200 words

Researchers have struggled to identify how certain states of mind influence
physical health. One biologist thinks he has an answer.

The pursuit of happiness

W hen Steve Cole was a postdoc, he had an unusual hobby: matching art buyers with artists that they might
like. The task made looking at art, something
he had always loved, even more enjoyable.
“There was an extra layer of purpose. I loved
the ability to help artists I thought were great to
find an appreciative audience,” he says.

At the time, it was nothing more than a
quirky sideline. But his latest findings have
caused Cole — now a professor at the Cousins
Center for Psychoneuroimmunology at the
University of California, Los Angeles — to
wonder whether the exhilaration and sense
of purpose that he felt during that period
might have done more than help him to find
homes for unloved pieces of art. It might have
benefited his immune system too.

At one time, most self-respecting molecu-
lar biologists would have scoffed at the idea.
Today, evidence from many studies suggests
that mental states such as stress can influence
health. Still, it has proved difficult to explain
how this happens at the molecular level — how
subjective moods connect with the vastly com-
plex physiology of the nervous and immune
systems. The field that searches for these expla-
nations, known as psychoneuro immunology
(PNI), is often criticized as lacking rigour.
Cole’s stated aim is to fix that, and his tool of
choice is genome-wide transcriptional analy-
sis: looking at broad patterns of gene expres-
sion in cells. “My job is to be a hard-core
tracker,” he says. “How do these mental states
get out into the rest of the body?”

With his colleagues, Cole has published a
string of studies suggesting that negative men-
tal states such as stress and loneliness guide
immune responses by driving broad programs
of gene expression, shaping our ability to fight
disease. If he is right, the way people see the
world could affect everything from their risk
of chronic illnesses such as diabetes and heart
disease to the progression of conditions such
as HIV and cancer. Now Cole has switched
tack, moving from negative moods into the
even more murky territory of happiness. It
is a risky strategy; his work has already been
criticized as wishful thinking and moralizing.
But the pay-off is nothing less than finding a
healthier way to live.

“If you talk to any high-quality neuro-
biologist or immunologist about PNI, it

will invariably generate a little snicker,” says
Stephen Smale, an immunologist at the Uni-
versity of California, Los Angeles, who is not
affiliated with the Cousins Center. “But this
doesn’t mean the topic should be ignored for-
ever. Someday we need to confront it and try
to understand how the immune system and
nervous system interact.”

THE BEST MEDICINE?
In 1964, magazine editor Norman Cousins was
diagnosed with ankylosing spondylitis, a life-
threatening autoimmune disease, and given a
1 in 500 chance of recovery. Cousins rejected
his doctors’ prognosis and embarked on his
own programme of happiness therapy, includ-
ing regular doses of Marx Brothers films, and
credited it with triggering a dramatic recovery.
He later established the Cousins Center, which
is dedicated to investigating whether psycho-
logical factors really can keep people healthy.

At the time, mainstream science rejected
the idea that any psychological state, positive
or negative, could affect physical well-being.
But studies during the 1980s and early 1990s
revealed that the brain is directly wired to the
immune system — portions of the nervous
system connect with immune-related organs
such as the thymus and bone marrow, and
immune cells have receptors for neurotrans-
mitters, suggesting that there is crosstalk.

These connections seem to have clinical
relevance, at least in the case of stress. One of
the first researchers to show this was virolo-
gist Ronald Glaser, now director of the Insti-
tute for Behavioral Medicine Research at the
Ohio State University in Columbus. “When I
started working on this in the 1980s, nobody
believed what stress could do, including me,”
he recalls. He and his colleagues sampled
blood from medical students, and found that
during a stressful exam period, they had lower
activity from virus-fighting immune cells1, and

higher levels of antibodies for the common
virus Epstein–Barr2, suggesting that stress
had compromised their immune systems and
allowed the normally latent virus to become
reactivated.

