Compassionate Intention As a Therapeutic Intervention by Partners of Cancer Patients: Effects of Distant Intention on the Patients' Autonomic Nervous System
This double-blind study investigated the effects of intention on the autonomic nervous system of a human "sender" and distant "receiver" of those intentions, and it explored the roles that motivation and training might have in modulating these effects. Skin conductance level was measured in each member of a couple, both of whom were asked to feel the presence of the other. While the receiving person relaxed in a distant shielded room for 30 minutes, the sending person directed intention toward the receiver during repeated 10-second epochs separated by random interepoch periods. Thirty-six couples participated in 38 test sessions. In 22 couples, one of the pair was a cancer patient. In 12 of those couples, the healthy person was trained to direct intention toward the patient and asked to practice that intention daily for three months prior to the experiment (trained group). In the other 10 couples, the pair was tested before the partner was trained (wait group). Fourteen healthy couples received no training (control group). Using nonparametric bootstrap procedures, normalized skin conductance means recorded during the intention epochs were compared with the same measures recorded during randomly selected interepoch periods, used as controls. The preplanned difference examined the intention versus control means at the end of the intention epoch. Overall, receivers' skin conductance increased during the intention epochs (z = 3.9; P = .00009, two-tailed). Planned differences in skin conductance among the three groups were not significant, but a post hoc analysis showed that peak deviations were largest and most sustained in the trained group, followed by more moderate effects in the wait group, and still smaller effects in the control group. Directing intention toward a distant person is correlated with activation of that person's autonomic nervous system. Strong motivation to heal and to be healed, and training on how to cultivate and direct compassionate intention, may further enhance this effect.
NTENTION ON THE
Dean Radin, PhD,
Jerome Stone, MA, RN,
Ellen Levine, PhD,
* Shahram Eskandarnejad, MD,
Marilyn Schlitz, PhD,
Leila Kozak, PhD,
Dorothy Mandel, PhD,
and Gail Hayssen
Objective: This double-blind study investigated the effects of
intention on the autonomic nervous system of a human
“sender” and distant “receiver” of those intentions, and it ex-
plored the roles that motivation and training might have in
modulating these effects.
Design: Skin conductance level was measured in each member
of a couple, both of whom were asked to feel the presence of the
other. While the receiving person relaxed in a distant shielded
room for 30 minutes, the sending person directed intention
toward the receiver during repeated 10-second epochs separated
by random interepoch periods. Thirty-six couples participated in
38 test sessions. In 22 couples, one of the pair was a cancer
patient. In 12 of those couples, the healthy person was trained to
direct intention toward the patient and asked to practice that
intention daily for three months prior to the experiment (trained
group). In the other 10 couples, the pair was tested before the
partner was trained (wait group). Fourteen healthy couples re-
ceived no training (control group).
Outcome measures: Using nonparametric bootstrap proce-
dures, normalized skin conductance means recorded during the
intention epochs were compared with the same measures re-
corded during randomly selected interepoch periods, used as
controls. The preplanned difference examined the intention ver-
sus control means at the end of the intention epoch.
Results: Overall, receivers’ skin conductance increased during
the intention epochs (z ⫽ 3.9; P ⫽ .00009, two-tailed). Planned
differences in skin conductance among the three groups were
not signiﬁcant, but a post hoc analysis showed that peak devia-
tions were largest and most sustained in the trained group, fol-
lowed by more moderate effects in the wait group, and still
smaller effects in the control group.
Conclusions: Directing intention toward a distant person is cor-
related with activation of that person’s autonomic nervous sys-
tem. Strong motivation to heal and to be healed, and training on
how to cultivate and direct compassionate intention, may fur-
ther enhance this effect.
Key words: Distant healing, autonomic nervous system, inten-
(Explore 2008; 4:235-243. © Elsevier Inc. 2008)
A 2004 government survey of adult Americans, conducted by
the US National Center for Health Statistics, showed that of the
top 10 complementary and alternative medicine healing prac-
tices, the most popular was prayer for self and the second was
prayer for others.
From a psychological perspective, the former
may be thought of as a coping mechanism in the face of uncer-
tainty or dire need. The possibility that prayer for self may pro-
mote one’s own healing is not considered controversial because
of the growing literature on the salutary effects of meditation
and placebo and the plausibility of psychoneuroimmunological
models of self-regulation.
Prayer for others is likewise understandable as a practical cop-
ing mechanism, but the idea that it might be efﬁcacious for
another person remains contentious. To avoid unnecessary reli-
gious connotations, the descriptive phrase distant healing inten-
tion (DHI) is sometimes used in the scientiﬁc and medical liter-
ature to refer to this practice.
Distant healing intention effects
are considered scientiﬁcally doubtful by some because the “dis-
tant” in DHI means shielded from all known causal interac-
Science is beginning to reconcile with the concept of
“spooky action at a distance” within fundamental physics, but so
far the idea that nonlocal effects might also exist in living sys-
and be pragmatically useful in some way, evokes as much
contempt as it does serious interest.
Because the mechanisms underlying postulated DHI effects
are unknown, most DHI experiments have focused on the
straightforward empirical question: does it work? Can DHI af-
fect medical symptoms and outcomes? Some clinical studies of
hospital inpatients and medical outpatients suggest that DHI
might be medically efﬁcacious,
but as a whole the clinical
evidence remains uncertain.
