• No results found

Blood Pressure in Late Adolescence and Very Low Birth Weight

N/A
N/A
Protected

Academic year: 2020

Share "Blood Pressure in Late Adolescence and Very Low Birth Weight"

Copied!
8
0
0

Loading.... (view fulltext now)

Full text

(1)

Blood Pressure in Late Adolescence and Very Low Birth Weight

Lex W. Doyle, MD, FRACP*‡; Brenda Faber, RN‡; Catherine Callanan, RN‡; and Ruth Morley, MB, BS§

ABSTRACT. Objectives. To determine whether blood pressure (BP) differed between very low birth weight (VLBW; birth weight<1500 g) subjects and normal birth

weight (NBW; birth weight >2499 g) subjects in late adolescence, and to determine whether growth restriction in utero was related to BP in VLBW survivors at this age.

Methods. This was a cohort study of 210 preterm sur-vivors with birth weights<1501 g born from January 1, 1977, to March 31, 1982, and 60 randomly selected NBW subjects from the Royal Women’s Hospital, Melbourne. BP was measured at 18years of age in 156 (74%) VLBW subjects and 38 (63%) NBW subjects with both a standard mercury sphygmomanometer and an ambulatory BP monitor.

Results. VLBW subjects had higher sphygmomanom-eter systolic and diastolic BPs than NBW subjects (mm Hg; mean difference [95% confidence interval]; systolic, 8.6 [3.4, 13.9]; diastolic, 4.3 [1.0, 7.6]). VLBW subjects also had significantly higher mean systolic ambulatory BPs (mm Hg; mean difference [95% confidence interval]) for the 24-hour period (4.7 [1.4, 8.0]), and for both the awake (5.0 [1.6, 8.5]) and asleep (3.6 [0.04, 7.1]) periods. There were no significant differences between the birth weight groups for any ambulatory diastolic BPs. Within the VLBW subjects, there was no significant relationship between birth weight standard deviation score and any measure of BP.

Conclusions. BP was significantly higher in late ado-lescence in VLBW survivors than in NBW subjects. Growth restriction in utero was not significantly related to BP in VLBW survivors. Pediatrics 2003;111:252–257;

very low birth weight, ambulatory blood pressure, hyper-tension, growth restriction in utero.

ABBREVIATIONS. BP, blood pressure; VLBW, very low birth weight; NBW, normal birth weight; ABP, ambulatory blood pres-sure; SD, standard deviation; CI, confidence interval; IQR, inter-quartile range.

H

igher blood pressure (BP) has been described in adult subjects of lower birth weight, and has been ascribed to growth restriction in utero.1 However, few of the subjects in the early studies were of very low birth weight (VLBW; birth weight ⱕ1500 g).2,3 Since the early studies, other

researchers have reviewed the literature and con-firmed the relationship between lower birth weight and higher later BP.4,5Others have described higher BP in childhood in VLBW survivors compared with normal birth weight (NBW; birth weight ⬎2499 g) subjects.6 However, it remains to be established whether the relationship between VLBW and later higher BP is related to growth restriction in utero.

The aim of this study was to determine whether BP differed between VLBW subjects and NBW subjects in late adolescence, and to determine whether growth restriction was related to BP in VLBW survi-vors at this age.

METHODS

The VLBW cohort comprised 210 subjects of birth weight

⬍1501 g who were all born in the Royal Women’s Hospital, Melbourne. From January 1, 1977, to March 31, 1982, 86 (33.2%) of 259 consecutive live births with birth weights 500 to 999 g sur-vived. From October 1, 1980, to March 31, 1982, 124 (90.5%) of 137 consecutive live births with birth weights 1000 to 1500 g survived. All VLBW subjects were preterm births (gestational age range: 24 to 36 completed weeks). There were 60 NBW subjects who were born in 1981–1982 within the Royal Women’s Hospital. They were selected randomly by the end digit of their medical record num-ber, and all were term (gestational age range: 37– 42 completed weeks). All of these cohorts were recruited in the newborn period with the intention of following them throughout childhood and beyond, primarily to compare outcomes for VLBW subjects with those of NBW. BP was one of the comparisons between the birth weight groups, which was always intended. The sample sizes of the respective groups were determined on outcomes earlier in childhood and not with respect to the power to detect differences in BP in later life.

Extensive perinatal data had been collected, including gesta-tional age and birth weight, gender, the occurrence of maternal hypertension in pregnancy, and antenatal corticosteroid therapy. Details of the early neonatal care of these cohorts have been described.7,8 No child received corticosteroids in the newborn period.

