Fantastic voyage: Chasing oxytocin from the bedside to the bench and back again

Abstract

This is the story of my 50-year career in medicine and research, and the people who influenced and helped me most along the way. I recount the way in which I became interested in oxytocin early in my career as a child psychiatrist, and how it led me back to Columbia University, my alma mater, to study oxytocin's role in mother-child innate behaviors. I recount how oxytocin/oxytocin receptor signaling was central to my basic and clinical research and present a new theory on mother-infant emotional behaviors that challenges 400 years of brain-centric science. My history underscores the important and unique perspective women bring to science and why women are especially needed in the sciences. I hope to inspire young women (and young men) who are beginning their careers in research.

Highlights

• •Colostrum and milk oxytocin are important in infant behavior, as well as in gut function and development.
• •Oxytocin/oxytocin receptor signaling is a modulator of mRNA translation, which explains in part why metabolic co-ordination between infant and mother is key to infant survival.
• •Autonomic states of mother and infant determine the perinatal autonomic socioemotional reflex (ASR), which determines approach or avoidant orienting reflex behaviors.
• •The mother-child ASR behaviors correlate with physiology as measured by vagal tone.

1. Background

In the 1970's, much like the scientists in the 1960's science fiction movie, I embarked on my own 50-year Fantastic Voyage, a thrilling chase after the illusive oxytocin receptor— into my patients' nervous systems, blood streams, cells, gut, brains and finally— into their hearts. I'd like to take you on my voyage, so that young women in science may be inspired to take one of their own.

First, however, I will briefly describe the academic research and clinical worlds in which my story is set, to provide some context for my 50-year history. Late 20th century psychiatric teaching was dominated by cortico-centric theories, genetic determinism and pharmaceutical interventions. The regulatory importance of brain-body feedback loops, especially the well-defined neural signaling pathways, were well known within autonomic neurophysiology, but those findings had not yet influenced mental health interventions. Autonomic research showed that regulation is not limited to the afferent or efferent signals influencing visceral organs, but relates to the dynamic regulation of the organs through well-defined neural pathways influenced by the social environment. Through my own clinical observations during this period, I discovered that the behaviors associated with the autonomic pathways were not fixed. The metabolically costly survival avoidance reactions displayed during family emotional encounters—which were routinely being treated with pharmaceuticals— could be transformed into spontaneous behaviors of emotional engagement via simple daily parent-child social exercises within the home. During this period, I started to suspect that oxytocin was playing a role in this dynamic regulatory relationship that I was witnessing. Parallel to my work, indices of vagal tone were being shown to provide an easily accessible measure of relational health. In the early 1990s before I joined the faculty, two dinstinguished Austrians, a psychiatrist and a physiologist, asked to meet with me and Columbia Psychiatry's Division of Psychobiology. Imagine my surprise and every-one's surprise when toward the end of the all-day meeting the visitors announced why they were there. They predicted that their measures of heart rate variability would prove the efficacy of my mother-child intervention.

My early intuitions about how co-regulatory autonomic state signaling can change neurophysiology proved to be accurate and were confirmed by our recent randomized controlled trials (RCTs) among preterm infants and preschool-age children and their mothers. Together with Steve Porges, we published data showing that the maturation of vagal development of the intervention preterm infants had a steeper curve when compared to standard care infants [1]. In an RCT that applied the mother-child intervention to a preschool child population, we found higher vagal tone six months following intervention that correlates with improved behavior.

Succinctly stated, co-regulation within the family promotes a physiology that supports health, growth, and restoration. The common theme of my work links sociality to homeostatic biological functions and co-regulation, which other researchers of oxytocin and the autonomic nervous system are currently describing as 'sociostasis' [2,3].

2. Beginnings

My three-year-old report card from nursery school quotes the teacher, “If someone is crying, Martha is there comforting.” From a young age, I felt that I was going to do some important work, because my mother would regularly take my small hands and say, “What important work are these hands going to do?” Inasmuch as my father was a research chemist and my mother's field was child development, it was probably a given that I would develop interests in both science and helping children and their parents. That led me to enroll in medical school at Columbia University College of Physicians and Surgeons to eventually become a psychiatrist.

However, the summer before the first year of medical school, I had a research job with the assignment to develop a microassay for chloramphenicol. Because the assay was meant to determine levels of this potentially lethal antibiotic in the bloodstream of preterm infants in the Neonatal Intensive Care Unit (NICU), I could only obtain a minute sample of blood. I succeeded in developing the assay, without receiving credit for the accomplishment. Shortly after that, the principal investigator left Columbia, taking my lab book with him. More on this brief event in my education later.

I went on to medical school, where I was trained by an outstanding and accessible clinical and research faculty who paid copious amounts of attention to students like me who showed extra interest in the work. It was assumed I would stay at Columbia for my residency. However, friends advised me to choose Albert Einstein College of Medicine's residency in General Psychiatry, which purportedly had a more eclectic, less gender-biased, program. After two years in residency at Einstein, I was offered a fellowship there in Child Psychiatry at Einstein, where I was assigned to a nursery for children diagnosed with autism. There, I developed an intervention that would change the course of my life and career. It was this intervention that led me to oxytocin research.

