VoL 160 No. 8
The Influence of Ambient Temperature on Some Commonly Measured Physiological Variables in Infants §
P. P. Fox, *M. O'Regan, §
G. Matthews
+Rotunda Hospital, Dublin; *Department of Statistics, Trinity College, Dublin 2. Summary Fifty-four infants underwent recording of heart rate, respiratory rate, long term and short term heart rate variability and end tidal carbon dioxide measurements at three different ambient temperatures (20~ 25-26~ 30~ under standardized conditions. The infants' skin and rectal temperature remained normal throughtout the study. The ambient temperature of the room where the study was conducted significantly influenced the infants' heart rate, respiratory rate and long and short term heart rate variability but not end tidal carbon dioxide values. The thermal environment in which an infant resides has a significant influence on some physiological variables and should be standardized in future physiological studies. Introduction The Sudden Infant Death Syndrome (S.I.D.S.) remains the major cause of infant deaths in the developed world and has been the subject of intense research over the past twenty years especially in the area of cardio-respiratory control~-~. Although lower ambient temperatures have been associated with increased rates of S.I.D.S), few physiological studies in infants have taken account of the ambient temperature. In this paper we report the effect of relatively modest changes in ambient temperature, well within the range in which infants are nursed at home4, on some commonly measured physiological variables in healthy infants. Subjects and Methods Study Groups (Table I). Apparent Life Threatening events (A.L.T.E.):- For this study the definition used was an infant being found lifeless and apparently dead or almost dead, requiring vigorous resuscitation including mouth-to-mouth resuscitation, in whom all investigations were negative and who was subsequently clinically normal. All infants had fully recovered clinically from the A.L.T.E. episode when our studies were carried out48 hrs. to 8 days later. Siblings of S.I.D.S.:- Infants in this group had a sibling who died of S.I.D.S., confirmed at necropsy. Control group:- Infants in this group were clinically normal and recruited at random from the Outpatients Delmrlment at the routine 6 week well-baby clinic or on the postnatal ward. A family history of S.I.D.S. or A.L.T.E. was excluded. Procedure All infants were admitted overnight to the Paediatric Unit TABLE I Characteristics of study groups Group
Mean (SD) B.W. (Kg.)
A.L.T.E (N=20) 3.58 (0.53) Siblings of S3.D.S. (N=21) 3.37 (0.41) Control (N=13) 3.54 (0.59)
Age (WK) 9.6 (2.4) 9.25 (2.7) 11.5 (1.3)
*Address for correspondence: Deparmaentof Paedialrics, Rotunda Hospital, Dublin 1. Tel. 730700. 249
and the study was carried out during normal nocturnal sleep. No medication was given. All investigations were carried out (by G.P.P.F.) between 00.00 and 04.00 hrs. in the same room. Each patient was in quiet sleep during the study period (i.e., regular breathing, no body movement, eyes closed and not moving, and chest movements in phase with abdominal movement) and results were included in the analysis only when the infant remained in this state during the study procedure. Each infant was placed in his or her normal sleep position, dressed as usual for sleep, using a similar amount of clothing and bedding in each case, and was studied at three different ambient temperatures. Tremperature measurement:- Core temperature was measured throughout the procedure by means of a rectal temperature probe placed 1 cm inside the infant's rectum, skin temperature with a skin probe attached to the infant's back, and environmental temperature with an air temperature probe placed within 60 cms of the infant's face. All three probes were attached to a Corometrics 512 neonatal cardiorespiratory monitor (Coromenics Medical Systems Inc., Wallingford, Connecticut, USA). The three readings were displayed continuously and recorded on paper strip. The neonatal monitor was also used to give a continuous recording of heart rate, including long-term and short-term variability, and chest wall movements on a paper sUnupmoving at a speed of 3 cm/min. Long term variability (LTV) was defined as the difference in the minium and maximum values of the neonatal heart rate over a period of 512 R waves expressed in beats per minute. Short term variability (STV) was def'med as the average change in event to event time over a period of 512 R waves expressed in milleseconds. Environmental temperature was confirmed with a mercury thermometer within 60 cms of the infant's face. Change in ambient temperature:- Each infant was assessed at ambient temperatures of 20~ 25-26~ and 30~ The normal temperature for the room used in the study was 2526~ To raise the temperature an electrical convector heater, 3 m from the cot, was used; there was no obvious noise from the fan in the heater and the infants remained in quiet sleep. The average time for the room temperature to rise to 30~ was 13 mins. The room tempelrature was lowered to 20~ by opening the upper part of the room window, the average time for the temperature to fall was 17 mins. Care was taken to
