INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Passive cigarette smoking is known to be a significant health risk (1). Infants exposed to passive cigarette smoking in the first year of life have an increased incidence of lower respiratory tract infections (2, 3). A large proportion of infants exposed to maternal passive cigarette smoke will also have been exposed in utero to cigarette smoke. Infants whose mothers smoked during pregnancy have been shown to have reduced lung function in the neonatal period (4). Despite this evidence there is often strong debate in the public domain about whether passive cigarette smoking in infants is in fact harmful. Although direct anatomic studies examining the airways of adult smokers have documented structural changes associated with this insult (7), there is no anatomic information on the effect of passive cigarette smoking on the infant airway.

The sudden infant death syndrome (SIDS) describes the unexpected and unexplained death of an apparently well infant (8). The diagnosis is one of exclusion and is made after a postmortem examination and detailed assessment of the environment in which death occurred. Most epidemiologic studies have linked maternal smoking with an increased risk of SIDS (9). As part of a larger study into lung structure in SIDS we elected to examine whether airways from infants who had died from SIDS and who had been exposed to high levels of maternal smoking were structurally different from airways from infants who had died from SIDS and who had not been exposed to maternal smoking. We postulated that infants who had been exposed to high levels of maternal smoking would show damage to their airways as a result of the passive cigarette smoke, independent of any changes associated with SIDS.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In 1991 the Victorian Sudden Infant Death Research Foundation commenced a 3-yr study into the epidemiology of SIDS in the State of Victoria, Australia. Families who had lost a child from SIDS were identified by the investigators using a variety of sources, including ambulance officers, pediatric emergency nursing staff, SIDS grief counsellors, and family members. The investigators then approached the families and invited them to participate in a formal interview where a detailed questionnaire covering various epidemiologic factors associated with SIDS was administered by a trained interviewer. The questionnaire included specific questions relating to maternal smoking. Mothers were asked to indicate whether they smoked before becoming pregnant and, separately, during the first, second, and third trimester of pregnancy. Mothers were also asked whether they smoked between the birth and the death of their child. For all questions, smoking was rated on a five-point scale with 0 indicating no smoking, 1 indicating less than 10 cigarettes a day, 2 indicating 10 to 20 cigarettes a day, 3 indicating 20 to 30 cigarettes a day, and 4 indicating a smoking habit of greater than 30 cigarettes a day.

As part of a larger study into lung structure in SIDS, the chief investigators of this present report were granted access to the raw smoking data from these questionnaires. The investigators were also granted access to the stored lung blocks from all of the infants who died from SIDS in the state of Victoria during the period 1991 to 1993. This allowed interpretation of lung histology in the light of the degree of smoke exposure of the infant.

Lung Tissue Analysis

Tissue preparation. All infants who died of SIDS involved in this study had had postmortem examinations performed by an experienced pediatric pathologist at the Victorian Institute of Forensic Medicine (VIFM). Permission was obtained from the Institute's ethics committee to access the stored lung tissue from those postmortems performed between 1991 and 1993. From each lung block one 5-µm section was taken and stained with hematoxylin-eosin. To preserve the block for possible future use by the VIFM, the investigators were limited to one section from each block. Slides were examined using a video-linked microscope Leica Laborlux D (Leica, Wetzlar, Germany) with the image being projected onto the monitor screen of a 486 DX computer. Images were assessed using the color image analysis program Quantimet 500+ (Leica Cambridge Ltd, Cambridge, UK). Airways were subjected to standard airway morphometric analysis, as described below.

Airway morphometry. On all airways cut in transverse section (defined as an even thickness of epithelium and an even thickness from the basement membrane to the outer smooth muscle layer), the following perimeters were measured; the inner airway wall perimeter (Pi) (defined by the luminal surface of the epithelial border), the perimeter of the outer border of the basement membrane (Pbm), the outer airway wall perimeter (Po) (defined by the outer border of the smooth muscle), and the total airway wall perimeter (Pt) (defined by the outer edge of the adventitia surrounding the airway) (Figure 1). The areas of airway smooth muscle, mucous glands, and cartilage within the airway wall were also measured. Airways that showed > 50% epithelial detachment or branching were excluded; however, when smaller sections of epithelium were missing, the border was interpolated between two intact areas.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Schematic representation of traced airway perimeters and the areas subsequently calculated. Standard nomenclature as described by Bai and colleagues (16) has been employed.

Calculations

Using the measured perimeters, the image analysis program calculated the areas defined by each of these perimeters: the epithelial wall area (WAe), the inner wall area (WAi), the outer wall area (WAo), and the total wall area (WAt). Because the wall areas and proportion of smooth muscle, mucous glands, and cartilage within the airway wall are influenced by the size, the areas were expressed in terms of the traced Pbm. Airways were divided into three size groups on the basis of the measured Pbm < 1, 1 to 2, and 2 to 4 mm.

