Article

Social Psychological and

Personality Science
Personality and Walking Speed Across 2018, Vol. 9(7) 773-780
ª The Author(s) 2017
Article reuse guidelines:
Adulthood: Prospective Evidence From sagepub.com/journals-permissions
DOI: 10.1177/1948550617725152
Five Samples journals.sagepub.com/home/spp

Yannick Stephan1, Angelina R. Sutin2, Gabriel Bovier-Lapierre1,

and Antonio Terracciano2

Abstract
Walking speed is one marker of health in adulthood. Although personality may contribute to gait speed, there is limited
longitudinal data on this association. Thus, the present study examined whether personality traits are prospectively associated
with walking speed among middle aged and older adults. Participants were adults aged from 25 to 100 years old (N > 15,000)
drawn from the Wisconsin Longitudinal Study Graduate and Sibling samples, the Midlife in the United States Survey, the
Health and Retirement Study, and the National Health and Aging Trends Survey. Across most samples and in a meta-analysis,
lower neuroticism and higher extraversion, conscientiousness, and openness at baseline were prospectively related to faster
gait speed. In the HRS, lower neuroticism and higher extraversion, conscientiousness, and openness were related to slower
gait speed decline. This study provides robust evidence that walking speed in adulthood reflects, in part, the individual’s
personality.

Keywords
personality, walking speed, adulthood

Walking (or gait) speed is considered one marker of an adulthood (Buchman et al., 2013; Canada, Stephan, Jaconelli,
individual’s health (Studenski, Perera, Wallace, et al., 2003; & Duberstein, 2016; Krueger, Wilson, Shah, Tang, & Bennett,
Studenski, Perera, Patel, et al., 2011). Longitudinal research 2006; Stephan, Sutin, Canada, & Terracciano, 2017; Suchy,
finds consistently that slower gait is predictive of a range Williams, Kraybill, Franchow, & Butner, 2010; Terracciano,
of deleterious outcomes, including poor mental health Stephan, Luchetti, Gonzalez-Rothi, & Sutin, in press; Tolea,
(Demakakos et al., 2013), higher risk of incident functional Costa et al., 2012). Higher neuroticism and lower extraversion,
limitations and disability (Cesari et al., 2009; Perera et al., openness, and conscientiousness are related to lower self-
2016), impaired cognition and incident dementia (Dumurgier reported physical functioning (Canada et al., 2016; Duberstein
et al., in press; Verghese, Wang, Lipton, Holtzer, & Xue, et al., 2003; Suchy et al., 2010) and are risk factors for objec-
2007), and ultimately higher mortality risk (Studenski, Perera, tively assessed poor muscle strength (Tolea, Terracciano, Mila-
Patel, et al., et al., 2011). A large range of factors that contrib- neschi, Metter, & Ferrucci, 2012), worse respiratory function
ute to gait speed have been identified from biological to social (Terracciano, Stephan, Luchetti, Gonzalez-Rothi, & Sutin, in
(Brunner et al., 2009; Rosso et al., 2015). The present study press), impaired aerobic capacity (Terracciano et al., 2013),
focused on the extent to which walking speed reflects individ- and steeper motor declines (Buchman et al., 2013). The find-
uals’ characteristic ways of thinking, feeling, and behaving, ings are more mixed for an association between agreeableness
that is, their personality traits. (the tendency to be cooperative and altruistic) and either self-
There is extensive evidence for the association between per- reported or objective measures of physical function (Canada
sonality and health across the life span (Friedman & Kern, et al., 2016; Terracciano et al., in press).
2014). Based upon the five-factor model (FFM; Digman,
1990), neuroticism (the tendency to experience distress and
1
negative emotions), extraversion (the tendency to be sociable Euromov, University of Montpellier, Montpellier, France
2
and to experience positive emotions), openness (the tendency College of Medicine, Florida State University, Tallahassee, FL, USA
to search for variety and to entertain new ideas), and conscien-
Corresponding Author:
tiousness (the tendency to be self-disciplined and organized) Yannick Stephan, EA 2991 Euromov, University of Montpellier, UFRSTAPS, 700
have been related to both self-reported and performance- Avenue du Pic St Loup, 34090 Montpellier, France.
based measures of physical and motor functions across Email: yannick.stephan@umontpellier.fr
774 Social Psychological and Personality Science 9(7)

