Tobacco Documents Online

TABLE 3-Associations of Selected Infant and Maternal Characteristics to Invasive Meningococcal Disease In Early Childhood: Metropolitan Atlanta, Ga, 1989-1996 Characteristic Risk Ratioa 95% Confidence Intervai Mothers age, y<20 vs 220 (reference) 1.52 0.71, 3.25 Mother's race: White vs Black (reference) 1.8 0.87, 3.58 Mother received Medicaid 1.54 0.80, 2.99 Mother married 0 70 0.33, 1.45 Mother smoked during pregnancy 2.93° 152, 5.66 Mother's education, y: <12 vs 212 (reference) 2.07° 1.02, 4.20 aRisk Ratio estimated by Cox proportional hazards model. All characteristics are adjusted for other characteristics in the table. TABLE 4-Attributable Fraction of Invasive Meningococcal Disease Among Children Due to Maternal Cigarette Smoking During Pregnancy: Metropolitan Atlanta, Ga, 1989-1996 Fraction Attributable to Maternal Smokinga All children 0.26 Children of White mothers 0.26 Children of African American mothers 0.25 aAttributable fraction was calculated with the following formula: attributable fraction = [(incidence among all children in the cohort) - (incidence among children of mothers who did not smoke)] /[incidence among all children in the cohort]. incidence= number of cases per 100 000 person-yearss during pregnancy. Approximately one quarter of invasive meningococcal disease among children could be attributed to passive expo- srtre to cigarette smoke- Several case--control studies" have linked exposure to cigarette smoke with inva- sive meningoeoccal disease. In one study in the United Kingdom, children younger than 5 years who were exposed to cigarettc smoke at home were 4.5 times more likely to acquire meningococcal disease than were controls.' These differences remained statistically sig- nificant even after control for social class A dose-response relation was found between the number of cigarettes smoked at home and the likelihood of invasive meningococcal disease. Fischer et al. have reported that dur- ing an outbreak of scrogroup B disease in the Pacific Northwest among children younger than 18 years, cigarette smoking by the mother was the strongest independent risk factor (odds rat9o=3.8) for invasive meningococcal disease e Mechanisms that may determine the increased risk of ineningococcal disease as a result of cigarette smoke exposure include the deleterious effects of cigarette smoke on mucosal integrity and the immune system.$ Cigarette smoke depresses respiratory mucus secretion and bronchial eiliary activ- ity and may reduce the effectiveness of the respiratory mucosa to act as a protective barrier against bacterial pathogens.e.te-2o Exposure to cigarette smoke adversely affects macrophage activity and neutrophi] func- tion"'-1a and may indirectly increase the risk for meningococca] infection by predispos- ing children to viral respiratory infections.2" Concurrent viral upper respiratory tract infections have been linked to meningo- coccal disease.2s-27 Several studies have implicated cigarette smoking as a risk factor for increased nasopharyngeal carriage of N meningilidis.1429-3b Therefore, children exposed to smokers may have a high likeli- hood of exposure to the pathogen, a high rate of carriage acquisition, or prolonged carriage of N meningilidis. "I he age-specific incidence we found (highest for infants 1 year and younger) was consistent with previous reports that meningococcal attack rates ara highest dur- ing the first year of life."' The incidence among I - to 2-year-old and 2- to 3-year-old children (2.0 and 2.9 per 100000 person- years, respectively) were also consistent with previous estimates." ' National esti- mates for annual incidence of ineningococ- cai disease for people of all ages is about I in 100 000 persons.' is Our finding that serogroups B and C accounted for most cases was also consistent with national esti- mates reported close to or during the time of our surveiliance.''3z Although meningococcal disease has been reported to have a higher incidence May 1999, Vol. 89, No. 5 Maternal Smoking and Meningococcal Disease among males than females and among Blacks than Whites,''3i we found no signifi- cant association by sex or race. Previous studies have reported an association between low socioeconomic status and meningococ- cat diseasec,9.3s,n because of an increased likelihood of acquiring meningococcal infec- tion through