Survivors were enrolled in a long-term follow-up program that was approved by the Research and Ethics Committees of the Royal Women’s Hospital. They had been assessed several times earlier in childhood, and were assessed again at 18⫹years of age. Written informed consent was obtained from all subjects. In sub-jects who were preterm, their age was corrected for prematurity, ie, calculated from their expected date of birth at term rather than their actual birth date, to be consistent with all of our previous reports for this cohort. A family history of elevated BP was re-corded if any first- or second-degree relatives had been treated for high BP.

BP was measured in 2 ways. The first way was with a standard mercury sphygmomanometer. Subjects were seated, after at least 5 minutes of rest, and the cuff size was two thirds of the upper arm length. The mean of 3 readings was recorded for both systolic and diastolic pressures. The diastolic BP was recorded as the pressure at the disappearance of sound (Korotkoff V). There were 2 re-search nurses who were trained how to measure sphygmoma-nometer BP and they were asked to record BP to the nearest 2 mm Hg. We used a normal and not a random zero sphygmomanom-eter, to replicate normal clinical circumstances for

sphygmoma-From the *Departments of Obstetrics and Gynecology, and Pediatrics, Uni-versity of Melbourne, Melbourne, Australia; ‡Division of Newborn Ser-vices, the Royal Women’s Hospital, Melbourne, Australia; and the §Depart-ment of Clinical Epidemiology and Biostatistics, the Royal Children’s Hospital, Melbourne, Australia.

Received for publication Sep 21, 2002; accepted Jul 1, 2002.

Address correspondence to Prof Lex W Doyle, Department of Obstetrics and Gynecology, University of Melbourne, Parkville 3052, Australia. E-mail: lwd@unimelb.edu.au

(2)

nometer BP measurement. The 2 research nurses who assessed the children did so without any knowledge of perinatal details, but they were aware of the birth weight of the subjects. Second, BP was measured with an ambulatory BP (ABP) monitor (Spacelabs 90207, Redmond, WA). The fitted cuff was a size appropriate for the mid-arm circumference; the subjects were given instructions in the monitor’s operation, and they were asked to wear it for 24 hours, if possible. Subjects had to remove the ABP for showers or vigorous sporting activities, and therefore they would not neces-sarily have worn it for 24 hours continuously. We averaged all valid readings and recorded this as just one value, regardless of the absolute number of individual observations that contributed to the average. The cuff was programmed to inflate every 30 minutes between 6amand midnight, and every hour between midnight and 6am. For a few subjects known to work at night, the timing was altered accordingly to match their period of antici-pated sleep. The subjects were asked to complete a diary of their activities over the 24-hour period, which was used to determine when they were awake and asleep. The ABPs were averaged for the whole 24-hour period, and separately for the awake and asleep periods. The change in ABPs when asleep compared with awake were calculated in absolute (mm Hg) and relative (%) terms. Elevated ABP was determined by means for systolic and diastolic above the height and sex-specific 95th centiles for ABP reported by Soergel et al9Systolic and diastolic loads were estimated as the percentage of measurements above the height and sex-specific 95th centiles for ABP.10

Height and weight were measured according to standard guidelines. Birth weight, height, and weight were converted to standard deviation (SD) scores relative to the British Growth Reference.11The reader is referred to Cole et al11for more details on the precise method. This method calculates the birth weight SD score relative to that expected for the child’s gender and gesta-tional age.

Data were edited and analyzed using SPSS for Windows pro-grams (SPSS Inc, Chicago, IL).12Dichotomous variables were com-pared by␹2analysis. Continuous variables were contrasted by the mean difference and 95% confidence intervals (CIs), or by Mann-WhitneyUtest if the data were skewed. The relationships between BPs and potential confounding variables were analyzed by step-wise multiple linear regression, and adjusted mean differences and 95% CIs between the birth weight subgroups were computed. Within the VLBW group, the relationship between BP and birth weight SD score was determined by linear regression, with and without adjustment for confounding variables. Although the orig-inal study groups were determined by outcomes earlier in child-hood, with data for 156 VLBW subjects and 38 NBW subjects we had 80% power to detect a difference in BP between groups of 0.5 SD, with a type I error of 5%.

RESULTS

BP was measured in 156 (74%) VLBW subjects and 38 (63%) NBW subjects. One VLBW subject had only sphygmomanometer BP measured, and 2 subjects (1 VLBW, 1 NBW) had only ABP BP measured. Of the 54 VLBW subjects without a BP measurement, 1 was assessed in a remote location but no BP was

re-corded, 8 were lost, 26 refused, 3 were too disabled, 7 were contacted by telephone only, and 9 were too remote (5 living in other countries, 4 living in other states). Of the 22 NBW subjects without a BP mea-surement, 4 were lost, 12 refused, 1 was too disabled, 2 were contacted by telephone only, and 3 were too remote (2 living in other countries, 1 living in another state). There were no significant differences in peri-natal variables between VLBW subjects with and without BPs (Table 1), or between NBW subjects with and without BPs (data not shown).

