Salivary Cortisol and Behavioral State Responses of Healthy ... - JOGNN

JOGNN

RESEARCH

Salivary Cortisol and Behavioral State Responses of Healthy Newborn Infants to Tactile-Only and Multisensory Interventions Rosemary C. White-Traut, Dorie Schwertz, Barbara McFarlin, and Joseph Kogan

Correspondence Rosemary C. White-Traut, PhD, RN, FAAN University of Illinois at Chicago, College of Nursing, 845 South Damen, Chicago, IL 60612. [email protected]

ABSTRACT Objective: To compare changes in stress reactivity (measured via the biomarker salivary cortisol) and behavioral state in healthy newborn infants immediately following 1 of 2 interventions: (1) tactile-only stimulation or (2) a multisensory, auditory, tactile, visual, and vestibular stimulation with a control group. Design: A randomized prospective design pilot study. Setting: Normal newborn nurseries of 2 midwestern perinatal centers.

Keywords sensory intervention salivary cortisol stress reduction in the neonate massage healthy infants

Participants: Forty healthy newborn infants receiving standard nursing care. Methods: Infants were randomly assigned to receive 15 minutes of tactile-only, auditory, tactile, visual, and vestibular, or no stimulation 30 minutes before feeding. Saliva samples were collected before, immediately following, and 10 minutes postintervention. Behavioral state was judged every minute. Results: Tactile-only group infants had the largest increase in cortisol levels, followed by control group infants. In contrast, infants who received the multisensory intervention showed a significant steady decline in cortisol. Asleep was the predominant state for all 3 groups and cry was minimal. Conclusions: Tactile-only stimulation may increase infant stress reactivity while the benefit of the multisensory auditory, tactile, visual, and vestibular intervention may be in the reduction of infant stress reactivity. Interventions appeared to have minimal effect on stress reactivity based on behavioral state.

JOGNN, 38, 22-34; 2009. DOI: 10.1111/j.1552-6909.2008.00307.x Accepted September 2008

Rosemary C. White-Traut, PhD, RN, FAAN, is a professor and department head in the Department of Women, Children and Family Health Science, College of Nursing, the University of Illinois at Chicago. Dorie Schwertz, PhD, RN, FAAN, is an associate professor and director of the Core Biologic Laboratory, Department of Biobehavioral Health Science, College of Nursing, the University of Illinois at Chicago. (Continued)

22

levated cortisol, a measure of stress reactivity, has been found to detrimentally a¡ect brain development (Chugani et al., 2001), whereas decreased stress reactivity increases infant learning and memory (Thompson & Trevathan, 2008). To counteract negative outcomes, maternal/caregiver behavior (talking, touching, eye-to-eye contact, and rocking) facilitates infant self-regulation and soothes infants, especially during times of stress (Braarud & Stormark, 2006; Jahromi, Putnam, & Stifter, 2004). Because cortisol assays have been validated as a noninvasive measure of stress reactivity with infants (Grauer, 1991;Jahromi et al. Kawakami, Tomonaga, & Suzuki, 2002; Kurihara et al.,1996; Spangler & Scheubeck,1993), this assay has been used to explore potential use of various behavioral interventions for stress reduction during painful and other stressful procedures (Braarud &

E

Stormark; Felt et al., 2000). The purpose of this pilot research was to compare changes in stress reactivity (measured via the biomarker salivary cortisol) and behavioral state in healthy newborn infants who received one of two interventions: a tactile-only stimulation (massage without additional human social interaction) or a multisensory auditory, tactile, visual, and vestibular (ATVV) stimulation with a control group who did not receive additional stimuli. In prior research on clinically stable hospitalized preterm infants it was found that in contrast to tactile-only stimulation, which leads to excessive increases in heart and respiratory rates and changes in behavioral state, ATVV produces positive autonomic and behavioral responses (WhiteTraut, Nelson, Silvestri, Cunningham, & Patel, 1997).

& 2009 AWHONN, the Association of Women’s Health, Obstetric and Neonatal Nurses

http://jognn.awhonn.org

RESEARCH

White-Traut, R. C., Schwertz, D., McFarlin, B., and Kogan, J.

In the current study the researchers sought to expand these previous ¢ndings by examining responsiveness to the ATVV intervention in healthy newborn infants receiving standard nursing care and by adding the measurement of salivary cortisol to quantify infant stress reactivity. With the use of animal models, researchers have shown that both the early environment and maternal behavior contribute to the development of the pituitary/adrenal response to stress. This response programs the nervous system, thereby in£uencing how the animal will respond to future environmental stressors (Caldji, Diorio, & Meaney, 2000; Francis et al., 1996). Sensory deprivation and maternal separation have been linked to an increased magnitude of stress response and vulnerability for stress-related illness later in life (Caldji et al.). Even brief maternal separation induces increased cortisol levels in guinea pig pups (Hennessy & Moorman, 1989). Events in the early postnatal environment, such as maternal separation, lack of handling, and painful procedures cause a downregulation of glucocorticoid receptor genes, which in turn may alter sensitivity to stressors (Ladd et al., 2000). Handling has been shown to directly alter the number of glucocorticoid receptor sites per cell in the hippocampus (Meaney & Aitken, 1985). In addition, newer evidence suggests that alteration in gene expression in response to maternal separation is resistant to subsequent environmental in£uences, thus the e¡ects may be irreversible (Francis, Diorio, Plotsky, & Meaney, 2002). In the human literature, researchers examined the e¡ect of additional sensory and behavioral interventions to reduce stress reactivity in preterm and full-term infants. Auditory stimulation using a maternal heartbeat was found to reduce cortisol after a heel stick when compared with infants who received either no additional auditory stimulation or the sound of a beating drum (Kurihara et al., 1996). In addition, vestibular stimulation with and without maternal vocalization has been used to soothe infants and reduce crying, which is associated with elevated cortisol levels (Jahromi et al., 2004). As previously mentioned, healthy preterm infants receiving sensory interventions have shown positive autonomic and behavioral outcomes (White-Traut et al.,1997). Auditory only and vestibular only stimuli produce minimal heart rate changes with little alteration in sleep, while moderate touch (tactile-only) stimulation produces larger increases in heart and respiratory rates (50% and 65%, respectively) above the limits of normal and also induces more rapid yet short-lived changes in behavioral state