The field of PNI has grown hugely since
then, with medical schools worldwide boasting
their own departments of mind–body medi-
cine, of which PNI is just one component. It is
now accepted that the body’s response to stress
can suppress parts of the immune system and,
over the long term, lead to damaging levels of
inflammation. Large epidemiological stud-
ies — including the Whitehall studies, which
have been following thousands of British civil
servants since 1967 — suggest3 that chronic
work stress increases the risk of coronary heart
disease and type 2 diabetes, for example. Low
socio-economic status increases susceptibility
to a wide range of infectious diseases, and there
is considerable evidence that stress increases
the rate of progression of HIV/AIDS. But
researchers have a long way to go before they
will understand exactly how signals from the
brain feed into physical health.

WORRIED SICK
PNI studies have mostly tended to look at levels
of individual immune-cell types or molecular
messengers — such as the stress hormone
cortisol and the immune messenger proteins
called cytokines — or the expression of indi-
vidual genes. But Cole wanted to get a sense of
how the whole system was working.

His first foray, published in 2007, looked at
loneliness4. Social isolation is one of the most
powerful known psychological risk factors for
poor health, but it is never certain whether it
causes the health problems, or whether a third
factor is involved: lonely people might be less
likely than others to eat well, for example, or to
visit their doctor regularly.

Cole and his colleagues looked at gene
expression in the white blood cells of six
chronically lonely people — people who had
said consistently over several years that they
felt lonely or isolated, and were fearful of other
people — and eight people who said that they
had great friends and social support. Out of the
roughly 22,000 genes in the human genome,
the researchers identified 209 that distin-
guished the lonely people from the sociable
ones: they were either regulated up to produce

“Mood matters. If we
change the psychology,
physiological changes

do parallel that.”

B Y J O M A R C H A N T

4 5 8 | N A T U R E | V O L 5 0 3 | 2 8 N O V E M B E R 2 0 1 3

FEATURENEWS

© 2013 Macmillan Publishers Limited. All rights reserved

more of an individual protein or regulated
down to produce less. Any individual gene
could easily look different by chance, but Cole
was struck by the overall pattern. A particu-
larly large proportion of the upregulated genes
in the lonely group turned out to be involved
in the inflammatory response, whereas many
of the downregulated genes had antiviral roles.
In sociable people, the reverse was true. It was
a small study, but one of the first to link a psy-
chological risk factor with a broad underlying
change in gene expression.

The researchers have since replicated that
result in a group of 93 people5. Cole says that
he has also seen a similar shift in gene expres-
sion in individuals exposed to various types of
social adversity, from imminent bereavement
to low socio-economic status.

The results make evolutionar y sense,
he says. Early humans in close-knit social
groups would have faced increased risk of
viral infections, so they would have benefited
from revved-up antiviral genes. By contrast,
people who were isolated and under stress
faced greater risk of injuries that could cause
bacterial infection — and thus would need to
respond by ramping up genes associated with
inflammation, to help heal wounds and fight
off those infections. But modern stresses lead
to chronic and unhelpful inflammation, which
over time damages the body’s tissues, increas-
ing the risk of chronic diseases such as athero-
sclerosis, cancer and diabetes.

To a classical immunologist such as Smale,

Cole’s results are “intriguing, wonderful obser-
vations”, but not yet completely convincing.
In future work, he wants to see the rest of the
physiological pathway nailed down. “Until
you put together a full understanding of that
mechanism, you have this level of uncertainty
and scepticism,” he says. That sentiment is ech-
oed by Alexander Tarakhovsky, an immunolo-
gist at the Rockefeller University in New York
City. Pinning down precise mechanisms — for
example, which neurotransmitters cause which
specific effects — is extremely difficult, he says,
because the brain and the immune system are
both so complex. Cole’s research “makes you
think about what the consequences of social
hardship could be, but it doesn’t really tell you
how it works”.

Greg Gibson, director of the Center for
Integrative Genomics at the Georgia Institute
of Technology in Atlanta, wants to see larger
studies but argues that the big-picture “genetic
architecture” that Cole is uncovering is worth
studying, even if not every detail of the mecha-
nism is yet understood. “A lot of people are tak-
ing a whole-genome approach, but they focus
only on a handful of ‘top hits’. They are missing
the wood for the trees.”