In 25 years of practice as a Child Psychiatrist, I had observed what I surmised were some very important biological underpinnings in mother-child affiliation and in inflammation. I treated a wide variety of children, from autism to severe behavioral dysregulation, to so-called “typically developing children” in my Mothering Center in Greenwich, CT. I concluded that the behavior of the children with their mothers in the original autistic nursery was remarkably like that of Nobel Laureate Niko Tinbergen's well-described approach-avoidance behavior in animals in conflict, for example choosing between two unfavorable options [4,5]. This was a surprise, because the children in the Mothering Center were not autistic. The very first observation I had made in the Einstein nursery was that the children with autism were clearly exhibiting avoidance behaviors: they pushed their mothers away, avoided eye contact, did not speak and sometimes even lashed out vocally and physically. The parents were unable to connect with their children. Their behavior toward the child was to either scolding, screaming or shutting down completely. And, here in my practice I was seeing the same kinds of behavior in ‘normal' families. In both sets of children, I came to realize that in addition to avoidance behaviors all the children were also exhibiting approach behaviors.

3. ‘Aha’ moment

This approach avoidance behavioral pattern I was observing in the full spectrum of children gave me my “aha moment' that changed me forever. Here is what happened.

On any given day, my assigned autistic patient would circle the program's large, enclosed play area, always keeping the mother in his peripheral vision (approach behavior). When the child came near the mother, she would reach out, he would jump away (avoidance behavior). Further, I noticed that the other children were focused on their mothers, too. I thought this phenomenon was significant and a potential opportunity to repair their relationship. One day, just after Mother's Day, I met with a child's mother. She told me that, because her mother had been involved in the treatment sessions, she decided for the first time in her life to take her mother a Mother's Day present. She said, “My mother grabbed me and held me close! I wanted her to let go immediately, and I wanted her to never let go”! I thought, aha! “Tinbergen's approach-avoidance conflict behavior [4].” I next thought, “If the mothers will take these children in their arms, the way the grandmother did on Mother's Day, and not let them go right away, it might help the child's ability to relate and develop. I had another mother try it with her child and within six weeks her child was talking and becoming empathic. Eventually, I developed some basic rules and protocol for this intervention.

• •The mother and child needed to be physically holding one another face to face for the duration of the session, which varied from 20 min to two or 3 h.
• •Both needed to fully release their own full range of emotions, whatever they were.
• •Both needed to tolerate the other's emotional release, whatever it was.

With this basic protocol, I started the treatment with two other families and saw the same kinds of results. After about six weeks, we began to see avoidance behaviors transformed into approach behaviors: the children made eye contact with their mothers, climbed into their laps without trying to get away, stroked their cheeks, listened, and spoke to them. The mothers were overcome with emotion and relief. I remember the first time a mother told me about her growing connection with her child, “I feel as though I just gave birth!” I thought, this is an oxytocin effect for sure!

Over 25 years in private practice treating behavioral, developmental problems in a wide range of patients— infant, toddler, preschooler, adolescent, parents, grandparents, male, female— I found that the basic rules for my dyadic parent-child emotional holding session that I had developed in the autistic nursery typically yielded the same dramatic results in improved gut function and improved behavior.

4. Life-altering decisions

At some point, in 1978, I decided to phone Nobel laureate Niko Tinbergen, who in 1973 had won the Nobel in Physiology or Medicine for his methods of observing animal behavior, including ‘pair bonding’. Niko and his wife and co-author Elizabeth (nicknamed Lies-pronounced ‘Lise’) Tinbergen had been studying autism by applying Niko's world-renowned method of animal observation to the behaviors of autism. At first, they expressed skepticism about my method. They thought their method, which involved remaining passive and non-threatening in the presence of the autistic child, was superior to my method, which involved encouraging the mother to express her full range of emotions directly to the child during a holding session. However, when they visited my Mothering Center and observed the dramatic phenotypical change in the children's behavior after a single session—especially how the children could socially engage and play nicely with their mother—they both changed their minds about my method. They adopted my approach into their practice. Niko and Lies devoted the rest his life to these ideas. Niko and Lies helped me disseminate my work to like-minded people internationally. In 1985, based on my ideas and method, they published Autistic Children –New Hope for a Cure (4). My ideas were the “new hope”. The Tinbergens' backing, which included helping me write my book ‘Holding Time’ [6], was crucial to my confidence in presenting my ideas to an international scientific and clinical audience. This collaboration endured until Niko died in 1988, and Lies shortly after.

My next breakthrough occurred after associating two different hormones with two distinct patient populations I was treating; autistic children and adopted orphans. To my surprise, I began to hear from adoptive mothers that they too felt as though they had just given birth or nursed, even if they had never given birth or nursed! I began thinking that after physically and emotionally connecting with their children, the mothers were experiencing rushes of oxytocin—the primary hormone of childbirth and nursing [7]. Hearing this changed my understanding of oxytocin's function. My first thought was that oxytocin must be extraordinarily powerful if adoptive or biological mothers felt that way. Since then of course many studies have supported the correlation between comforting interactions and OT increases in humans: Mother's voice [8,9] and skin-to-skin between mother/father and baby [10,11].

But here's where the findings became more interesting. The severely dysregulated children I was treating shared something else in common. They were riddled with gut problems—tummy aches, food aversions, constipation, diarrhea, or worse (i.e., inflammatory bowel disease or irritable bowel syndrome). If some of these psychiatric disorders were simply diseases of the mind, why did nearly all the children also present with severe GI issues? The link between their physical and behavioral symptoms seemed significant to me. I realized that this could not simply be a psychological problem. Somehow, there was a biological link or crossover between these two pathologies.