JOGNN 2009; Vol. 38, Issue 1

(from sleep to awake). In contrast, the ATVV (auditory, tactile, visual, and vestibular) intervention yields smaller physiological and yet more gradual prolonged changes in behavioral state (White-Traut et al.). In addition, Korean full-term infants 14 days postbirth housed in an orphanage who received the multimodal ATVV intervention twice daily, 5 days per week for 4 weeks showed improved growth and fewer illnesses when compared with controls (Kim, Shin, & White-Traut, 2003). Other reports document bene¢cial changes in behavioral state in response to di¡erent types of tactile stimulation including massage, combinations of sensory stimuli, and skin-to-skin contact (kangaroo mother care) in full-term and premature infants (Vickers, Ohlsson, Lacy, & Horsley, 2004; White-Traut, Nelson, Silvestri, Patel, & Kilgallon, 1993; White-Traut, Nelson et al., 2002; White-Traut, Studer et al., 2002). Behavioral interventions involving a combination of maternal soothing and auditory, visual, and vestibular stimuli also have been reported to reduce salivary cortisol during immunization (Felt et al., 2000). Furthermore, the cortisol level of infants receiving more sensitive and cooperative maternal behavior following a bath, returned to normal at a faster rate than infants receiving lower quality maternal behavior (Albers, Riksen-Walraven, Sweep, & de Weerth, 2008). Mixed results have been found in previous research on unimodal infant massage. Jutapakdeegul, Casalottia, Govitraponga, and Kotchabhakdia (2003) showed that postnatal tactile-only stimulation in the rat animal model increased corticosteroid levels. Furthermore, following tactile/kinesthetic stimulation (massage), preterm infants experienced increased cortisol, dopamine, norepinephrine, and epinephrine levels (Kuhn et al.,1991). In contrast, researchers examining the e¡ects of infant massage found that infants who received moderate-pressure massage are more likely to gain more weight, measure longer in length, and show better orientation and less excitability and depression on the Brazelton scale than infants who received light-pressure massage (Field et al., 2004). For infants of depressed mothers, massage therapy resulted in decreased salivary cortisol levels and an increase in active alert and active awake states, along with a decrease in crying, compared with infants who were rocked (Field, Grizzle, Sca¢di, Abrams, & Richardson,1996). In addition, infants receiving deep massage were able to discriminate between test and habituation stimuli while the control group failed to discriminate (Cigales, Field, Lundy, Cuadra, & Hart, 1997).

23

RESEARCH

Salivary Cortisol and Behavioral State Responses to Interventions

This pilot study compared changes in salivary cortisol and behavioral state in healthy newborn infants who received one of two sensory interventions. Furthermore, in a Cochrane review of infant massage, authors reported that based on a small number of studies, infant massage may positively e¡ect stress hormones and improve infant behavior and mother^infant interaction (Underdown, Barlow, Chung, & Stewart-Brown, 2006). Unfortunately, it is di⁄cult to interpret the above ¢ndings because the authors did not indicate whether infants also received human social interaction in conjunction with the massage protocols. The authors speculate that massage is rarely given without human social interaction such as eye contact and talking with the recipient. Although the underlying mechanism for these ¢ndings remains unknown, these results suggest that a combination of sensory stimuli may produce more subtle autonomic responses, be more e¡ective in facilitating positive behaviors, and may reduce stress reactivity in infants. This pilot study compared changes in salivary cortisol and behavioral state in healthy newborn infants who received one of two sensory interventions. It was hypothesized that compared with the control group receiving no intervention, the ATVV intervention would reduce stress reactivity as measured by salivary cortisol, while the tactile-only intervention would increase stress reactivity. Infants assigned to the control group would experience normal small variations in stress reactivity. It was further hypothesized that infants in both interventions would experience a change in behavioral state from sleep to alertness while infants in the control group would remain asleep.

Method Setting Barbara McFarlin, PhD, RN, CNM, RDMS, is a research assistant professor in the Department of Women, Children and Family Health Science, College of Nursing, the University of Illinois at Chicago.

The research was conducted at two midwestern perinatal centers. The infants were admitted to the normal newborn nurseries and cared for on the respective mother^baby units. The Institutional Review Boards at the university and the two clinical sites approved the research protocol. The research sta¡ obtained parental informed consent before enrollment.

gestational age as assessed by the resident physician at birth (Ballard et al., 1991) between 36 and 41 weeks. Infants were excluded if they were receiving antibiotics, phototherapy, intravenous therapy, oxygen therapy, or any other treatment other than standard nursery care. Infants also were excluded if they had documented or suspected chromosomal anomalies or a positive drug screen. Mothers had to be considered low risk based upon the Obstetrical Complications Scale (Littman & Parmelee, 1978). It was anticipated that by enrolling infants 24 hours postbirth, the e¡ects of analgesia and length of labor would be minimized. Descriptive statistics, one-way analyses of variance, two-tailed tests, and chi-square statistical techniques were conducted to determine the equivalency of the three groups before intervention.There were no di¡erences among the intervention groups in terms of birth weight, gestational age at birth by Ballard examination and by dates, gestational size, infant length, head circumference at birth, infant Apgar scores, infant postnatal complications (Littman & Parmelee), infant sex, maternal race, maternal parity, maternal age, and obstetric complications (Littman & Parmelee; see Table 1). All infants were delivered vaginally, with a mean age of 38.9 weeks gestation at birth and were clinically stable at birth with a mean Apgar score of 8.5 at

Table 1: Descriptive Data for Control, Tactile-Only, and Auditory, Tactile, Visual, and Vestibular (ATVV) Groups TactileControl

Only

ATVV

n 5 10

n 5 14

n 5 16

Variable

M (SD)

M (SD)

M (SD)

Gestational age by

38.9 (1.57)

38.8 (1.17)

39.3 (1.06)

38.3 (2.47)

38.9 (1.54)

39.2 (1.16)

exam (weeks) Gestational age by dates (weeks) Birth weight (g)

3222 (445) 3284 (497)

3306 (304)

Birth length (cm)

49.1 (2.0)

45.2 (11.5)

48.7 (3.1)

Head

33.5 (1.6)

33.8 (1.1)

34.1 (1.2)

9.6 (0.52)

9.9 (0.27)

9.8 (0.40)

8.8 (0.42)

8.1 (1.07)

8.5 (0.51)

circumference (cm)

Joseph Kogan, PhD, is a senior research specialist in the Department of Health Systems Science, College of Nursing, the University of Illinois at Chicago.

24

Participants A convenience sample of 60 vaginally delivered and medically stable infants between 24 and 48 hours old was recruited. Forty infants contributed to the ¢nal sample. An additional inclusion criterion was

Postnatal Complications Score (0-10) Apgar 1

JOGNN, 38, 22-34; 2009. DOI: 10.1111/j.1552-6909.2008.00307.x


RESEARCH


Table 1. Continued TactileControl

Only

ATVV

n 5 10

n 5 14

n 5 16

Variable

M (SD)

M (SD)

M (SD)

Apgar 5

8.9 (0.32)

9.0 (0.39)

9.0 (0.0)

Maternal age

22 (6.5)

28 (7.2)

26 (5.8)

Obstetrical

37.1 (2.56)

36.3 (2.20)

37.6 (1.55)

1 minute and 9 at 5 minutes. Mothers were healthy and without complications as suggested by a high mean Obstetrical Complications Scores (range 36^37.6, total possible score 41). Data collection sessions were conducted between 10 a.m. and 1 p.m. There were no di¡erences among the three groups on the mean time of day for the data collection session. Similarly, there were no di¡erences among the groups on time since last feeding (see Table 1).

Complications Score (0^41) Predata collection

139 (12.87)

131 (10.97)

134 (11.92)

session heart rate Hours old (hr:min)

34:29 (7:18) 35:06 (10:24) 36:32 (10:50)

Time since last

1:45 (0:58)

1:41 (0:44)

1:21 (0:47)

feeding (hr:min) Infant sex Male

3

5

10

Female

7

9

6

Primigravida

3

2

1

Multigravida

7

12

15

African American 7

6

10

White

1

3

2

Hispanic

2

5

3

Other

0

0

1

Gravidity

Maternal race

Size for gestational age AGA

9

10

15

SGA

0

1

0

1

3

1

Breast

0

3

1

Bottle

6

7

14

0

1

0

Private

2

4

6

Medicaid

6

6

6

HMO

1

2

2

Other

0

2

2

LGA Feeding method

Both Insurance

a

a

Note. a

Represents missing data.