BP was measured on 75.6% (65/86) of those of birth weight⬍1000 g, and 73.4% (91/124) of those of birth weight 1000 to 1500 g (␹2⫽0.13; P.72, not significant). There were no significant differences in any BP between those of birth weight ⬍1000 g and those of birth weight 1000 to 1500 g (mean difference [95% CI]; sphygmomanometer: systolic 2.8 [⫺2.0, 7.6], diastolic 1.8 [⫺1.3, 4.8]; 24-hour ambulatory: systolic⫺1.7 [⫺4.7, 1.4], mean⫺1.3, [⫺3.6, 1.0], dia-stolic⫺1.0, [⫺3.3, 1.2]), and hence they were consid-ered together as the VLBW group. The median age of BP measurement was 18.6 (interquartile range [IQR]: 18.0, 19.4) years in the VLBW group, and 18.5 (IQR: 18.3, 18.5) years in the NBW group (z⫽0.6,P⫽.56, not significant). Age at assessment was not related to BP within the age range studied.

In those with BP measurements, VLBW subjects were significantly less mature and lighter at birth, as expected, and their birth weight SD score was sub-stantially lower than the NBW group (Table 2). More of the mothers of VLBW subjects had hypertension in pregnancy and more were treated with antenatal corticosteroids for fetal lung maturity. More of the VLBW subjects were from multiple pregnancies. There were no significant differences between the birth weight groups in family history of hyperten-sion or gender. At 18 years of age, weight and height SD scores were significantly lower in VLBW subjects. Compared with NBW subjects, those of VLBW had significantly higher BPs for sphygmomanometer sys-tolic and diassys-tolic, ambulatory syssys-tolic over 24 hours, awake and asleep, and ambulatory mean over 24 hours and when awake (Table 3). The mean number of ABP readings per subject was similar in both groups (VLBW, 41.1; NBW, 40.7; mean difference, 0.4; 95% CI:⫺1.6, 2.4). The differences in ambulatory diastolic BPs and in mean BP when asleep were not

TABLE 1. Perinatal Variables in VLBW Subjects With and Without BP Measurements BP Statistical Significance

Yes n⫽156

No n⫽54

Gestational age (completed wk), mean (SD) 28.8 (2.0) 28.9 (2.6) ⫺0.1 (⫺0.7, 0.6)* Birth weight (g), mean (SD) 1098 (235) 1110 (246) ⫺12 (⫺87, 62)* Birth weight SD score, mean (SD) ⫺0.69 (1.03) ⫺0.70 (1.18) 0.01 (⫺0.32, 0.35)* Mother with hypertension in

pregnancy†,n(%)

37 (23.7) 7 (13.0) ␹22.80;P.09

Antenatal corticosteroid therapy,n(%) 82 (52.6) 21 (38.9) ␹23.00;P.08 Multiple birth,n(%) 31 (19.9) 10 (18.5) ␹20.05;P.83

Male,n(%) 73 (46.8) 33 (61.1) ␹23.29;P.07

* Mean difference (95% CI). † Includes preeclampsia.

(3)

statistically significant. Of the potentially confound-ing variables in Table 2, and includconfound-ing age when assessed, most BPs in Table 3 were significantly higher with a positive family history of hypertension (all except the ABP mean when asleep) and in males (all except ABP diastolics and ABP mean when awake). Weight SD score at 18 was positively related to some BPs, as was being a multiple birth. Sphyg-momanometer BP was significantly higher, but ABP systolic over 24 hours, systolic and mean when awake were significantly lower with increasing age at assessment. Adjusting for these significantly con-founding variables had little effect on the statistical conclusions concerning BP differences between VLBW and NBW subjects (Table 3). The remaining variables in Table 3 were not significantly related to BP. The sizes of the differences in BPs between the VLBW and NBW groups were similar in both males and females separately, but not all differences were statistically significant (Table 3).

The mean falls in ABPs with sleep were of similar size in VLBW and NBW subjects, both in absolute and percentage terms (Table 4).

The proportions of subjects with all ambulatory systolic BPs⬎95th centile were significantly higher in the VLBW group, but the proportions with all diastolic BPs ⬎95th centile were not significantly different between the groups (Table 5). The ABP loads for readings ⬎95th centile were significantly higher in the VLBW group for systolic BPs over 24 hours, and both awake and asleep, and for diastolic BP but only when awake (Table 5).

