a) Reading Comprehension

     L
     1
    0
-- I\      Test for differences between
             3 smoking categories after
    -1
$g -2          adjusting for other factors
            X2 (2 D.F.) = 130.1; P < 0.001

`5     '
rs     n = 8,545) (n = 1,981)  (n = i.489)
        1         I             r
       0      l-9         lO+
Amount smoked per day after 4th month of pregnancy

b) Mathematics Ability

Test for differences between
3 smoking categories after
adjusting for other factors
X2 (2 D.F.) = 198.0; P <O.OOl

if   -5.
E   -6-  -1,489,
-7 (n = 8,543 (n = 1,980)

    d     119       16+
Amount smoked per day after 4th month of pregnancy

c) Height

Test for differences between
Test for differences between
3 smoking categories
3 smoking categories
after adjusting for other factors
after adjusting for other factors
X2 (2 D.F.) = 60.4; P ( 0.001
X2 (2 D.F.) = 60.4; P ( 0.001

143" (n = 7,649) (n = 1,729)
143" (n = 7,649) (n = 1,729)   (n= 1,316)
                (n= 1,316)
    0 0      l-9
          l-9          lO+
                     lO+

   Amount smoked per day after 4th month of pregnancy
FIGURE 5 .-Tests of 11-year-old children by mothers' smoking
      habits after the fourth month of pregnancy

SOURCE: Butler, N.R. (16).

201


hearing mechanism had occurred in infants of smokers, possibly
due to a hypoxic effect of carbon monoxide on the cochlear
organ during development (132).

These studies suggest unfavorable effects of maternal smok-
ing during pregnancy on the child's long-term growth, intellec-
tual development, and behavioral characteristics. Although
these changes are difficult to study because of the vast complex-
ity of possible antecedent and confounding variables, high
priority should be given to obtaining conclusive answers about
the long-term consequences of fetal exposure to cigarette
smoke.

ROLE OF MATERNAL WEIGHT GAIN

In the search for mechanisms through which maternal smok-
ing reduces birth weight, the question has been asked whether
it might be an indirect result of reduced appetite, less intake of
food, and lower maternal weight gain (84,127). Several early
studies reported no differences between smoking and nonsmok-
ing women in intake of food or in weight gain, and concluded
that the effect of maternal smoking on birth weight was not
mediated in this way (146).

Meyer analyzed the relationships between maternal smoking,
birth weight, maternal weight gain, and gestation, using data
based on 31,788 births from the Ontario Perinatal Mortality
Study (106,107). She found a significant downward shift in the
distribution of birth weights as maternal smoking level in-
creased, but no similar shift in the distribution of maternal
weight gain with smoking. Whereas the usual strong relation-
ship between the proportion of births under 2,500 grams and
maternal smoking level was found, there was no similar trend
for the proportion of mothers who gained less than 10 pounds
during pregnancy. Finally, the proportion of infants weighing
less than 2,500 grams increased directly with the amount
smoked within each maternal weight gain group from less than
5 pounds to 40 pounds or more, as shown in Figure 6 (83). From
Figure 6, one might conclude that smoking has a more pro-
nounced effect on low birth weight when maternal weight gain
during pregnancy is less than 20 pounds.

Other studies have indicated a lack of relationship between
smoking and maternal weight gain, while demonstrating a di-
rect relationship between smoking and fetal growth rate. The
German prospective study of 6,200 pregnant women, examined
every month from the first trimester through delivery, showed
no significant association between ,smoking habit and weight
gain. The usual relationships were found between smoking and
202


E    25         Private Hospital Status
zg s"  20                    0 Nonsmoker
    15                    m c 1 Pack/Day
",g  10   8 B               1 + Pack/Day

r-
r    5
i5    0                                   40+

Maternal Weight Gain (pounds)

E    25         Public Hospital Status
a
8    20
G
4  15
ti  10
(on
f-
r    5
iii    0   o-4   5-9  lo-14 15-19 20-24 25-29 30-34 35-39  40 +

Maternal Weight Gain (pounds)

FIGURE 6.-Percentage of birth weights under 2,500 grams by
      maternal smoking level within maternal weight gain
      group (five-pound intervals) by hospital pay status.
      Births of 38 + weeks gestation (Ontario study)
SOURCE: Meyer, M.B. (83).

small-for-dates babies, with general retardation of weight,
length, and head circumference in proportion to the number of
cigarettes smoked during pregnancy (80). Miller and Hassanein
also found that the effects of smoking on fetal growth did not
appear to be related to maternal nutrition (93). Persson's study
showing retardation of fetal growth of smokers' babies by serial
measurement of biparietal diameters and by weight, length,
and other measurements at birth showed that the low birth
weights were independent of maternal weight gain. These au-
thors concluded that the fetal growth retardation resulted from
a direct pharmacological effect of smoking on the fetus "rather
than an influence resulting from nutritional deprivation" (114).

Hajeri and colleagues studied maternal weight gain in 105
smokers of 10 or more cigarettes a day with a control group of
nonsmokers who were similar with respect to gestation, age,
height, parity, and maternal weight at conception. Birth
weights, specific for sex, were significantly higher for infants of
                                       203


TABLE 3.-Birth weight under 2,500 gm by maternal smoking
      and prepregnant weight

Births ~2,500 gm per 100
    Total Births
Maternal Smoking        Ratio
  (Packs per day)    Smoker:Nonsmoker

Prepregnant       Total
Weight           Births

0      <l     If    `-Cl      1+

< 120 lb
(< 54 kg)
120-134 lb
(54-61 kg)
135+ lb
C-61 kg)

18,935     6.1     10.2    15.8     1.7      2.6


19,798     4.2     6.3     9.5     1.5      2.3


10,456     3.3     5.1     8.7     1.6      2.6

SOURCE: Meyer, M.B. (86).