250 F o x et al.
x.+.~.s.
August. 1991 TABLE H
AmbientTeanperamre*C Heart Rate (N=53) RespiratoryRate (N=50) S.T.V. (N=54) L.T.V. 0q=53) End Tidal CO~2(N=49) KPA
20
25
30
P
127 (11A) 26.3 (7.4) 12.3 (4.1) 28.8 (14.9) 4.5 (0.9)
131 (11.1) 35.7 (7.9) 10.9 (4.1) 24.0 (12.0) 4.5 (0.8)
133 (11.1) 43.5 (8.2) 11.6 (4.4) 28.2 (15.1) 4.4 (1.0)
< 0001 < 0.001 < 0.05 < 0.05 N.S.
L.T.V. - Long term heart rate variability S.T.V. - Short term h e m rate variability This table shows the effect of three different ambient temperatures on heart rate, respiratory rate, L.T.V., S.T.V. and end tidal CO 2 levels values
given are mean (SD). ensure that the infant waS not in a direct draught. The infants remained at the desired temperature fcr 20 mins to allow adaptationbeforephysiologicalvariables were recorded. The normal rectal temperature was taken as 37~ (_+0.5~ and infants were studied only if the rectal temperatt.~ remained within this range. The only variable changed throughout the study period was the air temperature, with each baby acting as his or her own control. End tidal carbon dioxide values were measured using a Gould capnograph Mark 3 CO2analyser (Gould Medical BV, P.O. Box 73, 3720 AB Bilthoven, The Netherlands) with the results recorded on a paper strip recorder. Sampling of gas from the anterior nares was performed by placing a soft silastic catheter (I.D. 1 mm. [0.062"] into a nostril for 5-10 rains). If the infant could not tolerate this, the catheter tip was held manually without touching the inside of the nose. Satisfactory recordings of end-expiratory CO2give a plateau from which the CO2 level is calculated, and expressed in Kilo pascals (KPa) adjusted to take account of the daily barometric pressure. The machine was recalibrated using 5% CO2 gas between each study. The dislance from anterior nares to the sampling point of the CO2 analyser was never more than 25 cms.
This study was approved by the Rolamda Research Committee and parental consent was obtained in each case. Statistical Analysis For each of the variables - heart rate, respiratory rate, short and long term heart variability, and end tidal carbon dioxide measurements- a 3 x 3 analysis of variance was carried out to examine whether: 1. there was a significant differencebetween groups regardless of temperature; 2. there was a significant difference between temperatures regardless of groups; 3. there was a difference in the profile of the three groups over time (group x temperaune effect).