The percent of smooth muscle shortening (PMS) was determined using the technique described by James and colleagues (17). Using this calculation the observed muscle "length" (Po) was compared with its calculated relaxed length (Por) using the equation,

(1)

where

(2)

and WAor, the relaxed outer wall area, is calculated by adding the measured area of wall between the basement membrane and the outer muscle border (WAo  WAbm) to WAbmr the "relaxed" lumen area. WAbmr is the area of a circle with a circumference equal to the measured Pbm, thus:

(3)

Data Analysis

To compare similar-size airways from different subjects, airways were divided into three arbitrary size groups using the Pbm < 1, 1 to 2, and 2 to 4 mm. Airway size groups were chosen to minimize the effects of growth and to enable comparisons with other studies (18, 19). All airway measurements of area were normalized for airway size by dividing by the basement membrane perimeter (17). For each variable, the mean value for each case was calculated. The differences between the means of size groups were tested using a one-way analysis of variance (ANOVA) as previously described by Carroll and colleagues (20).

Intraobserver error was expressed as the coefficient of variation, and was calculated for measurements made on 10 airways 10 times, as previously described by Carroll and colleagues (21). All measurements were made by the one observer who was blinded to the case classification.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A total of 83 cases of SIDS with completed questionnaire smoking data and lung blocks were available to the investigators. The mean age of the infants was 5.7 ± 4.3 mo (mean ± standard deviation). There were 48 male and 35 female infants. The mean time to interview was 31 ± 19 d (mean ± standard deviation) from the death of the infant. In 38 cases there was a consistent level of smoking through all four time periods assessed in the questionnaire. Nineteen mothers did not smoke either before, at any stage during, or after pregnancy (no smoke group). A further 19 mothers smoked more than 20 cigarettes a day in all these time periods (high smoke group).

In the remaining 45 cases, infants had been exposed to variable levels of cigarette smoke both during and after pregnancy. The mean age at death of these infants was 6.5 ± 5.4 mo (mean ± standard deviation). This was not significantly different from the 38 infants whose morphometry is reported in this study. As the variability in smoke exposure made the assessment of any relationship between observed histologic changes and smoking history impossible, this present report will concentrate on those two groups where smoking level was constant throughout all time periods.

The mean age at death of the infants in the high smoke exposure group was 4.9 ± 2.9 mo (mean ± standard deviation) and 5 ± 3.8 mo (mean ± standard deviation) in the no smoke exposure group. In each group there were 12 male and 7 female infants.

Airway Morphometry

Two hundred twenty-eight airways from the 19 infants in the high smoke exposure group were compared with 158 airways from the 19 infants in the no smoke exposure group (Table 1). There was no detectable difference in the airways from the two groups when viewed under simple light microscopy. Inner wall area expressed with relationship to Pbm was significantly greater in the high smoke exposure group in airways in the Pbm 2 to 4 mm group (Table 2). The epithelial thickness expressed in relationship to Pbm was also significantly increased in this airway-size group in the high smoke exposure group compared with that in the no smoke exposure group. The increased inner wall thickness was independent of the increase in the epithelial thickness. Total and outer wall areas were not significantly different in the larger airway size group.

There was no significant difference between the high and no smoke exposure groups in the two smaller airways size groups (Pbm < 1 mm and Pbm 1 to 2 mm) for epithelial, inner, outer, and total wall thickness.

Airway Smooth Muscle

There was no significant difference in the amount of measured smooth muscle within the two groups in all airway-size groups. The percentage of airway smooth muscle shortening showed marked variability both between airways in individual cases and between cases, ranging from zero to 40% throughout all cases. There was no significant difference in the degree of airway smooth muscle shortening between the high and no smoke exposure groups in any airway size. There was no significant difference in the amount of mucous gland or cartilage content of the airways between the two groups in any airway size.

The coefficient of variation of the observations was 2 ± 0.7%.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This present study has documented changes in airway wall dimensions in infants who have been exposed to consistently high levels of maternal smoking when compared with infants who have died from the same cause but had no exposure to maternal cigarette smoke. We believe this is the first reported study documenting anatomic abnormalities in the infant airway associated with passive cigarette smoking.

All lung blocks used in this present study were obtained from infants who had died from SIDS. This diagnosis was made after a careful postmortem examination and detailed assessment of the circumstances surrounding death in order to exclude other causes of death. No diagnostic pulmonary pathology has been described in SIDS; specifically, no previous studies of lung structure in SIDS utilizing direct measurements of airway thickness have documented increased airway wall thickness as a feature, and we are therefore confident that the findings of this present study reflect more on the degree of smoke exposure of the infants than the subsequent death from SIDS (19, 22, 23).