Independent Measures: The ATVV Intervention, Tactile-Only Stimulation, and Control Condition The 15 minute ATVV intervention was selected for this protocol because it is administered contingent on the infant’s responses and is composed of those sensory modalities that are the most often involved during mother^infant (or caregiver) interaction and maternal soothing behavior (Burns, Cunningham, White-Traut, Silvestri, & Nelson, 1994). Infants assigned to the ATVV intervention group (Group ATVV) received infant-directed talk via a soothing female voice with the infant (auditory stimulation) and social interaction via eye contact with the infant (visual stimulation) throughout the 15 minute period. For the ¢rst 10 minutes, the researcher massaged (using moderate pressure stroking) the infant (tactile stimulation), followed by 5 minutes of horizontal rocking (vestibular stimulation) (WhiteTraut, Nelson, Burns, & Cunningham, 1994). The tactile component consisted of placing the infant in a supine position and stroking over each body part in the following sequence: head, chest, abdomen, legs, and arms. The infant was then placed in the prone position for massage of the back. As noted in previous research, a decision tree was used to guide the application of the ATVV intervention (Burns et al., 1994) whereby the infants’ behavioral responses were monitored. The ATVV was continuously adapted based on the infants’ behavioral cues. For example, if the infant were to display disengagement cues such as hiccoughs or ¢nger splay, crying, fussing, or spitting/vomiting (Nursing Child Assessment Satellite Training Program, 1990) during the intervention, that step of the intervention was discontinued and the next portion was attempted (Burns et al.). If all the strokes were completed and time remained (within the 10 minute tactile stimulation period), the strokes where the infant exhibited more engagement cues would be repeated until 10 minutes passed and then the rocking initiated for 5 minutes. Thus, the ATVV intervention was administered contingent on the infant’s behavioral response.

25

RESEARCH


Contingency application has been deemed favorable because it provides opportunities for infants to interact with their environment and learn about predictability and the relationship of environmental stimuli to their behavior (White-Traut et al., 1994). In fact, during the study it was not necessary to discontinue the intervention for any infant because of potent negative behavioral cues. The research assistant was trained to criterion of greater than 85% reliability with the ATVV Reliability Checklist before the research. Continued reliability of the research assistant administering the ATVV intervention was periodically measured throughout the duration of the research. Continued reliability was assessed by calculating the amount of agreement of the administration of the intervention with the ATVV Reliability Checklist. The reliability score was maintained at 99%. Infants assigned to the tactile-only group (Group T) received 15 minutes of the tactile component of the ATVV intervention without any additional human social interaction (talking or eye contact). The researcher looked away from the baby’s face and did not talk to the infant during the intervention period. The massage was not modi¢ed in response to infant behavioral cues because the researcher did not look directly at the infant and therefore was unable to assess behavioral criteria. Previously, researchers have shown that tactile-only stimulation prompts excessive autonomic responses and immediate behavioral state changes from asleep to alert, with the alert state brie£y sustained (WhiteTraut et al.,1997). Thus, this intervention was chosen to further examine behavioral as well as hormonal responses to tactile-only interventions. Infants assigned to the control group (Group C) received routine hospital care that included rooming-in: the infants remained in their cribs, swaddled with no additional stimulation or care provided during the observation session. All infants were removed from their mothers’ room for data collection, which lasted for approximately 45 minutes.

Dependent Measures Stress Reactivity Stress reactivity (stress level) was measured via salivary cortisol. Salivary cortisol is a useful noninvasive measure of the neonate’s immediate stress reactivity response to both physical and environmental stimuli (Grauer, 1991; Gunnar, 1992; Herrington, Olomu, & Geller, 2004; Joyce, Keck, & Gerkensmeyer, 2001; Nelson, Arbring, & Theodorsson, 2001; Schmidt, 1998; White-Traut, Powlesland, Gelhar, Chatterton, & Morris, 1998; Whitten, Brock-

26

man, & Stavisky, 1998). In normal full-term infants, behavioral responses to stressors are correlated with elevated cortisol levels (Ahnert, Gunnar, Lamb, & Barthel, 2004;Joyce et al.Nelson et al.; Spangler & Scheubeck, 1993). Measurement of cortisol via whole saliva (saliva obtained from the parotid, submandibular, and sublingual glands) (Veerman, van den Keybus, Vissink, & Nieuw Amerongen,1996) is a simple and noninvasive measure that has facilitated the evaluation of the interactions among the hypothalamic-pituitary-adrenal axis, environmental events, and behavior within the context of research on health and development (Antonini, Jorge, & Moreira, 2000; Castro et al., 2000; Schwartz, Granger, Susman, Gunnar, & Laird,1998). Salivary cortisol level was measured at each time point using enzyme immunoassay (EIA) that depends on competition between an unknown amount of cortisol (from the sample) and enzymelinked cortisol (supplied by the kit) for a ¢xed number of antibody binding sites. Unbound cortisol is separated from bound cortisol and the amount of bound enzyme-linked cortisol is detected by absorbance at 405 nm after reaction with a chromagen. The EIA kit (Diagnostic Systems Laboratories Inc.) is speci¢cally modi¢ed for measurement of salivary cortisol. The EIA is reported by the manufacturer to be highly sensitive (minimal detection limit 5 0.072 mg/dl) with very little antibody cross reactivity with other endogenous steroids. Range of the assay is 0 to 10 mg/dl.The assay was validated in the laboratory by examining the Pearson correlation coe⁄cient between salivary samples (n 5 5) measured by both radioimmunoassay and EIA. The correlation was .99 (p 5 .01). Intraassay coe⁄cient of variation determined in the researchers’ laboratory was 6.2% 5.4% (n 5 4) and interassay coe⁄cient of variation was 5.3% 0.4% (n 5 5). Saliva samples were frozen immediately after collection and thawed at room temperature immediately before assay. The circadian rhythm of cortisol becomes established in term infants between 8 and 12 postnatal weeks (Castro et al., 2000; Klug et al., 2000). However, to control for any potential circadian patterns, the intervention and cortisol sampling were conducted at the same time of day over the course of the research (between 10 a.m. and 1 p.m.). Two posttest cortisol samples were obtained to validate the reactivity of the response (Ramsay & Lewis, 2003). Infant Behavioral State Seven categories of infant behavioral state were used: quiet sleep, active sleep, drowsy, quiet alert,

JOGNN, 38, 22-34; 2009. DOI: 10.1111/j.1552-6909.2008.00307.x


RESEARCH


active alert, crying, and indeterminate states (Thoman, 1987). Behavioral state was judged by a research assistant who was blinded to group assignment. To maintain blind ratings of infant behavioral state, auditory and visual cues were blocked from the research assistant who was judging behavioral state (primary rater) during the intervention/observation procedures. To assure reliability, a second research assistant (secondary rater) periodically judged behavioral state throughout the research for 25% of the observations. Interrater agreement between two observers remained above 88% for the duration of the research. For analysis, the ‘‘quiet’’ and ‘‘active sleep’’ categories were combined into one sleep category, and the ‘‘drowsy,’’ ‘‘quiet alert,’’ and ‘‘active alert’’ were combined into an awake category. ‘‘Cry’’ and ‘‘indeterminate’’ were maintained as separate categories because cry is indicative of increased stress and the indeterminate state is considered a transitional state, neither awake nor sleep.