Few studies have examined the association between person- Method

ality and walking speed. For example, Terracciano et al. (2013)
examined personality correlates of walking speed at normal
Participants
and maximal sustained speed on a 400-m course. Higher con- Participants were drawn from the Wisconsin Longitudinal
scientiousness, extraversion, and openness and lower neuroti- Study Graduate (WLSG) sample, the Wisconsin Longitudinal
cism were related to faster walking speed on both the normal Study Sibling (WLSS) sample, the Midlife in the United
and the maximal sustained pace; agreeableness was unrelated States Survey (MIDUS), the National Health and Aging
to test performance. To our knowledge, only one study has Trends Survey (NHATS), and the Health and Retirement
examined the association between personality and walking Study (HRS). In these five samples, personality was assessed
speed over time. Tolea et al. (2012) showed that conscientious- at baseline and walking speed was measured at follow-up. The
ness mitigated the decline in walking speed over 3 years in a present study included only participants from the five samples
sample of older adults, whereas no association was found for with complete data on baseline personality, demographic fac-
openness; the other FFM traits were not examined in the Tolea tors (age, sex, education, and race), and follow-up walking
and colleagues study. speed. The HRS and the NHATS also included a baseline
Taken as a whole, current knowledge on the association assessment of gait speed.
between personality and walking speed is rather scarce. In par- The WLS is a study of 1957 Wisconsin high schools gradu-
ticular, there is limited data on the extent to which personality ates and a selected sibling of some of the graduates. The WLS
prospectively predicts gait speed over time. Studies have either sample is broadly representative of older, White, non-Hispanic
relied on relatively small or selective samples or have not Americans who have completed at least a high school educa-
examined all five major dimensions of personality. When pre- tion. The present study used personality data collected in
dicting walking speed at follow-up or change in walking speed 1992–1993 for graduates and 1993–1994 for siblings. Walking
over time, we expect that all personality traits, with the excep- speed was measured in 2011 for both samples. The final ana-
tion of agreeableness, will be related to this marker of mobility lyzed sample was composed of 4,603 individuals in the WLSG
and health. Indeed, lower neuroticism and higher extraversion, and from 2,090 participants in the WLSS (see Table 1). More
openness, and conscientiousness are robust predictors of more information about both WLS samples and how to access the
frequent physical activity (Sutin et al., 2016) and contribute to data can be found at http://www.ssc.wisc.edu/wlsresearch/.
lower risk of chronic conditions over time (Weston, Hill, & The MIDUS is a longitudinal study of U.S. adults. Personality
Jackson, 2015). In turn, frequent physical activity and better data from the first wave (1994–1995, MIDUS I) were used, and
health are related to faster gait speed (Ip et al., 2013; Tolea walking speed was obtained from the Biomarker Project of the
et al., 2010). Furthermore, these personality traits are also asso- second wave (2004–2009, MIDUS II). The complete analyzed
ciated with less physiological dysregulation (Stephan, Sutin, sample was composed of 990 participants (Table 1). More
Lucchetti, & Terracciano, 2016; Sutin, Stephan, & Terrac- information about MIDUS and how to access the data can be
ciano, in press), better cardiorespiratory fitness (Terracciano found at http://www.midus.wisc.edu/.
et al., 2013), and better cognitive function (Luchetti, Terrac- The NHATS is a nationally representative prospective
ciano, Stephan, & Sutin, 2016; Pearman, 2009). This physiolo- cohort study of Medicare enrollees aged 65 years and older.
gical and cognitive profile has been found to mitigate the Participants were first interviewed in 2011 and are reinter-
decline in gait speed (Gale, Allerhand, Sayer, Cooper, & viewed annually. Personality was first assessed in 2013 for one
Deary, 2014; Richardson, Glynn, Ferrucci, & Mackey, 2015; third of the sample and in 2014 for the second third. The pres-
Verghese et al., 2011). The identification of a link between per- ent study used the combined personality data from these two
sonality and walking speed may contribute to a better under- waves and follow-up walking speed data were obtained in
standing of the motor signatures of the traits. Furthermore, 2015. The analyzed sample was composed of 1,919 individuals
this research can highlight a potential intermediary process in who provided complete data on personality at baseline and
the association between personality and health outcomes. walking speed at follow-up (see Table 1). Of this sample,
Finally, the identification of the personality correlates of walk- 1,776 participants also provided walking speed data at baseline.
ing speed may prove useful for advancing knowledge on the More information about NHATS and how to access the data
psychological makeup of individuals at risk of future incident can be found at http://www.nhats.org/.
mobility limitations. The HRS is a nationally representative longitudinal study of
Using five longitudinal samples of middle-aged and older adults aged 50 years and older. An enhanced face-to-face inter-
adults, the present study examined the association between per- view was implemented starting in 2006 that included a psycho-
sonality and walking speed over time. Based on the rationale social questionnaire with personality items. Half of the sample
described above, it was hypothesized that higher neuroticism answered this questionnaire in 2006 and the other half
would prospectively predict slower walking speed, whereas answered it in 2008. Follow-up walking speed was measured
higher extraversion, openness, and conscientiousness would in 2010 and 2012, respectively, for the 2006 and the 2008 sam-
predict faster walking. In addition, the extent to which these ples. Walking speed was measured only among respondents
traits were associated with changes in walking speed was aged 65 years or older. Data from both samples at baseline and
examined in two of the five samples. follow-up were combined. The final sample was composed of
Stephan et al. 775

Table 1. Baseline Characteristics of the Samples and Correlations of Study Variables With Walking Speed at Follow-Up.