By contrast, when DHI is tested under controlled laboratory
conditions, the evidence is less ambiguous. Meta-analyses indi-
1 Institute of Noetic Sciences, Petaluma, CA
2 Boulder Community Hospital, Boulder, CO
3 California Paciﬁc Medical Center, San Francisco, CA
4 University of Washington School of Medicine, Seattle, WA
5 Saybrook Graduate School, San Francisco, CA
* Currently at San Francisco State University, Calif.
# Corresponding Author. Address:
101 San Antonio Road, Petaluma, CA 94952
© 2008 by Elsevier Inc. Printed in the United States. All Rights Reserved EXPLORE July/August 2008, Vol. 4, No. 4
ISSN 1550-8307/08/$34.00 doi:10.1016/j.explore.2008.04.002
cate that DHI produces repeatable effects in the human auto-
nomic nervous system, detected typically by monitoring ﬂuctu-
ations in one person’s electrodermal activity (EDA) while a
distant person mentally attempts to inﬂuence the target person’s
emotions or attention.
The literature also indicates that
DHI effects can be detected in the central nervous system, as
measured in brain electrical activity
and also in the enteric nervous system.
The laboratory evidence may be clearer than the clinical evi-
dence because there are no “competing” intentions to interfere
with the test results, such as the prayers of clinical patients’ loved
ones, and also because physiological ﬂuctuations can be objec-
tively monitored in real time, whereas healing responses in the
clinic may progress over days or weeks. The context of labora-
tory studies is also quite different from that of clinical studies. In
the lab, the person assigned to “send” DHI (hereafter called the
sender) is typically a volunteer who is not especially motivated
or trained to provide DHI, and the person assigned to receive
DHI (the receiver) is often just curious to see what will happen
(the terms sender and receiver are used for expository reasons;
they do not imply a signaling model as the underlying mecha-
nism). Given these low motivational factors, it should not be
surprising that the magnitude of effects observed in such studies
is rather small (eg, the meta-analytic effect size estimate reported
by Schmidt et al
d ⫽ 0.11; P ⫽ .001).
The goal of the present study was to see what would happen
when the powerful, real-life motivations associated with clinical
trials of DHI were combined with the controlled context and
objective measures offered by laboratory protocols. In addition,
most previous DHI studies assigned the sender’s role to a labo-
ratory staff member, so the sender and receiver were often strang-
ers. The present study sought to enhance the ecological validity
of postulated DHI-type connections between couples by recruit-
ing long-term, bonded pairs and by exploring the role of training
and motivation in potentially modulating DHI effects. Given
the laboratory context, we did not test for distant healing per se,
but rather the physiological effects of distant intention. With
this caveat in mind, the term DHI will be used hereafter for ease
Pairs of friends, long-term partners, married couples, and mother-
child pairs were recruited to participate in one of three groups:
trained, wait, or control. Two of the groups were comprised of
adult couples, one of whom was healthy and the other was
undergoing treatment for cancer. The cancer patients and their
partners were recruited throughout the San Francisco Bay area
by healthcare provider referrals and newspaper advertisements.
The study design was explained to interested parties, including
the random assignment to different conditions, the data collec-
tion procedures, potential risks and beneﬁts, and their rights as
voluntary participants, including informed consent. Couples
were excluded if they were participating in family therapy, were
receiving any form of “energy healing,” if the partner was en-
rolled in a cancer support group, or if they chose at any time to
leave the study.
The healthy partner was assigned the role of the sender of DHI
and the patient the role of the receiver. In the trained group, the
sender attended a program involving discussion and practice of
a DHI technique based on the cultivation of compassionate inten-
tion, deﬁned as the act of directing selﬂess love and care toward
another person, with intention to relieve their suffering and
enhance their well-being.
The training program, developed and provided by second
author, consisted of a daylong, eight-hour, group workshop,
followed by a daily half-hour practice at home for three months.
The program included a lecture on the healing potential of
compassionate intention, discussion of common resistances to
positive expectations about DHI, guided instruction in several
meditation and mental focusing practices, and guided exercises
in breath-based techniques for enhancing compassion, as vari-
ously practiced in Tibetan Buddhism (the practice known as
After attending the training session and practicing the DHI
meditation daily for three months (healthy partners were asked
to keep a daily log to verify their practice), these couples were
tested in the laboratory. In the wait group, the couple was tested
before the healthy partner attended the training program. A
third group consisted of healthy couples who received no train-
ing (the control group). Of those recruited for the trained group
and wait group, 10 couples eventually dropped out. Reasons
provided included time constraints, dissolution of the couple’s
relationship, the couple was in search of a “quick ﬁx,” death of a
patient, spouse was not available, complications of cancer, or
because one or more concepts in the training program clashed
with the couple’s belief system.
