UC Riverside
UC Riverside Previously Published Works
Title
Personality, well-being, and health.
Permalink
https://escholarship.org/uc/item/8j36s5f2
Journal
Annual review of psychology, 65(1)
ISSN
0066-4308
Authors
Friedman, Howard S
Kern, Margaret L
Publication Date
2014
DOI
10.1146/annurev-psych-010213-115123
Peer reviewed
eScholarship.org Powered by the California Digital Library
University of California
PS65CH26-Friedman ARI 31 October 2013 17:18
Personality, Well-Being,
and Health
∗
Howard S. Friedman
1
and Margaret L. Kern
2
1
Department of Psychology, University of California, Riverside, California 92521;
email: Howard.Friedman@ucr.edu
2
Department of Psychology, University of Pennsylvania, Philadelphia, Pennsylvania 19104;
email: mkern@sas.upenn.edu
Annu. Rev. Psychol. 2014. 65:719–42
The Annual Review of Psychology is online at
http://psych.annualreviews.org
This article’s doi:
10.1146/annurev-psych-010213-115123
Copyright
c
2014 by Annual Reviews.
All rights reserved
∗
The authors contributed equally to this article.
Keywords
lifespan perspective, trajectories, conscientiousness
Abstract
A lifespan perspective on personality and health uncovers new causal path-
ways and provides a deeper, more nuanced approach to interventions. It is
unproven that happiness is a direct cause of good health or that negative emo-
tion, worry, and depression are significant direct causes of disease. Instead,
depression-related characteristics are likely often reflective of an already-
deteriorating trajectory. It is also unproven that challenging work in a de-
manding environment usually brings long-term health risks; on the contrary,
individual strivings for accomplishment and persistent dedication to one’s
career or community often are associated with sizeable health benefits. Over-
all, a substantial body of recent research reveals that conscientiousness plays
a very significant role in health, with implications across the lifespan. Much
more caution is warranted before policy makers offer narrow health recom-
mendations based on short-term or correlational findings. Attention should
be shifted to individual trajectories and pathways to health and well-being.
719
Annu. Rev. Psychol. 2014.65:719-742. Downloaded from www.annualreviews.org
by University of Pennsylvania on 01/06/14. For personal use only.
Click here for quick links to
Annual Reviews content online,
including:
• Other articles in this volume
• Top cited articles
• Top downloaded articles
• Our comprehensive search
Fur ther
ANNUAL
REVIEWS
PS65CH26-Friedman ARI 31 October 2013 17:18
Contents
INTRODUCTION............................................................... 720
OUTCOMES..................................................................... 721
Longevity...................................................................... 721
QualityofLife................................................................. 722
Multiple Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722
LimitsofBiomarkersasOutcomes.............................................. 723
HAPPINESS, SUBJECTIVE WELL-BEING, AND HEALTH . . . . . . . . . . . . . . . . . . . . 723
PowerofPositiveEmotion?..................................................... 724
MeaningandPurpose........................................................... 726
Optimism...................................................................... 726
NEUROTICISM, DEPRESSION, WORRY, AND DISEASE. . . . . . . . . . . . . . . . . . . . . . 727
Depression..................................................................... 727
ChallengeandHealth........................................................... 729
CONSCIENTIOUSNESS, MATURITY, AND LONGEVITY . . . . . . . . . . . . . . . . . . . . 731
EarlyLifeInfluences............................................................ 733
CONCLUSION: IMPLICATIONS FOR INTERVENTIONS . . . . . . . . . . . . . . . . . . . . 733
INTRODUCTION
Although the relationships among personality, well-being, and health have been studied for
millennia—since the days of the bodily humors proposed by Hippocrates and Galen—the
field remains riddled with conceptual confusion, method artifacts, and misleading conclusions.
When inferences drawn from this field are based on incomplete models, they lead to waste-
ful and even harmful interventions and treatments. Scientists and laypersons alike may over-
generalize from short-term personality correlates of health and overlook long-term causal
processes.