WLSG WLSS MIDUS HRS NHATS

M rWalking rWalking rWalking M rWalking M rWalking

Variables (%) SD Speed M (%) SD Speed M (%) SD Speed (%) SD Speed (%) SD Speed

Age (years) 53.19 0.62 .06** 52.53 6.84 .30** 46.23 11.76 .22** 71.19 6.80 .30** 78.26 6.97 .41**
Sex (% women) 54% — .14** 53% — .14** 55% — .11** 58% — .13** 57% — .16**
Race (% White) 100% — — 100% — — 95% — .01 89% — .17** 76% — .22**

Education 13.86 2.37 .22** 14.03 2.53 .23** 7.56 2.42 .18** 12.92 2.78 .23** 5.42 2.25 .32**
Walking speed 1.02 0.25 — 1.03 0.26 — 1.11 0.22 — 0.83 0.29 — 0.77 0.26 —
(m/s)
Neuroticism 3.17 0.98 .09** 3.22 0.95 .09** 2.20 0.66 .03 1.98 0.58 .08** 2.18 0.83 .05
Extraversion 3.86 0.89 .08** 3.78 0.91 .09** 3.21 0.55 .02 3.24 0.54 .12** 3.19 0.73 .11**
Openness 3.67 0.80 .11** 3.63 0.75 .09** 3.06 0.50 .02 2.96 0.54 .15** 2.87 0.81 .18**
Agreeableness 4.76 0.74 .02 4.71 0.72 .04 3.47 0.48 .09* 3.55 0.45 .00 3.60 0.51 .02
Conscientiousness 4.87 0.67 .06** 4.76 0.71 .04 3.46 0.43 .05 3.39 0.45 .14** 3.28 0.68 .18**

Note. See Method section for differences in the assessment and coding of walking speed, personality, and education in each sample. rWalking Speed ¼ Pearson cor-
relation between baseline variables and walking speed at follow-up. Wisconsin Longitudinal Study Graduate (WLSG): N ¼ 4,603; Wisconsin Longitudinal Study
Sibling (WLSS): N ¼ 2,090; Midlife in the United States Survey (MIDUS): N ¼ 990; Health and Retirement Study (HRS): N ¼ 5,966; National Health and Aging
Trends Survey (NHATS): N ¼ 1,919.
*p < .01. **p < .001.

5,966 participants (Table 1). In addition, 4,383 individuals also Covariates. In the five samples, age, sex, and education were
had walking speed data at baseline. More information about specified as control variables a priori. Education was reported
HRS and how to access the data can be found at http://hrsonli in years in the WLSG, the WLSS, and the HRS and measured
ne.isr.umich.edu/index.php. Attrition analysis are presented in on a scale ranging from ranging from 1 (no grade school) to
the Online Supplemental Material. 12 (doctoral level degree) in the MIDUS. In NHATS, a scale
ranging from 1 (no schooling completed) to 9 (master’s profes-
sional or doctoral degree) was used. Race was controlled in the
Personality. In the WLSG and the WLSS, a 29-item version of
MIDUS, the HRS, and the NHATS and was coded as 1 for
the Big Five Inventory (John, Donahue, & Kentle, 1991) was
White and 0 for Other.
used. Participants were asked whether they agreed or disagreed
with descriptive statements using a scale ranging from 1 (dis-
agree strongly) to 6 (agree strongly). In the MIDUS and the Data Analysis
HRS, personality traits were assessed using the Midlife Devel-
Regression analyses were conducted to examine the relation
opment Inventory (MIDI; Lachman & Weaver, 1997). In the
between personality and walking speed. In each sample, walk-
MIDUS, participants were asked how much 25 adjectives that
ing speed at follow-up was predicted by baseline personality,
assessed neuroticism, conscientiousness, extraversion, open-
controlling for demographic factors. Each trait was examined
ness, and agreeableness described themselves on a scale rang-
separately and then simultaneously in supplemental analysis.
ing from 1 (not at all) to 4 (a lot). The same scale was used in
A supplementary analysis was also conducted controlling for
the HRS, with 1 additional item. Cronbach as ranged from .58
disease burden. The results from the five samples were com-
to .80 across the four samples. A 10-item version of the MIDI
bined in a random effects meta-analysis using sample size and
was used in NHATS, using two adjectives for each of the five
p value from regression models with each trait examined sep-
traits. Participants answered on the same 4-point scale.
arately. Heterogeneity of results across the five samples was
examined using the Q test. The meta-analysis for each trait was
Walking speed. Walking speed was measured using a 2.5-m conducted with the Comprehensive Meta-Analysis software
course in the WLSG, the WLSS, and the HRS, a 15.24-m span (https://www.meta-analysis.com/). In the HRS and the
in the MIDUS, and a 3-m span in the NHATS. In each study, NHATS, we also examined the association between personality
participants were asked to walk at their normal pace. The best and change in walking speed. Residualized change analyses
of two trials was taken. Participants with at least one perfor- using regression were conducted to predict walking speed at
mance were included. Speed was calculated by dividing the follow-up from baseline personality, demographic factors, and
distance (in m) by the time recorded (in s). Participants with baseline walking speed. Significant findings would suggest that
values 3 standard deviations above and below the mean were personality is associated with changes in walking speed
removed before conducting the analysis (WLSG: N ¼ 27; because it represents the effect of personality after controlling
WLSS: N ¼ 2; MIDUS: N ¼ 9; HRS: N ¼ 6; NHATS: N ¼ 5). for earlier levels of walking speed.
776 Social Psychological and Personality Science 9(7)

Due to the number of statistical tests, a more conservative

Heterogeneity Q

Note. Coefficients are standardized regression coefficients. 95% confidence intervals are in parentheses. Wisconsin Longitudinal Study Graduate (WLSG): N ¼ 4,603; Wisconsin Longitudinal Study Sibling
approach was taken by setting p to <.01 for all analysis.