Results The results of the analyses of variance show that for each of the variables there was no significant difference between the groups and the profiles for the groups over the three temperatures. With ihe exception of end tidal carbon measurements, there was a significant difference in the responses at the three temperature levels. Table II presents the means and standard deviations for each of the variables at each of the temperatures. As there was no significant difference between groups
or group x temperature effect the means are presented for the total sample. Infants were included when results were available at each ambient temperature studied. All infants skin and rectal temperature measurements remained within the normal range during the study period. Discussion This study has shown that some commonly measured physiological variables are influenced by the ambient temperature of the room where the study is conducted. To minimise the effect of additional variables we studied all infants in the same room, at the same time of the night, used the same amount of bedding and clothing and used a standard study procedure conducted by the same person (G.P.P.F.). Obviously future studies looking for subtle differences in physiological variables between groups of infants should strictly control the thermal environment where the studies are carried out. The fact that in this study the resting respiratory rate increased from 26.3 at 20~ to 43.5 at 30~ with no 'change in the end tidal CO2levels (4.5 KPA at 20~ 4.4 KPA at 30~ implies that more carbon dioxide is being expired secondary to an increased metabolic rote despite the infants core and skin temperature remaining within the normal range. As the infant's head and neck is the only part directly exposed to the varying ambient temperatures it is posible that some of the observed effects are mediated by local temperature receptors. This paper reinforces the need to improve our understanding of the interaction between extrinsic "environmental" factors, such as ambient temperature and an infant's homeostatic control mechanisms.
Acknowledgements We wish to thank the Irish Sudden Infant Death Association and the Friends of the Rotunda for jointly funding this study and Audrey Dixon for typing the paper.
1. 2. 3. 4. 5.
RegePences Valdes-Dapena, M.A. SuddenInfantDeathSyndrome:Areviewofthe medical literature 1974-79. Pediatrics 1980: 66, 579-614 Suddell Infant Death Syndrome. Medical aspects and psychological management, Ed. Culbertson, Jl., Krous, H. F, Bendell, R. D. 1989. Edward Arnold, London. M ~ h y , M. F. G., C a m ~ i l , M. L Sudden Infant Death Syndrome and environmental temperature: an analysis using vital statistics. B. J. Epideminl. Commun. I/ealth 1987: 41, 63-71. Wailoo, M. P., Petersen, S. A., Whittaker, H., Goodenough, P. The thermal environment in which 3-4 month olds sleep at home. Arch. Dis. Child. 1989: 64, 6130-4. Prechil, H. F.R. The behavioural states of the newborn infant (a review). Brain Res. 1974: 76, 185-212.
The Influence of Ambient Temperature on Some Commonly Measured Physiological Variables in Infants §
P. P. Fox, *M. O'Regan, §
G. Matthews
+Rotunda Hospital, Dublin; *Department of Statistics, Trinity College, Dublin 2. Summary Fifty-four infants underwent recording of heart rate, respiratory rate, long term and short term heart rate variability and end tidal carbon dioxide measurements at three different ambient temperatures (20~ 25-26~ 30~ under standardized conditions. The infants' skin and rectal temperature remained normal throughtout the study. The ambient temperature of the room where the study was conducted significantly influenced the infants' heart rate, respiratory rate and long and short term heart rate variability but not end tidal carbon dioxide values. The thermal environment in which an infant resides has a significant influence on some physiological variables and should be standardized in future physiological studies. Introduction The Sudden Infant Death Syndrome (S.I.D.S.) remains the major cause of infant deaths in the developed world and has been the subject of intense research over the past twenty years especially in the area of cardio-respiratory control~-~. Although lower ambient temperatures have been associated with increased rates of S.I.D.S), few physiological studies in infants have taken account of the ambient temperature. In this paper we report the effect of relatively modest changes in ambient temperature, well within the range in which infants are nursed at home4, on some commonly measured physiological variables in healthy infants. Subjects and Methods Study Groups (Table I). Apparent Life Threatening events (A.L.T.E.):- For this study the definition used was an infant being found lifeless and apparently dead or almost dead, requiring vigorous resuscitation including mouth-to-mouth resuscitation, in whom all investigations were negative and who was subsequently clinically normal. All infants had fully recovered clinically from the A.L.T.E. episode when our studies were carried out48 hrs. to 8 days later. Siblings of S.I.D.S.:- Infants in this group had a sibling who died of S.I.D.S., confirmed at necropsy. Control group:- Infants in this group were clinically normal and recruited at random from the Outpatients Delmrlment at the routine 6 week well-baby clinic or on the postnatal ward. A family history of S.I.D.S. or A.L.T.E. was excluded. Procedure All infants were admitted overnight to the Paediatric Unit TABLE I Characteristics of study groups Group
Mean (SD) B.W. (Kg.)