In VLBW subjects, there were no significant rela-tionships between birth weight SD score and any BP measurements (Table 6, Fig 1). The proportion of variance in any BP variable explained by the birth weight SD score was ⬍0.3%. There were many growth-restricted VLBW survivors, but there was no birth weight SD score below which the BP was any higher (Fig 1). The regression coefficients for the

TABLE 2. Perinatal and Growth Variables in Subjects With BP Measurements at 18⫹Years of Age, Contrasting Birth Weight Groups Birth Weight Group Statistical

Significance* VLBW

n⫽156

NBW n⫽38

Perinatal variable

Gestational age (completed wk), mean (SD) 28.8 (2.0) 40.0 (1.1) ⫺11.2 (⫺11.8,⫺10.5) Birth weight (g), mean (SD) 1098 (235) 3493 (494) ⫺2395 (⫺2502,⫺2287) Birth weight SD score, mean (SD) ⫺0.70 (1.03) 0.02 (0.90) ⫺0.72 (⫺1.08,⫺0.36) Family history of hypertension,n(%) 68 (43.6) 12 (31.6) ␹21.81,P.18 Mother with hypertension in pregnancy,n(%) 37 (23.7) 1 (2.6) ␹2⫽8.6,P.003 Antenatal corticosteroid therapy,n(%) 82 (52.6) 0 (0) ␹234.6,P.0001

Male,n(%) 73 (46.8) 20 (52.6) ␹2⫽0.42,P.52

Multiple birth,n(%) 31 (19.9) 0 (0) ␹29.0,P.003

Growth at 18⫹years

Weight SD score, mean (SD) 0.12 (1.36)‡ 0.52 (1.41) ⫺0.40 (⫺0.89, 0.09) Height SD score, mean (SD) ⫺0.37 (1.11)‡ 0.29 (0.95) ⫺0.65 (⫺1.04,⫺0.26)

* Mean difference and 95% CI; unless otherwise stated. † Includes preeclampsia.

N⫽10 without weight or height data, either because nonambulant from cerebral palsy (n⫽2), or not measured (n⫽8).

TABLE 3. BP Data by Birth Weight Group and Gender

Birth Weight Group Mean Difference

(95% CI)

Adjusted Mean Difference*

(95% CI)

Males, n⫽73 Mean Difference

(95% CI)

Females, n⫽83

Mean Difference

(95% CI) VLBW

n⫽156

NBW n⫽38

Sphygmomanometer blood pressure (mm Hg), mean (SD)

Systolic 124.8 (14.8)† 116.1 (14.1)† 8.6 (3.4, 13.9) 10.6 (5.8, 15.5) 8.6 (1.2, 15.9) 9.5 (2.2, 16.8) Diastolic 72.4 (9.6)† 68.2 (7.2)† 4.3 (1.0, 7.6) 3.8 (0.8, 6.9) 3.6 (⫺1.0, 8.3) 5.2 (0.5, 9.9) Ambulatory BP (mm Hg),

mean (SD)

Systolic—24 h 122.1 (9.6) 117.4 (7.4) 4.7 (1.4, 8.0) 5.2 (2.1, 8.3) 6.4 (1.6, 11.1) 3.8 (⫺0.2, 7.8) Mean—24 h 87.6 (7.2)† 85.1 (4.8) 2.4 (0.02, 4.9) 2.2 (⫺0.1, 4.6) 3.1 (⫺0.7, 6.9) 2.0 (⫺1.0, 5.0) Diastolic—24 h 69.2 (7.0)† 68.0 (4.8) 1.1 (⫺1.2, 3.5) 0.8 (⫺1.5, 3.2) 1.5 (⫺2.4, 5.3) 0.6 (⫺2.3, 3.6) Systolic—awake 126.7 (9.9)‡ 121.6 (8.6) 5.0 (1.6, 8.5) 5.9 (2.8, 9.1) 6.7 (1.7, 11.7) 4.5 (0.2, 8.6) Mean—awake 91.7 (7.5)‡ 88.8 (5.6) 2.9 (0.3, 5.5) 2.7 (0.1, 5.3) 3.4 (⫺0.7, 7.4) 2.7 (⫺0.6, 6.1) Diastolic—awake 73.5 (7.4)‡ 72.0 (5.6) 1.5 (⫺1.0, 4.1) 1.3 (⫺1.2, 3.8) 1.6 (⫺2.4, 5.7) 1.3 (⫺2.0, 4.5) Systolic—asleep 109.8 (9.8)§ 106.3 (7.6)¶ 3.6 (0.05, 7.1) 3.8 (0.4, 7.2) 5.2 (0.5, 10.0) 3.2 (⫺1.7, 8.1) Mean—asleep 76.1 (10.0)§ 75.2 (6.0)¶ 0.9 (⫺2.5, 4.4) 1.4 (⫺2.0, 4.8) 2.2 (⫺1.8, 6.2) 0.4 (⫺5.3, 6.1) Diastolic—asleep 58.0 (7.3)§ 57.0 (6.0)¶ 0.9 (⫺1.7, 3.6) 0.7 (⫺1.9, 3.3) 1.2 (⫺2.8, 5.1) 0.6 (⫺3.0, 4.2)

* Adjusted for confounding variables. †N⫽1 with missing data.