DON’T WORRY, BE HAPPY
In 2010, Cole received an e-mail from Bar-
bara Fredrickson, a friend from graduate
school who was now studying emotional well-
being at the University of North Carolina in
Chapel Hill. “Remember me?” she said. She

was interested in the biological correlates of
happiness and other positive emotional states,
and suggested that the pair collaborate. After
years of looking at stress and adversity, Cole
loved the idea. “I was bored as hell with mis-
ery,” he says.

If PNI as a whole has credibility issues,
studying well-being is even trickier. It is more
slippery to measure than stress — there is no
biological marker such as cortisol to fall back
on and no simple way to induce it in the lab,
and mainstream biologists tend to look down
on fuzzy methods of data collection such as
questionnaires.

One approach is to test whether it is pos-
sible to reverse the adverse effects on gene
expression caused by stress. Cole has collabo-
rated in three small, randomized, controlled
trials that attempt to do this. Studies involving
45 stressed caregivers6 and 40 lonely adults7
respectively found that courses in medita-
tion shifted gene-expression profiles in the
participants’ white blood cells away from
inflammatory genes and towards antiviral
genes. A third trial8, led by psycho-oncologist
Michael Antoni at the University of Miami,
Florida, involved 200 women with early-stage
breast cancer. In those who completed a ten-
week stress-management programme, genes
associated with inflammation and metastasis
were downregulated compared with those of
women in the control group, who attended
a one-day educational seminar. Meanwhile,
genes involved in the type I interferon response

A
N

D
R

E
W

B
U

R
TO

N
/G

E
T
T
Y

A volunteer helps to bag meals for the homeless at Cathedral Kitchen in Camden, New Jersey.

2 8 N O V E M B E R 2 0 1 3 | V O L 5 0 3 | N A T U R E | 4 5 9
© 2013 Macmillan Publishers Limited. All rights reserved

(which fights tumours as well as viruses) were
upregulated in the women who took the stress-
management course. “Our conclusion was that
mood matters,” says Antoni. “If we change the
psychology, physiological changes do parallel
that.”

Cole and Fredrickson aspired to go further.
Instead of looking at the benefits of blocking
stress, they wanted to investigate what happens
in the body when people are happy. To that
end, they asked 80 participants 14 questions,
such as how often in the past week they had felt
happy or satisfied, and how often they felt that
their life had a sense of meaning9. The ques-
tions were designed to distinguish between
the two forms of happiness recognized by
psychologists: hedonic well-being (charac-
terized by material or bodily pleasures such
as eating well or having sex) and eudaimonic
well-being (deeper satisfaction from activi-
ties with a greater meaning or purpose, such
as intellectual pursuits, social relationships or
charity work).

The researchers were surprised to find that
the two types of happiness influenced gene
expression in different ways. People with a
meaning-based or purpose-based outlook had
favourable gene-expression profiles, whereas
hedonic well-being, when it occurred on its
own, was associated with profiles similar to
those seen in individuals facing adversity.

One interpretation is that eudaimonic well-
being benefits immune function directly. But
Cole prefers to explain it in terms of response
to stress. If someone is driven purely by hol-
low consumption, he argues, all of their happi-
ness depends on their personal circumstances.
If they run into adversity, they may become
very stressed. But if they care about things
beyond themselves — community, politics,
art — then everyday stresses will perhaps be

of less concern. Eudaimonia, in other words,
may help to buffer our sense of threat or uncer-
tainty, potentially improving our health. “It’s
fine to invest in yourself,” says Cole, “as long as
you invest in lots of other things as well.”

PERILS OF POSITIVE THINKING
This is just the kind of advice that attracts some
of the most vociferous criticisms of Cole’s work.
James Coyne, a health psychologist and emeri-
tus professor at the University of Pennsylvania
in Philadelphia, says that Cole and Frederick-
son’s well-being study is simply too small to
show anything useful. He also argues that the
measures of eudaimonic and hedonic happi-
ness are so highly correlated in the study as to
be essentially the same thing. Coyne says that
early results are being vastly over-sold. “They
claim that if you make the right choices, you’ll
be healthy. And if you don’t, you’ll die.”