Procedure

from swallowing the dental gauze, dental £oss was sewn through the gauze and secured around the researcher’s hand. Lemon and sugar (lemonade) crystals were used to stimulate saliva production (Schwartz et al., 1998). After the saliva specimens were obtained, they were immediately placed on ice and frozen at 70 1C until analyzed via an EIA. Assays were conducted when all the data were collected. Saliva samples were noted with ID numbers and were assayed by a laboratory blinded to group assignment. Infant behavioral state was judged by a research assistant blinded to group assignment. Judging of behavioral state occurred immediately before the intervention/observation session (baseline) and recorded at the end of every minute throughout the 15 minute intervention.When behavioral state was to be judged, the intervener stopped the ATVV intervention and covered the infant with a receiving blanket from the waist down. For infants in the control group, the intervener kept her hands at her side and the infants remained covered with the blanket throughout the observation period. For all infants, the intervener placed her hands on the side of the open crib for 10 seconds during the judging of behavioral state. To maintain a blind assessment of infant behavioral state, the research assistant (di¡erent than the intervener) looked away from the infant and wore a head set to block out sound during the 15 minute session. The research assistant viewed the infant only at the speci¢ed times for judging infant behavioral state.

After informed consent was obtained from the infants’ parents, eligible infants were randomly assigned (via random start in a random numbers table) to receive 15 minutes of tactile-only stimulation (Group T), 15 minutes of ATVV stimulation (Group ATVV), or 15 minutes of no stimulation (Group C). The intervention or control group observation was administered in a quiet room away from the mother’s room yet near the nursery 30 minutes before the next anticipated feeding. Thus all infants were removed from their mothers’ room for approximately 45 minutes.

Statistical Analysis

Saliva specimens were collected a total of three times: before (Time 1), immediately following (Time 2), and again 10 minutes (Time 3) postcompletion of the intervention. The timing of salivary cortisol measurement was selected to obtain a baseline level, a level immediately upon conclusion of the intervention, and at 10 minutes following conclusion of the intervention, at which time the infant was returned to the mother for feeding. This time frame was selected based on previous behavioral state data which showed that 10 minutes after completion of the ATVV intervention aroused infants would return to sleep (White-Traut, Studer et al., 2002). Also, researchers have shown that cortisol levels peak in infants 20 to 25 minutes postinitiation of a stressor (Ramsay & Lewis, 2003). Cotton dental gauze was used to collect the saliva specimens (cotton without wood). The use of cotton does not a¡ect the reliability of the cortisol assay (Shirtcli¡, Granger, Schwartz, & Curran, 2001). To prevent the infant

Infant behavioral state was compared among the groups by chi-square. The row frequency of behavioral state for each time point was converted to row percents for analysis. A mixed e¡ects model with repeated measures was used to determine the di¡erences in the mean cortisol levels over the course of the research (Time 1 through Time 3). A mixed e¡ects model was used for analysis because of missing cortisol values in all three groups. The mixed model allowed for the inclusion of all of the available data and also accounted for within-subject correlation. Estimates of ¢xed e¡ects (post hoc analysis) allowed for comparison of the di¡erences in mean cortisol levels between Groups C and ATVV and Groups T and ATVV with a t test. This model maintains the a at .05. The ¢xed e¡ects model accounted for the covariates of gestational age and Apgar score at 5 minutes. Infant gender was dropped from the ¢nal model because it was not a signi¢cant variable. A Pearson’s correlation was conducted to determine whether a relationship


27

RESEARCH


existed between infant behavioral state and cortisol levels.

Results Salivary Cortisol Levels Although data were obtained for 60 infants, there were several cases of insu⁄cient saliva to conduct EIA analysis. Forty infants contributed to the ¢nal saliva samples. For the baseline cortisol analyses, there were 10 su⁄cient aliquots for Group C, 14 for Group T, and 16 for Group ATVV. At Time 2, immediately after the intervention/observation period, there were 9 su⁄cient aliquots for Group C, 12 for Group T, and 10 for Group ATVV. For the 10 minute postintervention cortisol analyses, there were 8 su⁄cient aliquots for Group C, 5 for Group T, and 6 for Group ATVV (see Figure 1). Thus, there were 19 su⁄cient aliquots remaining at Time 3. Means and standard errors (SEs) are reported (see Figure 2). A post hoc power analysis was conducted on the data reported here to estimate what sample size would be needed to detect signi¢cant di¡erences among the three groups. A post hoc power analysis on these data determined that a sample size of at

least eight subjects per group was needed to ¢nd a signi¢cant e¡ect between baseline (Time 1) and 10 minutes postintervention (Time 3; power of .80 with a 5 .05). The range of salivary cortisol levels were within the range previously reported by Klug et al. (2000) for newly born full-term infants between 25 and 55 hours postbirth (0.24^6 mg/dl). There were no signi¢cant di¡erences in cortisol level among the three groups at baseline. Mean salivary cortisol ranged from 1.59 to 3.18 mg/dl for the three groups. Over the intervention period, cortisol levels varied among the three groups (Time 1 through Time 3; see Figure 2). Infants assigned to Group C experienced a slight increase in cortisol from baseline (SEs in parentheses) 2.43 mg/dl (0.72) to 3.00 mg/dl (1.3) (1.84 mg/dl [0.6] at Time 2). Infants assigned to Group T had a more noticeable increase from 1.59 mg/dl (0.6) at baseline to 3.82 mg/dl (0.9) at completion of the tactile-only intervention, stabilizing to 3.63 mg/dl (1.7) at 10 minutes poststimulation. In contrast, Group ATVV infants had a steady decline in cortisol from baseline at 3.18 mg/dl (0.7) to 3.11 mg/dl (1.0) immediately postintervention, fol-

Newborn assessed for eligibility by recruitment nurse

Mother approached by recruitment nurse Informed consent obtained by PI

60 infants randomly assigned

Control Group 20 observed and 20 saliva samples collected

Tactile-Only Group 20 received tactile intervention and 20 samples collected

ATVV Group 20 received ATVV intervention and 20 samples collected

All samples (n = 60) collected in gauze and sent to lab for centrifuge and analysis

Time 1 Sufficient aliquots (n = 10) Insufficient aliquots (n = 10)





Sufficient aliquots (n = 10) Insufficient aliquots (n = 6)




Time 2

Figure 1. Flow diagram of saliva specimens obtained from three groups of full-term infants. Two infant groups received behavioral interventions and one control group did not receive any additional stimuli.

28

JOGNN, 38, 22-34; 2009. DOI: 10.1111/j.1552-6909.2008.00307.x


RESEARCH


Salivary cortisol (µg/dl)

6 5

infants in the study (36 and 41 weeks gestation), suggesting near-term and late-term infants are vulnerable to stress, F(2, 27) 5 4.52, p 5 .02.