Remarkably, when the autistic or the adopted children connected emotionally with their mothers, not only did their behavioral issues improve, but their gut problems also went away. Positive mother-child behavioral interactions seemed to have a biological mechanism that decreased stress in both the child and mother and set the child on a normal developmental trajectory. I felt strongly that oxytocin was involved in physiological and behavioral changes, but what could account for gut changes? Then, I learned from the parents of one of my patients that their autistic child was being treated with the hormone secretin. They told me that a physician reported that three autistic boys he was testing with secretin for G.I. symptoms suddenly made eye contact and spoke [12,13].

At this point, I began to form the hypothesis for a biological mechanism underlying the efficacy of my treatment method. Exogenous administration of secretin improved gut function and improved behavior. Endogenous release of oxytocin through affiliation may improve gut function, feelings and behavior. Perhaps oxytocin and secretin might be acting synergistically to improve behavior and gastrointestinal function. And conversely, maybe behavior is affecting the release of the hormones. Both oxytocin and secretin might be released endogenously when the mother and child connected emotionally, thereby leading to simultaneous improved gut function and behavior. But would I ever be able to test my hypothesis?

Then, at my 25th medical school reunion, I ran into a distinguished research scientist, Harold Neu MD, who ran a neighboring lab where I developed the chloramphenicol assay as a student 25 years earlier. He told me, to my surprise, that the PI who took my lab book had just received a 25-year career award, based in part on my assay. My reaction surprised my friend. I wasn't mad. Instead, I was delighted. To me, it verified my capability to conduct basic research and gave me confidence to make the next big leap in my career.

5. From bedside to bench

The switch from clinical practice to a research career at Columbia University Medical Center highlighted both the benefits and the challenges of being a woman. I had witnessed some bias in medical school where I was one of 11 women amongst 127 men. In solo practice, I was not aware of this bias, but it turned out that navigating my academic research career required help all along the way. Importantly, there are a few notable women and men who helped me in my successful transition from the bedside to the bench.

The first woman to help me was a woman with remarkable political skills and vision, Anke Nolting, Associate Dean of Development and Alumni Relations at Columbia's medical school. Over the course of several decades, she conceived of and created many programs at Columbia by directing the paths of many researchers. I came to know Anke well through the medical school alumni organization. In the 1990s, she began to formulate a strategy to bring me back to Columbia to test and bring to practice my revolutionary ideas. Anke was convinced that I could find funding from donors based on my clinical work, and she persisted with that idea until it actually happened! Learning of the opportunity, a philanthropist friend said that he would give me the money to go back to Columbia. Thus, within a relatively short timeframe I found myself a member of the research faculty at Columbia University Medical Center.

Once at Columbia, Anke introduced me to her husband, David Ruggiero PhD, who was an expert on the visceral brain. David introduced me to his widely respected lab scientist, Muhammad Anwar. Anwar's knowledge and expertise in immunocytochemistry was unparalleled. Here was a once-in-a-lifetime chance to test my hormone hypothesis— that hormones such as oxytocin and secretin can overcome gut inflammation and behavioral symptoms— in an animal model.

We began by looking for secretin in the brain. Secretin was first identified in 1902 to be involved in the digestive process [14]. One hundred years later, it was still assumed the hormone was restricted to gut function. A connection between the gut and the brain was not widely accepted. With Anwar's help, we were the first to find secretin-producing neurons in the forebrain, an area responsible for diurnal changes in our body [15]. The established link between smell, emotional memory and our gut research supported my hypothesis that secretin and oxytocin would treat behavioral and developmental disorders, including autism [16]. This hypothesis is supported by my findings and by recent research demonstrating that oxytocin is produced and secreted from enterocytes in the intestinal epithelium in response to secretin stimulated by L. reuteri. Taken together, this work thereby identifies oxytocin as an intestinal hormone and provides mechanistic insight into avenues by which gut microbes promote host health [17].

It was clear to me from the start, however, that secretin and oxytocin might not be acting independently, largely because I had witnessed the powerful ameliorating gut-behavior effects in my early mother-child therapy. I had also demonstrated secretinergic and oxytocinergic cells enveloping the micro-vessels of the hypothalamus and sending processes into the lumen of the blood vessels [16]. They must be working together.

Together with David and Anwar, we devised an experiment to induce rodent gut inflammation. We treated three groups with the hormones. One group received secretin by itself. One group received oxytocin alone. A third group received secretin and oxytocin together. I hypothesized that the secretin-oxytocin combination would improve gut inflammation. There was no change in clinically relevant low-dose secretin or in the oxytocin cohorts. In combination, however, gut inflammation decreased significantly. The combination reduced the protein TNF-alpha, one of the main actors in conditions of inflammatory or autoimmune diseases. The combination also reduced the cytokine IL-6 in the amygdala.

Other research has shown that elevated maternal IL-6 predicts poor impulse control in the child, as well as increased IL-6 in the child [18]. This finding is thought to be related to the microbiome [19], which drew a direct line between gut and brain health. This connection was the basis upon which I was awarded a U.S. patent for the treatment autism and visceral inflammation with combined secretin and oxytocin [20].