15.39*
4.65

4.69

4.63

7.70
Meta-Analysis

(WLSS): N ¼ 2,090; Midlife in the United States Survey (MIDUS): N ¼ 990; Health and Retirement Study (HRS): N ¼ 5,966; National Health and Aging Trends Survey (NHATS): N ¼ 1,919.
Results
Table 1 presents the correlations between the demographic

[0.083, 0.048]
variables and personality traits at baseline with walking

Random Effect

[0.073, 0.108]

[0.018, 0.083]

[0.005, 0.040]

[0.060, 0.106]
speed at follow-up. Table 2 presents results of regression

.06**

.09**

.08**
.05*
analyses with personality traits at baseline predicting walking

.02
speed at follow-up controlling for the demographic covari-
ates. Comparison of Tables 1 and 2 suggests that the pattern

Adjusted for age, sex, and education. bAdjusted for age, sex, education, and race. cAdjusted for age, sex, education, race, and time elapsed between assessments.
of associations between personality and walking speed were
similar with or without the demographic factors included.

[0.073, 0.004]

[0.008, 0.069]
The results were mostly consistent with our hypothesis: The

[0.030, 0.105]

[0.047, 0.124]

[0.070, 0.146]
meta-analysis and findings from most of the individual sam-

c
NHATS

.07**

.09**

.11**
ples indicated that lower neuroticism and higher extraver-

.03

.03
sion, conscientiousness, and openness were related to faster
gait speed. Additional analysis controlling for disease burden
revealed no change in the overall pattern of relations in the
five samples (see Supplemental Table S1). When all traits

[0.104, 0.057]
were entered simultaneously, conscientiousness remained a

[0.088, 0.134]

[0.053, 0.101]

[0.007, 0.055]

[0.082, 0.129]
significant predictor of walking speed across all samples, and

.08**

.11**

.08**

.11**
higher extraversion was related to faster speed in all but one
Table 2. Summary of Regression Analysis Predicting Follow-Up Walking Speed From Baseline Personality.

b
HRS

.03
of the samples (see Supplemental Table S2).
The effect size of the significant associations between
personality traits and walking speed were about one third the
effect of age and were similar or larger than the effect of sex.
For example, in the HRS, the effect of conscientiousness (bs ¼
[0.120, 0.003]

[0.066, 0.056]

[0.099, 0.024]
[0.001, 0.121]

[0.001, 0.122]
.11) was about one third of the effect of age (bs ¼ .28), and 1
b

SD difference on conscientiousness corresponded roughly to

MIDUS

.06

.06

.00

.04

.06
an age difference of 2.5 years in walking speed. To provide
an additional illustration of effect sizes, Figure 1 depicts
walking speed among individuals with scores on extraversion
and conscientiousness 1 SD above and below the mean.
Adjusted for demographic factors, individuals with higher
[0.105, 0.024]

[0.034, 0.052]

[0.023, 0.058]

extraversion walked on average 0.06 m/s faster than those

[0.050, 0.130]

[0.023, 0.103]

with lower extraversion (.09 < d < .32; weighted average

.09**
a
WLSS

.06*

.06*
.01

.02

d ¼ .26). Similar differences were observed between indivi-

duals with higher and lower conscientiousness (.11 < d < .30;
weighted average d ¼ .25).
In additional analyses, the moderating role of sex and age in
the association between personality and walking speed was
[0.080, 0.023]

tested (see Online Supplemental Material). These analysis

[0.008, 0.049]
[0.045, 0.101]

[0.018, 0.078]

[0.030, 0.086]

revealed that, despite some significant interactions, neither sex

a

.05**

.07**

.06**
WLSG

.05*

nor age were consistent moderators across the five cohorts (see
.02

Online Supplemental Material).

Finally, the longitudinal analysis in the HRS revealed that
higher extraversion, conscientiousness, and openness were
related to slower decline over the 4-year follow-up, whereas
neuroticism was associated with steeper decline in walking
Conscientiousness

*p < .01. **p < .001.

speed (Table 3). The association between openness and change

Agreeableness
Extraversion
Neuroticism

in walking speed was reduced to nonsignificant in a model with

Openness
Variables

all traits included simultaneously (see Online Supplemental

Material; Table 3). In the NHATS, none of the traits was asso-
ciated with walking speed changes from baseline to follow-up.
a
Stephan et al. 777

Discussion
A
1.2 High Extraversion (1 Based on five large longitudinal samples of middle-aged and
SD above the mean)
1.1
older adults, the present study found that personality prospec-
Low Extraversion (1 SD
tively predicts walking speed. Consistent with our hypotheses,
Walking Speed (m/s)

below the mean)