(4)

change in any BP measurement with 1 SD increase in birth weight SD score were all ⬍1 mm Hg; some coefficients were negative and some were positive (Table 6). Of the possible confounding variables in Table 2, all BPs in Table 6 were significantly higher with a positive family history of hypertension (in-crease ranging from 4.0 –7.0 mm Hg), and all except the ambulatory diastolic pressure were significantly higher in males (increase ranging from 2.9 –9.6 mm Hg) and with increasing weight SD score at 18 years of age (increase ranging from 0.9 –3.6 mm Hg per 1 SD increase in weight SD score). No other potential confounding variables were statistically significant, including multiple birth, gestational age, or birth

weight. Height SD score was not statistically signif-icant after weight SD score entered the analysis. Ad-justing for the statistically significant confounding variables had little effect on the nonsignificant rela-tionship between any BP and birth weight SD score (Table 6). No statistical conclusions were altered if all of the independent variables in Table 2 were simul-taneously forced into the regression analysis.

DISCUSSION

In our study, VLBW subjects had higher sphygmo-manometer systolic and diastolic BPs than NBW sub-jects, and higher mean systolic ambulatory BPs. There were no significant differences between the

TABLE 4. Change in ABP With Sleep

Birth Weight Group Mean Difference

(95% CI) VLBW

n⫽145

NBW n⫽35

Change in ABP with sleep, mean (SD) Systolic

Absolute change in mm Hg ⫺16.9 (6.9) ⫺16.4 (6.8) ⫺0.5 (⫺3.1, 2.0) % change ⫺13.3 (5.2) ⫺13.3 (5.2) 0.0 (⫺2.0, 1.9) Mean

Absolute change in mm Hg ⫺15.6 (8.3) ⫺15.3 (5.8) ⫺0.3 (⫺2.6, 2.0) % change ⫺21.0 (8.2) ⫺21.0 (7.5) 0.0 (⫺2.9, 3.0) Diastolic

Absolute change in mm Hg ⫺15.6 (8.5) ⫺14.2 (5.3) ⫺1.4 (⫺4.4, 1.5) % change ⫺16.9 (9.5) ⫺15.8 (5.7) ⫺1.1 (⫺4.4, 2.2)

TABLE 5. Proportions Above 95th Centile for ABP and ABP Loads in Birth Weight Subgroups Birth Weight Group Statistical

Significance VLBW

n⫽145

NBW n⫽38

ABP⬎95th centile,n(%) ␹2⫽,P

Systolic—24 h 49 (33.8) 4 (10.5) 7.9, .005

Diastolic—24 h 25 (17.2) 2 (5.3) 3.4, .064

Systolic—awake 36 (26.1) 1 (2.6) 9.9, .002

Diastolic—awake 7 (5.1) 0 (0) 2.0, .16

Systolic—asleep 29/137 (21.2) 1/35 (2.9) 6.5, .011 Diastolic—asleep 13/137 (9.5) 3/35 (8.6) 0.03, .87

ABP load, %-median (IQR) z⫽,P

Systolic—24 h 34.9 (19.0, 57.2) 22.0 (8.2, 40.2) 3.2, .002* Diastolic—24 h 25.0 (12.5, 40.2) 24.4 (13.7, 38.4) 0.64, .52* Systolic—awake 25.6 (8.9, 48.3) 10.0 (3.0, 31.0) 3.4, .001* Diastolic—awake 10.9 (3.1, 24.1) 6.7 (1.7, 9.8) 2.1, .040* Systolic—asleep 15.4 (0, 43.0)† 7.7 (0, 25.0)‡ 2.7, .007* Diastolic—asleep 10 (0, 22.2)† 8.3 (0, 16.7)‡ 0.80, .42*

* Mann-WhitneyUtest. †n⫽8 with missing data. ‡n⫽3 with missing data.