factors such as increased house- hold crowding and poor nutrition.b'9" The level of educational attainment is commonly used as a proxy for socioeconomic status.01 In our cohort, low maternal education level was independently associated with increased risk for invasive meningococcal disease for young children. Young maternal age at deliv- ery, being an unmarried mother, and receiv- ing Medicaid are also associated with low socioeconomic status, and in our cohort these factors were univariately associated with menutgococcal disease in all children and in children of White mothers. One limitation of this study is that the number of cases of invasive meningococcal disease was small, considering the breadth of the study population. We did not account for out-migration from the study area and may',..+' have missed counting some meningococcal cases. However, because the incidence of tlre disease was very low and the median follow- up duration for most children was 3 years (most cases occurred within the first year of life), movement of participants out of the study area is unlikely to have strongly influ- enced our findings. Another limitation is that cigarette smoking by the mother during preg- Dancy was identified through self-reported information acquired from birth certificates. A total of 10.3% of all mothers were reported to have smoked during pregnancy, and the rate was higher among White mothers than Black mothers (12.2% vs 8.3%). Using data from a postdelivery mail survey of a sample of women who gave birth in Georgia, epi- demiologists recently estimated that about 34% of women who smoke during pregnancy do not have this fact documented on the.y child's birth certificate.'- In addition, under-~ reporting of smoking may be greater among Black women and women with higher educa- tion levels.42 However, underreporting of maternal smoking would bias the potential association between smoking and meningo- coccal disease toward the nuil- Because we could not directly measure the passive exposure of children to cigarette smoke, we assumed that mothers who reported smoking during pregnancy continued to smoke after giving birth. Studies have shown that most women who smoke at any time during pregnancy do so throughout preg- nancy, and among those who quit smoking during pregnancy, most relapse shortly after giving birth."s Relapsers who report they American Journal of Public Health 715 2505586046

Yusuf et al. TABLE 2-Univariate Relation Between Selected Infant and Maternal Characteristics and Invasive Meningococcal Disease: Metropolitan Atlanta, Ga, 1989-1996 Characteristic 11 i ~ Sex of child Female Male Birthweight, g <2500 ?2500 Gestation at birth, wk <37 ?37 Abnormal conditions in newborne Yes No Mother's race White Black Other/unknown Mother's age at delivery, y <20 >20 MotheYs marital status Married Not married Mother received Medicaid Yes No Mother's education, y <12 ?12 Mother smoked during pregnancy Yes No All Children (n =283 291) White Children (n = 163 501) Black Children (n = 110 770) No. of No. With Meningo- No. of No. With Meningo- No. of No. With Meningo- Children coccal Disease (%) Children coccal Disease (%) Children coccal Disease (%) 138 367 19 (0.01) 79 506 144 924 28 (0.02) 83 995 13 (0.02) 54 480 6(0,01) 18 (0.02) 56 290 10 (0.02) 23 397 6(0.03) 9137 259 894 41 (0.02) 154 364 29 517 7(0.02) 12 893 253 774 40 (0.02) 150 608 10 460 3(0.03) 5196 272 831 44 (0.02) 158 305 4(0.04) 13 664 2(0.01) 27 (0.02) 97 106 14 (0.01) 4(0.03) 15 962 3 (0.02) 27 (0.02) 94 808 13 (0.01) 2(0,04) 5024 1 (0.02) 29 (0.02) 105 746 15 (0.01) 163 501 31 (0.02) 110770 16(0.01) 9020 0(0.00) 34 398 13 (0.04)' 12 550 248 893 34 (0.01) 150 951 10 (0.08)s 21 367 3 (0.01) 21 (0.01) 89403 13(0.01) 191 614 23 (0.01)' 140 918 91 677 24 (0.03) 22 583 76 712 23 (0.03)' 25 834 206 579 24 (0.01) 137 667 48 668 21 (0.04)a 23 766 234 623 26 (0.01) 139 735 18(0.01)' 42708 5(0.01) 13 (0.06) 68 062 11 (0.02) 17 (0.07)a 49 059 6(0.01) 14 (0.01) 61 711 10 (0.02) 16 (0.07)' 23 280 5(0.02) 15 (0.01) 87 490 11 (0.01) 29 267 16 (0.05)' 19 883 11 (0.06)a 9234 5(0.05)a 254024 31 (0.01) 143618 20(0.01) 101 536 11 (0.01) aProportions of children acquiring meningococcal infection are significantly different between at least 2 levels of the characteristic (Fisher exact test, P<.05). ' °Abnormal conditions diagnosed in the newborn included