In my quest to understand the biological role of oxytocin and secretin, I now had several clues. Oxytocin, a brain hormone, brought the mothers and children together. Secretin, which was thought to be a gut hormone, turned out to be a brain hormone, too. And, it treated inflammatory bowel disease in a rodent model, but only in the presence of oxytocin. When I put oxytocin and secretin together, it amplified their protective effects. When the secretin/oxytocin treatment was so effective in combination, I thought, there must be oxytocin receptors in the gut. At the time, it was still widely believed that oxytocin was cleaved by peptidases in the stomach. Therefore, oxytocin should be inactive and there should be no oxytocin receptors in the gut. To me, however, that made zero sense. By then, it was well known that milk contains anti-peptidases. Therefore, there must oxytocin receptors in the gut. They just haven't been found. Anwar whole-heartedly took up the challenge me find them.

Fortuitously, this task led me to the second woman who contributed to my oxytocin research; the world-renowned oxytocin expert C. Sue Carter (See article in this issue) whose research on oxytocin I had read [21,22]. When Lynne P. Hofer, a psychoanalyst, and wife of Myron Hofer MD, recognized the novelty and importance of my clinical ideas when she and Myron visited my Mothering Center in the 1990's, she urged Myron to help me. Myron is considered a dean of the field of mother-infant psychobiology. Myron called Sue and asked her to mentor me. Sue jokingly told me later that receiving an assignment from Myron was like getting a call from God. How could she refuse? Sue's help proved invaluable. She connected me to many helpful people, one of whom was Gloria Hoffman PhD, whom Sue had engaged 20 years earlier to develop an antibody to the oxytocin receptor. Gloria still had some of that antibody, and she generously shared it with me.

Anwar had tested commercially available antibodies to the oxytocin receptor without result. We were running out of time and money. Gloria's antibody was our last chance. One day, Anwar strutted into my office, his shaking hands holding a slide of stained gut tissue, chock-a-block full of oxytocin receptor fluorescent staining, lit up like a Christmas tree. (Fig. 1).

Fig. 1.

Oxytocin receptor expressed in the epithelium (gut mucosa) of the post-natal day 7 baby rat gut.

At that time, Mike Shelanski MD PhD, Chairman of Pathology and Cell Biology (now the Vice Dean of Research for Columbia University Medical Center) became a major mentor and supporter of mine. This was when the Department of Psychiatry leadership felt my gut research was not relevant to Psychiatry. When I presented the idea of establishing the BrainGut Initiative, Mike appointed me to his department and accepted my program into his department as well. Crucially, he defended me in an infamous debate between the leadership of the Department of Pathology & Cell Biology and the Department of Psychiatry over whether the gut has any influence on the brain. Needless to say, Psychiatry lost the debate. Mike was a supporter of mine and gave me very important advice, both in my research and politically.

With Mike Shelanski's support, I co-created the BrainGut Initiative with Michael Gershon, MD, a world-renowned expert on serotonin and the enteric nervous system. Anwar and I discovered that the oxytocin receptor in the gut mucosal lining. Mike and I found oxytocin receptors in the enteric nervous system and showed that the oxytocin receptor is developmentally regulated, maintains barrier function, opposes inflammation, and regulates the number and function of serotonergic neurons in the ENS. We also showed that the oxytocin receptor determined height of villi and depth of crypts, and relocated to the crypts at weaning, where it has a role in cell growth and proliferation [23,24].

Now that I knew the gut has oxytocin receptors, I wanted to know their function. That led me to pursue the oxytocin signaling pathway at the cellular level. I turned to Hadassah Tamir PhD, who had helped Mike Gershon with his early serotonin studies. Hadassah was one of the first women neuroscientists at Columbia. Together, we conducted some very informative basic cell research on oxytocin. Hadassah enlisted her friend Benjamin (Benny) Klein MD, PhD, an expert in cell signaling in a variety of tissues, including foundational research on breast cancer and breastmilk. With Benny and Anwar, I investigated oxytocin signaling in mucosal cells and in brainstem nuclei [[25], [26], [27]]. Together, we were able to demonstrate the anti-inflammatory effects at the cellular level that I had originally seen reflected in mother-child interactions in my clinical practice. We showed that oxytocin acts on the mTORC1 pathway, which interestingly has several signaling molecules with genes abnormal in autism [28].

We found that oxytocin modulates protein folding that occurs during protein synthesis along the mTORC1 pathway [25]. If proteins synthesize too fast, the misfolded proteins cannot fold quickly enough, causing apoptosis (cell death). We discovered when misfolded proteins start to accumulate, oxytocin slows the folding process, thus promoting cell-survival. In particular, oxytocin controls the rate-limiting step for cap-dependent translation. In other words, oxytocin plays an important role in modulating the expression of many genes that respond to endogenous or exogenous signals such as nutrient supply or stress.

Not only does oxytocin determine cell survival, but it also synchronizes mom and baby's metabolisms. When babies nurse, they ingest oxytocin from the milk and stimulate mom's body to make more oxytocin. Oxytocin/oxytocin receptor signaling is critical to nutrient absorption and gut physiology. In our studies, oxytocin appeared to determine the height of the infant villi—tiny, finger-like structures that increase the surface area of the gut [24]. The height of the villi is responsive to the available nourishment in the environment, although the adaptive response mechanism are complex. Some animal research has shown that if you restrict calories, the villi will shorten, while other research has shown that villi will shorten or lengthen based upon various medical conditions [29,30].