TABLE 6. Growth Restriction in Utero and BP in VLBW Subjects Regression Coefficient*

and 95% CI

Adjusted Regression Coefficient*† and 95% CI

Sphygmomanometer blood pressure (mm Hg)

Systolic ⫺0.57 (⫺2.89, 1.75) ⫺1.08 (⫺3.43, 1.27) Diastolic ⫺0.05 (⫺1.56, 1.46) ⫺0.38 (⫺1.91, 1.14) ABP (mm Hg)

Systolic ⫺0.36 (⫺1.86, 1.12) ⫺0.14 (⫺1.62, 1.35)

Mean ⫺0.16 (⫺1.29, 0.96) 0.16 (⫺1.05, 1.37)

Diastolic 0.12 (⫺0.98, 1.22) 0.54 (⫺0.66, 1.75)

* Represents change in BP for 1 SD increase in birth weight SD score.

(5)

birth weight groups for any ambulatory diastolic BPs. Within the VLBW subjects, there was no signif-icant relationship between birth weight SD score and any measure of BP.

In the original studies reporting on the relation-ship between low birth weight and higher BP in later life, few subjects would have been VLBW. In the first of Barker et al’s original papers2 birth weight sub-groups were determined by tertiles, and the sample sizes in each birth weight stratum were not specified. There were⬃500 males and 500 females at 10 years old in the lowest tertile with birth weights approxi-mately ⬍3000 g from an original sample of 9921 children born in 1970, and⬍237 males and females at 36 years of age of similar birth weight born in 1946. In their next study of the relationship between birth weight and BP, there were only 45 of 449 subjects at ages 46 to 54 years with birth weights⬍512pounds (2500 g) born before 1945.3 Given the diminishing number of births with lower birth weight and the low survival rates for VLBW infants at the time of birth, few subjects in either study would have been ⬍1500 g birth weight.

Since the original studies, other researchers have described higher BP in childhood in VLBW survi-vors. Pharoah et al6 reported that systolic BP mea-sured with a Dinamap (an oscillometric BP device; Critikon, Tampa, FL) was significantly higher at 15 years of age in VLBW subjects (birth weight range: 650 g to 1500 g) compared with controls (birth weight range: 2098 – 4550 g); the mean difference was 3.2 mm Hg (95% CI: 0.4 – 6.0). They did not find a sig-nificant difference, however, in diastolic BP (mean difference: 1.1; 95% CI: ⫺0.7–2.9). In contrast, the sizes of the mean differences in systolic BPs between VLBW and NBW subjects in our study were greater, whether measured with a sphygmomanometer or with the ABP monitor. Moreover, there was a signif-icantly higher diastolic BP in VLBW subjects in our study, but only when measured by the sphygmoma-nometer. A possible reason for the greater

differ-ences in BP between VLBW and NBW subjects in our study compared with Pharoah et al6is related to the older age at assessment in our study, as others have described a bigger influence of low birth weight on BP with increasing age.13

The advantages of ABP measurement over the conventional sphygmomanometer include improved objectivity in measurement (the ABP does not know the birth weight of the subject), as well as avoidance of “white coat” hypertension.14 It must be recog-nized, however, that the 2 methods of measuring BP are not identical. For example, the Spacelabs 90207 ABP monitor measures only the mean BP, and com-putes the systolic and diastolic pressures according to an internal algorithm. This may explain, in part, why there were no significant differences on the ABP diastolic pressures between groups but there was with the sphygmomanometer BP. Not only can the overall means of ABPs be compared, but the propor-tions with ABP readings above the 95% centile for height and gender9 can also be compared. In our study, the proportion of VLBW subjects with systolic ABP readings⬎95th centile was substantially higher than the NBW subjects.

Apart from avoiding expectation bias, the ABP offers other advantages over conventional BP moni-toring. One advantage is the ability to compare awake with asleep BPs. Normally BP should fall at least 10% with sleep.10 Reduction or an absence of the fall with sleep is associated with end-organ dam-age in hypertensive states,15 and occurs in various conditions, such as renal disease and diabetes melli-tus.10 In our study, there was no difference in the diurnal variation in BP between VLBW and NBW subjects. ABP monitoring also allows for calculation of the BP loads, both systolic and diastolic, which may also be more predictive of the stress of hyper-tension on end-organs.10 In our study, VLBW sub-jects had significantly increased systolic BP loads compared with NBW subjects, as well as an in-creased diastolic load when awake. Another advan-tage of ABP monitoring might include an improve-ment in tracking of BP over time in adolescence,16 but it is unknown whether tracking occurs in VLBW subjects.

