E. J., Wing, R. R., & McCaffery, J. M. (2014). Associations
impairs daytime functioning but not sleep duration. Health Psy- of self-reported and actigraphy-assessed sleep characteristics with
chology, 29, 124–129. doi:10.1037/a0018646. body mass index and waist circumference in adults: Moderation by
92 Social Psychological and Personality Science 8(1)

gender. Sleep Medicine, 15, 64–70. Retrieved from http://doi.org/ Simpson, J. A., Rholes, W. S., & Nelligan, J. S. (1992). Support seek-
10.1016/j.sleep.2013.08.784 ing and support giving within couples in an anxiety-provoking sit-
Mikulincer, M., & Shaver, P. R. (2007). Attachment patterns in adult- uation: The role of attachment styles. Journal of Personality and
hood: Structure, dynamics, and change. New York, NY: Guilford Social Psychology, 62, 434–446. Retrieved from http://doi.org/
Press. 10.1037/0022-3514.62.3.434
National Sleep Foundation. (2013). International bedroom poll: Sum- Slatcher, R. B. (2010). Marital functioning and physical health: Impli-
mary of Findings. Arlington, VA: Author. cations for social and personality psychology. Social and Person-
Ong, A. D., Exner-Cortens, D., Riffin, C., Steptoe, A., Zautra, A., & ality Psychology Compass, 4, 455–469. doi:10.1111/j.1751-9004.
Almeida, D. M. (2013). Linking stable and dynamic features of 2010.00273.x
positive affect to sleep. Annals of Behavioral Medicine, 46, Slatcher, R. B., & Selcuk, E. (in press). A social psychological per-
52–61. Retrieved from http://doi.org/10.1007/s12160-013-9484-8 spective on the links between close relationships and health. Cur-
Reid, K. J., Martinovich, Z., Finkel, S., Statsinger, J., Golden, R., Har- rent Directions in Psychological Science.
ter, K., & Zee, P. C. (2006). Sleep: A marker of physical and men- Slatcher, R. B., Selcuk, E., & Ong, A. D. (2015). Perceived partner
tal health in the elderly. The American Journal of Geriatric responsiveness predicts diurnal cortisol profiles 10 years later. Psy-
Psychiatry, 14, 860–866. Retrieved from http://doi.org/10.1097/ chological Science, 26, 972–982. Retrieved from http://doi.org/10.
01.JGP.0000206164.56404.ba 1177/0956797615575022
Reis, H. T. (2007). Steps toward the ripening of relationship science. Stepanski, E. J., & Rybarczyk, B. (2006). Emerging research on the
Personal Relationships, 14, 1–23. Retrieved from http://doi.org/ treatment and etiology of secondary or comorbid insomnia. Sleep
10.1111/j.1475-6811.2006.00139.x Medicine Reviews, 10, 7–18. doi:10.1016/j.smrv.2005.08.002
Reis, H. T., & Patrick, B. C. (1996). Attachment and intimacy: Com- Troxel, W. M. (2010). It’s more than sex: Exploring the dyadic nature
ponent processes. In E. T. Higgins & A. W. Kruglanski (Eds.), of sleep and implications for health. Psychosomatic Medicine, 72,
Social psychology: Handbook of basic principles (pp. 523–563). 578–586. Retrieved from http://doi.org/10.1097/PSY.
New York, NY: Guilford Press. 0b013e3181de7ff8.It
Revenson, T. A., Marı́n-Chollom, A. M., Rundle, A. G., Wisnivesky, Troxel, W. M., Buysse, D. J., Hall, M., & Matthews, K. A. (2009).
J., & Neugut, A. I. (2016). Hey Mr. Sandman: Dyadic effects of Marital happiness and sleep disturbances in a multi-ethnic sample
anxiety, depressive symptoms and sleep among married couples. of middle-aged women. Behavioral Sleep Medicine, 7, 2–19.
Journal of Behavioral Medicine, 39, 225–232. Retrieved from http://doi.org/10.1080/15402000802577744
Rossi, A. S. (2001). Caring and doing for others: Social responsibility Watson, D., Weber, K., Assenheimer, J. S., Clark, L. A., Strauss, M. E.,
in the domains of family, work, and community. Chicago, IL: Uni- & McCormick, R. A. (1995). Testing a tripartite model: I.
versity of Chicago Press. Evaluating the convergent and discriminant validity of anxiety
Ryff, C. D., Almeida, D. M., Ayanian, J. S., Carr, D. S., Cleary, P. D., and depression symptom scales. Journal of Abnormal Psychol-
Coe, C., . . . Williams, D. (2007) Midlife Development in the United ogy, 104, 3–14. Retrieved from http://doi.org/10.1037/0021-
States (MIDUS II), 2004–2006 [Computer file]. ICPSR04652-v1. 843X.104.1.3
Ann Arbor, MI: Inter-university Consortium for Political and
Social Research [distributor]. doi:10.3886/ICPSR04652 Author Biographies
Selcuk, E., Gunaydin, G., Ong, A. D., & Almeida, D. M. (2016). Does
partner responsiveness predict hedonic and eudaimonic well- Emre Selcuk is an assistant professor in the Department of Psychol-
being? A 10-year longitudinal study. Journal of Marriage and ogy at Middle East Technical University, Turkey. His research focuses
Family, 78, 311–325. on the formation, maintenance, and functions of attachment relation-
Selcuk, E., & Ong, A. D. (2013). Perceived partner responsiveness ships across the life span.
moderates the association between received emotional support and Sarah C. E. Stanton is a postdoctoral fellow at Wayne State Univer-
all-cause mortality. Health Psychology, 32, 231–235. Retrieved sity. She uses a social psychological approach to understand the cog-
from http://doi.org/10.1037/a0028276 nitive and affective aspects of close relationships and their effects on
Selcuk, E., Zayas, V., Gunaydin, G., Hazan, C., & Kross, E. (2012). behavior, physiology, and health and well-being.
Mental representations of attachment figures facilitate recovery
following upsetting autobiographical memory recall. Journal of Richard B. Slatcher is an associate professor of social psychology at
Personality and Social Psychology, 103, 362–378. doi:10.1037/ Wayne State University. His research has two main facets: basic
a0028125 research on close relationship processes and investigations of the links
Selcuk, E., Zayas, V., & Hazan, C. (2010). Beyond satisfaction: The between close relationships, biological processes, and physical health.
role of attachment in marital functioning. Journal of Family The-
Anthony D. Ong is an associate professor of human development at
ory & Review, 2, 258–279. Retrieved from http://doi.wiley.com/
Cornell University and an associate professor of gerontology in med-
10.1111/j.1756-2589.2010.00061.x
icine at Weill Cornell Medical College. His research focuses on the
Shen, X., Wu, Y., & Zhang, D. (2016) Nighttime sleep duration, 24-
dynamic processes that underlie expressions of vulnerability and
hour sleep duration and risk of all-cause mortality among adults:
adaptation across the life span.
A metaanalysis of prospective cohort studies. Scientific Reports,
6, 21480. doi:10.1038/srep21480 Handling Editor: Nickola Overall