1 higher neuroticism was associated with slower gait speed,
0.9
whereas higher extraversion, openness, and conscientiousness
were related to faster speed over time, with or without control-
0.8 ling for demographic factors. The results were robust despite
0.7
differences among the five cohorts in terms of age and other
demographic characteristics, the personality and walking speed
0.6 measures, and for length of follow-up. It is of note that similar
WLSG WLSS MIDUS HRS NHATS
associations were observed over both short-term (1 year) and
Sample
long-term (up to 18 years) intervals. These traits were also
related to changes in gait speed over the 4-year follow-up in the
B High Conscienousness
HRS but not the 1-year follow-up in the NHATS. These find-
1.2 (1 SD above the mean)
Low Conscienousness ings point to the relevance of psychological traits for the phys-
1.1 (1 SD below the mean) ical functioning of older adults. In particular, it provides the
Walking Speed (m/s)

1 most comprehensive account to date of the link between per-

sonality and walking speed.
0.9 Extraversion and conscientiousness were the most consis-
0.8 tent personality correlates of walking speed. Active and enthu-
siastic individuals and those with self-discipline and
0.7
organization walked faster at follow-up and declined less in
0.6 gait speed over time in the HRS. This finding extends past
WLSG WLSS MIDUS HRS NHATS reports of a cross-sectional association between higher extra-
Samples version and faster speed (Terracciano et al., 2013; Tolea
et al., 2010) by showing that this association is not dependent
Figure 1. Walking speed for high (1 SD above the mean) and low on the two variables measured at the same time. As expected,
(1 SD below the mean) extraversion (Panel A), and conscientiousness neuroticism was associated with slower gait speed at follow-
(Panel B) in the five samples, adjusted for demographic factors within up and steeper decline over time in the HRS. Also consistent
samples. In the Wisconsin Longitudinal Study Graduate (WLSG), the
with our hypothesis, openness to experience was related to
Wisconsin Longitudinal Study Sibling (WLSS), and the Health and
Retirement Study (HRS), walking speed was measured using a 2.5-m walking faster. There was, however, heterogeneity across sam-
course. It was measured with a 15.24-m span in the Midlife in the ples, perhaps due to differences in the personality measures
United States Survey (MIDUS) and with a 3-m span in the National used or the characteristics of the samples. Past research has
Health and Aging Trends Survey (NHATS). Error bars are standard found a cross-sectional association between openness and
errors. faster walking speed (Tolea et al., 2010); this study found that
this relation persists over an extended period of time. As
Table 3. Summary of Regression Analysis Predicting Walking Speed expected, we found no consistent relation between agreeable-
Changes From Baseline Personality. ness and gait speed across the five samples.
These findings add to existing research on the association
Variables HRSa NHATSb
between neuroticism, extraversion, openness, conscientious-
Neuroticism .05** .03 ness and broader measures of mobility and physical function-
(0.075, 0.023) (0.057, 0.000) ing (Agmon & Armon, 2016; Buchman et al., 2013; Canada
Extraversion .08** .01 et al., 2016; Krueger et al., 2006). These traits are related to risk
(0.050, 0.103) (0.020, 0.036) of frailty, a geriatric syndrome characterized by loss of muscle
Openness .04* .03
mass and declines in energy and physical function (Stephan
(0.013, 0.068) (0.003, 0.060)
Agreeableness .01 .01 et al., in press). The behavioral, cognitive, and health profiles
(0.018, 0.036) (0.015, 0.041) of extraverted, open and conscientious individuals have been
Conscientiousness .06** .03 found to contribute to the preservation of mobility across adult-
(0.032, 0.085) (0.000, 0.056) hood, whereas individuals with higher neuroticism have more
risky health-related profiles. Extroverted, open, and conscien-
Note. Health and Retirement Study (HRS): N ¼ 4,383; National Health and
Aging Trends Survey (NHATS): N ¼ 1,776. tious individuals, for example, are more physically active
a
Adjusted for age, sex, education, race, and baseline walking speed. bAdjusted (Sutin et al., 2016) and are less likely to suffer from either
for age, sex, education, race, baseline walking speed, and time elapsed between chronic physical health conditions (Weston et al., 2015) or
assessments.
depressive symptoms (Hakulinen et al., 2015). In contrast,
*p < .01. **p < .001.
778 Social Psychological and Personality Science 9(7)

