RESULTS

VIRAL INFECTIONS AND VERY PRETERM BIRTH (I)

Viremia was more often detected in first trimester blood samples from pregnant women who delivered very preterm, compared to women who delivered at term (Table 7). Any viremia was found in 10 women with very preterm deliveries and in 5 women with term deliveries. Of the positive samples, Parvovirus B19 was detected in 7 preterm cases and in 3 term controls. Although non-significant, risk estimates indicated that any viremia and Parvovirus B19 was associated with a twofold increase in risk of very preterm delivery.

Table 7. Prevalences of viral infections and odds ratios of very preterm delivery.

Cytomegalovirus, found in one preterm case, is not included in the table below.

Virus + Virus - Odds ratios (95% CI)

No. (%) No. (%) crude adjusted^†

Any viral infection

Very preterm delivery 10 (3.7) 259 (96.3) 2.28 (0.77-6.77) 2.21 (0.71-6.84)

Term delivery* 5 (1.7) 296 (98.3) 1.00 1.00

Parvovirus B19^‡

Very preterm delivery 7 (2.6) 258 (97.4) 2.66 (0.68-10.4) 2.66 (0.64-11.1)

Term delivery* 3 (1.0) 294 (99.0) 1.00 1.00

Human herpes virus 6

Very preterm delivery 2 (0.7) 267 (99.3) 1.12 (0.16-8.00) 1.08 (0.14-8.08)

Term delivery* 2 (0.7) 299 (99.3) 1.00 1.00

* reference

† adjusted for maternal age and parity

‡ not analyzed in 4 cases and 4 controls

One may speculate that early pregnancy viremia would increase the risk especially for spontaneous preterm labour and extremely preterm birth (delivery before 28 weeks).

However, the group of very preterm infants to virus positive mothers were not different from the group of very preterm infants to virus negative mothers. Etiologies of very preterm delivery were similarly distributed in the two groups, as were gestational ages and birth weights.

PRETERM DELIVERY, LEVEL OF CARE AND INFANT MORTALITY (II) During the years 1992 to 1998, 2,253 infants were born before 32 completed weeks of gestation, of whom 267 (12.6 percent) died during the first year of life. The majority of deaths occurred during the neonatal period (n=232).

In the univariate analysis, infants born in general hospitals had a reduced risk of infant mortality compared with infants born in university hospitals (Table 8). In the multivariate analysis, infants born in general hospitals faced a 33 percent increased risk of infant mortality. This shift was primarily due to different gestational age distributions in university and general hospitals. Besides level of care, gestational age, birth weight for gestational age and infant sex were also associated with risk of infant mortality.

Table 8. Risk of infant mortality associated with level of care.

Infant mortality Odds ratios (95% CI)

(%) unadjusted adjusted^† Hospital type

- university hospitals* 14.2 1.00 1.00

- general hospitals 10.3 0.70 (0.54 – 0.90) 1.33 (0.88 – 2.02)

Gestational week

- 24 55.6 23.96 (13.48 – 42.62) 41.48 (18.23 – 94.38)

- 25 35.5 10.78 (6.54 – 17.79) 16.67 (8.78 – 31.62)

- 26 19.9 4.75 (2.76 – 8.18) 6.12 (2.87 – 13.03)

- 27 13.5 3.01 (1.76 – 5.15) 3.46 (1.89 – 6.35)

- 28 9.7 2.08 (1.20 – 3.62) 2.46 (1.37 – 4.41)

- 29 6.8 1.40 (0.81 – 2.44) 1.42 (0.75 – 2.67)

- 30 6.4 1.32 (0.77 – 2.26) 1.41 (0.73 – 2.74)

- 31* 5.0 1.00 1.00

Birth weight for gestational age

- large (LGA) 44.4 7.54 (3.46 – 16.43) 10.90 (3.18 – 37.38)

- appropriate (AGA)* 9.6 1.00 1.00

- small (SGA) 14.7 1.63 (1.24 – 2.16) 3.08 (1.94 – 4.89)

Infant sex

- male 13.2 1.33 (1.02 – 1.73) 1.64 (1.13 – 2.38)

- female* 10.3 1.00 1.00

*reference

† adjusted for mode of delivery, hospital type, gestational age, birth weight for gestational age, infant sex, foetal presentation, placental complications, and maternal hypertensive illness There was a significant interaction between gestational age and hospital type with regard to infant mortality (p=0.049). Stratified analyses showed that the increased mortality risk in general hospitals was confined to extremely preterm infants (Table 9).

Table 9. Infant mortality by hospital type, stratified by gestational age.

Infant mortality Odds ratios (95% CI)

(%) unadjusted adjusted^†

24–27 weeks

- university hospital* 22.8 1.00 1.00

- general hospital 32.2 1.60 (1.02 – 2.50) 2.00 (1.15 – 3.49) 28 –31 weeks

- university hospital* 8.1 1.00 1.00

- general hospital 5.8 0.69 (0.45 – 1.07) 0.83 (0.51 – 1.33)

*reference

† adjusted for the covariates as in Table 8.

PRETERM BIRTH AND BLOOD PRESSURE IN ADULTHOOD (III) Characteristics of the conscripted men (n=329,495) are presented in Table 10.

