Father Trials: Hormonal Experiments on Prenatal and Postnatal Parenting

In animals, parenting is under strong hormonal control, and it would be naive
to suppose that this is very different for human parents. Indeed, several
studies in humans have shown associations of oxytocin and vasopressin levels
with parent-child interaction (Apter-Levi, Zagoory-Sharon, & Feldman, 2014;
Atzil, Hendler, Zagoory-Sharon, Winetraub, & Feldman, 2012; Bick & Dozier,
2010; Feldman, Gordon, Schneiderman, Weisman, & Zagoory-Sharon, 2010).
Furthermore, experimental manipulation of oxytocin and vasopressin levels
through nasal administration affects parenting behavior and the neural
processing of infant signals (Cohen-Bendahan, Beijers, van Doornen, & de
Weerth, 2015; Naber, van IJzendoorn, Deschamps, van Engeland, &
Bakermans-Kranenburg, 2010; M. M. Riem et al., 2011) .
We propose to employ a randomized, double-blind, placebo-controlled
within-subject design to gain insight into the hormonal and behavioral dynamics
of the paternal role in a critical phase of parenthood: the transition to
having the first baby. The focus lies on the 50% of parents who have tended to
be neglected in research and *until recently- in family policies: fathers. In
most western countries fathers have increased their participation in parenting
over the past decades. And even though in most families the participation of
fathers in child rearing is modest, this does not mean that their parental role
is irrelevant for child development (Kok et al., 2015), or that this role is
carved in stone and resistant to change.
A special focus is on a dimension of parenting that has received
considerable attention in animal research but, despite its evolutionary
importance, not in studies on humans: the role of the parent as protector.
Protection is a crucial aspect of human parenting. This is perhaps demonstrated
most convincingly when we are confronted with the absence of parental
protection, i.e. neglect, or child physical abuse. Neglect has the highest
prevalence of all categories of child maltreatment and has serious consequences
for many domains of child development (Gilbert et al., 2009). Physical abuse
has its peaks in early childhood, and infant crying is a documented trigger of
early physical abuse (Barr, Trent, & Cross, 2006).
In a seminal paper, Shelley Taylor and her colleagues proposed the
tend-and-befriend model as an alternative to the fight-or-flight model of
behavioral responses to stress (Taylor et al., 2000). Tending, the protection
and care of offspring, and befriending, the formation and maintenance of
interpersonal relationships with conspecifics, were proposed as strategies that
females use in times of stress to defend themselves and their offspring. A
central role in this model is attributed to oxytocin, which provides the
neuroendocrinal basis for affiliation with social groups. A second hormone that
may play a role in protective fathering is vasopressin. In men, vasopressin
levels have been associated with aggressive behavior (Coccaro, Kavoussi,
Hauger, Cooper, & Ferris, 1998). Moreover, while testosterone has been
associated with aggressive behavior in general, vasopressin has been suggested
to be associated specifically with protective aggressive behavior (van Anders,
Goldey, & Kuo, 2011). Indeed, several studies have reported associations
between vasopressin levels and parental aggression towards a male intruder
(Nephew & Bridges, 2008; Nephew, Byrnes, & Bridges, 2010).
In research on parenting quality, parental sensitivity is a key
construct. It refers to the ability to attend to infant signals and to respond
promptly and appropriately (Ainsworth, Bell, & Stayton, 1974). Studies
consistently show that fathers are less sensitive towards their infants and
toddlers than mothers (Barnett, Deng, Mills-Koonce, Willoughby, & Cox, 2008;
Hallers-Haalboom et al., 2014; Schoppe-Sullivan et al., 2006; Volling,
McElwain, Notaro, & Herrera, 2002), although seeing infants has similar
motivational salience to men and women (Parsons et al., 2011). Similarly to the
pattern of associations for mothers, higher levels of paternal sensitivity
predict more favorable child outcomes (Lewis & Lamb, 2003). Sensitive parenting
starts with the processing of infant signals, which has been shown to be
affected by oxytocin levels (M. M. Riem et al., 2011), but may also be affected
by vasopressin levels. These hormones may play a role in fathers* processing of
infant signals as well, but the direction of the association is as yet unclear.
Mapping parental brain responses to infant stimuli using fMRI has
increased our knowledge of brain processes involved in parenting sensitivity.
Brain regions expected to be important to parenting are circuitries related to
(1) arousal/salience (amygdala, ventral striatum), (2) reflexive care
(hypothalamus), (3) emotion regulation (insula, medial prefrontal cortex,
anterior cingulate cortex), and cognitive / empathic processing (insula,
inferior frontal and orbitofrontal gyri, temporoparietal junction) (Parsons,
Young, Murray, Stein, & Kringelbach, 2010; Swain et al., 2014). Exposing
parents to child stimuli in fMRI studies activates neural systems involved in
these regions. Effects of oxytocin and vasopressin on the amygdala and the
insula, medial prefrontal cortex, and inferior frontal gyrus have been
established in females (Atzil et al., 2012; M. M. Riem et al., 2011).
Replication in fathers and fathers-to-be is badly needed.
For protective parenting responses, the amygdala may play a central
role. The amygdala functions as an *alarm* to relay signals of threat. Infants
are rewarding attachment-objects, motivating parental care and attention, which
intensifies protection of the child against potential threats (Szechtman &
Woody, 2004). Moreover, oxytocin intensifies the reward associations of the
infant by increasing the release of opiates (Depue & Morrone-Strupinsky, 2005),
and higher reward triggers heightened distress and protection in the presence
of threat.
Connectivity analyses focusing on circuits involving coordinated
activity of distant brain areas are important for examining how brain regions
work in concert in response to infant stimuli. Bos, Panksepp, et al. proposed a
neural model for hormonal effects on connectivity, according to which oxytocin,
by acting on the amygdala, shifts activity to prefrontal regions, whereas
vasopressin activates connectivity with the brainstem (Bos, Panksepp, Bluthe, &
van Honk, 2012). In the current proposal we will test this model adapted to the
processing of infant cry signals and protective parenting. In parents, their
own infants* crying elicits more brain activation than the sounds of unfamiliar
babies (Parsons et al., 2010; Swain et al., 2014), with mothers showing a more
intense level of activation than fathers.
In the proposed series of RCTs fathers will be observed prenatally and
postnatally, both with their own child (postnatally) and an infant simulator
(prenatally), and using experimentally manipulated infant stimuli (prenatally
and postnatally). We will test the effects of hormone administration (oxytocin
and vasopressin) on the processing of infant crying, on protective parenting,
and on the quantity (involvement) and quality (sensitivity) of father-child
interaction. Testing the effects of hormone administration on the processing of
infant crying, protection, involvement, and sensitivity may help unravel the
mechanisms of effective parenting.