nonsmokers, with a mean difference for boys of 330 grams and
for girls of 320 grams (pc.01). Mean extrauteral weight gain,
calculated as the difference between maternal weight gain and
the weights of fetus and placenta, was 7,044 grams for smokers
and 6,899 grams for nonsmokers (49).
Garn has compared mean birth weights, specific for gesta-
tional age, of babies of obese smokers, all nonsmokers, and all
smokers, using data from the Collaborative Perinatal Project of
the National Institute of Neurological and Communicative Dis-
orders and Stroke (NINCDS). Obesity was defined as the top 15
percent of the distribution of prepregnant weights, shown sepa-
rately for black and white women. Babies of the 1,383 obese
white smokers had mean birth weights similar to the total
group of white nonsmokers and higher than the total group of
white smokers. The 1,001 obese black smoking mothers had
babies whose mean birth weights were generally higher than
those of all black nonsmokers, leading Garn to conclude that
"maternal obesity (weight-defined) apparently counteracts the
smoking effect on the conceptus" (43). Because birth weight is
strongly correlated with maternal size, a more appropriate
comparison would have been between mean birth weights of the
babies of obese smokers and the babies of obese nonsmokers.
That such a comparison would show the usual relationship to
maternal smoking level is suggested by Meyer's analysis of
birth weight by maternal smoking and prepregnancy weight
(Table 3). The correlation between maternal weight and the
proportion of low-birth-weight babies is clear at each smoking
level, and the independent relationship between smoking level
and low birth weight is clear at each level of maternal weight.
204


TABLE 4 .-Mean birth weights in successive pregnancies to the
      same women, by smoking habit

                                 Mean

Smoking Habits   Smoking Habits   Birth Weight (gm)

   First          Second                       Difference
  pregnancy       pregnancy     N     Xl    #2   end-1st (gm)

Smoker        Smoker         886    3204   3228     +24
Nonsmoker      Nonsmoker       988    3356    3388      +32
________________________________________-----------------------------------------------------------------------
Difference: Nonsmoker - Smoker (gm)        +152   +160

Smoker      Nonsmoker       119    3271    3381      +110
Nonsmoker      Smoker         108 3323    3265      -58
__________________________________________________________________---------------------------------------------------
Difference: Nonsmoker - Smoker (gm)        +52 +116

SOURCE: Naeye, k. (93).

The relative increases in the proportion of low-weight births
with light and with heavy smoking are almost identical in the
three strata of prepregnant weight (86).

Studies of birth weight, maternal weight, and maternal
weight gain should also be carefully controlled for maternal age
and parity. In studies of successive births to the same mother
included in the Collaborative Perinatal Project of the NINCDS,
Garn found that prepregnancy weights increased with succes-
sive pregnancies by similar amounts for smokers and nonsmok-
ers (44). Naeye, using the same data base, reported that mater-
nal weight gain was less in the second pregnancy than in the
first pregnancy for smokers, for nonsmokers, and for women
who changed habits between pregnancies in either direction
(93). Second babies weighed on the average 24 grams more than
first babies if the mother smoked both times, and 32 grams more
if the mother smoked neither time (Table 4). If the mother
smoked during the first and not during the second pregnancy,
the second baby weighed an average of 110 grams more than the
first baby; in women who smoked during the second pregnancy
but not during the first pregnancy, second babies averaged 58
grams less than first babies (93).

The most careful analyses indicate that the effect of maternal
smoking is a direct one not mediated through an effect on mat-
ernal appetite, eating, or weight gain. In conclusion, as stated in
a Lancet editorial, "the appeal of the nutritional hypothesis is
that women might be more readily encouraged to eat more dur-
ing pregnancy than discouraged from smoking. . . . However, if,
                                        205


as now seems more likely, the growth-retarding effect of smok-
ing is due to fetal hypoxia, there is no short-cut to removing
this adverse influence" (63). This conclusion in no way obviates
the enormous importance of dietary factors during pregnancy.

Overt maternal malnutrition is associated with inadequate
growth. Recently, it has been suggested that more subtle alter-
ations in the maternal supply of essential nutrients combined
with compromised uteroplacental circulation may contribute to
reduced fetal growth. Crosby, et al. (26) observed that the con-
centrations of each of 14 amino acids and carotene were reduced
significantly in the blood of smoking mothers, These workers
postulated that, while these differences were on the order of 10
or `20 percent, they could be an important factor in producing
the small-for-gestational-age infants associated with maternal
smoking. In a study of over 1,100 pregnant women, Schorah, et
al. (135) noted an inverse correlation between the number of
cigarettes smoked and the leukocyte ascorbic acid concentra-
tion. For instance, the leukocyte ascorbic acid concentration
was about 22 percent less in the blood of women who smoked
more than 20 cigarettes a day as compared with controls. De-
spite a 15 percent increase in the number of circulating leuko-
cytes in the blood of smokers, the blood ascorbic acid concentra-
tion was still 10 percent less than in controls. These differences
were even more marked in women from lower socioeconomic
groups. The authors suggested that in addition to the role of
ascorbic acid in fetal nutrition, these lowered concentrations
might be related to the increased incidence of premature rup-
ture of the amniotic membranes in smoking women.

Smoking, Fetal and Infant Mortality, and Morbidity

SPONTANEOUS ABORTION

Past studies have demonstrated a statistically significant as-
sociation between maternal cigarette smoking and spontaneous
abortion (55,61,104), some showing a strong dose-response re-
lationship (110,144,162). Spontaneous abortions are difficult to
study because of problems of ascertainment. In prospective
studies, early abortions may be missed, and bias may occur if
one group tends to register earlier than the other. Retrospect-
ive studies allow more complete ascertainment but are subject
to errors of recall. Nevertheless, higher rates of spontaneous
abortion have been associated with maternal smoking in both
types of studies (61,104,162).