Table 10. Characteristics of men conscripted for military service.

Gestational age (weeks)

24-28 29-32 33-36 37-41 42-43

At birth

Birth weight, grams 1,192 (270) 1,825 (426) 2,745 (507) 3,590 (484) 3,815 (479) At conscription

Age, years 18.2 (0.4) 18.2 (0.4) 18.2 (0.4) 18.2 (0.4) 18.2 (0.4) BMI, kg/m² 22.1 (3.2) 22.1 (3.1) 22.3 (3.3) 22.3 (3.2) 22.4 (3.3) Systolic BP, mm Hg 132 (13) 131 (12) 130 (11) 129 (11) 128 (11) Values are mean (SD).

The proportion of men with high systolic blood pressure varied from 32 percent among men born extremely preterm (<28 weeks) to 19 percent among men born postterm (>42 weeks) (Table 11). Linear regression analyses showed that systolic blood pressure increased with decreasing gestational week (0.31 mm Hg/week, p<0.001). Compared to men born at term (37-41 weeks), men born extremely preterm faced a two-fold increased risk of high systolic blood pressure (Table 11). Being born SGA was associated with a 10 percent increased risk of high systolic blood pressure, compared to being born AGA.

Table 11. Perinatal characteristics and risk of high systolic blood pressure at conscription.

High systolic BP Odds ratios (95% CI)

(%) crude adjusted^†

Gestational age (weeks)

24 – 28 31.5 1.81 (1.30-2.52) 1.88 (1.33-2.68)

29 – 32 26.6 1.44 (1.27-1.62) 1.45 (1.28-1.64)

33 – 36 23.5 1.21 (1.16-1.26) 1.24 (1.19-1.30)

37 – 41* 20.2 1.00 1.00

42 – 43 18.7 0.91 (0.89-0.94) 0.90 (0.88-0.93)

Birth weight for gestational age

SGA 21.5 1.09 (1.04-1.14) 1.10 (1.05-1.15)

AGA* 20.2 1.00 1.00

LGA 19.2 0.94 (0.89-0.99) 0.90 (0.85-0.95)

*reference

† adjusted for gestational age, birth weight for gestational age, age and parity of the mother, height at conscription, conscription year, family situation of the conscriptor, and parents’

educational level and socio-economic status.

The risk of high systolic blood pressure (>140 mm Hg) per one-week decrease in gestational age increased both within and between families (adjusted odds ratios were 1.03 [1.01-1.05] and 1.06 [1.04-1.08], respectively), indicating that an association between gestational age and systolic blood pressure exists after controlling for common genetic and shared environmental factors after birth.

Finally, there was a significant interaction between gestational age and birth weight for gestational age (p=0.03) with regard to high systolic blood pressure (• 140 mm Hg).

To enable stratified analyses by gestational age, extremely and very preterm births were collapsed into one category (<32 weeks). SGA was not a risk factor of high systolic

PRETERM BIRTH, BIRTH WEIGHT, AND TYPE 2 DIABETES (IV) In the cohort of 18,230 like-sexed twins, rates of type 2 diabetes consistently increased with decreasing birth weight (Table 12). Compared to twins born at term (37 to 41 gestational weeks), type 2 diabetes was slightly more common among preterm twins.

Table 12. Rates of type 2 diabetes by perinatal characteristics.

Rate of type 2 diabetes Birth weight (grams) (%)

-1,999 5.3

2,000 to 2,499 3.9

2,500 to 2,999 2.7

3,000 to 3,499 2.6

3,500 - 2.4

Gestational age (weeks)

31 – 34 3.7

35 – 36 3.3

37 – 41 3.0

42 – 45 3.3

In cohort analyses, there was no association between gestational age and type 2 diabetes, while the risk of type 2 diabetes increased with decreasing birth weight.

Compared to twins with birth weight from 2,500 to 2,999 grams (reference), twins with birth weight less than 2,000 grams had a twofold increase in risk of type 2 diabetes, after adjustment for perinatal and adult covariates. When birth weight was used as a continuous measure, a 500-gram decrease in birth weight was associated with a 44 percent increase in risk.

To elucidate whether the association between low birth weight and type 2 diabetes was confounded by genetic factors, risks were calculated in twin pairs discordant for type 2 diabetes (Table 13). An increased risk of type 2 diabetes with lower birth weight was found within dizygotic but not within monozygotic twin pairs.

Table 13. Risks for type 2 diabetes among twin pairs, discordant for type 2 diabetes, per 500 grams difference in birth weight.

Co-twin case-control odds ratio (95% CI) ^† Dizygotic twin pairs (N=206)

twin with type 2 diabetes 1.38 (1.02-1.85)

non-diabetic twin* 1.00

Monozygotic twin pairs (N=97)

type 2 diabetes 1.02 (0.63-1.64)

non-diabetic twin* 1.00

* Reference

† Twin pairs are by definition matched for shared environmental and common genetic factors.

Finally, the association between birth weight and type 2 diabetes was explored in additional co-twin case-control analyses, using subsets of discordant twin pairs.

Genetic confounding was found also within twin pairs in which the diabetic twin was treated with an antidiabetic drugs, within twin pairs born at term, and within twin pairs with moderate as well as large differences in birth weight.