Coyne wants researchers across the field
of PNI to stop publicizing claims about
health benefits until the science is more solid.
“They’re turning it into books and workshops,
telling people how to live their lives.”

Fredrickson, for example, is the author
of two popular books, including Positivity
(Crown Archetype, 2009), which posits that
a specific ratio of positive to negative emo-
tions (2.9013, to be precise) is linked to good
health. The book has been praised by eminent
psychologists such as Daniel Goleman and
Martin Seligman, but the set of equations
behind the ratio was criticized this year10 by
Alan Sokal, a physicist at New York Univer-
sity (who famously published a deliberately
nonsensical paper in the journal Social Text
in 1996, intended to expose the lack of rigour
in the field of cultural studies). He pointed
out that the equations are based on param-
eters from a 1962 paper on air flow, with no

connection to psychological data at all. Fre-
drickson acknowledges problems with the
maths, which she based on a peer-reviewed
paper on the complex dynamics of teams11, but
says that she stands by the fundamental princi-
ples described in the book. “There seems good
enough evidence to suggest that emotions con-
tribute to health.”

Cole and Fredrickson agree that their study
is small and needs to be repeated. But they say
that extensive previous research has validated
the questionnaire they used and confirmed
that it measures two distinct, albeit highly cor-
related, emotional states. They also note that
correlation does not necessarily mean that two
states are the same: height and weight are also
highly correlated, for example, yet describe dif-
ferent things. Each type of happiness tends to
encourage the other, says Fredrickson, “but we
can try to understand which is leading the way
towards health”.

The researchers are not the first from the
PNI community to face accusations of wish-
ful thinking. Indeed, the story of the field’s
founder — hailed in the press as proof of
the power of positive emotions — has been
questioned. Immunologists have suggested
that Cousins was not suffering from ankylos-
ing spondylitis at all, but from polymyalgia
rheumatica, which often clears up on its own.
His “health probably coincidentally remitted”,
says Cole.

Despite the criticisms, and the fact that his
work is in its early days, Cole says that he is
struck by the evidence that positive emotions
can override the biological effects of adversity
— enough to make changes in his own life.
Although he no longer has time to engage in
the art trade, he has embraced the ways that
his hobby helped him. “I have spent most of
my career and personal life trying to avoid or
overcome bad things,” he says. “I spend a lot
more time now thinking about what I really
want to do with my life, and where I’d like to
go with whatever years remain.” ■

Jo Marchant is a freelance science journalist
based in London.

1. Kiecolt-Glaser, J. K. et al. Psychosom. Med. 46, 7–14
(1984).

2. Kiecolt-Glaser, J. K., Speicher, C. E., Holliday, J. E. &
Glaser, R. J. Behav. Med. 7, 1–12 (1984).

3. Cohen. S., Janicki-Deverts, D. & Miller, G. E. J. Am.
Med. Assoc. 298, 1685–1687 (2007).

4. Cole, S. W. et al. Genome Biol. 8, R189 (2007).
5. Cole, S. W., Hawkley, L. C., Arevalo, J. M. G. &

Cacioppo, J. T. Proc. Natl Acad. Sci. USA 108,
3080–3085 (2011).

6. Black, D. S. et al. Psychoneuroendocrinology 38,
348–355 (2012).

7. Creswell, J. D. et al. Brain Behav. Immun. 26,
1095–1101 (2012).

8. Antoni, M. H. et al. Biol. Psychiatry 71, 366–372
(2012).

9. Fredrickson, B. L. et al. Proc. Natl Acad. Sci. USA
110, 13684–13689 (2013).

10. Brown, N. J. L., Sokal, A. D. & Friedman, H. L. Am.
Psychol. http://dx.doi.org/10.1037/a0032850
(2013).

11. Losada, M. Math. Comput. Model. 30, 179–192
(1999).

Psychoneuroimmunologist Steve Cole studies how stress and happiness affect health.

N
O

A
H

B
E
R

G
E
R

4 6 0 | N A T U R E | V O L 5 0 3 | 2 8 N O V E M B E R 2 0 1 3
© 2013 Macmillan Publishers Limited. All rights reserved