Although growth restriction in utero is related to higher BP in term and near term subjects with birth weights ⬎1500 g,17 it remains to be established whether the relationship between VLBW and later BP is related more to growth restriction in utero than just to prematurity per se. The only other study of which we are aware that relates growth restriction in utero with BP within VLBW subjects was by Steven-son et al.18They reported that systolic BP at 15 years of age was significantly higher in VLBW subjects compared with those of greater birth weight. They also assessed the additive effect of birth weight ratio (subject’s birth weight divided by the expected birth weight for gestational age) on BP but reported that there was no significant correlation in the VLBW group. They did not, however, provide any data for the relationship. In another study of subjects of birth weight⬍2000 g, Irving et al19reported that systolic BP was 7 mm Hg higher at 24 years of age compared

(6)

with those of birth weight⬎2000 g. They also did not find any relationship between BP and growth restric-tion at birth in those of birth weight⬍2000 g. In our study, growth restriction in utero, as assessed by the birth weight SD score, was not significantly associ-ated with any BP, either unadjusted or adjusted for confounding variables of family history of hyperten-sion, gender, and weight SD score. Our observation is consistent with Stevenson et al.18The direction of the effect of growth restriction on sphygmomanom-eter systolic BP in our study was, however, in the same direction as expected by the fetal origins hy-pothesis. Given that we had relatively few subjects compared with the larger epidemiologic studies that formed the basis for the fetal origins hypothesis, it is possible that we were underpowered to find a sta-tistically significant relationship between birth weight SD score and BP in the VLBW subjects. An-other consideration is the questionable validity of birth weight norms for infants who are born prema-turely and who are VLBW, because these may well differ from weights of infants who remain in utero until term. It is possible that preterm children who are born VLBW are growth restricted relative to fe-tuses of a similar gestation remaining in the uterus to term, and hence the growth restriction hypothesis might apply more generally to all VLBW subjects.

Regardless of the lack of statistical significance of growth restriction and BP, the VLBW subjects in our study had substantially higher BPs, more were above the 95th centile, and they had higher BP loads than the NBW subjects, at a relatively young age. This raises 2 issues. The first issue is to ask why was their BP so much higher, if it was not caused by growth restriction in utero? Perhaps preterm birth itself, or the early days of postnatal life expose VLBW subjects outside the uterus to systematically different condi-tions than experienced by those still inside the uterus. These differences may be sufficient to “pro-gram” later higher BP. Indeed, such influences may be operating outside the uterus at a time and in much the same way as for more mature fetuses that remain in the uterus through the third trimester, but who subsequently have higher BP with lower birth weight.17

The second issue is the long-term health for the surviving VLBW subjects. Already at a relatively young age, they have higher BPs than their peers, and proportionally more have BPs in clinically im-portant ranges. Although VLBW subjects comprise approximately only 1% of all births, they may be destined to comprise a much higher proportion of

adults with hypertensive disease. This effect will only be accentuated by the higher survival rates of VLBW subjects in the current millenium, compared with the era when our subjects were born.

ACKNOWLEDGMENTS

This study was supported in part by a grant from the Royal Women’s Hospital Research Foundation and by VicHealth (the Victorian Health Promotion Foundation) (to Ruth Morley).

REFERENCES

1. Barker DJP.Mothers, Babies and Disease in Later Life. Vol 341. London, United Kingdom: BMJ Publishing Group; 1998

2. Barker DJ, Osmond C, Golding J, Kuh D, Wadsworth ME. Growth in utero, blood pressure in childhood and adult life, and mortality from cardiovascular disease.Br Med J.1989;298:564 –567

3. Barker DJ, Bull AR, Osmond C, Simmonds SJ. Fetal and placental size and risk of hypertension in adult life.Br Med J.1990;301:259 –262 4. Law CM, Shiell AW. Is blood pressure inversely related to birth weight?

The strength of the evidence from a systematic review of the literature.

J Hypertens.1995;14:935–941

5. Huxley RR, Shiell AW, Law CM. The role of size at birth and postnatal catch-up growth in determining systolic blood pressure: a systematic review of the literature.J Hypertens.2000;18:815– 831

6. Pharoah PO, Stevenson CJ, West CR. Association of blood pressure in adolescence with birthweight.Arch Dis Child Fetal Neonatal Ed.1998;79: F114 –F118

7. Kitchen WH, Yu VY, Lissenden JV, Bajuk B. Collaborative study of very-low-birthweight infants: techniques of perinatal care and mortal-ity.Lancet.1982;1:1454 –1457

8. Kitchen WH, Ryan MM, Rickards A, et al. Changing outcome over 13 years of very low birthweight infants.Semin Perinatol.1982;6:373–389 9. Soergel M, Kirschstein M, Busch C, et al. Oscillometric

twenty-four-hour ambulatory blood pressure values in healthy children and adolescents: a multicenter trial including 1141 subjects.J Pediatr.1997; 130:178 –184