anemia, injury during birth, fetal alcohol syndrome, respiratory distress syndrome, meconium aspiration syndrome, seizures, and other or unclassified conditions. received Medicaid, or had fewer than 12 years of education (Table 2). For children of White mothers and African American mothers, the proportion acquiring meningococcal disease was significantly `ncreased by maternal cigarette smoking ' -4uring pregnancy. Cox proportional hazards analysis indicated that children of mothers who smoked during pregnancy were 2.9 times more likely to acquire meningococcal dis- ease than were children of mothers who did not smoke during pregnancy (Table 3). No significant interaction between race and maternal smoking status was found in determining risk for meningococcal dis- ease (data not shown). In addition, exami- nation of maternal smoking as a time-vary- ing covariate (data not shown) indicated that the association between maternal ciga- rette smoking and risk for meningococcal disease did not significantly differ (P=.552) between a child's first year of 714 American Journal of Public Health life and the second and third years of life. After the effect of smoking during preg- nancy was accounted for, no significant dose-response relation between the omn- ber of cigarettes smoked and risk for meningococcal disease was found (data not shown). Compared with children of moth- ers with 12 or more years of education, children of mothers with fewer than 12 years of education were 2.1 times more likely to acquire meningococcal disease. Although interaction between race and maternal education level was significant in determining the risk for meningococcal disease (data not shown), the number of cases was too small to allow reliable deter- mination of risk by race. The fraction of meningococcal disease attributable to exposure to cigarette smoke (for which maternal smoking during preg- nancy was a proxy) was similar for all chil- dren, of both White mothers and African American mothers (Table 4). Discussion Our findings are consistent with the results of previous studiesf-'i and.indicate that exposure to maternal cigarette smoking is a risk factor for sporadic meningococcal disease in young children. The rate of inva- sive meningococcal disease was 5 times higher for children of mothers who reported smoking during pregnancy than for children of mothers who did not smoke during preg- nancy. This relation persisted when children of White mothers and children of African American mothers were examined separately. Maternal cigarette smoking during preg- nancy was also independently associated with invasive meningococcal disease in mul- tivariate analysis. After factors reflecting social and economic status were adjusted for, children of mothers who smoked during pregnancy were 2.9 times more likely to acquire invasive meningococcal disease than were children of mothers who did not smoke May 1999, VoL 89, No. 5 2505586045

Maternal Smoking and 34eningococcal Disease rneningococcal and influenza infections. NErglJMer! 1971;287:5-9. _;,.7. Cxrtwright KA, Jones DM, Smith Al, Stuart 35. _ 1M, Kaczmarski EB, Palmer SR. Influenza A and meningococcal di+ease. Lancet. 1991;338554-557. 36. - Moore PS, Hierholzer 1, DeWitt W, et al. Respi- ratory viruses and Mycoplasma as cofacmrs for .- epidemic Group A meningococcal meningitis. JAMA. 1990;264:1271-1275. 37. '.'29. Stuart JM, Cartwright KAV, Robinson PA, ' Noah ND. Effect of smoking on meningococcal carriage.It+ncee 1989;2:723-72i 'Ihomas JC, Bendana NS, Waterman SH, et al. 38. - Risk factors for carriagc of ineningococcus in - the Los Angeles County men'sjai I systenr- Am J EpidemioJ. 1991;t33:286-295. 39. - 11 Blacku'ell CC, Tzanakaki G, Kremastinou 1, et rz al. Factors affecting carriage of Areisseria meningitidis among Greek military recruits. Epirlemtollnfect 1992;108:441-448. Cartwright K. Meningococcal carriage and dis- easc. In: Carlwright K, ed. Meningococcal Dis- 40_ 1. ease. New York, NY: John Wiley & Sons; 1995:115-146. ..33. Caugant DA, Hoiby EA, Magnus P, et al. 42 - Asymptonratic carriage of Neisseria meningi- tidis in a randomly sampled population. J Clin Microbiol 1994;32 J 23-330- Kremastinou J, Blackwcll C, Tzanakaki C, 43. Kallergi C, Elton R, Wcir D_ Parental smoking and carnage of Neisseria rneningitidis among - Greek schoolchildren_ Scund J b,fect D6-_ 1994:26:719-723. Schuchat A, Robinson K, Wenger JD, et al. Bacterial meningitis in the United States 1995. N EnglJMrd 1997;337:970-976. Foster MT, Sanders E, Gintcr M. Epidemiology ofsulphonamidc resistant meningococcalinfcc- tions in a civilian population. Atn J Epidemiol. 