Kullander and Kallen found higher rates of "spontaneous
abortion" among smoking women, but noted that many of these
206


pregnancies were unwanted. Analysis of their data showed that
the relative risk of spontaneous abortion of smokers compared
with nonsmokers was 1.20 for wanted and 1.35 for unwanted
pregnancies (61). A case-control study of spontaneous abortion
with important variables held constant reported an 80 percent
increase in the odds of smoking among the cases compared with
controls (60).

Recent studies corroborated the finding of associations be-
tween smoking and spontaneous abortion risk. In a small retro-
spective study in New Zealand, Fergusson found that women
who smoked more than 20 cigarettes a day had almost twice the
nonsmoker risk of having had a previous spontaneous abortion,
and that the association could not be explained by differences in
maternal age, educational level, parity, race, socioeconomic
status or marital status (42). In a study of 12,013 consecutive
pregnancies in Dublin, Ireland, Murphy and Mulcahy found a
positive association between the number of cigarettes smoked
and the rates of spontaneous abortion, independent of the ef-
fects of maternal age and parity. The authors stated that in-
duced abortions are a negligible factor in Ireland and concluded
that maternal smoking leads to reduced reproductive efficiency
at all stages of pregnancy (92). Himmelberger and colleagues
surveyed a group of professional women in medicine concerning
the influence of maternal smoking on their 12,194 pregnancies
(54). After controlling for interfering variables, the risk of spon-
taneous abortion for certain subgroups of heavy smokers was
estimated to be as much as 1.7 times that for nonsmokers. Spon-
taneous abortion rates were lowest in the 25 to 29 year old cate-
gory, increasing with age to levels of 33 and 36 percent for
nonsmokers and smokers, respectively, at age 40 plus. The rela-
tive increase in risk associated with maternal smoking was
highest at the youngest ages and decreased with increasing age
(54).
An editorial in the British Medical Journal summarized these
findings and stated: "Cigarette smoking, one of the first man-
ifestations of women's social emancipation, is emerging as a
possible threat to her procreative role." The proportion of ab-
normal karyotypes in abortuses of women who smoke appears
to be reduced rather than increased (1). The mechanism under-
lying the smoking-related excess appears to be due to complica-
tions of pregnancy rather than to any fetal abnormality (13).

CONGENITAL MALFORMATIONS
Several studies have reported perinatal, fetal, or neonatal
mortality rates by cause. In these comparisons, death rates due
                                        207


TABLE 5.--Incidence of congenital abnormality (all single births)

Nonsmokers       Smokers

Number Percent Number Percent

Total abnormal infants
Type of abnormality
Anencephaly
Spina bifida
Other C.N.S. abnormality
Cardiovascular abnormality
Gut abnormality
Genito-urinary abnormality
Bone abnormality
Cleft palate and/or hare lip
Other abnormality

      2.37            2.73


18      0.2       15      0.2
20      0.22      23      0.3
38      0.42      36     0.47
34      0.37      32      0.42
21      0.23      24     0.32
39      0.43      25      0.33
65      0.72      52      0.68
10      0.11      20     0.26
19      0.21      18      0.24

x2 (all abnormalities) = 2.22, ~~0.05.
x2 (cleft palate and hare lip) = 5.36, 0.01~ ~(0.05.
SOURCE: Andrews, J. (3).

to congenital malformations have usually been lower for smok-
ers' than for nonsmokers' infants (3,22,46,87). This is compatible
with the finding that smoking-related spontaneous abortions
have a lower frequency of abnormal karyotypes and tend to
occur later than spontaneous abortions in nonsmokers. As pre-
viously described, increased losses of conceptus associated with
maternal smoking appear to be due to pregnancy problems and
complications rather than to abnormalities of the embryo or
fetus (41). Andrews and McGarry, in a community study of
18,631 pregnancies in Cardiff, Wales, reported that smokers' in-
fants had lower mortality rates from malformations than those
of nonsmokers. Rates of stillbirths due to congenital malforma-
tions were 0.32 and 0.27 per 100 nonsmokers and smokers re-
spectively. Corresponding rates for neonatal deaths were 0.33
and 0.31 per 100 babies of nonsmoking and smoking mothers. On
the other hand, the incidence of congenital malformations
among all single births in Andrews' population was higher
among smokers' babies, overall, and specifically higher for cleft
palate and lip. Among other sites, some were higher for smokers
and some for nonsmokers, as is shown in Table 5 (3).
A significant positive association between cardiac malforma-
tions and maternal smoking was shown by Fedrick and col-
leagues, based on firm diagnoses among stillbirths, neonatal
deaths, and survivors to age 7 from the British Perinatal Mor-
tality Survey. However, this difference was largely due to the
inclusion of patent ductus arteriosus, which may or may not be
classified as a malformation (80).
208


Heavy Smoker

a  0~""""""""`~
     18    22    26    30   34    38   42    46    50
               Maternal Age

FIGURE 7.-Risk of congenital abnormality according to age and
       smoking habit

SOURCE: Himmel'berger, D.U. (54).