A.L.T.E (N=20) 3.58 (0.53) Siblings of S3.D.S. (N=21) 3.37 (0.41) Control (N=13) 3.54 (0.59)
Age (WK) 9.6 (2.4) 9.25 (2.7) 11.5 (1.3)
*Address for correspondence: Deparmaentof Paedialrics, Rotunda Hospital, Dublin 1. Tel. 730700. 249
and the study was carried out during normal nocturnal sleep. No medication was given. All investigations were carried out (by G.P.P.F.) between 00.00 and 04.00 hrs. in the same room. Each patient was in quiet sleep during the study period (i.e., regular breathing, no body movement, eyes closed and not moving, and chest movements in phase with abdominal movement) and results were included in the analysis only when the infant remained in this state during the study procedure. Each infant was placed in his or her normal sleep position, dressed as usual for sleep, using a similar amount of clothing and bedding in each case, and was studied at three different ambient temperatures. Tremperature measurement:- Core temperature was measured throughout the procedure by means of a rectal temperature probe placed 1 cm inside the infant's rectum, skin temperature with a skin probe attached to the infant's back, and environmental temperature with an air temperature probe placed within 60 cms of the infant's face. All three probes were attached to a Corometrics 512 neonatal cardiorespiratory monitor (Coromenics Medical Systems Inc., Wallingford, Connecticut, USA). The three readings were displayed continuously and recorded on paper strip. The neonatal monitor was also used to give a continuous recording of heart rate, including long-term and short-term variability, and chest wall movements on a paper sUnupmoving at a speed of 3 cm/min. Long term variability (LTV) was defined as the difference in the minium and maximum values of the neonatal heart rate over a period of 512 R waves expressed in beats per minute. Short term variability (STV) was def'med as the average change in event to event time over a period of 512 R waves expressed in milleseconds. Environmental temperature was confirmed with a mercury thermometer within 60 cms of the infant's face. Change in ambient temperature:- Each infant was assessed at ambient temperatures of 20~ 25-26~ and 30~ The normal temperature for the room used in the study was 2526~ To raise the temperature an electrical convector heater, 3 m from the cot, was used; there was no obvious noise from the fan in the heater and the infants remained in quiet sleep. The average time for the room temperature to rise to 30~ was 13 mins. The room tempelrature was lowered to 20~ by opening the upper part of the room window, the average time for the temperature to fall was 17 mins. Care was taken to
250 F o x et al.
x.+.~.s.
August. 1991 TABLE H
AmbientTeanperamre*C Heart Rate (N=53) RespiratoryRate (N=50) S.T.V. (N=54) L.T.V. 0q=53) End Tidal CO~2(N=49) KPA
20
25
30
P
127 (11A) 26.3 (7.4) 12.3 (4.1) 28.8 (14.9) 4.5 (0.9)
131 (11.1) 35.7 (7.9) 10.9 (4.1) 24.0 (12.0) 4.5 (0.8)
133 (11.1) 43.5 (8.2) 11.6 (4.4) 28.2 (15.1) 4.4 (1.0)
< 0001 < 0.001 < 0.05 < 0.05 N.S.