There is nevertheless excellent evidence that individual characteristics from earlier in life are
reliable predictors and likely causal elements of health later in life. An especially striking finding
to emerge in recent years is that a host of characteristics and behaviors associated with the broad
personality dimension of conscientiousness is predictive of health and longevity, from childhood
through old age. The reasons for these associations are complex and sometimes appear paradoxical,
as there are multiple simultaneous causal links to health. The modern study of personality, how-
ever, provides many of the concepts, tools, and models necessary for a deeper and more accurate
understanding of health, well-being, and long life.
In particular, there is considerable misapprehension concerning the pathways to good
health. In this article, we review many of the causes and consequences of the associations
among personality, behavior, well-being, and health and longevity. We do this in the context
of expanded models and perspectives. Because much of the confusion in the area of personality
and health arises from ambiguous definitions, weak measurement, and overlapping constructs
of health, we begin with health outcomes. We then review and scrutinize the connections
among happiness and health, and among depression, worry, and disease, which likely are
not what they first appear to be. Finally, we explain and evaluate the emerging consensus
on the significance of conscientiousness across the lifespan and offer suggestions for health
interventions.
720 Friedman
·
Kern
Annu. Rev. Psychol. 2014.65:719-742. Downloaded from www.annualreviews.org
by University of Pennsylvania on 01/06/14. For personal use only.
PS65CH26-Friedman ARI 31 October 2013 17:18
OUTCOMES
Study of personality—an individual’s relatively stable predispositions and patterns of thinking,
feeling, and acting—and its relationships to well-being and health continues to be plagued by an
overreliance on self-report measures. This is a special problem because many of the questions (or
items) used to assess personality are the same questions used to assess health and well-being. Much
better assessment strategies are needed.
Outcome measures of well-being may ask individuals how good they feel, how well they cope,
and how satisfied they are with life. These are very similar to personality measures of low neu-
roticism (“am relaxed most of the time”; “am calm”; “am not angry or depressed”) and high
agreeableness (“am on good terms with others”; “am warm and sympathetic”). Thus it is not sur-
prising that people who report having a joyful, cheerful, relaxed, and agreeable personality also
report life satisfaction, emotional thriving, and well-being. Such correlations have little to say
about achieving well-being. Relatedly, studies of patient populations often suffer from personality
selection artifacts (biases) because neurotic individuals are more likely to report symptoms (such as
chest pain) and to seek medical care than nonneurotics, even when there is little or no discernable
organic disease. Although such serious measurement artifacts have been recognized for decades
(Watson & Pennebaker 1989), erroneous causal deductions are still common.
Analogous issues plague self-report measures of physical health. The commonly used multi-
purpose Short-Form (36) Health Survey (SF-36), or the closely related RAND 36-Item Health
Survey (RAND-36), can be very useful for assessing overall disease burden. However, the
SF-36 contains multiple dimensions, including behavioral dysfunction, objective reports, sub-
jective rating, and distress and well-being (Ware 2004). So employing the full SF-36—without
sufficient attention to its components—as an outcome measure of health in studies of personal-
ity and health again confounds the predictor with the outcome because individuals who report
a neurotic, distressed personality also report pain, feeling sick, and a poor sense of well-being.
Sometimes this flaw is obfuscated by invoking the significant well-established finding that self-
rated health predicts mortality risk (Idler & Benyamini 1997). That is, the argument asserts that
self-reported personality predicts self-reported health, and self-reported health predicts mortal-
ity, and so therefore a study of self-reported personality and self-reported health is really a study
of personality and physical health. A valuable scientific approach, however, necessitates multi-
method assessments of personality and behavior coupled with more objective measures of health
outcomes.
Longevity
Longevity is, for most purposes, the single best measure of health. First, it is highly reliable and
valid. Although there is some unreliability of public records such as birth certificates and death
certificates, it is generally the case that if a death certificate shows that a man died on April 15,
2013 at age 80 from septicemia, then it is very likely that he lived eight decades. It is also very
likely that he is currently in terrible “health,” and so health validity is strong. Life expectancy is
thus one of the key measures of public health used worldwide.