neuroticism is associated with a range of health-damaging regression to test for the link between personality and changes
behaviors, such as smoking (Hakulinen et al., 2015) and seden- in walking speed, other ways of modeling change over two
tary behavior (Sutin et al., 2016) and higher disease burden waves, such as latent change score models within a structural
(Weston et al., 2015) and depressive symptoms (Hakulinen equation model framework, may prove useful.
et al., 2015) that may result in walking limitations and slower In conclusion, this study provides evidence that personality
speed. In addition, higher conscientiousness and openness are traits prospectively predict walking speed and walking speed
associated with better cognitive function, whereas neuroticism decline among middle-aged and older adults. The findings
is related to worse cognition (Luchetti et al., 2016) and extra- were mostly consistent across the five cohorts and personality
version is related to faster processing speed (Pearman, 2009). was a prognostic factor for walking speed performance across
Neuroticism, extraversion, openness, and conscientiousness short and long-term follow-up, up to 18 years. Thus, person-
may also be related to gait speed through biological pathways. ality assessments may provide some information about who
Individuals with lower neuroticism, higher extraversion, open- is at risk of poor physical function across adulthood, which
ness, and conscientiousness, for example, experience less phy- may be targeted by preventive programs designed to reduce
siological dysregulation (Stephan et al., 2016) and have better mobility limitations and ultimately promote independent
cardiorespiratory fitness and energy (Terracciano et al., 2013), living in older age.
which are associated with faster walking speed (Richardson
et al., 2015; Rosso et al., 2015). Declaration of Conflicting Interests
The association between personality traits and walking The author(s) declared no potential conflicts of interest with respect to
speed among middle-aged and older adults suggest that how the research, authorship, and/or publication of this article.
fast an individual walks can be considered an expression or a
motor signature of personality. Specifically, extroverted, con- Funding
scientious, open, and emotionally stable people are character-
The author(s) received no financial support for the research, author-
ized by faster gait. Furthermore, walking speed may function
ship, and/or publication of this article.
as an intermediate, motor marker of the risk of poor outcomes
associated with some personality traits. Indeed, slow gait is pre-
Supplemental Material
dictive of a higher risk of functional limitations (Cesari et al.,
2009), incident dementia (Dumurgier et al., in press), and The supplemental material is available in the online version of the
higher mortality risk (Studenski, Perera, Patel, et al., et al., article.
2011). Personality traits in general, and high neuroticism and
low conscientiousness in particular, are related to difficulties References
in activities of daily living (Suchy et al., 2010) and risk of Agmon, M., & Armon, G. (2016). A cross-sectional study of the asso-
Alzheimer’s disease (Terracciano et al., 2014) and mortality ciation between mobility test performance and personality among
(Jokela et al., 2013). Thus, it is likely that the risk of poor out- older adults. BMC Geriatrics, 16, 105. doi:10.1186/s12877-016-
comes associated with high neuroticism and low conscientious- 0272-8
ness may manifest at the motor level, through slower gait. Brunner, E., Shipley, M., Spencer, V., Kivimaki, M., Chandola, T.,
The present study has several strengths including the Gimeno, D., . . . Marmot, M. (2009). Social inequality in walking
examination of five large samples, prospective analyses, speed in early old age in the Whitehall II study. The Journals of
long-term follow-up, and the investigation of all five major Gerontology Series A: Biological Sciences and Medical Sciences,
dimensions of personality. Each sample was large enough to 64, 1082–1089. doi:10.1093/gerona/glp078
provide sufficient power to detect small effect sizes. How- Buchman, A. S., Boyle, P. A., Wilson, R. S., Leurgans, S. E., Arnold,
ever, there are also several limitations. The observational S. E., & Bennett, D. A. (2013). Neuroticism, extraversion, and
design of this study limits the possibility of determining cau- motor function in community-dwelling older persons. The Ameri-
sal relations. Although personality was considered a predictor can Journal of Geriatric Psychiatry, 21, 145–154. doi:10.1016/j.
of walking speed, future research may consider the potential jagp.2012.10.015
reciprocal relation between personality and gait speed. Canada, B., Stephan, Y., Jaconelli, A., & Duberstein, P. R. (2016). The
Despite the finding of a consistent association between per- moderating effect of chronological age on the relation between
sonality and walking speed over different periods, from 1 to neuroticism and physical functioning: Cross-sectional evidence
almost 20 years, the size of this association was relatively from two French samples. The Journals of Gerontology Series B:
small. However, walking is a complex behavior that reflects Psychological Sciences and Social Sciences, 71, 35–40. doi:10.
a wide range of biological, clinical, behavioral, cognitive, and 1093/geronb/gbu083
environmental factors, and each factor has a small effect. Cesari, M., Kritchevsky, S. B., Newman, A. B., Simonsick, E. M.,
Therefore, personality is unlikely to have a strong effect. In Harris, T. B., Penninx, B. W., . . . Pahor, M. (2009). Added value
addition, the personality measures were different across sam- of physical performance measures in predicting adverse health-
ples and all of them were brief. Detailed personality question- related events: Results from the health, aging, and body composi-
naires are needed to examine the personality facets associated tion study. Journal of the American Geriatrics Society, 57,
with walking speed. Finally, although the present study used 251–259. doi:10.1111/j.1532-5415.2008.02126.x
Stephan et al. 779