Control Tactile ATVV

4 3 2 1 0 Time 1

Time 2

Time 3

Figure 2. Mean salivary cortisol (with SE bars) by group at baseline, immediate postintervention and 10 minutes postintervention. Note: Baseline 5 Time 1, immediately postintervention 5 Time 2, and 10 minutes postintervention 5 Time 3. Control Group n at Time 1 5 10; Time 2 5 9; Time 3 5 8. Tactile Group n at Time 1 5 14; Time 2 5 12; Time 3 5 5. ATVV Group n at Time 1 5 16; Time 2 5 10;

To further explore the group by time ¢ndings in the overall model, estimates of ¢xed e¡ects comparing Group ATVV (as a base) to Group C and then separately for Group T con¢rmed signi¢cantly higher levels of cortisol in Group T compared with Group ATVV, t(57) 5 3.08, p 5 .003, two-tailed (see Table 2). Nearly signi¢cant di¡erences were identi¢ed for Group C, t(57) 5 1.98, p 5 .052, two-tailed. Controlling for groups, cortisol increased at the lowest (36 weeks) and the highest (41 weeks) gestational ages and for the lowest Apgar score of 8 at 5 minutes. Gestational age by group e¡ects revealed that the older infants in the tactile-only group had higher cortisol levels than younger infants. Group comparisons con¢rmed that Group T had signi¢cantly higher cortisol levels at both Time 2 and Time 3.

Time 3 5 6. Cortisol was significantly different between Groups T (Tactile) and ATVV (po.01) and over time

Infant Behavioral State

among all three groups (po.01).

The distribution of infant behavioral state was similar among the groups at baseline and over the course of the intervention as shown in Table 3. The predominant state at baseline was sleep for ATVV infants and awake for Tactile and Control groups. At completion of the intervention, all three groups

lowed by 1.74 mg/dl (1.1) at 10 minutes postintervention. When the three groups were compared over time via a mixed model (general linear mixed model with Type III tests of ¢xed e¡ects), signi¢cant di¡erences in mean cortisol levels were identi¢ed, F(2, 27) 5 4.75, p 5 .02. Speci¢cally, the control group experienced a slight decrease in mean cortisol at Time 2 followed by an increase over baseline at Time 3, while the tactile-only group experienced an immediate increase in mean cortisol at Time 2 which was sustained at Time 3 (Figure 2). In contrast, the ATVV group experienced a steady decline from Time 1 toTime 3. To account for changes in cortisol level in the three groups over time, the following variables were introduced in a repeated measures mixed model: gestational age, Apgar score, gestational age by group and experimental group (control, tactile-only, and ATVV). As previously reported, in this model cortisol increased for the control and tactile-only groups and decreased for the ATVV group, F(2, 56) 5 4.90, p 5 .01. Mean cortisol was lowest for infants between 37 and 40 weeks gestation and highest for infants at 36 and 41 weeks gestation, F(1, 26) 5 4.93, p 5 .03. Although the 5 minute Apgar scores ranged from 8 to 10, the highest mean cortisol level was identi¢ed for infants with an Apgar score of 8 and the lowest cortisol for infants with an Apgar score of 10, F(1, 22) 5 4.46, p 5 .05. Cortisol was increased for the youngest and the oldest


Table 2: Estimates of Fixed Effects p

Standard Parameter Intercepta

Estimate

Error

18.3854 4.9827

df

t Value

Value

23.065

3.690 .001

group Control

6.9215

5.2373

28.199

1.322

.197

Tactile-

16.2717

5.4412

23.172

2.990 .007

0.2966 0.1094

23.147

2.711

.012

0.5959 0.2823

22.197

2.111

.046

0.3962 0.1389

22.908

2.853 .009

Control

0.3823 0.1929

56.957

1.982

Tactile-

0.5893 0.1912

57.097

3.082 .003

only Gestational age Apgar 5

Gestational age by group Tactileonly Over time .052

only Note. a

Intercept 5 Time 0.

29

RESEARCH


Salivary cortisol levels increased for the control and tactile-only group while infants assigned to the multisensory intervention exhibited a steady decline.

were predominately in the sleep state. There was very little crying among the three groups during the intervention period. No infants were observed in the indeterminate state category, and thus it was dropped from the analysis. The frequencies of the seven behavioral state categories (row frequencies for each time point) were combined into three categories: asleep, awake, and cry. These frequencies were then converted into row percents for analysis. Pearson’s correlations between cortisol levels and the percents of infant behavioral state were not signi¢cant (Time 1, r 5 .23, n 5 36, p 5 .183; Time 2, r 5 .08, n 5 35, p 5 .66, and Time 3, r 5 .01, n 5 19, p 5 .99, twotailed). For Time 1 this is most likely due to the small sample size.

Discussion In the present pilot research, stress reactivity increased as evidenced by elevated salivary cortisol levels for the control (C) and tactile-only (T) group infants, while the stress reactivity of infants assigned to the multimodal intervention (Group ATVV) exhibited a steady decline in salivary cortisol over the course of the intervention. The increase in cortisol for the control and tactile-only groups (Figure 2) cannot be explained; however, infants in both groups were removed from their mothers’ room and did not experience human social interaction (or any additional nursing care) over the short data collection session. Such a ¢nding is similar to previous research results on both animals and humans in which separation from the mother or maternal un-

Table 3: Distribution of Behavioral State by Time Control

Tactile-Only

ATVV

n 5 10

n 5 14

n 5 16

AS/AW/CR

AS/AW/CR

AS/AW/CR

Baseline

2/8/0

6/8/0

10/6/0

Midintervention

6/4/0

6/7/1

10/5/1

Immediate

6/4/0

11/3/0

12/4/0

Behavioral state

postintervention Note. AS 5 Asleep; AW 5 Awake; CR 5 Cry.

30

availability (e.g., lack of social interaction) resulted in increased cortisol levels (Bugental, Martorell, & Barraza, 2003; Hennessy & Moorman, 1989). The largest increase and overall change in cortisol was observed in Group T infants, which suggests that a moderate touch tactile-only stimulation can increase stress reactivity and thus may be more stressful than the ATVV intervention that incorporates human social interaction and soothing vestibular stimuli with a moderate touch massage. The current ¢ndings are similar to those of Jutapakdeegul et al. (2003) who showed that postnatal touch stimulation in the rat animal model increased corticosteroid levels when tactile-only stimulation was provided. The ¢ndings also support earlier research with preterm infants in whom tactile/kinesthetic stimulation (massage) increased cortisol, dopamine, norepinephrine, and epinephrine levels (Kuhn et al., 1991). The present results and those of Kuhn et al. contradict Acolet et al. (1993) who reported a decrease in serum cortisol in preterm infants following massage alone, as well as Field et al. (1996) who found a decrease in salivary cortisol among infants of depressed mothers following massage. Without a better understanding of the potential human social interaction involved in the various massage protocols, it is di⁄cult to compare the results of the present study with Kuhn, Acolet, or Field. This is the ¢rst study in which researchers examine stress reactivity in response to the ATVV intervention and a moderate touch massage that includes no human social interaction. Compared with previous research, the increase in cortisol levels of infants in the tactile-only group rival levels previously observed in infants undergoing painful procedures. Short-term stressors such as physical exams, inoculations, heel sticks, and circumcision have been shown to increase cortisol levels in newborns and young infants (Gunnar, 1992; Gunnar, Connors, Isensee, & Wall, 1988; Lewis & Ramsay, 1995; Ramsay & Lewis, 2003). Between 57 and 80 hours after birth, circumcision increases cortisol levels three times baseline levels (to a mean of 24 mg/dl) (Gunnar, 1992) while heel sticks also increase mean cortisol three times greater than baseline (Gunnar, Porter, Wolf, Rigatuso, & Larson, 1995; Gunnar et al., 1988). In the present research, a 2.2-fold increase in cortisol was identi¢ed for infants assigned to the tactile-only group while a 1.8-fold decrease was identi¢ed for infants who received the ATVV intervention. The highest cortisol level recorded postmassage was 7.1 mg/dl for an infant in the tactile-only group, levels similar to that