1971;93:346r353. Farries JS, Dickson W, Greenwood E, Malhoha TR, Abbott 1D, Joncs DM. Meningococczl infections in Bolton, 1971-74. Lancet 1975;2' IIg-120. Rhan KM, Bhatt SM, Mirva NB. Mcningococ- cal mcningitis_ Easr Afr Med J 199G 73:35-39. Moore PS_ Meningococcal meningitis in sub- Saharan Africa: a model for the epidemic proccss. Clin lnfect Dls_ 1992;14:515-525. Kaiser AR, Hennekens CH, Saslan MS, Iiayes PS, Bennett 1V. Scroepidcmiology and chemo- prophylaxis of disease due to sulphonanride- resistant Neisseria menbigtridis in a civilian population. J Infect Dis. 1974;130:217--224. Liberatos P, Link BG, Kelsey JL. The measure- ment ofsocial class in epidemiology. Epirlemiof Rev 1988;10:87-121. Dietz PM, Adams MM, Rochat RW, Mathis MP prenatal smoking in two consecutive pregnan- cies: Georgia, I989-1992_ Maternrd Child Health 1 1997;1:43-51. Kendrick 1S, Zahniser SC, Miller N, et al_ Inte- grafing smoking cessation into routine public prenatal care: the smoking cessation in preg- nancy projeca. Anr J Public Hcaltk. 1995`85: 217-222. 44. Fingerhul LA, Kleinman JC, Kendrick JS. Smoking before, dm'ing, and after pregnancy. Am J Public Health 1990;80:541-544. 45. Mullen PD, Quinn VP, Ershoff DH. Mainte- nance of nonsmoking postpartum by women who stopped during pregnancy Am J Public Health. 1990; 8 D:992-924. 46. Office of Smoking and Health. Reducing the Health Corseguences of Stnaking: 25 Yeurs of Progress. A Report of the Surgeon General. Washington, DC: US Dept oPllealth and Human Serviccs; 1989. DHHS publication (CDC)88-406. 47. Standacrt SM, Lekowit-z LB Jr, Horan 1M, Hutcheson RIl, Schaffner W'Ihe reporting of communicable diseases: a controlled study of Neisseria meningitidi.s and Haeniophilu.e inf7rren- zae infections. Clin lnfect Dis. 1995;20:30-36. 48. Marier R. The reporting of communicable dis- cases. Ant J EpidemioL 1977;105587-590. 49. Berg AT, Shapiro ED, Capobianco LA_ Group day care and risk for serious infectious ill- nesses Am J EpidemioL 1991;133: 154-163_ 50. Cooley JRT, Holland W W. Corkhill RT. Influ- ence of passive smoking and parental phlegm on pneumonia and bronchitis in early child- hood.Lancet_1974;L1031-1034. 51 Flinton AE. Surgery for otitis media with efit~ sion in children and its relationship with parental smoking. J Laringol Oto! 1989;103559 561. - 1999, Vol. 89, No. 5 American Joumal of Public Health 717

h4eternal Srnnking and Afeningococcal Disease who died through Septembcr 31, 1996, were linked to the corresponding birth certificates. initial linking was based on last names and birth dates of children, and fiuther matching and validation were done by children's first names, mothers' maiden names, and fathers' last names. The birth and death certificates were matched for more than 90% of the eligi- ble childrea Data from prospective surveil- lance for invasive meningococcal disease" were used to identify cases of invasive meningococcal disease in these children. Cases of meningococcal disease were linked to the birth-death file by matching first and last names and date of birth. Children for whorn information was missing were excluded from analysis. The study cohort consisted of 283291 children who were assumed to have resided in metropolitan Atlanta throughout the study period_ The duration of follow-up was defined as the period between birth and diag- nosis of meningococcal disease for children who developed the disease; the period between birth and death for children who died during follow-up without developing meningococcal disease; and the period between birth and September 3 t, 1996, or the third birthday, whichever came first, for the rest of the cohort. Children whose birth certificates were not matched to either death certificates or surveillance records for the meningococcaldisease group were assumed to be alive and not to have devel- oped meningococcal disease. Identification ofCases A case of invasive meningococcal dis- ease was defined by the isolation ofNmenin- gitidls from either the blood or the cere- brospinal fluid of the patient. Isolates were serogrouped by standard laboratory