L.T.V. - Long term heart rate variability S.T.V. - Short term h e m rate variability This table shows the effect of three different ambient temperatures on heart rate, respiratory rate, L.T.V., S.T.V. and end tidal CO 2 levels values
given are mean (SD). ensure that the infant waS not in a direct draught. The infants remained at the desired temperature fcr 20 mins to allow adaptationbeforephysiologicalvariables were recorded. The normal rectal temperature was taken as 37~ (_+0.5~ and infants were studied only if the rectal temperatt.~ remained within this range. The only variable changed throughout the study period was the air temperature, with each baby acting as his or her own control. End tidal carbon dioxide values were measured using a Gould capnograph Mark 3 CO2analyser (Gould Medical BV, P.O. Box 73, 3720 AB Bilthoven, The Netherlands) with the results recorded on a paper strip recorder. Sampling of gas from the anterior nares was performed by placing a soft silastic catheter (I.D. 1 mm. [0.062"] into a nostril for 5-10 rains). If the infant could not tolerate this, the catheter tip was held manually without touching the inside of the nose. Satisfactory recordings of end-expiratory CO2give a plateau from which the CO2 level is calculated, and expressed in Kilo pascals (KPa) adjusted to take account of the daily barometric pressure. The machine was recalibrated using 5% CO2 gas between each study. The dislance from anterior nares to the sampling point of the CO2 analyser was never more than 25 cms.
This study was approved by the Rolamda Research Committee and parental consent was obtained in each case. Statistical Analysis For each of the variables - heart rate, respiratory rate, short and long term heart variability, and end tidal carbon dioxide measurements- a 3 x 3 analysis of variance was carried out to examine whether: 1. there was a significant differencebetween groups regardless of temperature; 2. there was a significant difference between temperatures regardless of groups; 3. there was a difference in the profile of the three groups over time (group x temperaune effect).
Results The results of the analyses of variance show that for each of the variables there was no significant difference between the groups and the profiles for the groups over the three temperatures. With ihe exception of end tidal carbon measurements, there was a significant difference in the responses at the three temperature levels. Table II presents the means and standard deviations for each of the variables at each of the temperatures. As there was no significant difference between groups
or group x temperature effect the means are presented for the total sample. Infants were included when results were available at each ambient temperature studied. All infants skin and rectal temperature measurements remained within the normal range during the study period. Discussion This study has shown that some commonly measured physiological variables are influenced by the ambient temperature of the room where the study is conducted. To minimise the effect of additional variables we studied all infants in the same room, at the same time of the night, used the same amount of bedding and clothing and used a standard study procedure conducted by the same person (G.P.P.F.). Obviously future studies looking for subtle differences in physiological variables between groups of infants should strictly control the thermal environment where the studies are carried out. The fact that in this study the resting respiratory rate increased from 26.3 at 20~ to 43.5 at 30~ with no 'change in the end tidal CO2levels (4.5 KPA at 20~ 4.4 KPA at 30~ implies that more carbon dioxide is being expired secondary to an increased metabolic rote despite the infants core and skin temperature remaining within the normal range. As the infant's head and neck is the only part directly exposed to the varying ambient temperatures it is posible that some of the observed effects are mediated by local temperature receptors. This paper reinforces the need to improve our understanding of the interaction between extrinsic "environmental" factors, such as ambient temperature and an infant's homeostatic control mechanisms.
Acknowledgements We wish to thank the Irish Sudden Infant Death Association and the Friends of the Rotunda for jointly funding this study and Audrey Dixon for typing the paper.
1. 2. 3. 4. 5.
RegePences Valdes-Dapena, M.A. SuddenInfantDeathSyndrome:Areviewofthe medical literature 1974-79. Pediatrics 1980: 66, 579-614 Suddell Infant Death Syndrome. Medical aspects and psychological management, Ed. Culbertson, Jl., Krous, H. F, Bendell, R. D. 1989. Edward Arnold, London. M ~ h y , M. F. G., C a m ~ i l , M. L Sudden Infant Death Syndrome and environmental temperature: an analysis using vital statistics. B. J. Epideminl. Commun. I/ealth 1987: 41, 63-71. Wailoo, M. P., Petersen, S. A., Whittaker, H., Goodenough, P. The thermal environment in which 3-4 month olds sleep at home. Arch. Dis. Child. 1989: 64, 6130-4. Prechil, H. F.R. The behavioural states of the newborn infant (a review). Brain Res. 1974: 76, 185-212.