Some recent studies have shown a positive association be-
tween maternal smoking and congenital malformations, defined
in a variety of ways. Himmelberger and colleagues carried out a
mail survey of professional women in medicine (54). They were
interested in exposure to anesthetic gases in the operating
room, and evaluated possible effects on pregnancy outcome of a
number of factors including cigarette smoking. Information was
obtained and analyzed by a multiple logistic regression based
on 12,914 pregnancies, including 10,523 live births, which repre-
sented a response rate of 53.2 percent. After the effects of age,
exposure to anesthetic gases, and pregnancy history were con-
trolled, the risk of congenital abnormalities for babies of
mothers who smoke was estimated. A statistically significant
risk (p< .05) for maternal smoking was found. Figure 7 shows
the estimated risk of congenital abnormality as a function of
maternal age for nonsmokers, moderate smokers (1 to 19 per
day), and heavy smokers (20 plus per day). Relative risks for
heavy smokers compared with nonsmokers were as high as 2.3.
Rates of abnormalities in each general category were higher for
the children of smokers (see Table 6). The significant increase in
cardiovascular abnormalities among smokers' children is in
agreement with Fedrick's findings (40) and in general agree-

209


TABLE 6.-Comparison of congenital abnormality rates for
      babies born of smokers and nonsmokers, by type of
      abnormality

Abnormality

Smokers     Nonsmokers p*
%     No.     %     No.

Cardiovascular                19.07+   (68)    13.65    (95)   0.02
Respiratory                   15.15    (54)    12.07    (84) 0.10
Musculoskeletal                23.84    (85)    19.69   (137) 0.08
Gastrointestinal                13.46    (48)     9.48    (66)   0.04
Central nervous system          11.50    (41)    10.20    (71) 0.27
Urogenital                    21.32    (76)    15.81   (110)   0.02

*One-tail significance level for the test of the difference between two
proportions.
+Rate is number of congenital abnormalities per 1,000 live births. Rates
based upon 3,565 live births among the smokers and 6,958 live births among
the nonsmokers.
SOURCE: Himmelberger, D.U. (54).

ment with the study of Andrews and McGarry (3). Himmel-
berger, et al. point out that their findings are based on retro-
spective survey data, obtained by mail, and therefore subject to
bias from various sources, including that of a high nonresponse
rate. However, the study methods have been designed to elimi-
nate those effects (54).
A recent study by Borlee and Lechat controlled for confound-
ing variables by matching births with congenital malformations
to control births according to hospital and time of birth, mater-
nal age, sex of child, and socioeconomic level of parents. Two
hundred and two children with malformations diagnosed at
birth were compared with 175 controls, from a total of 17,970
consecutive births studied from June 1972 through May 1974.
No differences were found between cases and controls in the
distribution of smoking habits, including the number of cigar-
ettes smoked with or without filters. Sixty-six percent of
mothers of malformed infants and 68 percent of mothers of con-
trols were nonsmokers. Fathers' smoking habits were also simi-
lar among cases and controls. Significantly more mothers of
malformed infants were heavy coffee drinkers (8 plus cups per
day). Because of the frequent association between heavy coffee
drinking and smoking, both habits should be included in studies
of environmental factors possibly related to the risk of congeni-
tal malformations (10). The same is true for consumption of al-
cohol in populations where drinking is prevalent.

Mau and Netter have reported births by gestation, birth
weight, perinatal mortality, and the incidence of congenital
malformations by smoking habits of fathers in 3,696 cases in
210


which the mother was a nonsmoker. Trends toward lower birth
weights and more preterm births with increasing levels of pat-
ernal smoking were not statistically significant. In the total
study of 5,200 births, regardless of maternal smoking habits,
there was a significant increase in the incidence of severe mal-
formations with increasing levels of paternal smoking; children
of heavily smoking fathers had about twice the expected inci-
dence. Although malformations in all systems were more fre-
quent if the father smoked over 10 cigarettes per day, only the
differences in facial malformations were significantly different
(pc.01) by smoking level. The authors state that the trends
with paternal smoking were independent of maternal smoking
level, maternal and paternal age, and social class (120).

More studies of these possible relationships are urgently
needed. As serious malformations are relatively rare, the case-
control approach is probably the method of choice, with careful
matching of cases with suitable controls.

PERINATAL MORTALITY

The 1973 report, The Health Consequences of Smoking and
the 1979 Report have summarized studies demonstrating a di-
rect relationship between level of maternal smoking and risk of
perinatal loss. The reports have also clarified reasons for the
variation in risk observed in these studies (146,147).

Two important reasons for variability between studies have
been demonstrated. First, other important variables such as
age, parity, race, and socioeconomic status influence the results
if they are unequally distributed between comparison groups of
smokers and nonsmokers (89). Second, cigarette smoking is
more harmful to the pregnancies of certain women than to
those of others. In general, women with other risk factors were
at greater risk from smoking than otherwise low-risk women
(3,15,22,128,144,159).

Table 7 illustrates these points. It shows that women charac-
terized by low social class, low level of education, being very
young or old during pregnancy, or being black, have higher
risks of perinatal mortality than their counterparts. Their in-
crease in risk due to smoking is relatively greater. Meyer, et al.
measured the perinatal mortality risks of light smokers (less
than a pack of cigarettes per day) and of heavy smokers (one
pack or more per day) relative to nonsmoker risks within sub-
groups of the population. The increased risk of perinatal mortal-
ity for light smokers who were young, low-parity, and non-
anemic was less than 10 percent. At the other extreme, mothers
characterized by high-parity, public hospital status, previous

211


E  TABLE 7.-Examples of perinatal mortality by maternal smoking status related to other subgroup
Es          characteristics


                                                         Perinatal or neonatal
                             No. of births                   deaths/l,000 births

  Study                     Non-                             Non-                  Relative
  Population                 Smokers     Smokers     Category      smokers      Smokers        risk*

  British Perinatal Mortality      11,145       4,660      Social class
  Survey, England, all births                       1,2 (high)        25.8         26.3          1.02
                                                3-5           33.5         46.6          1.39

  Washington Co. Maryland,       7,646       4,641      Father's
    white                                      education
                                       9+ years        14.4t        16.lt         1.12
                                       S 8 years        17.6t        38.0t         2.16

  Northern Finland, white        8,898       2,346                     23.2         23.4          1.01