In a series of randomized control trials (RCTs) the following hypothesis will
be tested: Intranasal administration of oxytocin and vasopressin affect neural
and behavioral responses to infant signals and threat to the infant.
* Oxytocin and vasopressin modulate activation in the *parental brain* systems
related to arousal, reflexive care, and cognitive/empathic processing.
* Oxytocin will increase connectivity of the amygdala with prefrontal regions,
and vasopressin will increase amygdala connectivity with the brainstem.
* Oxytocin will promote an affiliative response to threat to the infant, and
vasopressin will promote an defensive response.

We will employ a randomized, double-blind, placebo-controlled within-subject
design. The within- subject experiments will be conducted in three phases of
parenthood: prenatal (child*s age = -3 months), early postnatal (child*s age =
+2 months), and late postnatal (child*s age = +7 months). In each phase of
parenthood, 35 40 fathers (30 plus 15% oversampling to compensate for
attrition) will visit the Leiden University Medical Center (LUMC) for three
experimental sessions, which take place with intervening periods of 1 week.

All within-subject experiments include three conditions: intranasal
administration of 24 IU of (1) oxytocin, (2) 20 IU vasopressin, and (3) a
placebo. The three conditions imply six possible counterbalanced orders of
conditions, and assignment of participants to order of administration will be
random. Administration will be double-blind.

The data collection is expected to take approximately 1.5 years. The study will
be conducted at the fMRI facilities at the LUMC.
Before the start of the study, we will conduct a pilot study. The research
methods will be similar to those described in this protocol, with two
exceptions. First, instead of three conditions (oxytocin, vasopressin,
placebo), the pilot will include only two conditions: vasopressin and placebo.
We will focus on vasopressin because our research group already has ample
experience with oxytocin administration studies. Second, instead of three
phases of parenthood (prenatal, early postnatal, late postnatal), the pilot
will include only one phase of parenthood: prenatal. Specifically, we plan to
invite 26 fathers expecting their first baby. The pilot study is expected to
take approximately three months.
In order to assess the effects of the transition to fatherhood on
neurobiological and behavioral responses, the pilot sample will be invited for
a postnatal follow-up session when their child is 4 months of age. This
follow-up session will be similar to the placebo pilot session, with
adjustments in accordance with the postnatal phase. Follow-up of the pilot
study is expected to take approximately 2 months.

Fathers will visit the LUMC for three experimental sessions: administration of
(1) oxytocin, (2) vasopressin, and (3) a placebo. Sessions will take place with
intervening periods of 1 week. Three types of measurements will be conducted
during each session: (1) hormonal (assessment of oxytocin, testosterone,
vasopressin, and estradiol levels in salvia and blood), (2) neural (fMRI:
processing infants signals and processing threat to infant tasks), and (3)
behavioral (handgrip during infant cry, quantity of care, quality of care, and
protection tasks).The three conditions (oxytocin, vasopressin, placebo) imply
six possible counterbalanced orders of conditions, and assignment of
participants by the LUMC pharmacist to order of administration will be random.
Administration will be double-blind: Neither the subjects nor the investigator
conducting the experiment will know which of the substances is administered.

There are no risks associated with the assessments used in this study. Possible
side effects of oxytocin and vasopressin are negligible. No adverse effects
have been reported in participants/patients undergoing MRI at the currently
available field strengths. Once we understand the neurobiological underpinnings
of good-enough and poor parental sensitivity and protection, better attempts
can be made to improve parenting and reduce the adverse effects of poor
parenting. The proposed set of studies is ground-breaking in that it includes
paternal protection, an important dimension of parenting that has been
neglected in all imaging studies and virtually all behavioral studies of
parenting to date, maybe because of the almost exclusive focus on mothers.
Furthermore, real-time measures using apps and mobile phones have been
underused in research and may provide data with high ecological validity, that
may differ considerably from data collected retrospectively using traditional
questionnaires on time spending. Thus, the importance of the benefits gained
from this research outweighs the minimal risks involved.