When a couple arrived at the lab, informed consents were
signed and then the experimenters attached electrodes to each
person to monitor ﬁve physiological variables. The principal
measurement was EDA, speciﬁcally, skin conductance level
(SCL), as this is the variable most frequently employed in simi-
lar, previous studies. Electrodermal activity was monitored with
two electrodes, each ﬁlled with an isotonic electrode gel (Biopac
GEL101, Biopac, Goleta, Calif) and attached to the left palm by
using double-sided adhesive collars (Biopac type TSD203, 8-mm
Ag/AgCl electrodes). These electrodes were attached to a Biopac
GSR-100C EDA ampliﬁer set to the 0- to 2-
S range (2
volt, or 0-20
S for the full 10-volt range of the EDA ampliﬁer).
For exploratory purposes, we also monitored one channel of
electroencephalogram at C
, ﬁngertip blood volume on the left
thumb, electrocardiogram, and abdominal respiration. Results
of those measurements will be reported in other publications.
All signals were recorded at either 500 or 1000 samples per
second, and each person was monitored by a separate physiolog-
ical recording system (Biopac M150, Biopac). To assist in the
computational process, all raw physiological data were down-
sampled to 100 samples per second before analysis.
The couple was asked to maintain a “feeling of connected-
ness” with each other. To assist with this intentional focus, each
person was asked to exchange a personal item, like a ring or
watch, and to hold that object in his/her free right hand for the
duration of the session. In the control group, couples were asked
to decide which of the two might be more receptive, and that
236 EXPLORE July/August 2008, Vol. 4, No. 4 Effects of Distant Intention on Patients’ Autonomic Nervous System
person was assigned the role of the receiver. In the trained group
and wait group, the cancer patient was always the receiver.
The receiver was asked to relax in a reclining chair inside a
double steel-walled, electromagnetically and acoustically
shielded chamber, as illustrated in Figure 1 (series 81 solid cell,
ETS-Lindgren, Cedar Park, Tex). The receiver was informed that
the sender would be viewing his or her live video image at
random times from a distant location, and that during those
periods the sender would try to make a special intentional effort
to mentally connect with him or her. No one involved in the
experiment knew exactly when those random periods would
occur, as they were selected by a computer (described below).
A low-light video camera was focused on the receiver’s face,
and the interior of the shielded room was illuminated with a
25-watt incandescent bulb. The physiology and video signals
were routed outside the shielded room via optical ﬁber to two
computers (SI Tech Models 2809/2010 and Model 2550, Bata-
via, Ill), one dedicated to recording the receiver’s physiological
signals and the other used to automatically run the experimental
session, including switching the video image to the sender’s
location at random times.
To test for possible sensory cues between the sender and re-
ceiver locations, audio tests were conducted to check whether
tones as loud as 110 dB at 1,000 Hz sounded in the sender’s
room could be detected inside the receiver’s shielded chamber.
Subjective hearing tests along with quantitative audio tests using
a digital sound level meter conﬁrmed that the test tones were
indistinguishable from background noise inside the chamber
(model 840028, Sper Scientiﬁc, Scottsdale, Ariz). To further
isolate the shielded room from potential infrasound cues, the
chamber rested upon a vibration-dampening vinyl mat in the
basement of a building.
After the receiver was settled in the shielded room, the sender
was led through two closed doors to a dimly lit room 20 meters
away and asked to sit in a chair about a half meter in front of a
video monitor. An experimenter explained that when the video
monitor showed the receiver’s image, the sender was to try to
mentally “connect” with the receiver with as much intensity as
possible. The principal experimenter (D.R.) was blind to
whether a couple was in the trained or wait group but was aware
of the condition for the control group participants, the majority
of whom were recruited by and guided through the experimental
sessions by L.K., D.M., and G.H.
The sender’s electrodes were connected to the same model
Biopac system as the receiver’s, using the same type of ampliﬁ-
ers, settings, and data sampling rates. The digitized outputs from
both Biopac systems were transmitted over a local area network
and streamed to two Windows-based PCs, each running
Biopac’s Acknowledge 3.7.1 data collection software (Figure 1).
The timing of the viewing periods was controlled by a Windows
PC running a program written by D.R. in Microsoft Visual Basic
6.0. When that program was launched, it created a random
timing schedule for either 25 (control group) or 36 (motivated
groups) 10-second visual stimulus epochs. Epochs were sepa-
rated from one another by a randomly determined ﬁve to 40
second interepoch interval (Figure 2; the random source in both
cases was based on Visual Basic 6.0’s pseudorandom algorithm,
seeded by the PC’s CPU clock at the beginning of each session).
To synchronize the sender and receiver physiological signals, at
the beginning of each stimulus epoch the computer switched the
video signal from the receiver’s chamber to the video monitor in
front of the sender and simultaneously sent onset marker signals
to both the sender and receiver Biopac systems (using signals
generated by an analog to digital circuit; model ADR-100, On-
Figure 1. Laboratory layout. The experimenter’s workstation (E) consisted of three computers: two recorded the physiological data from the
sender (PC-S) and the receiver (PC-R) Biopac systems. The third (PC-E) controlled the random timing of the stimuli and a video switch. The receiver
was in an electromagnetically and acoustically shielded room; the sender was in a distant room behind two doors and a double wall.