JOGNN, 38, 22-34; 2009. DOI: 10.1111/j.1552-6909.2008.00307.x


RESEARCH


reported for newborns after restraint on a circumcision board (6.1 mg/dl) or handling during a physical exam (9.2 mg/dl) (Gunnar,1992). Infants assigned to Group ATVV exhibited a decrease in stress reactivity as their salivary cortisol levels steadily declined. The results of previous research support the use of combined behavioral and sensory approaches to reducing infant stress reactivity. For example, a behavioral/multisensory intervention of holding combined with sweet taste reduced crying after painful episodes (Gormally et al., 2001). During painful episodes, rocking in conjunction with maternal vocalization reduced infant stress and crying more than rocking alone (Jahromi et al., 2004) while other behavioral/multisensory interventions using a combination of maternal soothing behaviors (auditory, visual, and vestibular stimuli) reduced salivary cortisol during immunization (Braarud & Stormark, 2006; Felt et al., 2000;Jahromi et al.). More recently, maternal touch, vocalizations, and a¡ection showed a greater contribution to reducing infant distress during inoculation when compared with distraction, presenting the maternal face or use of a paci¢er (Jahromi & Stifter, 2007). Furthermore, that Korean full-term infants housed in an orphanage who received the ATVV intervention showed better growth and fewer illnesses requiring clinic visits when compared with controls suggests that long-term use of the ATVV intervention, which includes contingent human social interaction and rocking, may indeed reduce stress and improve infant well-being (Kim et al., 2003). Infant behavioral state, speci¢cally cry, is also an indicator of infant stress reactivity. Yet in the research currently reported, infants showed minor £uctuations in state and very little crying was observed. Previously researchers working with full-term and preterm infants found stable hemoglobin oxygen saturation (between 85% and 90%), heart (between 140 and 180 beats/minute), and respiratory rates (between 30 and 60 breaths/minute) during the multisensory ATVV intervention (White-Traut, Studer et al., 2002; White-Traut et al., 2004). However, during moderate touch tactile-only stimulation, as used in the current study, researchers found that well premature infants responded with elevated proportions of heart and respiratory rates outside the limits of normal (greater than 180 beats/ minute and greater than 60 breaths/minute) and rapid changes in behavioral state all of which quickly subsided by conclusion of the intervention (White-Traut et al., 1997). In the present research, there was minimal crying during either of the inter-


ventions. Thus, the two interventions did not appear to a¡ect behavioral stress reactivity. Salivary cortisol is considered an acceptable measure of infant stress reactivity; however, additional measures that reliably quantify infant stress reactivity are needed. In previous research a relationship between increased crying and increased cortisol during circumcision and blood sampling was documented (Gunnar et al., 1988). However, infant behavior has not been consistently correlated with change in cortisol level in response to stressors making it di⁄cult for clinicians to assess the extent of the infant’s stress reactivity (Gunnar & Donzella, 2002; Gunnar et al., 1988, 1995). Other reports indicate that additional measures, including behavioral and autonomic nervous system indicators, are needed to better understand infant stress reactivity (Gunnar et al.,1995).

Limitations Small sample size due to inadequate saliva samples from participating infants was a limitation of the current research. Although signi¢cant ¢ndings were identi¢ed, the sample size for Times 2 and 3 did not meet the criteria in a post hoc power analysis. Thus, the ¢ndings should be interpreted with caution before implementing the intervention as a standard of care. Ethical concerns preclude obtaining serum samples for neurohormonal assays in well full-term or premature infants and an adequate volume of each aliquot from the saliva samples was expected to be obtained. In previous research, su⁄cient aliquots were obtained without saliva stimulation in 3- to 4-day-old newborn infants (Klug et al., 2000; Nelson et al., 2001). However in the present research, only 66% of the aliquots had a su⁄cient amount of saliva for baseline analysis and even fewer for the postintervention analysis. Although others have not reported di⁄culty with obtaining adequate aliquots (Gunnar,1992;Klug et al.; Nelson et al.), it remains undetermined whether it was the method or timing of obtaining saliva or the inability of some newly born infants to produce enough saliva that resulted in the limitation. The majority of the infants were bottle-fed, thus they were not likely to be dehydrated. The limitations identi¢ed can be addressed in future research and do not preclude further testing of salivary cortisol responses of premature infants (who may have less saliva) to the ATVV intervention.

Future Research Further research with a larger sample is needed to replicate these ¢ndings and to further evaluate the

31

RESEARCH


The benefit of the multisensory intervention may be in the reduction of infant stress.

relationship between behavioral state and cortisol level, as there appears to be a complex relationship. To address the inadequate sample size and methodological issues, future research should consider expanding the time interval between each saliva collection to a minimum of 30 minutes to allow for more saliva production and saliva should be collected for 8 minutes instead of 5 minutes (Herrington et al., 2004). In addition, behavioral state data should be collected more frequently and throughout the collection of saliva to better describe the range of behavioral state change. Additionally, research with infants 36 and 41 weeks gestation and other at-risk or preterm infants is warranted to document whether the ATVV intervention reduces cortisol levels, how long-term administration might in£uence the levels, and whether the interventions are bene¢cial for other infant populations. In order to assist clinicians in assessing infant response, further research is warranted to identify the relationship of cortisol levels to additional behavioral indicators of stress reactivity (other than infant crying alone). Furthermore, research is needed to document whether responses to the ATVV intervention are related to the context of the intervention, the routine modi¢cation based on the infant’s behavioral cues, or additional infant behavioral indicators, such as infant temperament. In the meantime, the current preliminary ¢ndings reported extend previous reports of the bene¢ts of the ATVV intervention for full-term infants as well as at-risk premature and full-term infants (e.g., more quiet alert and sleep states, faster feeding progression, more weight gain, and shorter length of hospital stay) and suggest that the bene¢ts may be due to lower stress reactivity. Additional research is needed to better understand relationships between these indicators and potential mechanisms responsible for the ATVV group’s lower cortisol levels.

and may bene¢t from maternal/caregiver soothing e¡orts that use a multisensory approach following these stressors. In addition, infants at 36 and 41 weeks gestation (near and late term) and those with Apgar scores of 8 or less may be more vulnerable to sensory stimuli and thus require additional consideration. Identifying contingent interventions that alleviate infant stress will assist clinicians in implementing evidence-based interventions for their patients. The ¢ndings also emphasize the importance of identifying whether o¡ering tactile-only stimulation within the context of massage interventions may have potentially negative consequences in newly born infants and may be less bene¢cial in reducing infant stress than the ATVV intervention. The increase in cortisol among infants in the tactile-only group provides some evidence for concern with such a form of intervention delivery. On the other hand, the steady decline in cortisol for the ATVV intervention group underscores the importance of human social contact (eye contact and talking) and soothing vestibular stimuli in conjunction with moderate touch stimulation via stroking for newborn infants and suggests that the bene¢t of the multisensory intervention may be in the reduction of infant stress.