Second, using longevity as the outcome helps avoid what we call the “all-cause dilemma”
artifact. These are cases in which a person has a disease such as cancer, and, for example, the prostate
or breast is removed, and then soon after the individual dies not of cancer but of something else.
If the focus of the study is on cancer survival (as a function of personality, coping, and treatment),
the death may not be picked up; that is, the cancer did not progress and/or the person did not die
of cancer. The patient is considered to be “cured” of cancer even if the patient dies of a different
cause. In other words, much research on personality and health is limited and even distorted by the
www.annualreviews.org
•
Personality, Well-Being, and Health 721
Annu. Rev. Psychol. 2014.65:719-742. Downloaded from www.annualreviews.org
by University of Pennsylvania on 01/06/14. For personal use only.
PS65CH26-Friedman ARI 31 October 2013 17:18
still-common focus on single-disease conditions, with insufficient attention to overall outcomes,
especially overall mortality risk.
Relatedly, it is misleading to speak of personality traits or coping styles that predict cancer risk
or heart disease risk (e.g., type A personality) if such factors equally predict (are equally relevant
to) other diseases. And, in fact, the basic five-factor personality dimensions (particularly consci-
entiousness, neuroticism, and extraversion, but also often agreeableness and openness) do predict
multiple diseases (Friedman 2007, Goodwin & Friedman 2006). This issue was noted many years
ago (Friedman & Booth-Kewley 1987), but studies of personality predictors of particular diseases,
without sufficient regard for the broader context, are still common. Rigorous research programs
on personality, well-being, and health would do better to employ multidimensional assessments
of both personality and health and, whenever possible, to include follow-ups to measure all-cause
mortality or multiple hard outcomes of disease.
Quality of Life
General health is well captured by longevity because the people who live the longest usually are
not those who have been struggling with diabetes, cancer, heart disease, and other chronic dis-
orders. But measures that also directly consider the quality of life—such as the number of years
that one lives without significant impairment—are of increasing interest. The World Health
Organization uses healthy life expectancy (HALE), defined as years lived without significant im-
pairment from disease or injury. The European Union has developed an indicator of disability-
adjusted life expectancy (“Healthy Life Years”). Health psychologists such as Robert Kaplan (2002)
have advocated for health-related quality-of-life measures that take into account years of life and
the amount of disability while minimizing the value of any “benefits” that come from curing one
disease only to have it be replaced by another. Such robust measures include rigorous definitions
of disability—such as inability to work, walk, dress, converse, and remember—rather than simply
self-report measures of how one feels.
Multiple Outcomes
Consistent with the World Health Organization’s definition of health as composed of physical,
mental, and social components, we have found (in our own research) that it is empirically and
heuristically useful to distinguish and use at least five core health outcomes in addition to longevity
(Friedman et al. 2010, Friedman & Martin 2011; see also Aldwin et al. 2006, Baltes & Baltes 1990,
Rowe & Kahn 1987). In brief, they encompass the following:
(a) Physical health (the ability and energy to complete a range of daily tasks; either diagnosed
or not diagnosed with organic disease such as heart disease or cancer). Physical health is defined
by an evaluation or evidence-based judgment by a health professional, such as an exam that might
be used to qualify for medical treatment or disability payments. (b) Subjective well-being (positive
mood; life satisfaction). Subjective well-being is often seen as having both an emotional component
(frequency of positive and negative emotions) and a cognitive component of self-perceived life
satisfaction (Diener et al. 2013). (c) Social competence (successful engagement in activities with
others). Social competence includes the ability to maintain close relationships, to have a supportive
social and/or community network, and to support others. (d ) Productivity (continued achievement;
contributing to society). Productivity involves work that has potentially monetary/economic (paid)
value or contributions of recognizable artistic, intellectual, or humanitarian value. With an aging
population in many countries, productivity is taking on new meanings and importance (Fried
2012). (e) Cognitive function (the ability to think clearly and remember) is defined in terms of
722 Friedman
·
Kern
Annu. Rev. Psychol. 2014.65:719-742. Downloaded from www.annualreviews.org
by University of Pennsylvania on 01/06/14. For personal use only.