Effects of Distant Intention on Patients’ Autonomic Nervous System EXPLORE July/August 2008, Vol. 4, No. 4
trak Control Systems, Sudbury, Ontario, Canada). At the end of
each stimulus epoch, the computer switched the video signal off
and sent offset markers to the two Biopac systems. After both
participants were secured in their respective rooms, the experi-
menter checked to see if the physiological recordings, marker
signals, and video switch were operating properly. When every-
thing was in order, an experimenter started the controlling pro-
gram and attended to other tasks while waiting for the session to
Hypotheses and Analyses
The principal hypothesis was that the sender’s DHI directed
toward the distant, isolated receiver would cause the receiver’s
autonomic nervous system to become activated. A secondary
analysis explored whether the factors of motivation and training
modulated the postulated effect.
In the following description, the term epoch refers to the 20-
second period from ﬁve seconds before stimulus onset to ﬁve
seconds after stimulus offset (this range was used to examine the
physiological responses in temporal context), and stimulus epoch
refers to the 10-second DHI period between stimulus onset and
offset. The analysis examined changes in SCLs averaged across
epochs (ensemble average) to see how the sender and the receiver
responded to DHI in time synchrony.
To determine the statistical signiﬁcance of the observed re-
sults, the following procedures were independently applied to
the receiver and the sender SCL data. This bootstrap analysis, a
common method within the larger domain of computational
statistics, is a nonparametric way of analyzing physiological data
because it makes no assumptions about the underlying structure
of the data, and because it answers precisely what we wish to
know: did SCL change in an unexpected way during the actual
stimulus epochs, as compared with other, randomly selected
times? All SCL data were smoothed using a one-second sliding
average window (⫾500 msec), then a sequence of steps was
applied as follows:
● for each SCL sample in a given session, subtract the SCL value
at stimulus onset to form a measure of change in SCL during
● calculate the ensemble mean of the baseline-subtracted ep-
ochs in step one for each session
● calculate the grand ensemble mean across all sessions of inter-
est (eg, trained group sessions)
● select random starting points in each session, one for each
epoch in the original session, and from those create new,
20-second random epochs; subtract the baseline from each of
these random epochs as in step one
● form the ensemble mean of the random epochs
● do the same for the other sessions, then calculate a grand
ensemble random epoch mean
● repeat steps 4 to 6 10,000 times to build up a bootstrap distri-
bution of ensemble random epochs that could have occurred in
the experiment if the original epochs had occurred at different
times than they did in the actual experiment
● normalize each sample in the original ensemble average curve
(20 seconds ⫻ 100 samples/second ⫽ 2,000 samples) by using
the mean and standard deviation of the bootstrap distribution
formed in step seven, as z
, where i ranges from
1 to 2,000 samples, x
is sample i from the original ensemble
is the mean and
the standard deviation of the
associated sample from the bootstrap distribution; this step
essentially creates a z score for each sample in the original
ensemble epoch; basing the results on normalized scores
weights each epoch equally
Under the null hypothesis, the precise timing of the epochs
should not matter because the receiver was thoroughly isolated
from the sender. Thus, if at stimulus offset the normalized en-
semble epoch for the receiver signiﬁcantly deviated from chance
(as determined by the bootstrap process), it would suggest that
the receiver had responded to, or more generally was correlated
with, the sender’s DHI. To avoid multiple testing problems, the
preplanned hypothesis examined the normalized deviation only
at stimulus offset.
Seventy-two people participated in the study (Table 1), includ-
ing two minors (a mother-son and mother-daughter pair: 36
couples who together conducted a total of 40 sessions, 38 of
which were usable (two control sessions could not be analyzed
because the senders’ physiological data failed to record prop-
erly). Ideally, the three groups of participants would have been
matched by gender and age, but in practice this was difﬁcult to
achieve as the clinical groups mostly involved women with
breast cancer, and this tended to skew the age and gender of
Figure 2. Protocol for sender stimulus. Random interepoch intervals ranged from ﬁve to 40 seconds, separated by 10-second distant healing
intention periods. The receiver’s live video image appeared on the monitor during the sending periods, otherwise the monitor was black.
238 EXPLORE July/August 2008, Vol. 4, No. 4 Effects of Distant Intention on Patients’ Autonomic Nervous System
those groups. In addition, two couples in the control group
switched roles as the sender and the receiver, and all individuals
in the control dyads were healthy.
All participants in the trained group and wait group ﬁlled out
demographics questionnaires upon beginning the study, then
before and after the training periods they ﬁlled out question-
naires on mood,
and spiritual well-be-
and the patients only ﬁlled out the Functional Assessment
of Chronic Illness Therapy (FACIT, Version 4), a self-report
measure designed for cancer patients to assess various factors
associated with well-being.
Analysis of the demographics
indicated that the trained group and wait group were well
matched in terms of gender, ethnicity, family history of cancer,
income, prior participation in a cancer therapy group, and in-
volvement in a religious practice.
Analysis of the psychosocial data showed no signiﬁcant dif-
ferences in well-being, mood, or quality of life between healthy
partners in the trained group or wait group. However, trained
group patients showed both a decline in physical well-being as
compared with wait group patients (P ⬍ .01, two-tailed), and an
improvement in spiritual well-being (P ⬍ .01, two-tailed; both
results survive Bonferroni corrections for multiple tests). We
might interpret this apparently contradictory outcome to imply,
purely as a metaphor, that distant intentions might act as church
bells that are rung incessantly to assist in healing the ill. The
benevolent intentions associated with such chimes may be per-
ceived and appreciated by the mind, thereby raising one’s spirits,
but they may also prevent the body from getting the rest it needs,
making the body feel worse.