It is now well established that oxytocin promotes general well-being throughout the body. As we and others have shown, it acts as a neuro-protective agent by preventing neuro-apoptosis, neuro-inflammation, and neuronal oxidative stress, and by restoring mitochondrial function [[24], [25], [26],28,31]. And emotional connection increases the release of oxytocin. Thus, it's no wonder that meaningful social interactions can boost our moods, improve our behavior, and bolster our health [22]. It also appears to be true that not having meaningful emotional connections hurts our ability to produce beneficial hormones. In a study by my friend Lane Strathearn MD, Director of the Division of Developmental and Behavioral Pediatrics at the University of Iowa, mothers who experienced emotional neglect in their own childhoods showed significantly lower levels of oxytocin in their spinal fluid as adults [32]. This long-term effect speaks to the importance of connecting emotionally in childhood. The power of nurture could not be clearer. Emotional connection between mothers and children stimulates oxytocin pathways, promoting healthy growth and development over the life span [21].

Taken together, our oxytocin findings help explain the importance of metabolic coordination between mother and baby, which I think is a fundamental part of autonomic co-regulation. Facilitating this coordination between mother and infant promotes survival. It also promotes optimal development. And best of all—our studies show establishing family autonomic emotional connection in the family is accessible and free to everyone, no matter what background or life circumstances.

6. Back to the bedside

While I was pursuing oxytocin function, my donors urged me to translate our basic research findings back to the bedside. Understandably, they were most interested in helping families struggling with the behavioral and gut symptoms. At this point, our team, together with Anke Nolting and our chief benefactor, Jenn Hoos Rothberg (executive director of the Einhorn Family Charitable Trust), white-boarded a program that transformed the BrainGut Initiative in Psychiatry and Pathology & Cell Biology into the Nurture Science Program in Pediatrics.

At last, I was able to rigorously test my hypothesized connection between behavior and hormones, and between behavior and emotions, that I developed years before in the autistic nursery. I had no experience treating prematurely born infants. Nonetheless, based on the powerful effects I observed in older populations, I reasoned that connecting a mother and baby emotionally in the NICU would improve both short- and long-term physiological and behavioral outcomes.

7. Randomized controlled trials of Family Nurture Intervention

This brings me to my most important male collaborator, Michael Myers PhD, who had become Chief of Developmental Neuroscience Division in Psychiatry when Myron Hofer decided to step down. Michael was the perfect person to help me in our clinical research since he had a deep understanding of interactions between physiological and behavioral processes from his early animal research with Myron. Michael then brought into our collaboration Raymond “Buddy” Stark MD, a neonatologist who coincidentally had been my anatomy partner in medical school. Buddy and Michael had studied electroencephalography (EEG) in fetal sheep and in babies. Buddy was certain that we would not be able to show a difference in EEG in just six weeks of intervention. He was greatly surprised and delighted when we did. We went on to publish seven papers showing profound differences in the EEGs of the intervention infants by term-equivalent age [[33], [34], [35], [36], [37], [38], [39]]. Remarkably, we were able to replicate the brain findings in a multisite replication RCT at Columbia, and in a totally different setting at the University of Texas San Antonio, run by Distinguished Professor of Neonatology Alice Gong MD [35].

Our team conducted two randomized controlled trials over 9 years, comparing standard neonatal intensive care unit (NICU) care (SC) with SC plus Family Nurture Intervention (FNI). FNI included ∼6 weeks (24–36 h) of facilitated mother-infant ‘calming sessions’ aimed at ‘emotional connection’. My primary goal and objective with the newly born preterm infants was to connect them and their mothers emotionally during the NICU stay. To date, we have published over 17 papers comparing the SC and FNI group results. Compared to controls, FNI infants showed significant higher power on EEG, more term-like cortical network function, correlated with improvement in neurobehavioral functioning, autonomic health and development, FNI mothers showed significant improvement in depressive symptoms and the dyad scored higher on relational health at key assessment points through age 5. Key results showing improved FNI-treated infant outcomes were: Lower intervention-dose-related HR over NICU stay [40]; Better autonomic regulation at term age [1]; Increased cortical activity [37]; Better child social relatedness, language, cognition, attention and lower risk for autism at 18 months [41]; and better autonomic regulation in mothers and children at 4–5 years of age [42].

At this point in the follow-up of the trial cohorts, we can confidently say, on average, FNI infants had significantly better short- and long-term neurobehavioral functioning, autonomic health and developmental trajectories, and FNI mother-infant dyads had significantly better emotional relationships through the five year follow-up.

8. Independent validation and extension of our brain findings

Because of our EEG findings at Columbia, Sampsa Vanhatalo MD, PhD joined our group for a six-month sabbatical. Sampsa was developing innovative research techniques in brain science at the University of Helsinki, Finland, where he has created new methods of analyzing brain function using large EEG data sets. In a landmark study published in Science Translational Medicine [39], Sampsa first confirmed our prior EEG findings, an important step in validating our study results. Then, he and his team were able to show that our intervention significantly affects entire brain networks, and that those findings correlated with neurodevelopment at 18 months corrected age. Just as importantly, Sampsa's team compared the brain networks of babies born at term age to the networks of our study vs standard care infants. The intervention group and the full-term cohort were virtually identical. Our intervention was able to restore the infant's brain network function such that it was indistinguishable from a group of infants born at full term age. This was very exciting, for it suggests that FNI corrected the negative effects that prematurity has on preterm infant brain network function. It was a powerful conclusion: Sampsa's findings, together with our EEG and other findings, strongly support the idea that connecting infants and mothers emotionally at the time of birth has positive short-term benefits that are predictive of long-term outcomes.