  California, middle to                              Race
   upper middle class           6,067       3,726      White         11.0t        11.3t         1.03
                        2,219       1,071       Black          17.1t        21.5t          1.26

  Boston City Hospital                               Race
    Prenatal Clinic                 513         892      White          29.2         31.4
                        1,225                                                     1.08
                                       636      Black           28.6         64.1          1.89

  Quebec, 10% sample of          3,912       2,967     Maternal age
   registered births                                  ~25           12.1         16.1          1.33
                                                  25-34         12.6         13.2          1.05
                                            35+          23.0         41.7          1.81


Social Class

20 -
IO-#
  Para   Para   Para
1,2,3 0     4+

I -Under 35 years-

Smoker

Non-
smoker

Para   Para   Para
1,2,3 0     4+

-35 years +-

OL

FIGURE &--Theoretical cumulative mortality risk according to
      smoking habit, in mothers of different age, parity,
      and social class groups

SOURCE: Butler, N.R. (15).

low-weight births, or anemia had an increased perinatal mortal-
ity risk of 70 to 100 percent when they were heavy smokers (88).
To help visualize the interacting effects of maternal smoking
and of other factors on perinatal mortality risk, Butler has cal-
culated theoretical mortality risks based on data from the
British Perinatal Mortality Study. In Figure 8, perinatal mor-
tality risks by social class, maternal age, and parity are ar-
ranged in order of increasing magnitude. The differences be-
tween smokers' and nonsmokers' risks are represented by the
height of the bars, which varies depending on other risk factors
(15).
These studies show that the risk of spontaneous abortion, of
fetal death, and of neonatal death increases directly with in-
creasing levels of maternal smoking during pregnancy. Studies
of smoking during pregnancy show a range of perinatal mortal-
ity risk ratios (smokers versus nonsmokers) from a low of 1.01 to
a high of 2.42. Variability between risk ratios in different study
populations may be due to lack of comparability between smok-
ers and nonsmokers in other respects, or to interaction between
smoking and other pregnancy risk factors. Studies failing to

213


take account of other important variables may show unusually
high or unusually low risk ratios.

CAUSE OF DEATH

The increased perinatal mortality associated with maternal
smoking is concentrated within a few cause-specific categories.
Excess stillbirths have been associated with antepartum
hemorrhage or abruptio placentae and with "unknown cause"
(3,46). Excess neonatal deaths were associated with immaturity,
asphyxia, atelectasis (23), and with the respiratory distress
syndrome (3).

Meyer and Tonascia (87) analyzed fetal and neonatal deaths
to identify causes of death which showed an excess if the mother
smoked. Fetal and neonatal deaths by coded cause and mater-
nal smoking habit are shown in Table 8: For each cause the
observed numbers for smokers were compared with the number
expected at nonsmoker rates. The differences between observed
and expected numbers indicate the number of deaths in each
category attributable to maternal smoking.

Fetal deaths showed a major smoking-related excess in the
category of "unknown" cause and some increase from "anoxia"
and "maternal cause." By contrast, neonatal deaths related to
smoking were in the category of "prematurity alone," or in the
related category of "respiratory difficulty." The tentative con-
clusion to be drawn here is that fetuses and neonates whose
deaths were related to maternal smoking had no recognizable
pathology, but had died in utero from anoxia, maternal cause, or
unknown cause, or had suffered the consequences of preterm
delivery.

Complications of Pregnancy and Labor

Studies have consistently found a direct relationship between
maternal smoking level and the incidence of placenta previa,
abruptio placentae, bleeding during pregnancy,and premature
rupture of membranes (3,24,46,61,86,87,94,95,130,144,145). The
association is independent of socioeconomic and racial back-
ground (144), parity (3) and many other factors (86) (Figure 8).

These complications carry with them a high risk of fetal and
neonata loss, and are frequently cited as the cause of death among
the offspring of women who smoke. Kullander and Kallen found a
significant increase in the.frequency of abruptio placentae among
smokers' children dying before the age of 1 week (61). In a prospec-
tive study of 9,169 pregnancies by Goujard and colleagues, a large
proportion of the increase in stillbirths among smokers was
caused by abrutio placentae (46).
214


TABLE 8.-Fetal and neonatal deaths by coded cause and
      maternal smoking habit (Canadian English-speaking
       mothers)

Coded cause

   Observed             Observed
           Expected Expected p+
Nonsmoker Smoker smoker* difference value

Fetal deaths
Unknown
Malformations
Hemolytic disease
Anoxia
Maternal cause
All others

Total

Neonatal deaths
Unknown
Malformations
Hemolytic disease
Respiratory difficulty
Prematurity alone
Maternal cause
All others

Total

Total Births

75
32
11
16
31
8

125
24
15
29
45
13

173

52
22
7
46
33
2
16

251

51
24
8
63
65
6
16

178

15,240

233

16,549

81.4      43.6
34.7      - 10.7
11.9       3.1
17.4      11.6
33.u      11.3
a.7       4.3

187.9      63.1

56.5      -5.5
23.9       0.1
7.6       0.4
50.0      13.0
35.8      29.2
2.2       3.8
17.4      -1.4

193.3      39.6

0.003
N.S.
N.S.
N.S.
N.S.
N.P.

0.003

N.S.
N.S.
N.S.
N.S.
0.005
N.S.
N.S.

0.06

N.S. = not significant.
*Based on nonsmoker rate.

p+ value derived from chi square based on a null hypothesis of no difference
between smokers and nonsmokers.
SOURCE: Meyer, M.B. (87).

Naeye reviewed the clinical and postmortem material from
the 3,897 fetal and infant deaths in the Collaborative Perinatal
Project of the NINCDS (102) and reported an association be-
tween perinatal mortality rates caused by abruptio placentae
and number of cigarettes smoked by the mother (95). Abruptio
placentae was the underlying cause identified in 11 percent of
all the deaths in this large study (94).