Demakakos, P., Cooper, R., Hamer, M., de Oliveira, C., Hardy, R., Luchetti, M., Terracciano, A., Stephan, Y., & Sutin, A. R. (2016).
& Breeze, E. (2013). The bidirectional association between Personality and cognitive decline in older adults: Data from a
depressive symptoms and gait speed: Evidence from the longitudinal sample and meta-analysis. The Journals of Gerontol-
English Longitudinal Study of Ageing (ELSA). PLoS One, 8, ogy Series B: Psychological Sciences and Social Sciences, 71,
e68632. 591–601. doi:10.1093/geronb/gbu184
Digman, J. M. (1990). Personality structure: Emergence of the Pearman, A. (2009). Basic cognition in adulthood: Combined effects
five-factor model. Annual Review of Psychology, 41, 417–440. of sex and personality. Personality and Individual Differences,
doi:10.1146/annurev.ps.41.020190.002221 47, 357–362. doi:10.1016/j.paid.2009.04.003
Duberstein, P. R., Sörensen, S., Lyness, J. M., King, D. A., Conwell, Perera, S., Patel, K. V., Rosano, C., Rubin, S. M., Satterfield, S.,
Y., Seidlitz, L., & Caine, E. D. (2003). Personality is associated Harris, T., . . . Newman, A. B. (2016). Gait speed predicts incident
with perceived health and functional status in older primary care disability: A pooled analysis. The Journals of Gerontology Series
patients. Psychology and Aging, 18, 25–37. doi:10.1037/0882- A: Biological Sciences and Medical Sciences, 71, 63–71. doi:10.
7974.18.1.25 1093/gerona/glv126
Dumurgier, J., Artaud, F., Touraine, C., Rouaud, O., Tavernier, B., Richardson, C. A., Glynn, N. W., Ferrucci, L. G., & Mackey, D. C.
Dufouil, C., . . . Elbaz, A. (2017). Gait speed and decline in gait (2015). Walking energetics, fatigability, and fatigue in older
speed as predictors of incident dementia. The Journals of Gerontol- adults: The study of energy and aging pilot. The Journals of
ogy Series A: Biological Sciences and Medical Sciences, 72, Gerontology Series A: Biological Sciences and Medical Sciences,
655–661. doi:10.1093/gerona/glw110 70, 487–494. doi:10.1093/gerona/glu146
Friedman, H. S., & Kern, M. L. (2014). Personality, well-being, and Rosso, A. L., Sanders, J. L., Arnold, A. M., Boudreau, R. M., Hirsch,
health. Annual Review of Psychology, 65, 719–742. doi:10.1146/ C. H., Carlson, M. C., . . . Newman, A. B. (2015). Multisystem
annurev-psych-010213-115123 physiologic impairments and changes in gait speed of older adults.
Gale, C. R., Allerhand, M., Sayer, A. A., Cooper, C., & Deary, I. The Journals of Gerontology Series A: Biological Sciences and
J. (2014). The dynamic relationship between cognitive func- Medical Sciences, 70, 317–322. doi:10.1093/gerona/glu176
tion and walking speed: The English Longitudinal Study of Stephan, Y., Sutin, A. R., Canada, B., & Terracciano, A. (2017). Per-
Ageing. Age, 36, 9682. doi:http://doi.org/10.1007/s11357- sonality and frailty: Evidence from four samples. Journal of
014-9682-8 Research in Personality, 66, 46–53. doi:10.1016/j.jrp.2016.12.006
Hakulinen, C., Elovainio, M., Pulkki-Raback, L., Virtanen, M., Stephan, Y., Sutin, A. R., Luchetti, M., & Terracciano, A. (2016).
Kivimaki, M., & Jokela, M. (2015). Personality and depressive Allostatic load and personality: A 4-year longitudinal study. Psy-
symptoms: Individual participant meta-analysis of 10 cohort chosomatic Medicine, 78, 302–310. doi:10.1097/PSY.0000000
studies. Depression and Anxiety, 32, 461–470. doi:10.1002/da. 000000281
22376 Studenski, S., Perera, S., Patel, K., Rosano, C., Faulkner, K., Inzitari,
Hakulinen, C., Hintsanen, M., Munafo, M. R., Virtanen, M., Kivimäki, M., . . . Nevitt, M. (2011). Gait speed and survival in older adults.
M., Batty, G. D., Ip, E. H., Church, T., Marshall, S. A., Zhang, Q., Journal of the American Medical Association, 305, 50–58.
Marsh, A. P., & Guralnik, J., . . . LIFE-P Study Investigators. Studenski, S., Perera, S., Wallace, D., Chandler, J. M., Duncan, P. W.,
(2013). Physical activity increases gains in and prevents loss of Rooney, E., . . . Guralnik, J. M. (2003). Physical performance mea-
physical function: Results from the lifestyle interventions and sures in the clinical setting. Journal of the American Geriatrics
independence for elders pilot study. The Journals of Gerontology Society, 51, 314–322.
Series A: Biological Sciences and Medical Sciences, 68, Suchy, Y., Williams, P. G., Kraybill, M. L., Franchow, E., & Butner, J.
426–432. doi:10.1093/gerona/gls186 (2010). Instrumental activities of daily living among community-
John, O. P., Donahue, E. M., & Kentle, R. L. (1991). The Big Five dwelling older adults: Personality associations with self-report,
Inventory—Versions 4a and 54. Berkeley, CA: Institute of performance, and awareness of functional difficulties. The
Personality and Social Research, University of California. Journals of Gerontology Series B: Psychological Sciences and
Jokela, M. (2015). Personality and smoking: Individual-participant Social Sciences, 65, 542–550. doi:10.1093/geronb/gbq037
meta-analysis of nine cohort studies. Addiction, 110, 1844–1852. Sutin, A. R., Stephan, Y., Luchetti, M., Artese, A., Oshio, A., &
doi:10.1111/add.13079 Terracciano, A. (2016). The five-factor model of personality and
Jokela, M., Batty, G. D., Nyberg, S. T., Virtanen, M., Nabi, H., Singh- physical inactivity: A meta-analysis of 16 samples. Journal of
Manoux, A., & Kivimäki, M. (2013). Personality and all-cause Research in Personality, 63, 22–28. doi:10.1016/j.jrp.2016.05.001
mortality: Individual-participant meta-analysis of 3,947 deaths in Sutin, A. R., Stephan, Y., & Terracciano, A. (in press). Personality and
76,150 adults. American Journal of Epidemiology, 178, 667–675. metabolic dysfunction in young adulthood: A cross-sectional
doi:10.1093/aje/kwt170 study. Journal of Health Psychology. doi:10.1177/13591053
Krueger, K. R., Wilson, R. S., Shah, R. C., Tang, Y., & Bennett, D. A. 16677294
(2006). Personality and incident disability in older persons. Age Terracciano, A., Schrack, J. A., Sutin, A. R., Chan, W., Simonsick, E.
and Ageing, 35, 428–433. doi:10.1093/ageing/afl028 M., & Ferrucci, L. (2013). Personality, metabolic rate and aerobic
Lachman, M. E., & Weaver, S. L. (1997). The Midlife Development capacity. PLoS One, 8, e54746. doi:10.1371/journal.pone.0054746
Inventory (MIDI) Personality Scales: Scale construction and scor- Terracciano, A., Stephan, Y., Luchetti, M., Gonzalez-Rothi, R., &
ing. Technical report. Sutin, A. R. (in press). Personality and lung function in older
780 Social Psychological and Personality Science 9(7)