9. The universal Welch Emotional Connection Screen (uWECS)

At the beginning of our clinical research, I began a collaboration with another remarkable woman in science, Amie Hane, PhD, Professor of Psychology at Williams College. Amie joined our faculty and our research group at Columbia as a Research Scientist. Amie is an expert in child stress-responding and the coding of mother-child behavior. With Amie, we validated the original Welch Emotional Connection Screen (oWECS), a behavioral assessment instrument that measures ‘emotional connection’. Importantly, whether the subject pair was emotionally connected or not connected correlated with simultaneous autonomic measures of heart rate and respiratory sinus arrhythmia (vagal tone) [43]. Amie has been part of all of our clinical research.

As luck would have it, Sampsa's spouse, Ulla Vanhatalo, who accompanied him on his NYC sabbatical, is a linguist with a PhD in synonymy, which is how native language speakers understand the tiny differences in meanings of synonyms. Ulla had conducted large studies looking at the differences in the meanings of words that convey feelings. Even though complex vocabulary is connected to our primary language, there are semantic primes that are common to all languages, i.e., ‘universal’. About 65 primes can be translated into every language, because they are the simplest way of expressing essential thoughts or emotions. Examples of primes are: you, me, same, other, good, bad, want, feel, see, hear, touch, live, die, etc. Regardless of our primary language, we all share these semantic primes. Ulla together with Cliff Goddard helped us translate the original WECS (oWECS) into semantic primes, and thus into a universal language WECS (uWECS) [44].

To illustrate the difference between the original and the universal WECS, the first behavioral modality ATTRACTION is described on the original oWECS form as follows:

Cozy proximity, Mutual gaze, Warm touch, Drawn to each other.

The same behavioral modality ATTRACTION is described on the universal uWECS form as follows:

MOM AND CHILD BOTH WANT TO BE VERY CLOSE. Child feels something very good because Mom’s body is touching his/her body. Mom feels the same. Child very much wants to be close to Mom. Mom very much wants to be close to child. Mom often looks at child's face. Mom feels something very good when she sees child’s face. Child often looks at Mom’s face. Child feels something very good when he/she sees Mom’s face. It is often like this for some time: Mom looks at the child’s eyes, at the same time child looks at Mom’s eyes. Mom feels something very good because of it. Child feels something very good because of it. Mom often touches child with her hands, she often touches child’s face with her face. When Mom does this, child feels something very good. At the same time Mom feels something very good. Child often touches Mom with his/her hands, child often touches Mom’s face with his/her face. When child does this, Mom feels something very good. At the same time, child feels something very good.

While the translation is considerably longer than the original, the “prime” language makes it possible to translate the description into any language in the world. In this way, the Universal WECS (uWECS) validates the universality of social emotions. At the same time serves as a rigorous research instrument and as a simple practical clinical guide for establishing and tracking family emotional connection.

10. Theoretical breakthroughs

As impressive as our clinical trial brain results were, I did not think the results were brain mediated. Our other findings confirmed another hypothesis emerging from our work— the physiological changes seen throughout the body, particularly the heart, are driven by external signaling to the autonomic nervous system, independent of the brain. Indeed, we showed that the intervention lowered heart rate [40] and led to more mature autonomic regulation at term age in preterm infants [1]. Both mother and preterm child had better physiologic regulation at the five-year follow-up, as measured by vagal tone [42]. This is very important, since both mother and child are at risk for cardiovascular disease. Later, we were able to show similar statistically significant vagal tone results in a preschool population.

Just when I was starting my research at Columbia, I found a work and life partner, Robert Ludwig, who was primed for an equal collaboration with me. He spent 15 years researching an historical basis that would explain why some professionals took such strong issue with my ideas and my clinical work. After reading scores of papers and books in physiology and psychology going back 400 years to Descartes, he concluded that my ideas challenge some of the longest-held beliefs and theories in the science of behavior [45]. We began publishing a new theory that begins with novel assumptions and explains emotional behavior from my perspective [46,47].

Since my earliest clinical work, I have looked at family emotional behavior from a very different perspective from that of conventional psychiatry, which aims at child independence and self-regulation. To my astonishment, when I started treating preterm infants, the conventional goal of NICU care was to help the preterm infants self-regulate! I could not imagine how an infant that was conditioned by continuous co-regulation within the mother's womb could self-regulate outside the womb. I knew from my clinical work treating children with autism and adopted orphans that the most effective way to change symptomatic emotional behavior and physiologic dysregulation was to help the child and parent co-regulate one another's emotional and autonomic state.

Robert realized my work challenges three pillars of conventional theories on mother-infant emotional behavior: 1) MEMORY: How babies learn [48], 2) SENSORY PATHWAY: How environmental signaling between mother and infant is communicated through the nervous system and consummates in socio-emotional behavior [49], and 3) CONTROL OF EMOTION: How mother-infant socioemotional behavior is controlled [45]. We built a new theory based on new assumptions, and a Copernican-sized paradigm-shifting model has emerged. Copernicus placed the sun and not the earth at the center of our celestial orbit. Similarly, our theory and findings place the autonomic nervous system and not the brain at the center of our socioemotional orbit.