Acknowledgments Funded by the Harris Foundation. The authors acknowledge Drs. Ruth London and Richard Belkengren (posthumous) and sta¡ at Mount Sinai and Rush University Medical Centers, Chicago.

REFERENCES Acolet, D., Modi, N., Giannakouloloulos, X., Bond, C., Weg, W., Clow, A., et al. (1993). Changes in plasma cortisol and catecholamine concentrations in response to massage in preterm infants. Archives in Diseases of Childhood, 68, 29-31. Ahnert, L., Gunnar, M. R., Lamb, M. E., & Barthel, M. (2004). Transition to child care: Associations with infant-mother attachment, infant negative emotion, and cortisol elevations. Child Development, 75, 639-650. Albers, E. M., Riksen-Walraven, J. M., Sweep, F., & de Weerth, C. (2008). Maternal behavior predicts infant cortisol recovery from a mild everyday stressor. Journal of Child Psychology and Psychiatry, 49, 97-103.

Clinical Implications These results provide indirect support for the important role that contingent multisensory maternal/caregiver comforting behaviors can play in reducing infant stress in a variety of clinical situations. Healthy infants are often exposed to stressors, for example, blood draws, injections, circumcision or short separation from their mothers

32

Antonini, S. R., Jorge, S. M., & Moreira, A. C. (2000). The emergence of salivary cortisol circadian rhythm and its relationship to sleep activity in preterm infants. Clinical Endocrinology, 52, 423-426. Ballard, J. L., Khoury, J. C., Wedig, K., Wang, L., Eilers-Walsman, B. L., & Lipp, R. (1991). New Ballard Score, expanded to include extremely premature infants. Journal of Pediatrics, 119, 417-423. Braarud, H. C., & Stormark, K. M. (2006). Maternal soothing and infant stress responses: Soothing, crying and adrenocortical activity during inoculation. Infant Behavior and Development, 29, 70-79.

JOGNN, 38, 22-34; 2009. DOI: 10.1111/j.1552-6909.2008.00307.x


RESEARCH


Bugental, D. B., Martorell, G. A., & Barraza, V. (2003). The hormonal costs

Jahromi, L. B., Putnam, S. P., & Stifter, C. A. (2004). Maternal regulation of

of subtle forms of infant maltreatment. Hormones and Behavior,

infant reactivity from 2 to 6 months. Developmental Psychology,

43, 237-244.

40, 477-487.

Burns, K., Cunningham, N., White-Traut, R. C., Silvestri, J. M., & Nelson, M.

Jahromi, L. B., & Stifter, C. A. (2007). Individual di¡erences in the contribu-

N. (1994). Infant stimulation: Modi¢cation of an intervention based

tion of maternal soothing to infant distress reduction. Infancy, 11,

on physiological and behavioral cues. Journal of Obstetric, Gynecologic, & Neonatal Nursing, 23, 581-589. Caldji, C., Diorio, J., & Meaney, M. J. (2000). Variations in maternal care in infancy regulate the development of stress reactivity. Biological Psychiatry, 48, 1164-1174.

255-269. Joyce, B. A., Keck, J. F., & Gerkensmeyer, J. (2001). Evaluation of pain management interventions for neonatal circumcision pain. Journal of Pediatric Health Care, 15, 105-114. Jutapakdeegula, N., Casalottia, S. O., Govitraponga, P., & Kotchabhakdia,

Castro, M., Elias, P. C., Martinelli Jr., C. E., Antonini, S. R., Santiago, L., &

N. (2003). Postnatal touch stimulation acutely alters corticosterone

Moreira, A. C. (2000). Salivary cortisol as a tool for physiological

levels and glucocorticoid receptor gene expression in the neona-

studies and diagnostic strategies. Brazilian Journal of Medical and Biological Research, 33,1171-1175.

tal rat. Developmental Neuroscience, 25, 26-33. Kawakami, K., Tomonaga, M., & Suzuki, J. (2002). The calming e¡ect of

Chugani, H. T., Behen, M. E., Muzik, O., Juhasz, C., Nagy, F., & Chugani, D. C.

stimuli presentation on infant Japanese Macaques (Macaca

(2001). Local brain functional activity following early deprivation: A

fuscata) under stress situation: A preliminary study. Primates:

study of postinstitutionalised Romanian orphans. Neuroimage, 14, 1290-1301. Cigales, M., Field, T., Lundy, B., Cuadra, A., & Hart, S. (1997). Massage enhances recovery from habituation in normal infants. Infant Behavior and Development, 20, 29-34.

Journal of Primatology, 43, 73-85. Kim, T. I., Shin, Y. H., & White-Traut, R. C. (2003). Multisensory intervention improves physical growth and illness rates in Korean orphaned newborn infants. Research in Nursing and Health, 26, 424-433. Klug, I., Dressendorfer, R., Strasburger, C., Kuhl, G. P., Reiter, H. L., Reich, A.,

Felt, B. T., Mollen, E., Diaz, S., Renaud, E., Zeglis, M., Wheatcroft, G., et al.

et al. (2000). Cortisol and 17-hydroxyprogesterone levels in saliva

(2000). Behavioral interventions reduce infant distress at immuni-

of healthy neonates: Normative data and relation to body mass

zation. Archives of Pediatrics and Adolescent Medicine, 154,

index, arterial cord blood ph and time of sampling after birth.

719-724.

Biology of the Neonate, 78, 22-26.

Field, T., Grizzle, N., Sca¢di, F., Abrams, S., & Richardson, S. (1996). Mas-

Kuhn, C. M., Schanberg, S. M., Field, T., Symanski, R., Zimmerman, E., Sca-

sage therapy for infants of depressed mothers. Infant Behavior

¢di, F., et al. (1991). Tactile-kinesthetic stimulation e¡ects on

and Development, 19, 107-112.

sympathetic and adrenocortical function in preterm infants. Jour-

Field, T., Hernandez-Reif, M., Diego, M., Feijo, L., Vera, Y., & Gil, K. (2004). Massage therapy by parents improves early growth and development. Infant Behavior and Development, 20, 435-442. Francis, D., Diorio, J., LaPlante, P., Weaver, S., Seckl, J. R., & Meaney, M. J. (1996). The role of early environmental events in regulating neuro-

nal of Pediatrics, 119, 434-440. Kurihara, H., Chiba, H., Shimizu, Y., Yanaihara, T.,Takeda, M., Kawakami, K., et al. (1996). Behavioral and adrenocortical responses to stress in neonates and the stabilizing e¡ects of maternal heartbeat on them. Early Human Development, 46, 117-127.

endocrine development: Moms, pups, stress, and glucocorticoid

Ladd, C. O., Huot, R. L., Thrivikraman, K. V., Nemero¡, C. B., Meaney, M. J., &

receptors. Understanding Aggressive Behavior in Children, 794,

Plotsky, P. M. (2000). Long-term behavioral and neuroendocrine

136-152.

adaptations to adverse early experience. Progress in Brain

Francis, D. D., Diorio, J., Plotsky, P. M., & Meaney, M. J. (2002). Environmental enrichment reverses the e¡ects of maternal separation on stress reactivity. Journal of Neuroscience, 22, 7840-7843. Gormally, S., Barr, R. G., Wertheim, L., Alkawaf, R., Calinoiu, N., & Young, S. N. (2001). Contact and nutrient caregiving e¡ects on newborn infant pain responses. Developmental Medicine and Child Neurology, 43, 28-38. Grauer, T. T. (1991). The relationship of predominant behavioral state and salivary cortisol in the human neonate. The Journal of the New York State Nurses’ Association, 22, 9-6.