adults. The Journals of Gerontology Series B: Psychological Verghese, J., Wang, C., Lipton, R. B., Holtzer, R., & Xue, X. (2007).
Sciences and Social Sciences. doi:10.1093/geronb/gbv161 Quantitative gait dysfunction and risk of cognitive decline and
Terracciano, A., Sutin, A. R., An, Y., O’Brien, R. J., Ferrucci, L., Zon- dementia. Journal of Neurology, Neurosurgery & Psychiatry, 78,
derman, A. B., & Resnick, S. M. (2014). Personality and risk of 929–935. doi:10.1136/jnnp.2006.106914
Alzheimer’s disease: New data and meta-analysis. Alzheimer’s & Weston, S. J., Hill, P. L., & Jackson, J. J. (2015). Personality traits pre-
Dementia: The Journal of the Alzheimer’s Association, 10, dict the onset of disease. Social Psychological and Personality Sci-
179–186. doi:10.1016/j.jalz.2013.03.002 ence, 6, 309–317. doi:10.1177/1948550614553248
Tolea, M. I., Costa, P. T., Terracciano, A., Ferrucci, L., Faulkner, K.,
Coday, M. M. C., . . . Simonsick, E. M. (2012). Associations Author Biographies
of openness and conscientiousness with walking speed decline:
Findings from the Health, Aging, and Body Composition Yannick Stephan is an assistant professor of health and aging psy-
Study. The Journals of Gerontology. Series B, Psychological chology at Montpellier University in Montpellier, France. His
Sciences and Social Sciences, 67, 705–711. doi:10.1093/ger research focuses on the consequences of personality traits across the
onb/gbs030 lifespan and on personality development.
Tolea, M. I., Costa, P. T., Terracciano, A., Griswold, M., Simonsick, Angelina R. Sutin is an assistant professor in the department of beha-
E. M., Najjar, S. S., . . . Ferrucci, L. (2010). Sex-specific corre- vioral sciences and social medicine at the College of Medicine of Flor-
lates of walking speed in a wide age-ranged population. The ida State University in Tallahassee, USA. Her research addresses how
Journals of Gerontology Series B: Psychological Sciences and personality traits are associated with physical and mental health across
Social Sciences, 65B, 174–184. doi:http://doi.org/10.1093/ger adulthood.
onb/gbp130
Tolea, M. I., Terracciano, A., Milaneschi, Y., Metter, E. J., & Ferrucci, Gabriel Bovier-Lapierre is a currently a graduate student at Mon-
L. (2012). Personality typology in relation to muscle strength. tpellier University, in Montpellier, France. His research focuses on the
International Journal of Behavioral Medicine, 19, 382–390. doi: consequences of personality for health across the lifespan.
10.1007/s12529-011-9166-5
Antonio Terracciano is an assistant professor in the department of
Verghese, J., Holtzer, R., Oh-Park, M., Derby, C. A., Lipton, R. B., &
geriatrics at the College of Medicine of Florida State University in
Wang, C. (2011). Inflammatory markers and gait speed decline in
Tallahassee, USA. His research focuses on the relationships between
older adults. The Journals of Gerontology Series A: Biological
personality traits and physical and mental health across the lifespan.
Sciences and Medical Sciences, 66A, 1083–1089. doi:10.1093/ger-
ona/glr099 Handling Editor: Simine Vazire