Our autonomic theory of emotions posits that the emotional relationship between mother and fetus starts in utero by way of autonomic conditioning [48]. Following birth, the mother-infant relationship develops via two types of conditioning: conditioning of the central nervous system and brain and unconscious conditioning of the autonomic nervous system. Conventional theory states that the brain controls emotional behavior in a top-down executive manner through internal feedback mechanisms in the individual [45]. In my experience, infants, especially preterm infants, cannot self-regulate. Self-regulation eventually devolves into dysregulation. A fact that surprises most people is that 90 % of the information carried by the vagus nerve, which forms a feedback loop between gut and brain, is transmitted ‘up’ from the gut to the brain. Only 10 % of the traffic is down from brain to the viscera. For optimal function, the brain needs to be receiving ‘homeostasis’ or ‘all-is-well’ signals from body. The fact that the anti-inflammatory hormone oxytocin is so prevalent and important in the gut and heart, as well as in the visceral brain, is further evidence that brain function in many ways depends on visceral function. We showed copious oxytocin receptors in the nodose ganglion, a vagal relay station between the viscera, including gut and heart, and the visceral brain [23].

The various findings of our research efforts support an alternative learning model for innate or instinctive emotional behavior. Socioemotional behavior of infants and preschool-aged children is not based on thinking. Rather, it is based on a primitive ‘innate’ reflex, which Pavlov called the ‘social’ or ‘cardiac’ reflex. Based on his observations in dogs, Pavlov named the phenomenon ‘effect of person’. In his experiments, the heart rate of dogs was subject to external ‘control’ (calming) when the owner/trainer maintained physical contact (petting). Pavlov's human-animal effect was later shown to also exist between two humans. This research led to our theoretical advance. Simply stated, an individual's heart rate (emotional state) is subject to external ‘control’ (calming) when the individual is in direct contact with another trusted individual. For example, an autonomically conditioned memory of a beloved grandparent can trigger an approach response and lower heart rate when they come together. An autonomically conditioned memory of an abusive parent or grandparent can trigger an avoidant response and higher heart rate. From Pavlov, we learned that in addition to conscious ‘internal’ cortical control, emotions and heart rate are subject to unconscious ‘external’ autonomic memory control through interaction with another person [50,51].

In our theoretical model, social signals are relayed to the primitive brainstem. It is generally accepted that the signals from the brainstem are transmitted via the reticular activating system upward to the higher areas of the brain. At the same time, it is largely ignored that signals are transmitted more rapidly downward via the gigantocellular reticular nucleus, which is made up of giant single cell neurons, to the medulla oblongata at the very base of the brain stem and thence directly to the heart via the vagus nerve [49]. As the signals pass through the giant cells, they are ‘activated based on conditioned cellular memory. In the case of an ‘approach’ reflex, for instance, these changes lead to unconscious neuronal reflexes that trigger hormonal release in the heart via intrinsic inflammatory or anti-inflammatory (among other functions) oxytocin/oxytocin receptor system in the aortic arch [52,53]. The heart then circulates the hormones throughout the body in the blood and initiates the corresponding behavioral responses.

Theoretically, changes in the hormone levels affect the baby's orienting reflex, a developmentally critical infant behavior. During the orienting reflex to the mother, the baby's heart and lungs react quickly, via the autonomic (automatic) branch of the body's nervous system. Hormones, including oxytocin, spread through the brain and body through circulation in less than a second [54]. By comparison, it takes a number of seconds for the ascending signals from the brainstem to reach the cortex. This causes one of two autonomic state reactions, depending on prior autonomic conditioning in the relationship: approach or avoidance.

As noted above, there is much research showing that oxytocin levels are linked to social encounters between parents and children. New research among preterm infants is showing that skin-to-skin contact helps to ameliorate the hormonal stress in preterm infants by activating the oxytocin release with simultaneous reduction of cortisol secretion [10]. Our randomized control trials revealed significantly lower heart rate over the NICU hospitalization in the intervention group. However, there is no consensus on the mechanisms underlying the phenomenon. The gut-brain axis continues to be the main area of interest and focus of research. Gut microbes are capable of producing most neurotransmitters found in the human brain. And evidence is accumulating to support the view that gut microbes influence central neurochemistry and behavior [55]. Manipulating the gut microbiota with psychobiotics, prebiotics, or even antibiotics offers a novel approach to altering brain function and treating gut-brain axis disorders, such as depression and autism.

Aside from these promising advances, however, I believe the field of medicine needs to focus more on the natural social mechanisms involved in the control of oxytocin levels. To that end, our calming cycle theory offers a simple and straight-forward explanation of how social interactions within the family can condition natural release of oxytocin and other hormones to promote adaptive behavior and well-being [45].

An emotionally connected encounter with a mother triggers a ‘calming’ reflex. Heart rate and breathing slow. The body returns to homeostasis. On the other hand, an emotionally disconnected encounter triggers an avoidance reflex. Our randomized control trials revealed significantly lower heart rate over the NICU hospitalization in the intervention preterm infants [40], higher vagal tone in both mother and preterm child long-term [42], as well as significantly higher vagal tone in preschool children after an intervention to establish emotional connection.