Analysis of data from the Ontario Perinatal Mortality Study
corroborated these findings. Increasing levels of smoking re-
sulted in a highly significant increase in the risks of placental
abruptions, placenta previa, bleeding in pregnancy, and prema-
ture and prolonged rupture of membranes. Fetal and neonatal
deaths were analyzed for associations between them and
smoking-related excesses of various coded complications of
pregnancy and labor. Although most diagnoses showed no asso-
ciation with excess mortality for smokers' babies, a few stood
                                        215


0 Nonsmokers
o Smokers

I
12

I     I       I         I      I     1
16     20      25      30    34    36
    Gestation (weeks)

FIGURE 9.- Mean plasma volume in nonsmokers and smokers

SOURCE: Pirani, B.B.K. (117).

out as highly significant. Excess fetal deaths of smokers' babies
were strongly associated with bleeding during pregnancy,
either before (P = 0.01) or after (p = 0.0005) 20 weeks gestation.
In other coded categories, a significant excess of fetal deaths
occurred among smoking mothers with abruptio placentae
(p = 0.0001) or other obstetrical problems. Similar comparisons
were made for neonatal deaths. A strong, significant relation-
ship between smoking-related excess neonatal deaths and a his-
tory of bleeding before 20 weeks of gestation was found
(p = 0.0001). Other categories that showed significant increases
of smoking-associated neonatal deaths were the admission
status of rupture of membranes only, other obstetrical compli-
cations, and duration of rupture of membranes over 48 hours
(8'7).
216


PREECLAMPSIA
Several published studies have reported that the incidence of
preeclampsia is declining as the number of cigarettes smoked
increases (109,145). Data from the British Prenatal Mortality
Study were cross-tabulated by parity, severity of preeclampsia,
and maternal smoking status. Smokers had lower rates of all
grades of preeclampsia than nonsmokers, whether they were
primiparae or multiparae (15). Andrews and McGarry showed
that the inverse relationship between cigarette smoking and
preeclamptic toxemia was independent of social class, maternal
weight before pregnancy, and maternal weight gain during
pregnancy (3). Despite this effect of smoking on the incidence of
preeclampsia, there is a greatly increased risk of perinatal mor-
tality if preeclampsia does develop in a smoker (3,34,129). Sev-
eral authors have suggested that this negative association may
be due to the hypotensive effect of thiocyanate, which is derived
from the cyanide present in cigarette smoke and is regularly
found in the blood of smokers (3,109). Because preeclampsia is
predominantly a complication of first pregnancies, it is possible
that the occasional finding of reduced rates of perinatal mortal-
ity in young, primiparous, light smokers who are otherwise
healthy is due to this relationship.
Pirani and MacGillivray performed seven serial mea-
surements from the end of the second trimester until term in 31
nonsmokers and 29 smokers. After 25 weeks gestation the
plasma volume of smokers failed to keep pace with that for
nonsmokers, the increases in volume being 25 percent less in
smokers (Figure 9). Plasma volume and total body water expan-
sion are related to birthweight, at least in primigravidas. After
30 weeks of gestation, total body water in smokers plateaued in
contrast to nonsmokers, so that by term their body water vol-
ume increase was about 25 percent less. Serum heat-stable al-
kaline phosphatase levels in smokers significantly exceeded the
concentration in nonsmokers from the 37th week of pregnancy
onward. This enzyme is of placental origin, and cigarette smok-
ing may contribute to this change by its effects on the placenta
(117).
Whether the reduction in the incidence of preeclampsia with
maternal smoking is due to the hypotensive effects of thiocyanate,
to the reduced size of the baby, to a smaller increase in maternal
blood volume, or to another process requires further study.
PRETERM DELIVERY, PREGNANCY COMPLICATIONS,
AND PERINATAL MORTALITY BY GESTATION
Studies of large numbers of births to measure mean gestation
by smoking habit have demonstrated differences of only a day
                                        217


or two. This finding led to the conclusion that maternal smoking
does not affect gestation (14,52,74,102,146,159). On the other
hand, abundant evidence has been presented that a smoking-
related increase in preterm delivery plays an important role in
the increased risk of neonatal death for infants of smokers.
When the proportion of preterm births is measured, rather
than the mean gestation, smokers have shown consistently
higher rates than nonsmokers, as illustrated in Table 9. In four
studies in which all births and perinatal deaths were included,
the risk of early delivery increased from 36 to 47 percent if the
mother smoked, and 11 to 14 percent of all preterm births could
be attributed to maternal smoking (3,15,38).
Figure 10, using data from the Ontario Perinatal Mortality
Study, shows percentage distributions by gestational age of
births to nonsmokers, light smokers, and heavy smokers, plot-
ted on a semilogarithmic scale to emphasize differences be-
tween smoking-level groups in very preterm births. There is
little difference between the means of these curves because the
great majority of births occur around term in all groups. There
is, however, a significant and dose-related increase in the pro-
portions of preterm babies born to women who smoke. These
preterm deliveries account for a small proportion of total births
but for a large proportion of the deaths (82,146).
As previously reviewed, Meyer and Tonascia have related the
excess fetal and neonatal mortality of smokers' infants and the
excess incidence of pregnancy complications among women who
smoke to the gestational age of occurrence, using a life-table
approach. A starting population of all pregnancies in utero at 20
weeks was used to calculate the probabilities of fetal death, live
delivery followed by survival or death, or the occurrence of a
complication followed by fetal death or delivery. At 28 weeks
(the next point defined by the data), the population at risk in-
cluded those remaining in utero at that point. Figure 11 shows
the probability of perinatal death during each period of gesta-
tional age starting at 20 weeks. Risks for smokers' infants were
significantly greater in the earlier weeks, but not different after
38 weeks gestation (87,146).
A similar approach was applied to determine the risk by ges-
tation of abruptio placentae, placenta previa, and premature
rupture of membranes for smokers and nonsmokers. The risk of
all these complications was higher for smokers throughout ges-
tation, but in all the differences were most significant in the
weeks of pregnancy from 20 to 32 or 36 weeks (87,146). The lower
limit of 20 weeks was built into the study design, which included
all single births of at least 20 weeks gestation (106,107).
These studies show that excess deaths. of smokers' infants are
218