To reduce the potential biasing effects of movement artifacts, all
data were visually inspected, and SCL epochs with artifacts were
eliminated from further consideration (artifacts were identiﬁed
by D.R., who was not blind to each epoch’s underlying condi-
tion). This analysis slightly reduced the potential total of 1,170
to 1,140 epochs (97%) as follows: 387 of 410 trained epochs
(94%), 360 of 360 wait epochs (100%), and 393 of 400 control
epochs (98%). The ﬁrst two trained trials consisted of 25 epochs,
the last 10 consisted of 36 epochs, for 410 possible epochs.
The sender’s SCL across all epochs, sessions, and groups in-
creased substantially after stimulus onset, conﬁrming the ex-
pected activation in the sender’s sympathetic nervous system as
a result of the increased mental effort associated with providing
DHI (Figure 3). About two seconds after stimulus onset, the
sender’s SCL began to increase, peaking three seconds later at
more than z ⫽ 12 standard errors above the baseline. In addi-
tion, as predicted by the DHI hypothesis, the receiver’s SCL also
signiﬁcantly increased. A half second after stimulus onset, the
receiver’s SCL began to rise, peaking by stimulus offset at z ⫽ 3.9
standard errors over the baseline (P ⫽ .00009; all P values cited
Motivated Versus Control Groups
Figure 4 is the same analysis applied to just the motivated group
(trained group, N ⫽ 387 epochs; wait group, N ⫽ 360 epochs;
747 epochs combined, 22 participants). The receiver SCL signif-
icantly increased to z ⫽ 3.45 (P ⫽ .0006) at stimulus offset,
peaking at 7.8 seconds at z ⫽ 4.481 (P ⫽ 7.4 ⫻ 10
By comparison, Figure 5 shows that the receiver SCL for the
control group (16 sessions, 393 epochs) increased to z ⫽ 2.4 (P ⫽
.02) at stimulus offset. The difference between the motivated
and control group outcomes at stimulus offset was not signiﬁ-
cant (z ⫽ 0.73; P ⫽ .46). When comparing effect sizes per stim-
ulus epoch (where e ⫽ z⁄
N, N being the number of epochs) as
shown in Figure 6, the receivers’ SCL at stimulus offset was
observed to be about the same magnitude in all of the groups.
Table 1. Participant Demographics
Group Sessions Couples
Age of Sender,
Age of Receiver,
Control 16 14 7–71 (41) 24–58 (39) 11M/7F 5M/13F
Trained 12 12 37–84 (55) 38–78 (54) 7M/5F 4M/8F
Wait 10 10 42–77 (57) 41–79 (53) 9M/1F 1M/9F
M, male; F, female.
Two of the 18 control sessions did not produce usable data.
Figure 3. Sender (thin line) and receiver (bold line) normalized mean
skin conductance levels across all 38 sessions (N ⫽ 1,140 epochs),
from ﬁve seconds before stimulus onset (at 0 seconds) to ﬁve seconds
after stimulus offset (at 10 seconds), to show the effect in context.
EDA, electrodermal activity.
Effects of Distant Intention on Patients’ Autonomic Nervous System EXPLORE July/August 2008, Vol. 4, No. 4
Comparison of the receiver SCL time course among three
groups reveals a more interesting trend, as shown in Figure 7.
Receivers in all three groups responded quickly at stimulus on-
set, but (1) the control group’s response subsided after four
seconds, (2) the wait group’s response was initially stronger and
subsided after ﬁve seconds, and (3) the trained group’s response
continued to progressively rise for eight seconds, reaching the
maximum deviation among all three groups (in this comparison,
the number of epochs in each curve is approximately the same,
so the normalized curves are not biased by differences in sample
size: trained ⫽ 387 epochs, wait ⫽ 360 epochs, control ⫽ 393
epochs). These differences were not predicted in advance so they
must be interpreted with caution. However, if future replications
continue to show similar patterns, then training plus motivation
would appear to enhance receiver’s response over motivation
alone, and motivation would appear to enhance the response
over interest alone.
Analysis of all skin conductance data indicates that the sender’s
DHI had a measurable effect on the receiver’s autonomic ner-
vous system. Inspection of the time course of SCL over the
average epoch suggests that the trained group had a slower but
more sustained effect, followed by a more moderate wait group
response, and an even smaller control group response. The over-
all SCL effect size per session for the motivated groups was
e ⫽ 3.45⁄
22 ⫽ 0.74, some 6.7 times larger than the earlier
meta-analytic estimate of e
⫽ 0.11, and surprisingly even the
control group effect size was some 5.4 times larger than e
16 ⫽ 0.60). It might be noted that the absolute magni-
tudes of the observed effects were still rather small, for example,
for the receiver’s SCL, the peak changes over baseline amounted
to fractions of a microSiemen (of course, small magnitude effects
do not imply no effects; statistically speaking the results are
Figure 4. Sender (thin line) and receiver (bold line) normalized mean
skin conductance levels for all motivated sessions (n ⫽ 736 epochs).
EDA, electrodermal activity.