Research, 122, 81-103. Lewis, M., & Ramsay, D. (1995). Stability and change in cortisol and behavioral response to stress during the ¢rst 18 months of life. Developmental Psychobiology, 28, 419-428. Littman, B., & Parmelee, A. H. (1978). Medical correlates of infant development. Pediatrics, 61, 470-474. Meaney, M. J., & Aitken, D. H. (1985). The e¡ects of early postnatal handling on hippocampal glucocorticoid receptor concentrations: Temporal parameters. Brain Research, 354, 301-304. Nelson, N., Arbring, K., & Theodorsson, E. (2001). Neonatal salivary corti-

Gunnar, M. R. (1992). Reactivity of the hypothalamic-pituitary-adrenocor-

sol in response to heelstick: Method modi¢cations enable analysis

tical system to stressors in normal infants and children. Pediatrics,

of low concentrations and small sample volumes. Scandinavian

90, 491-497.

Journal of Clinical and Laboratory Investigation, 61, 287-291.

Gunnar, M. R., Connors, J., Isensee, J., & Wall, L. (1988). Adrenocortical

Nursing Child Assessment Satellite Training Program. (1990). Keys to

activity and behavioral distress in human newborns. Developmen-

caregiving: Study guide. Seattle, WA: NCAST Publications, Univer-

tal Psychobiology, 21, 297-310.

sity of Washington.

Gunnar, M. R., & Donzella, B. (2002). Social regulation of the cortisol lev-

Ramsay, D., & Lewis, M. (2003). Reactivity and regulation in cortisol

els in early human development. Psychoneuroendocrinology, 27,

and behavioral responses to stress. Child Development, 74,

199-220. Gunnar, M. R., Porter, F. L., Wolf, C. M., Rigatuso, J., & Larson, M. C. (1995). Neonatal stress reactivity: Predictions to later emotional temperament. Child Development, 66, 1-13. Hennessy, M. B., & Moorman, L. (1989). Factors in£uencing cortisol and behavioral responses to maternal separation in guinea pigs. Behavioral Neuroscience, 103, 378-385.

456-464. Schmidt, N. A. (1998). Salivary cortisol testing in children. Issues in Comprehensive Pediatric Nursing, 20, 183-190. Schwartz, E. B., Granger, D. A., Susman, E. J., Gunnar, M. R., & Laird, B. (1998). Assessing salivary cortisol in studies of child development. Child Development, 69, 1503-1513. Shirtcli¡, E. A., Granger, D. A., Schwartz, E., & Curran, M. J. (2001). Use of

Herrington, C. J., Olomu, I. N., & Geller, S. M. (2004). Salivary cortisol as in-

salivary biomarkers in biobehavioral research: Cotton-based

dicators of pain in preterm infants: A pilot study. Clinical Nursing

sample collection methods can interfere with salivary immunoas-

Research, 13, 53-68.

say results. Psychoneuroendocrinology, 26, 165-173.


33

RESEARCH


Spangler, G., & Scheubeck, R. (1993). Behavioral organization in new-

White-Traut, R. C., Nelson, M. N., Silvestri, J. M., Patel, M., Berbaum, M., Gu,

borns and its relation to adrenocortical and cardiac activity. Child

G. G., et al. (2004). Developmental patterns of physiological re-

Development, 64, 622-633.

sponse to a multisensory intervention in extremely premature and

Thoman, E. B. (1987). The sleeping and waking states of infants: Correlations across time and person. Physiology and Behavior, 41, 531-537.

high-risk infants. Journal of Obstetric, Gynecologic, & Neonatal Nursing, 33, 266-275. White-Traut, R. C., Nelson, M. N., Silvestri, J. M., Patel, M., & Kilgallon, D.

Thompson, L. A., & Trevathan, W. R. (2008). Cortisol reactivity, maternal

(1993). Patterns of physiologic and behavioral response of inter-

sensitivity, and learning in 3-month-old infants. Infant Behavior

mediate care preterm infants to intervention. Pediatric Nursing, 19,

and Development, 31, 92-106.

625-629.

Underdown, A., Barlow, J., Chung, V., & Stewart-Brown, S. (2006). Mas-

White-Traut, R. C., Nelson, M. N., Silvestri, J. M., Vasan, U., Littau, S., Melee-

sage intervention for promoting mental and physical health in

dy-Rey, P., et al. (2002). E¡ect of auditory, tactile, visual, and

infants aged under six months. Cochrane Database Systemic

vestibular intervention on length of stay, alertness, and feeding

Review, CD005038.

progression in preterm infants. Developmental Medicine and Child

Veerman, E. C., van den Keybus, P. A., Vissink, A., & Nieuw Amerongen, A. V. (1996). Human glandular salivas: Their separate collection and analysis. European Journal of Oral Science, 104, 346-352. Vickers, A., Ohlsson, A., Lacy, J. B., & Horsley, A. (2004). Massage for pro-

Neurology, 44, 91-97. White-Traut, R. C., Powlesland, J., Gelhar, D., Chatterton, R., & Morris, M. (1998). Methodologic issues in the measurement of oxytocin in human neonates. Journal of Nursing Measurement, 6, 155-173.

moting growth and development of preterm and/or low birth-

White-Traut, R., Studer, T., Meleedy-Rey, P., Murray, P., Labovsky, S., &

weight infants. Cochrane Database of Systematic Reviews, 2,

Kahn, J. (2002). Pulse rate and behavioral state correlates after au-

CD000390. DOI: 10.1002/14651858.CD000390.pub2.

ditory, tactile, visual, and vestibular intervention in drug-exposed

White-Traut, R. C., Nelson, M. N., Burns, K., & Cunningham, N. (1994). Environmental in£uences on the developing premature infant: Theoretical issues and applications to practice. Journal of Obstetric, Gynecologic, & Neonatal Nursing, 23, 393-401. White-Traut, R. C., Nelson, M. N., Silvestri, J. M., Cunningham, N., & Patel, M. (1997). Responses of preterm infants to unimodal and multimodal

neonates. Journal of Perinatology: O⁄cial Journal of the California Perinatal Association, 22, 291-299. Whitten, P. L., Brockman, D. K., & Stavisky, R. C. (1998). Recent advances in noninvasive techniques to monitor hormone-behavior interactions. American Journal of Physical Anthropology, 27(Suppl.), 1-23.

sensory intervention. Pediatric Nursing, 23, 169-175.

34

JOGNN, 38, 22-34; 2009. DOI: 10.1111/j.1552-6909.2008.00307.x