TABLE 9.-Preterm births by maternal smoking habit: relative and attributable risks, derived from published
       studies

Study

Smokers
(proportion)

     Preterm Births*             Relative
       per 100                  Risk
      Total Births             Smokers/Non-
Nonsmokers        Smokers         smokers

Attributable
  Risk
   %

Cardiff                 .465             6.7              9.2             1.36              14
Great Britain            .274             4.7              6.9             1.47              11
Montreal               .432             7.7             10.6             1.38              14
Ontario                .435             7.4             10.1            1.36              14

o Cardiff and Ontario data are for < 38 weeks. All others are for < 37 weeks.
SOURCE: Andrews, J. (3), Campbell, J.M. (15), Fabia, J. (38), Meyer, M.B. (86), U.S. Department of Health, Education, and Welfare
       (146).


60.0

60.0--

40.0-=

20.0-w

10.0-m
3
6.0..

4.0--

2.0.-

1.0::

0.6--


0.4 --

0.2--




O.l-

1 = 959/o Cl

I  :  :  :  :  :  :  :  :  :  i  1  1
20    24     26     32     36    40    44+

          Gestation: Weeks

FIGURE lo--Percentage distribution by weeks of gestation of
       births to nonsmokers, smokers of less than one
      pack per day, and smokers of one pack per day or
            more

SOURCE: Meyer, M.B. (82).

found mainly in the coded cause categories of "unknown" and
"anoxia" for fetal deaths, and in the categories of "prematurity
alone" and "respiratory difficulty" for neonatal deaths. This
finding indicates that the excess deaths result not from abnor-
malities of the fetus or neonate, but from problems related to
220


the pregnancy. Increasing levels of maternal smoking result in
a highly significant increase in the risks of placental abrup-
tions, placenta previa, bleeding early or late in pregnancy, pre-
mature and prolonged rupture of membranes, and preterm de-
livery, all of which carry high risks of perinatal loss. Although
there is little effect of maternal smoking on mean gestation, the
proportion of fetal deaths and live births that occur before term
increases directly with maternal smoking level. Up to 14 per-
cent of all preterm deliveries in the United States may be at-
tributable to maternal smoking. According to the results of one
large study, the most significant difference between smokers'
and nonsmokers' risk of perinatal mortality and pregnancy
complication occurs at the gestational ages from 20 to 32 or 36
weeks.

These findings lead to the conclusion that maternal smoking
can be a direct cause of fetal or neonatal death in an otherwise
normal infant. The immediate cause of most smoking-related
fetal deaths is probably anoxia, which can be attributed to pla-
cental complications with antepartum bleeding in 30 percent or
more of the cases. In other cases, the oxygen supply may simply
fail from reduced carrying capacity and reduced unloading
pressures for oxygen caused by the presence of carbon
monoxide in maternal and fetal blood. Neonatal deaths occur as
a result of the increased risk of early delivery among smokers,
which may be secondarily related to bleeding early in preg-
nancy and premature rupture of membranes (146).

Long-Term Morbidity and Mortality

Studies of infant and child morbidity and mortality by the
mother's smoking habits usually cannot distinguish between
the effects of smoking during pregnancy and the effect of the
infant's or child's passive exposure to cigarette smoke after
birth. Several studies have found that hospitalization rates for
pneumonia and bronchitis were higher during the first year of
life for infants of smoking mothers (20,21,53). Rates in children
were higher if the smoking parents also had cough and phlegm.
Harlap and Davies found that the risk of contracting
pneumonia or bronchitis in the first year of life more than dou-
bled if the parents smoked more than 24 cigarettes a day (53).
A unique and important study of morbidity and mortality in
smokers' and nonsmokers' children up to the age of five has now
been published by Rantakallio (119). The experience up to age 5
of over 12,000 children born in 1966 in Northern Finland, com-
prising 96 percent of all births in two provinces, was ascertained
through hospital and death records and questionnaires. Smok-
                                        221


           Nonsmokers   Smokers

Total Births       27,420    21,465
Total Deaths         634       624
Probability of Death    .023      .029

z   0.06      I= 95% Cl
z
2   0.04 - -

z
5   0.02 - -

`C

Total Deaths -  (263)       (198)  (121) (110) (152)(149) (167) (78)
Smokersand ) ; ; ; : : : : : : : 1 1
Nonsmokers 20 24    28    32 36    40 42+
                     Gestation: Weeks

FIGURE ll.-Probability of perinatal death for smoking and
      nonsmoking mothers, by period of gestational age
       (bars show 95% confidence intervals)

SOURCE: Meyer, M.B. (87).