11. The tough road behind me

In a significant way, being a woman has been an advantage in understanding mother-infant and mother-child interactions. As a woman and mother, I can see emotional behavior through a lens that is very different from the way men view such behavior. I have experienced the biology of co-regulation between mother-fetus and mother-infant and mother-child, which I saw extended to father-infant, parent-parent and grandparent-parent, and grandparent-grandchild interaction.

Robert contends that behavioral sciences have for too long been dominated by men and male thinking, which has prevented therapeutic breakthroughs seen in other scientific disciplines. In any case, women are bringing a different perspective to science and medicine and asking new questions and testing new hypotheses. This is good for science, and this is good for society.

What have we accomplished on behalf of families to date? We have shown that emotional connection, unlike individual attachment traits or styles, is an autonomically conditioned state between two people. The state is conditional on the relationship of the two people (e.g., mother and child), and the state is conditionable. I believe the conditional nature of the state provides a clinical breakthrough, in that autonomic state can be used to predict long-term development. And, since autonomic state is malleable and plastic, it can be changed slowly over time or, as we have shown, relatively quickly. A traumatic event in in-fancy or childhood can have life-long effects, but the effects can be counter conditioned by emotional connection. At any moment, an emotional connection can occur through mutually expressed deep emotion, particularly when it involves mothers with infants and young children, whose survival instincts are powerful and conditionable. We have shown that our mother-infant FNI, which is focused on emotional connection, profoundly improved autonomic regulation in the baby [1,[33], [34], [35],[37], [38], [39], [40]]. We have shown that the change can alter developmental trajectories of infants and children through age five. A study of the mothers, as well as of preterm infants, confirmed that the FNI intervention benefits both mother and child [42]. We have extended our mother-child intervention to an older preschool population and have confirmed our physiologic regulation hypothesis [41,42,56]. We have advanced knowledge of oxytocin in the gut-brain connection.

We have a new theory of mother-child emotional behavior. We have defined the mediating variable that leads to emotional connection to be deep mutual expression of emotions. We have begun to reveal the biological underpinnings of the beneficial anti-inflammatory mechanisms of mother-child emotional connection, especially oxy-tocin/oxytocin receptor signaling in the gut and brain, adding to our understanding of oxytocin and cardiac function. We have an assessment tool for measuring emotional connection which correlates with autonomic state. We have shown that the behaviors of the original Welch Emotional Connection Screen (oWECS) [43] correlate with autonomic physiology; so not only is it a behavioral measure, but it also reflects physiological changes in a pair during emotional connection and physiological co-regulating [43]. We have created a universal version of the Welch Emotional Connection Screen (uWECS) that can be easily translated into every language in the world 44 and can be used as a therapeutic guide [44].

12. Financial support and more

Being a woman and representing paradigm-shifting ideas was certainly not an issue with my private funders, both women and men. I am profoundly grateful for the support of Albert and Kathy Gordon, Fleur Fairman and Tim Wallach, Kathy Emmett and David Golub, Morgan and Tara Rutman, John and Rainy Erwin, Mary Stephenson, Mary Catherine and Don Huffines. My largest, most important source of funding, which has allowed me to accomplish an unusually broad span of investigation from molecular to animal to human research, including testing treatment methods and developing novel theories, was from David Einhorn and his private philanthropic organization, the Einhorn Collaborative. Without the past 15-year close strategic and thought partnership with the Einhorn Collaborative's executive director, Jenn Hoos Rothberg, and lead strategist on bonding, Ira Hillman, the fundamental evidence and myriad discoveries would not have been possible.

13. Closing thoughts

With the help of a great many people I have been able to overcome barriers and show the benefits of being a woman in science. All my collaborators, staff, patient-families, funders, as well as my friends and family, my son Bram Welch-Horan MD, my incredibly supportive mother and father and sister and brother-in-law have played and continue to play a role in that journey. These efforts have resulted in a body of work across multiple fields supporting the pioneering ideas I came to Columbia to test. I am especially indebted to my friend Sue Ann Power, who helped in the preparation of this manuscript and who has provided feedback, editing, bold encouragement and wise counsel throughout my clinical and research careers. My life-long friend Judith Vogel has been with me in this work the entire 50 years, starting with creating the conditions that led to my first insights. I must also acknowledge the contributions to my thinking, presentations and communications by my dear friend Jim Chesnutt and his team. As of now, the accomplishments are more than I could have ever expected.

Looking back, I repeatedly ask myself would I change anything, if I could do it all again? A little voice in me says I would not do it at all. It took so much determination and tenacity, and the cooperation and contribution of so many amazing people. I have a profound sense of gratitude for the help of others, including my patient-families. I don't see how I could do it again.

Yet, I don't feel my work is finished. The motivations that drove me in the past are the same that drive me today, the childhood drive that inspired me over seven decades ago when I was a three-year old—a desire to help other people overcome their upsets. Now, I am excited by the opportunities our new non-profit organization, the Martha G. Welch Center (emotionalconnection.org), will bring. I hope to provide help directly to all families in their homes on their mobile devices through our new emotionalconnection app and website.

In closing, I wish to convey my sense of fulfillment from this 50-year fantastic journey into clinical practice and research, and I hope that my story motivates young people, especially women, to pursue their own fantastic voyage in medicine and science.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.