tech- niques. Cases were identified as part of a lab- oratory-based surveillance project for meningococcal and other invasive bacterial infections in metropolitan Atlanta,O-15 an area with 32 hospitals and a population of 2344514 (1990 census). Initial case reports were obtained from hospital laboratories and infection-conlrol practitioncrs. Laboratories were audited every 6 months to identify wire- ported cases and validate reported cases. Medical records of all audited patients were also reviewed. lnformation Regarding lnfant and Maternal Characteristics Characteristics of infants and mothers were obtained from the birth certificate data file. Selected infant characteristics included sex, birthweight (<250t1 g, ? 2500 g), and ges- tation at birth (<37 weeks, ?37 weeks). Assessed maternal characteristics included age at delivery (<20 years, ?20 years), race (White, Black, other), education (<12 years, ?12 years), marital status (married, not mar- ried), and smoking of cigarettes during preg- nancy (yes/no; if yes, the number of ciga- rettes smoked daily). The birth certificate and Georgia Medicaid claims databases were linked to determine dre source of payment for prenatal care and delivery (Medicaid or other source). These variables were chosen because they may be confounders of an association between maternal smoking and sporadic meningococcal disease among children. Analysis Age-specific incidence was determined by the number of cases per total person-years of follow-up for each year of age. Unlvariate relationships between selected infant and maternal characteristics and invasive meningococcal disease were assessed by using the Fisher exact test. Variables found to be significantly associated with invasive tneningococcal disease were entered in a multivariate model. To account for variable lengths of follow-up, Cox proportional haz- ards analysis was used to idenfify indepen- dent associations between selected risk factors and meningococcal disease. The proponional hazard assumption was assessed graphically by the SAS procedure L(F>i"IESV6 and by a test for the interaction between maternal smoking and time since the infant's birth." Models were fitted by means of the SAS procedure PHREG.'° E'otential interaciions between maternal race and maternal smok- ing dtuing pregnancy and between maternal race and maternal education were also assessed. Because no cases occurred in chil- dren of ntothers whose race was classified as "other;" information from children of White women and Black women was used in haz- ards analysis. Results Fifty-five cases were identified. Two cases were in children born outside metro- politan Atlanta, and these cases were exchtded. Four cases could not be linked to births in the study area or in Georgia, per- haps because of migration into the study area from another state. Information on maternal smoking during pregnancy was not available for these children. Information on maternal smoking during pregnancy or matemal edu- cation level was missing for 2 other cases. After we made these exclusions, we ;~ lyzed data from 47 cases. Serogroup B N meningitidis accounted for 17 of the 47 cases (36%); 10 cases (21%) were serogroup C, 5 cases (11%) were serogroup Y, and I case (2%) was serogroup W 135. Serogroup information was not avail- able for 14 cases. Meningitis was diagnosed for 18 cases (38%). Four of the 47 cases (9%) died of meningococcal disease. No meningococcal disease outbreaks or secondary cases were noted inc metropolitan Atlanta during the study period. The age-specific incidence was highest for children I year or younger (Table 1). The proportion of children who acquired invasive meningococcal disease did not differ significantly by sex, gestation at birth, abnor- mal conditions, or mother's race, but several variables were significant in univariate analy- ses (Table 2)- The proportion of children acquiring meningococcal disease was grea- for teenaged mothers than older moth.,.j (0.04% vs 0.01 %q P=.003); tuunarried moth- ers than married mothers (0.03% vs 0.01 %. P=.008); mothers whose prenatal care and/or delivery was paid for by Medicaid th:ut moth- crs whose prenatal care and/or delivery was paid for by another insurer (0.03% vs 0.01%. P=.002); mothers with fewer than 12 years of education than mothers with more than 12 years of education (0.04°/a vs <0.01%, P=.00001); and mothers who reported smok- ing during pregnancy than mothers who reported not smoking during pregnancy (0-05% vs 0.0 t%, P =.00001). Among children of Wlrite mothers, the proportion acquiring meningococcal dis- ease was significantly higher for those whose mothers were young, not married, TABLE I-Incidence of Invasive Meningococcal Disease by Age of Infant: Metropolitan Atlanta, Ga, 1989-1996 Age Interval No. of Person-Years No. of Incidence Rate of Children, y Children of Foilow-Up Cases (95 % Confidence Interval) 0-1 283 291 279 761 36 12.8 (9.3, 17.8) 1-2 270 191 248 985 5 2.0 (0.8, 4.8) 2-3 228 534 207 692 6 2.9 (1.3, 6.4) I 'Incidence of invasive meningococcal disease per 100 000 person-years, May [999, Vol. 89, No. 5 American fournal of Public Health 713 2505586044 Moil,

Maternal Cigarette Sn7oking and Invasive Meningococcal Disease: A Cohort Study Among Young Children in Metropolitan Atlanta, 1989-1996 Hussain R. Yusuj, MBBS. MPH. Rnger W Rochat. MD, i-Vendv S- Bnugh.nart, MSPH, Paul M. Garg7u(!o. PfiD. Brudlet-A - PerkAu. MD- A9ars D. Brnntlef. MPH- and Datdd S. Stepherrs, MD Neisseria mexfngitidrs is a major cause of bacterial meningitis and septicemia in the United States. Approximatelv 2600 cases of meningococcal disease occur annually in the United States. and attack rates are highest among young children.'-~ The disease is often devastatine: case-fatal- ity rates range from 10% to I$u rvi vors may have long-term sequelac such a: hydrocephalus and mental and physical handicaps." In dre United Stales. mosl meningococeal disease occurs spor<tdicalband is caused by organisms of capsular serogroups B, C, Y, and W135 -''' The cur- rently available polysaccharide meningo- coccal vaccines do not offer protectton against serogroup B meningococci and are poorly immunogenic in children younger than 2 years.s Identification of groups at high risk for disease and identification of modifiable risk factors for invasive meningococcal disease are imporlam for developing effective prevention strategie, for this illness. In recent years, at least "t studies in Norway, England, and the United Statcs have reported a link between exposure to tobacco smoke and meningococcal dis- ease.°"" These studies have suggested that passive exposure of children to cigareuc smoke increases their likelihood nf acquir- ing invasive meningococcal disease rceer- alfold. We report here the results of a cohort study to determine Ihe association between cigarette smoking during preg- nancy and other maternal and infant char- acteristics and the risk of developing spo- radic invasive meningococcai disease during early childhood. In this study. we assumed that women who smoked during pregnancy continued to do so after g3ving birth, thereby exposing their children to cigarette smoke. Alethods Data Sotures and Subjectc We used a retrospective cohort design in tr'hich the past characterisucs of the cohort ucre idenutied and their subseqaent diseasc experience was reconstructed to the present. The occurrence of sporadic menin_eoeoeeal disease in a cohort of children born in the metropolitan Atlanta, Ga, area (Claytnn. Cobb. DeKalb. Douglas. Ftdton. Gwinnett. NeNNton, and Rockdale counties) was recon- structed tip to rhe completion of the child's third year of life or the end of the study period (September 31- 1996) - ~`,e used Geor- ~ia's birth cenificale database to identify all children bom to mothers residing in the met- ropolitan Atlanta area between January I. 1959, and December 31, 1995. The death cer- lificates of all children 3 years and younger I-lussam R Yusof Rueer W'. Ro<hat. Paul ]4 Garetullo. and Marv D Brantlc~ are with the office of Pennatal P.prlerniology. Di\ision of Public Hcalth, Gr,ruin Ocparimcnr of Hu .. Resources, Atlanta, Ilussain R. Yu>uf, Roger W Rochar and Paul M Gargiullo are also .vith the Dici,ion of Reproducuve Health. \ational Center for Chronic Disease Prevention and Health Promo- tioe. Centers for Di.ceavc Cantrol and Prevention. Atlanta Wendy S f3auehman anct David S. Sicphens are uith tho Veterans 4lratrs Medicnl Center-Ananta: David S Stcphens i; also with the Division or Infectious Diseases. Deparmtent of Medicine. Emorc Universuy School of:\ledicine, Anantn f3radley A- Pcrkiaa is with the Division of Bactedal and Alccoi~c Diseases- A"ational Centcr ror hifenious Diseases, Centers for Disease Con- trol and Prcvention, Atlanla. Rcquests I'or repruns should be scm to Hussain R. 1'usuf. Naionai Immmnizaion Procr.mi. Ccnrers for Discasc Control and Prevcntion. \4ailstop E-52, 4-'o t3uford Hrrv-Atlanta, G.A 3034 1- This paper uas accepted Derember 2, 1998. Viay 1999-Vol-89,No-5 2505586043