ing was rare in this population, and the smokers tended to be
young and otherwise healthy. Fourteen percent of pregnant
women smoked fewer than 10 cigarettes per day (mean number
after the second month of pregnancy 3.9) and 3 percent smoked
more than 10 cigarettes per day (mean number 12.2); the re-
maining 83 percent of the population were nonsmokers. It was
therefore possible to remove the usual problems of confounding
variables by close individual matching of 1,750 smokers to
nonsmoking "controls". Matching factors included marital
status, maternal age within 2 years, and place of residence, with
the latter category including many socioeconomic variables to
equalize the probable use of medical facilities and other dif-
ferences. Although the author states that perinatal mortality
did not show a statistically significant increase for smokers,
222


rates were 24 per thousand for controls, 26 per thousand for
light smokers, and 33 per thousand for "heavy" smokers (de-
fined as smoking 10 plus cigarettes per day). These rates are
similar to those found in other studies in which differences were
statistically significant. Postneonatal mortality, from 28 days to
5 years, was higher for smokers' children with rates of 11.1 and
3.9 per thousand for smokers' and nonsmokers' children respec-
tively. Overall death rates of 24.7 per thousand births in smok-
ing women and 16.5 per thousand births in nonsmoking women
were reported for children under the age of 5, of which 12.6 and
8.8 were neonatal.
In addition, the children of the smokers were hospitalized
more frequently, had more visits to doctors, and had longer av-
erage durations of hospital stays than children of nonsmokers.
Respiratory diseases caused significantly more hospitalizations
among smokers' children. It is of great interest that the chil-
dren born to a subgroup of women who stopped smoking during
the last 3 months of pregnancy showed no increase of post-
neonatal mortality or morbidity up to the age of 5, compared
with controls. However, these women had been very light smok-
ers before quitting. Table 10, derived from Rantakallio's study,
shows that the various outcomes measured show increasing
rates of morbidity and mortality with increasing levels of smok-
ing. However, it may not be possible to distinguish between the
adverse effects of maternal smoking during pregnancy and the
adverse effects on infants and children exposed to cigarette
smoke in the home, because women who smoked during preg-
nancy probably also continued to smoke after pregnancy.
Because of the known carcinogenic potential of tobacco smoke
and the evidence that benzo(a)pyrene reaches the placenta,
Neutel and Buck investigated the relationship of maternal
smoking during pregnancy to the incidence of cancer in children
aged 7 to 10. A combined population of 89,302 births from the
Ontario Perinatal Mortality Study and the British Perinatal
Mortality Survey was used as a base population for a prospec-
tive study in which 65 cancer deaths and 32 cancer survivors
were identified. For cancer of all sites, the children of smokers
had a relative risk of 1.3, with 95 percent confidence limits of 0.8
to 2.2. A dose-response relationship was not observed. The num-
bers were not large enough to determine significant differences
by site. Excess cancer rates for children of mothers who smoke
and a possible dose-related progression were concentrated at
ages 0 to 24 months, but these rates were based on small num-
bers of cases. The authors conclude that "although a significant
excess is not demonstrable, a doubling of the cancer risk for
children of smokers cannot be ruled out." Their equivocal re-
                                         223


TABLE lO.-Long term effects of morbidity and mortality by level of maternal smoking

A. Mortality

Number of children
Doctor visits per child
(mean number)
Hospitalizations per child
(mean number)
<Age 1
  Age 1-5

      Nonsmokers             Light Smokers     Heavy Smokers
Control 1          Control 2       (l-10 per day)      (10 + per day)


1300              258              1302              252


  .71             .61               .76               .83


   .19              .15               .22               .39
   .14              .08               .17               .30
   .15              .17               .22              .25

B. Perinatal and postneonatal mortality (28 days to 5 years) per 100 births, by maternal smoking


                               Nonsmokers

                                 Control                Light

Smokers

Total

Heavy

Total births number                    1844                                1844
Perinatal mortality 1,000 births
              per                   23.9                       25.7         26.0          32.6
Postneonatal mortality                    3.9                                  11.1
All mortality 1,000 live births
         per                        16.5                                24.7

SOURCE: Rantakallio, P. (119).


sults were reported to encourage other workers to add to the
data (99). This should certainly be done, with particular empha-
sis on the first 2 years of life.

Rantakallio, et al. also analyzed the use rates of ophthal-
mological services in their follow-up study of approximately
12,000 children, relating these rates of prenatal factors ascer-
tained during pregnancy. The incidence of squint among smok-
ers' children was 22.5 per thousand, compared with 11.5 per
thousand among the children of matched, nonsmoking controls
(p < .05). On the other hand, rates of dacryostenosis and of other
congenital ocular malformations were higher among the chil-
dren of controls. The authors state that squint was inversely
correlated with birth weight and was more common among chil-
dren with other diseases, especially nervous or mental diseases
(121).

SUDDEN INFANT DEATH SYNDROME

Maternal smoking habits have been ascertained in several
studies of the sudden infant death syndrome (SIDS). In all of
these, an association has been found between maternal smoking
during pregnancy and the incidence of sudden infant death.
Steele and Langworth, in a study of 80 cases, each with two
matched controls, which were traced back to the Ontario
Perinatal Mortality Study population of 1960-61, found that
sudden infant deaths were strongly associated with the fre-
quency and level of maternal smoking during pregnancy
(p < .OOl>. Thirty-nine percent of the cases were nonsmokers ver-
sus 60 percent of controls; 36 percent of the cases and 27 percent
of the controls smoked less than a pack per day; 24 percent of
the cases and 10 percent of the controls smoked a pack per day
or more. The habits of the remaining 1 to 2 percent of mothers
were unknown (139).

Bergman and Wiesner studied 56 families who lost babies to
the sudden infant death syndrome and 86 control families. They
reported that a higher proportion of SIDS mothers smoked dur-
ing pregnancy than controls (61 percent versus 42 percent),
more smoked after pregnancy (59 percent versus 42 percent),
and SIDS mothers smoked a significantly greater number of
cigarettes than controls. These authors indicate that exposure
to cigarette smoke (passive smoking) appears to enhance the
risk for SIDS for reasons not yet known (8). However, whether
prenatal or postnatal exposure is more important cannot be de-
termined.

Naeye, et al., in their analysis of 125 SIDS victims from the
population of the Collaborative Perinatal Project of the
                                         225