Figure 5. Sender (thin line) and receiver (bold line) normalized mean
skin conductance levels for control sessions (n ⫽ 393 epochs).
Figure 6. Comparison of sender and receiver effect sizes (per epoch)
measured at stimulus offset (with ⫾2 standard error conﬁdence
intervals) for all sessions, motivated sessions (trained group and wait
group combined), and trained, wait, and control groups separately.
EDA, electrodermal activity.
Figure 7. Normalized comparison of receiver skin conductance lev-
els in the three groups. EDA, electrodermal activity.
240 EXPLORE July/August 2008, Vol. 4, No. 4 Effects of Distant Intention on Patients’ Autonomic Nervous System
Many artifacts can produce outcomes that mimic DHI effects. In
order of decreasing likelihood, they include measurement arti-
facts, sensory leakage between the sender and receiver, the re-
ceiver’s anticipation of the timing of stimulus epochs, software
artifacts, violation of statistical assumptions, selective data re-
porting, and collusion between the sender and the receiver.
Potential measurement artifacts include electrical crosstalk
that can arise between ampliﬁers in the same monitoring device,
or artifacts induced into the equipment due to electromagnetic
(EM) pulses associated with switching the video monitor signal
at stimulus onset and offset. The possibility of crosstalk was
precluded by using two independent Biopac monitors located
20 meters apart, each with its own data recording computer.
Potential effects of EM pulses were signiﬁcantly diminished by
the use of a double steel-walled EM-shielded room and 20-meter
separation between the sender and the receiver. The shielded
chamber was designed to effectively block all EM radiation
above 10 kHz, but it did not block extremely low frequency EM
or magnetic ﬁelds. Living systems are known to be sensitive to
weak EM and magnetic ﬁelds, so bioelectromagnetic factors
cannot be absolutely ruled out as potential artifacts.
prior successful studies (in distant perception tasks) conducted
with the sender on land and the receiver in a submarine under
many meters of seawater raise doubts that extremely low fre-
quency signals are sufﬁcient to explain this type of “nonlocal”
Further doubts are raised because of evidence sug-
gesting that DHI effects can be observed even when the sender
and the receiver are displaced in time.
Sensory leakage artifacts can include conscious or uncon-
scious visual, auditory, or vibratory cues that might pass between
the sender and the receiver. Such artifacts were precluded from
the present experiment through the use of separate rooms for the
sender and receiver, the latter being a heavily shielded chamber,
and through prior sound leakage tests. In addition, the experi-
menters were located between the sender and receiver locations,
with no other points of access between the two sites, so any
attempt by the couple to communicate through ordinary means
would have been detected. Moreover, the physiological condi-
tion of both participants was continually monitored during the
experiment, allowing detection of the smallest bodily move-
ments in either person. No gross motor movements consistent
with attempts at surreptitious signaling were detected in any of
For other potential artifacts, could the receiver have antici-
pated when the stimulus epochs were about to occur and then
respond accordingly? This possibility was prevented through the
use of random interepoch timing and double-blind conditions.
No one knew in advance of a session when each DHI epoch
would begin. The random timing and blinded design is also
relevant to assessing the impact of a potential bias due to the
awareness of D.R. of which couples were in the control group,
and to the fact that most of the data from the control couples
were collected by L.K., D.M., and G.H. Could different inter-
personal styles among these four investigators have inﬂuenced
how couples responded in this experiment? The question arises
because investigators holding different a priori opinions about
the likelihood of DHI effects have reported results, even in
jointly run experiments, that fell into alignment with their indi-
However, a replication study designed to exam-
ine the role of interpersonal interactions in more detail failed to
support the earlier results, thus the inﬂuences that different in-
vestigators may have on DHI outcomes remain uncertain.
any case, all investigators in the present study were open to the
concept of DHI, so although some interpersonal bias cannot be
ruled out, it seems unlikely that the knowledge of D.R. on which
couples participated in the control group would have had much
impact on the outcome. Indeed, all three groups showed signif-
icant results in the preplanned outcomes.
Potential violations of the assumptions underlying parametric
statistics were avoided by using a nonparametric, computational
bootstrap procedure to normalize the ensemble averages. To
prevent selective reporting biases, data from all usable epochs
across all sessions were analyzed and reported for the measure of
principal interest (SCL). Collusion between the sender and the
receiver would have been exceedingly difﬁcult to carry out, not
only because the EM shielding prevented obvious signaling
methods such as sounds and cell phones, but because almost all
of the couples participated in only one session, so they did not
know what to expect in advance about the laboratory setup or
the experimental protocol.