Yusuf et al. did not smoke during pregnancy would decrease the apparent magnitude of the identi- fied risk between passive tobacco smoke exposure and meningococcal disease. Thus, a concerned mother may stop smoking during pregnancy and be reported as a nonsmoker, but she may resume smoking after delivery and unknowingly place her infant at inereased risk for invasive meningococcal disease. These findings, in addition to the fact that adverse health effects during infancy can result from exposure to cigarette smoke in utero,°6 suggest that our findings are valid. Further, this study did not account for the exposure of children to cigarette smoke from persons in the household other than the mother. One strength of our study is that all resi- dent cases of invasive meningococcal disease were identified through a laboratory-based active surveillance system, and periodic audits were used to optimize the accuracy of detection. Studies have shown that, compared with passive surveillance systems, laboratory- based active surveillance for meningococcal 'isease is more accurate and able to identify L-nvice as many cases.°i'48The use of birth cer- tificate data allowed the examination of sev- eral infant and maternal characteristics; the reporting accuracy of a number of these has already been evaluated. Our study illustrates how the linking of information from birth and death certificate databases to data from a pop- ulation-based surveillance system for disease can be used to follow a birth cohort. For low- incidence diseases, this methodology may provide a simple, low-cost, yet powerful approach for assessing population-based rates by selected characteristics, computing relative risks, and examining trends. Moreover, this methodology permits an efficient approach to evaluating the effect of strategies to reduce riskk factors such as matemal smoking and any changes in their association with meningo- coccal or other disease. Matemal smoking during pregnancy has ^n associated with several adverse health effects on the fetus and infant, including pre- maturity, low birthweight, and sudden infant death syndrome.°6 Smdies have shown that passive exposure to cigarette smoke increases a child's susceptibility to various infec- tions.49-51 Our findings support recent reports linking exposure to cigarette smoke with increased risk forinvasive meningococcal disease in young children, and they add to the impetus for promoting smoking cessation among pregnant women and mothers. Health care providers and the general public need to be informed that cigarette smoking may be a strong risk factor for meningococcal disease. Smoking prevention can form an integral part of efforts to prevent meningococcal dis- ease. At least 2 Web sites (University of Illi- nois at Urbana and National Meningitis Trust Fund) provide information on meningitis pre- vention and indicate that avoidance of expo- sure to cigarette smoke may reduce the risk for acquiring the disease. Because the associ- ation between exposure to cigarette smoke and invasive meningococcal disease has implications for public health, this research should be replicated in other geographic areas where meningococcal disease surveil- lance is occurring. O Contributors I{.R.Yusufplanned the study, analyzcd the data, and wrote the paper. R. W Rochat and D. S. Stephens co- supervised study planning and implementation. W S. Baughman developed the meningococcal disease active surveillance database and assisted with study design. P. M. Gargiullo assisted with statistical analysis and study design. B. A. Perkins assisted with planning of study and interpretation of results_ M. D. Brantley assisted with matching birth and death record databases to develop a follow-up cohort, matching meningococcal cases to the cohort, and designing the study. R. W. Rochat, D. S. Stephens, W. S. Baughman, P. M. Gargiullo, B. A. Perkins, and M. D. Brantley contributed to writing the paper. All 7 authors are guarantors for thc integrity of the research. Acknowledgments We gratefully thank the following institutions and individuals for the assistance and support they pro- vided to this study: Georgia Emerging Infectious Diseases Program,Atlanta, and hospitals and staffof the Atlanta Active Surveillance Program. References I. Jackson LA, Wenger JD. Laboratory-based sur- veillance for meningococcal disease in selected areas, United States, 1989-1991. MMWR Morh Monal Wkfy Rep. 1993;42(SS-2)21-30. 