10. Sorof JM, Portman RJ. Ambulatory blood pressure monitoring in the pediatric patient.J Pediatr.2000;136:578 –586

11. Cole TJ, Freeman JV, Preece MA. British 1990 growth reference centiles for weight, height, body mass index and head circumference fitted by maximum penalized likelihood.Stat Med.1998;17:407– 429

12.SPSS for Windows. Version 9.0.1. Chicago, IL: SPSS Inc; 1999 13. Moore VM, Cockington RA, Ryan P, Robinson JS. The relationship

between birth weight and blood pressure amplifies from childhood to adulthood.J Hypertens.1999;17:883– 888

14. Sorof JM, Portman RJ. Ambulatory blood pressure measurements.Curr Opin Pediatr.2001;13:133–137

15. Verdecchia P, Clement D, Fagard R, Palatini P, Parati G. Blood pressure monitoring. Task force III: target-organ damage, morbidity and mortal-ity.Blood Press Monit.1999;4:303–317

16. O’Sullivan JJ, Derrick G, Foxall RJ. Tracking of 24-hour and casual blood pressure: a 1-year follow-up study in adolescents.J Hypertens.2000;18: 1193–1196

17. Leon DA, Johansson M, Rasmussen F. Gestational age and growth rate of fetal mass are inversely associated with systolic blood pressure in young adults: an epidemiologic study of 165,136 Swedish men aged 18 years.Am J Epidemiol.2000;152:597– 604

18. Stevenson CJ, West CR, Pharoah PO. Dermatoglyphic patterns, very low birth weight, and blood pressure in adolescence.Arch Dis Child Fetal Neonatal Ed.2001;84:F18 –F22

(7)

DOI: 10.1542/peds.111.2.252

2003;111;252

Pediatrics

Lex W. Doyle, Brenda Faber, Catherine Callanan and Ruth Morley

Blood Pressure in Late Adolescence and Very Low Birth Weight

Services

Updated Information &

http://pediatrics.aappublications.org/content/111/2/252

including high resolution figures, can be found at:

References

http://pediatrics.aappublications.org/content/111/2/252#BIBL

This article cites 17 articles, 4 of which you can access for free at:

Subspecialty Collections

sub

http://www.aappublications.org/cgi/collection/fetus:newborn_infant_ Fetus/Newborn Infant

following collection(s):

This article, along with others on similar topics, appears in the

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtml

in its entirety can be found online at:

Information about reproducing this article in parts (figures, tables) or

Reprints

http://www.aappublications.org/site/misc/reprints.xhtml

(8)

DOI: 10.1542/peds.111.2.252

2003;111;252

Pediatrics

Lex W. Doyle, Brenda Faber, Catherine Callanan and Ruth Morley

Blood Pressure in Late Adolescence and Very Low Birth Weight

http://pediatrics.aappublications.org/content/111/2/252

located on the World Wide Web at:

The online version of this article, along with updated information and services, is

by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.

Figure

TABLE 1.Perinatal Variables in VLBW Subjects With and Without BP Measurements
TABLE 2.Perinatal and Growth Variables in Subjects With BP Measurements at 18� Years of Age, Contrasting Birth Weight Groups
TABLE 5.Proportions Above 95th Centile for ABP and ABP Loads in Birth Weight Subgroups
Fig 1. Systolic BP (sphygmomanometer) and birth weight SDscore. Regression line and its 95% CI shown

References

Related documents

quantitative information and real time images o f fluid distribution in incontinence pads... 2:

The purpose of this study was to use HCAHPS survey data to examine relationships among patient-reported perceptions of health care provider communication, whether new medications

They investigated the HOXD9 function in glioma cell lines, and their results suggested that silencing HOXD9 could effectively suppress the proliferation of U87 glioma cells and

Conclusion: These results indicate that expression of UBE4B was higher in most NPC tis- sues compared to adjacent non-tumoral tissues, and that knockdown of UBE4B inhibited the

The tendency towards degradation is attributed to the large /«rerligand repulsions between the methyl substituents on the pyrazolyl groups and the hydrogen atoms

We examined the expression of the top four prognosis-related DEGs (NPAS2, GNG7, CHIA and SLC2A1) in lung cancer cell lines using qRT-PCR to validate the results of gene

This dissertation describes studies of G protein-coupled receptors (GPCRs) and ligand-gated ion channels (LGICs) using unnatural amino acid mutagenesis to gain high precision

As demonstrated in Figure 3A, the luciferase activities were substantially increased in a dose-dependent fashion and the activities were promoted to nearly 6-fold after 30 μ M