If not due to conventional explanations, then how do we inter-
pret these results? Sloan and Ramakrishnan
have asserted that
“Nothing in our contemporary scientiﬁc views of the universe or
consciousness can account for how the ‘healing intentions’ or
prayers of distant intercessors could possibly inﬂuence the
[physiology] of patients even nearby let alone at a great dis-
Is it really true that nothing in science suggests the presence of
connections between apparently isolated objects? Quantum en-
tanglement, a far from common sense effect predicted by quan-
tum theory and later demonstrated as fact in the laboratory,
shows that under certain conditions, elementary particles that
were once connected appear to remain connected after they
separate, regardless of distance in space or time. If this property
is truly as fundamental as it appears to be, then in principle
everything in the universe might be entangled.
jects and humans certainly do not appear to show such entan-
glements, and there are nontrivial arguments for why small-scale
entanglement would be difﬁcult to sustain in large, living sys-
tems. But still, one cannot help wondering what if this concept
did apply to humans? In a casual, indifferent, unmotivated cou-
ple, entanglements between their minds and bodies may be dif-
ﬁcult to detect, not only in a fundamental physics sense, but
even in an ordinary psychodynamic sense. By comparison, in a
long-term, highly motivated, bonded couple, and with the
sender speciﬁcally trained to provide compassionate intention,
the underlying correlations might be far more evident. Such a
relational model is appealing because it does not require any-
thing (force, energy, or signals) to pass between the sender and
the receiver. Instead, it postulates a physical correlation that is
always present between people (and everything else) due to the
“nonlocal threads” from which the fabric of reality is woven.
Effects of Distant Intention on Patients’ Autonomic Nervous System EXPLORE July/August 2008, Vol. 4, No. 4
Another possible interpretation is that the outcomes of this
and similar experiments are due to precognition on the part of
the investigators, who manage to begin each session at just the
right time so as to match natural ﬂuctuations in the receiver’s
physiology with the randomly determined moments of stimulus
onset and offset. Although such an explanation may seem im-
plausible, independent evidence in favor of retrocausal effects in
humans continues to accumulate,
so it is not inconceivable.
Indeed, because there are as yet no adequate theoretical models
that would predict macroscopic correlations akin to DHI, we are
obliged to remain open to a wide range of possible explanations.
A key limitation in the present study was the lack of closely
matched demographics among the three groups. Given this lim-
itation, it would be imprudent to draw strong conclusions about
performance differences among the groups. However, based on
the overall support of the formal hypothesis, it is possible to
draw one conclusion: directing one’s attention toward a distant
person apparently causes measurable changes in that person’s
nervous system. This suggests that DHI provides more than a
psychological coping mechanism, and that prayer for others is
the second most popular complementary and alternative medi-
cine modality for a very simple reason: it has an effect on the
human body, presumably an effect that is usually perceived as
beneﬁcial in some way. Whether it speciﬁcally promotes healing
remains to be seen.
This study was supported in part by grants from The Institute for
Research on Unlimited Love and the Institute of Noetic Sci-
ences. We thank Larry Dossey, Mitchell Krucoff, John Astin,
Fred Luskin, Brother David Stendl-Rast, Jessica Utts, Russell
Targ, Edwin May, and the staff of the Institute of Noetic Sci-
ences for helping to raise funds to complete this experiment.
This study is dedicated to Elisabeth Targ.
1. Barnes P, Powell-Griner E, McFann K, Nahin R. CDC Advance
Data Report #343. Complementary and Alternative Medicine Use
Among Adults: United States. May 27, 2004. Available at http://
nccam.nih.gov/news/2004/052704.htm. Accessed June 27, 2005.
2. Kiecolt-Glaser JK, McGuire L, Robles TF, Glaser R. Psychoneuro-
immunology and psychosomatic medicine:back to the future. Psy-
chosom Med. 2002;64:15-28.
3. Schlitz M, Radin DI, Malle BF, et al. Distant healing intention:
deﬁnitions and evolving guidelines for laboratory studies. Alt Ther-
apies Health Med. 2003;9:A31-A43.
4. Wallis C. Faith and healing. Time. June 24, 1996:58-64.
5. Sloan RP, Ramakrishnan R. The MANTRA II study. Lancet. 2005;
6. Walach H. Generalized entanglement: a new theoretical model for
understanding the effects of complementary and alternative medi-
cine. J Alt Comp Med. 2005;11:549-559.
7. Astin JA, Harkness E, Ernst E. The efﬁcacy of “distant healing”: a
systematic review of randomized trials. Ann Intern Med. 2000;132:
8. Krucoff MW, Crater SW, Green CL, et al. Integrative noetic thera-
pies as adjuncts to percutaneous intervention during unstable coro-
nary syndromes: Monitoring and Actualization of Noetic Training
(MANTRA) feasibility pilot. Am Heart J. 2001;142:760-769.
9. Krucoff MW, Crater SW, Gallup D, et al. Music, imagery, touch,
and prayer as adjuncts to interventional cardiac care: The Monitor-
ing and Actualisation of Noetic Trainings (MANTRA) II random-
ised study. Lancet. 2005;366:211-217.
10. Benson H, Dusek JA, Sherwood JB, et al. Study of the therapeutic
effects of intercessory prayer (STEP) in cardiac bypass patients: a
multicenter randomized trial of uncertainty and certainty of receiv-
ing intercessory prayer. Am Heart J. 2006;151:934-942.
11. Schlitz M, Braud WG. Distant intentionality and healing: assessing
the evidence. Alt Therapies Health Med. 1997;3(6):62-73.
12. Schmidt S, Schneider R, Utts J, Walach H. Distant intentionality
and the feeling of being stared at: two meta-analyses. Br J Psychol.
13. Duane TD, Behrendt T. Extrasensory electroencephalographic in-
duction between identical twins. Science. 1965;150:367.
14. Radin DI. Event-related EEG correlations between isolated human
subjects. J Alt Comp Med. 2004;10:315-324.