2. Fortnum HM, DavisAC. Epidemiology ofbac- terial meningitis. Arch Dis Child. 1993;68: 763-767. 3. Pinner RW, Gellin BG, Bibb WF, et al. Meningococcal disease in the United States, 1986. Jlnfect Dis. 1991;164:368-374. 4. Swartz B, Moore PS, Broome CV Global epi- demiology ofineningococcal disease. Clin Microbiol Rev 1989;2:S118-S t24. 5. Centers for Disease Control and Preventiort Meningococcal vaccines_ MMWR Morb Mortaf Wldy Rep. 1997;46(RR-5):255-259. 6- Stuart JM, Cari.vright KA, Dawson JA, Rickard 1, Noah ND. Risk factors for meningococcal dis- ease: a case control study in south west England. CammunityMed. t988;10:139-146. 7. Haneberg B, Tonjum T, Rodahl K, Gedde-Dahl TW. Factors preceding the onset ofineningo- coccal disease, with special emphasis on pas- sive smoking, stressful events, physical fitness and general symptoms of ill health. NlPHAnn. 198316a69-173. 8. Fischer M, Hedberg K, Cardosi P, et al. Tobacco smoke as a risk factor for meningococcal dis- ease. Pediatr Infect Dis J. 1997;16:979-983. 9. Smnwell-Smith RE, Stuart JM, Hughes AO. Robinson R GritTin MB. Cartwright K. Smok- ing, the environment and mcningococcal dis- ease: a case control study Epidernrol Infect. 1994;112:315-328. 10. Bredfeldt RC, Cain SR, Schutze GE, Holmes TM, McGhee LA. Relation between passive tobacco smoke exposure and the development ofbacterial meningitis in children. JAm 13oam' Fmn Pract 1995;8:95-98. 11. Imrey PB, Jackson LA, Ludwinski PH, et al_ Outbreak ofserogrnup C meningococcal dis- ease associated with campus bar patronage. Ant J Epidential. 1996;143:624-630. 12. Tappero JW, Reporter R, Wenger JD, et al. Meningococcal disease in Los Angeles County, California, and among men in the county jails. N Eng! J Med 1996;335:833-840- 13. Stephens DS, Hajjch RA, Baughmmn W S, Har- vey RC, Wenger 1D, Farley MM. Sporadic meningococcal disease in adults: results af a five year population based study Ann bttern Med 1995;123:937-940. 14. Jurado R, Fadcy MM, Periera E, et al. Increased risk of meningitis and bacteremia due to G'ste- ria nianocytogeues in patients with human immunodeficiency syndrome. C/in Infect Dis.1993;17:224-227 _ 15- Farley MM, Harvey RC, Stull T, et al. A popula- tion-based assessment ofinvasive disease due to group B strcptococcus in non-pregnant adults. N Engl J Med. 1993;328:1807-181 I. 16, SAS/StatSoftware' Changesand Enhancemems Through Release 6 12. Cary, NC: SAS tnstitute Inc;1997. 17. Collett D. Modelling survival data in medical research. London, England: Chapman & Hall; 1994. 18. Crolton J, Douglas A. Respiratory Diseases. Oxford, England: Blackwell Scientific Publica- tions; 7981:353. 19. Royal College of Physicians of London. Health or Sneoking7 Edinburgh, Scotland: Churehill Livingstone; 1986. 20. Weber A. Annoyance and irritation by passive smoki ng. Ptrv Med. 1984; I 3:618-625. 21. Cosio FG, Hoidal JR, Douglas SD, Michael AE Binding of solubic immune complexes by human monocytes and pulmonary macro- phages: effects ofcigareue smoking. JLab Cl[n Med I982;100:469.476. 22- Johnson SD, Houchens DP, Kluwe WM, Craig DK, Fisher GL. Effects of mainstream and environmental tobacco smoke on the immune system in animals and humans: a review. Crit Rev Tasicol. 1990;20:369-395. 23- Tardif J, Borgeat P, Laviolette M. Inhibition of human alveolar macrophage production of leukotrine B4 by acute in vitro and in vivo exposure to tobacco smoke. Am J Respir Cell .ddal BioL 1990;2:155-161. 24. Pedreira FA, Guzndolo MD, Feroli EJ, Metla GW, Weiss IP Involuntary smoking and inci- dence of respiratory illness during the first year oflifc Pediatrics.1985;75:594-597. 25. Kmsinski K, Nelson JD, Butler S, Luby JP, Kus- miesz. Possible association of mycoplasma and viral respiratory infections with bacterial meningitis. Am J Epidemiol. 1987;125: 499-508. 26. Young LS, LaForce FM, Head JJ, Freely JC, Bennett JV. A simultaneous outbreak of 7I6 American Journal of Public Health May 1999, Vol. 89, No. 5 2505586047