The infants from the high smoke exposure group had mothers who smoked more than 20 cigarettes a day both during pregnancy and postnatally. In utero smoke exposure represents a different insult to the growing airways than does postnatal smoke exposure, as in utero smoke exposure is not associated with direct contact between inhaled smoke and the airway. Postnatal smoke exposure is a true inhalational insult, with the infant directly inhaling sidestream cigarette smoke. Although increased inner airway wall thickness in infants of smoking mothers was found in this present study, we are unable to determine from our findings whether this alteration in airway morphometry is due to an in utero smoke exposure effect or a postnatal smoke exposure effect. An indirect effect of in utero smoke exposure may be to alter the structure of the inner airway wall such as increasing the amount of collagen in the basement membrane. Alternatively, direct irritation of the inner airway wall by passive smoke exposure postnatally could explain the findings of our present study.

Morphometric analysis of airway size is based on the traced perimeter of the basement membrane, as are the calculated airway wall areas (17). An increase in a certain airway wall area is interpreted as an increase in airway wall thickness as has been previously described (16). The alterations in airway wall thickness seen in this study were present in airways with a Pbm between 2 and 4 mm. This corresponds to airways with a luminal diameter between 0.6 and 1.2 mm. The previous work of Boyden and Tompsett (24) on lung development would suggest that the airways in which these changes were observed are small respiratory and terminal bronchioles.

The exact mechanism by which maternal smoking significantly increases the risk of SIDS is as yet undetermined. Some investigators have suggested that direct toxic effect on lung growth may occur secondary to in utero smoke exposure and that the altered lung growth may predispose infants to impaired lung function postnatally and, subsequently, an increased risk of SIDS. Maternal smoking has also been identified to be associated with alterations in the development of areas of the nervous system associated with respiration. Abnormalities in these areas may then produce secondary abnormalities in either in utero or postnatal lung development and/ or function, which may predispose to SIDS. The findings of this present study showing increased inner airway wall thickness in infants whose mothers smoked during pregnancy are unable to differentiate between a direct effect of the cigarette smoke exposure on lung development and whether smoke exposure produced abnormalities in other systems such as the nervous system that then led to the increased airway wall thickness.

Several epidemiologic studies have documented abnormalities in pulmonary function in infants in the first month of life born to smoking mothers (4). Most investigators have considered that these changes are a result of the detrimental effects of in utero smoke exposure. Hogg and colleagues (25) have previously shown that the small airways of young children (defined by them as those less than 2 mm in internal diameter) are responsible for a large part of the total airway resistance. Alterations to the airway wall structure, particularly increased inner airway wall thickness in airways this size, could then have major effects on airway physiology and indeed explain the observed alterations in increased airway reactivity that have been described in infants with a history of exposure to maternal cigarette smoke (6). Although this study has shown an increase in inner airway wall thickness that is independent of the observed increase in epithelial thickness, we are unable to determine from this study whether this increase in subbasement membrane thickness is due to tissue edema or indeed whether foreign material or excessive structural fibers have been laid down in this area.

Maternal smoking has been shown to be associated with an increased incidence of SIDS (9). Could the increase in the incidence of SIDS in infants with heavy-smoking mothers be explainable on the basis of an increased airway wall thickness as observed in this study? Moreno and colleagues (26) have previously shown that a small increase in inner airway wall thickness may result in a much larger increase in airway resistance for a standard level of airway smooth muscle contraction. Martinez (27) has postulated that the pathophysiology of SIDS may be related to small airway closure. Small airway closure would be promoted by an insult that produced an increase in inner airway wall thickness. Clearly, while this postulated mechanism may explain the course of death in infants who died from SIDS and who were exposed to high levels of maternal smoking, they do not explain the cause of death in infants whose mothers did not smoke. This may simply reflect the spectrum of pathophysiology of death in infants who die from SIDS (8).

Exposure to passive cigarette smoke in the infant has also been shown to be associated with an increased frequency of lower respiratory tract infections (2, 3). Although a relationship between SIDS and lower airway infection has been postulated, no direct evidence from the postmortems involved in this present study identified any evidence of lower airway infection. Lower airway infection by itself would not explain the observed increased inner airway wall thickness as infection-derived inflammatory processes would be expected to produce transmural increases in wall thickness rather than the observed inner airway wall thickness.

In summary, this present study found histologic changes in the airways of infants exposed to high levels of maternal smoking when compared with infants dying from the same cause but who had no history of maternal smoking. We conclude that passive cigarette smoking produces significant alterations in the structure of the developing infant airway, which may have significant physiologic sequela and may be related to the cause of death in SIDS in infants with a history of maternal smoking. These findings also emphasize the dangers of passive cigarette smoke exposure to infants and highlight the importance of antismoking measures aimed at reducing this insult.