15. Standish LJ, Kozak L, Johnson LC, Richards TJ. Electroencephalo-
graphic evidence of correlated event-related signals between the
brains of spatially and sensory isolated human subjects. J Alt Comp
16. Wackermann J, Seiter C, Keibel H, Walach H. Correlations between
brain electrical activities of two spatially separated human subjects.
Neurosci Lett. 2003;336:60-64.
17. Grinberg-Zylberbaum J, Delaﬂor M, Attie L, Goswami A. The Ein-
stein-Podolsky-Rosen paradox in the brain: the transferred poten-
tial. Phys Essays. 1994;7:422-427.
18. Richards TL, Kozak L, Johnson LC, Standish LJ. Replicable func-
tional magnetic resonance imaging evidence of correlated brain sig-
nals between physically and sensory isolated subjects. J Alt Comp
19. Achterberg J, Cooke K, Richards T, et al. Evidence for correlations
between distant intentionality and brain function in recipients: a
functional magnetic resonance imaging analysis. J Alt Comp Med.
20. Radin DI, Schlitz MJ. Gut feelings, intuition, and emotions: an
exploratory study. J Alt Comp Med. 2005;11:85-91.
21. Rosenthal R. Parametric measures of effect size. In: Cooper H,
Hedges LV, eds. The Handbook of Research Synthesis. New York, NY:
Russell Sage Foundation; 1994:231-244.
22. Chodron P. Awakening Loving-Kindness. Boston, Mass: Shambhala; 1996.
23. Rinpoche S. The Tibetan Book of Living and Dying. San Francisco,
Calif: Harper San Francisco; 1994.
24. Davis A, Meditation From the Heart of Judaism. Woodstock, NY: Jew-
ish Lights; 1997.
25. Lerner M. Jewish Renewal: A Path to Healing and Transformation. San
Francisco, Calif: Perrenial; 1995.
26. Krieger D. Accepting Your Power to Heal. Rochester, NY: Bear & Co; 1993.
27. Blair RC, Karniski W. An alternative method for signiﬁcance testing of
waveform difference potentials. Psychophysiology. 1993;30:518-524.
28. McNair DM, Lorr M, Droppleman IF. Proﬁle of Mood States. San
Diego, Calif: Educational and Industrial Testing Service; 1971.
29. Hudson WW. Walmyr Assessment Scales Scoring Manual. Tallahassee,
Fla: Walmyr Publishing Company; 1997.
30. Brady M, Peterman AH, Fitchett G, Mo M, Cella D. A case for
including spirituality in quality of life measurement in oncology.
31. Cella DF, Tulsky DS, Gray G, et al. The Functional Assessment of
Cancer Therapy (FACT) scale: development and validation of the
general measure. J Clin Oncol. 1993;11:570-579.
32. Webster K, Odom L, Peterman A, Lent L, Cella D. The Functional
Assessment of Chronic Illness Therapy (FACIT) measurement sys-
tem: validation of version 4 of the core questionnaire. Qual Life Res.
242 EXPLORE July/August 2008, Vol. 4, No. 4 Effects of Distant Intention on Patients’ Autonomic Nervous System
33. Lin JC, ed. Advances in Electromagnetic Fields in Living Systems. Vol. 4.
New York, NY: Springer Science⫹Business Media; 2005.
34. Schwartz SA. Opening to the Inﬁnite. Virginia Beach, VA: Nemoseen
35. Braud W. Wellness implications of retroactive intentional inﬂu-
ence: exploring an outrageous hypothesis. Alt Therapies Health Med.
36. Radin DI, Machado F, Zangari W. Effects of distant healing inten-
tion through time and space: two exploratory studies. Subtle Energies
Energy Med. 2000;11:207-240.
37. Leibovici L. Effects of remote, retroactive intercessory prayer on
outcomes in patients with bloodstream infection: randomised con-
trolled trial. BMJ. 2001;323:1450-1451.
38. Wiseman R, Schlitz M. Experimenter effects and the remote detec-
tion of staring. J Parapsychol. 1997;61:197-207.
39. Schiltz M, Wiseman R, Watt C, Radin DI. Of two minds: skeptic-
proponent collaboration within parapsychology. Brit J Psychol. 2006;
40. Radin DI. Entangled Minds. New York, NY: Simon & Schuster; 2006.
41. Radin DI. Unconscious perception of future emotions: an experi-
ment in presentiment. J Sci Explor. 1997;11 (2):163-180.
42. McCraty R, Atkinson M, Bradley RT. Electrophysiological evidence
of intuition: part 1. The surprising role of the heart. J Alt Comp Med.
43. McCraty R, Atkinson M, Bradley RT. Electrophysiological evidence
of intuition: part 2. A system-wide process? J Alt Comp Med. 2004;
44. May EC, Paulinyi T, Vassy Z. Anomalous anticipatory skin con-
ductance response to acoustic stimuli: experimental results and
speculation about a mechanism. J Alt Comp Med. 2005;11:587-
45. Spottiswoode SJP, May EC. Skin conductance prestimulus re-
sponse: analyses, artifacts and a pilot study. J Sci Explor. 2003;17:
46. Radin DI. Electrodermal presentiments of future emotions. J Sci