Exposure to Children and Risk of Active Trachoma in Tanzanian Women

Transcription

1 American Journal Epidemiology Vol. 137, 3 Copyright 1993 by The Johns Hopkins University School Hygiene and Pubfc Health Printed in U.S.A AD rights reserved Exposure to Children and Risk Active Trachoma in Tanzanian Women Nathan Congdon, 1 Sheila West, 1 Susan Vitale, 1 Sidney Katala, 2 and B. B.. Mmbaga 2 The authors surveyed the trachoma status 5 women aged 18-6 years and 527 children aged 1-7 years in the trachoma hyperendemic region Kongwa, Tanzania, in 1989 to further describe the importance exposure to young children as a risk factor for active trachoma in women. The women were identified as caretakers, who currently cared for children aged 1-7 years; noncaretakers, who lived with, but did not care for, children aged 1-7; or those without children aged 1-7 in the household. The age-adjusted odds ratios for active trachoma seemed to rise with greater exposure to young children, from 1. for women without such children, to 1.63 for noncaretakers and 2.43 for caretakers (trend test, p =.8). Among those who lived in households with young children, the prevalence active trachoma in women increased with the total number young children cared for and with the number infected children cared for. The prevalence active trachoma was 4% (6 ) for caretakers three or more infected children, compared with ( 88) for caretakers with no infected children (p <.1). Caring for infected children also appeared to be associated with signs chronic trachoma in caretakers. Noncaretakers who lived with infected children were not at a significantly increased risk for trachoma compared with noncaretakers who were not exposed to such children (5.4% (three 56) vs. 5.6% (one 18); p >.4). None the facial signs observed in the children (flies on the face, nasal discharge, etc.) appeared to increase the odds ratio active trachoma in caretakers beyond the increase associated with trachoma alone in the child. These data support the hypothesis that active disease in women is associated with direct caretaking young children with active disease. Strategies that interrupt household transmission may affect the blinding sequelae trachoma in women. Am J Epidemiol 1993;137: child care; eye infections; trachoma; women Trachoma is a leading cause blindness in the world, ranking behind only cataract as a threat to sight (1). Trachoma is an infectious process caused by the obligate intracellular agent Chlamydia trachomatis, characterized in its acute phase by inflam- Received for publication August 26, 1991, and in final form September 22, Abbreviations: Q, confidence interval; OR, odds ratio. 1 Dana Center for Preventive Ophthalmology, The Wilmer Institute, The Johns Hopkins University, Baltimore, Maryland. 2 Kongwa Primary Eye Care Project, Kongwa, Tanzania, and Heten Keller International, New York, NY. Reprint requests to Dr. Sheila West, Wilmer Bidding 116, The Johns Hopkins Hospital, 6 N. Wolfe St., Baltimore, MD Supported by a grant from the Edna McConnell Clark Foundation, a short-term research training grant (NIH 5T35 HL766) to Mr. Congdon, and National Institutes Health grant S1-RR The authors thank all those who assisted with this project, in particular, Dr. J. M. Temba, Director Preventive Services, Ministry Health; Dr. Upunda, Regional Medical Officer, Dodoma, Tanzania; Mr. Mlwande, District CCM Chairman, Mpwapwa, Tanzania, Dr. Gavyole, District Medical Officer, Mpwapwa, Tanzania; Mr. Sirna, Administrative Officer-in-Charge, Kongwa Subdistnct; and all the village chairmen and secretaries the villages that were surveyed. Project team member Matt Lynch and the rest the survey team deserve special commendation. Special thanks are also due to Dr. Virginia M. Turner for her central role In all phases the fieldwork and her assistance In the preparation this manuscript.

2 Exposure to Children and Risk Active Trachoma in Women 367 mation, and, after repeated reinfection, by scarring the tarsal conjunctiva. In the later stages, usually seen in somewhat older individuals, trichiasis can result from the scarring process, leading to abrasion and eventual opacification the cornea. In many areas, both active trachoma and its potentially blinding sequelae are more common among women than among men the same age (2-6). Since preschool-aged children (ages 1-7 years are preschool in Tanzania) are the main source active infection, the excess risk for women has been attributed to their traditional role as primary caretakers for this age group (4-7). The aims this study are to examine the evidence for this hypothesis on at least two levels. First, do women who live with and care for preschool-aged children have a higher prevalence active trachoma compared with women who live with, but do not care for, such children, or compared with women who do not live with young children at all? Second, among women who are caretakers, is the risk active trachoma higher if the children in their care have active disease? Last, we wished to identify factors that increase the risk intrafamilial transmission trachoma. Previous research has suggested that the facial cleanliness a child may play a role in acquiring and transmitting trachoma (8-1). This study compares specific elements a child's unclean face with the trachoma status the caretaker. MATERIALS AND METHODS An estimated 5 preschool-age children (and an estimated 25 caretakers) were needed for this study on the association disease in children and women. Two villages were selected for study from the trachoma hyperendemic region Kongwa, Tanzania, in The villages selected and the women chosen were part a larger study on risk factors for a relatively rare complication, trichiasis, in women. A cluster sampling scheme used for a previous trachoma survey in this area was used for this study (8). In essence, villages are divided into geopolitical units called "balozis," each consisting 1-3 houses. These balozis were geographically clustered into units 3-5 balozi, and clusters were sampled randomly to meet the target approximately 25 children per village. In village 1, three clusters (11 balozi) were randomly selected; this yielded less than 25 preschool children, so two additional balozi were randomly selected from the remaining 22 balozi, to yield a total 296 preschool children in village 1. A total four clusters (18 balozi) in village 2 were randomly selected, which yielded 231 preschool children. A census was conducted the sample clusters. Age, sex, and identity the caretaker were recorded for all children aged 1-7 years; age and relationship to the head household were recorded for all eligible women. Trained local workers who carried out the census were also instructed to observe the faces children for the presence or absence three signs deemed to comprise an unclean face: ocular discharge, nasal discharge, and flies on the face over a 3-second interval. Observations were carried out unobtrusively so as not to promote facewashing behavior. The reliability observations made in this way has been demonstrated previously (1). A clean face was defined as the absence all three signs. The interviewer also collected data on other potential risk factors for trachoma as part the larger study. These factors include presence cattle pens, distance to water, type homes and ros, and religion. Within 3 days the completion the census in each village, a clinic was held. At this clinic, trachoma grading children and women was carried out by a single experienced grader (S. K.) according to the World Health Organization simplified grading scheme (11, 12). In this scheme, active trachoma is defined as follicular trachoma (five or more follicles greater than.5 mm in diameter over the central tarsal area) and/ or severe, intense trachoma (more than 5 percent normal deep tarsal vessels obscured). Chronic sequelae trachoma are the following signs: trachomatous scarring (the presence scarring in the tarsal con-

3 368 Congdon et al. junctiva), trachomatous trichiasis or entropion (at least one eyelash rubbing on the globe), and corneal opacity (any easily visible corneal opacity over the pupil, in the presence trichiasis/entropion). Both eyes were graded for all subjects, and a subject was assigned a given grade if either eye was affected. The different grades were not mutually exclusive, as a subject may have scars and active disease simultaneously. Data were entered on a Toshiba 12 microcomputer (Toshiba 12 XE, Toshiba America Information Systems, Irvine, California) in thefieldand brought to Baltimore, Maryland, for analysis on a Microvax 36 (Digital Equipment Corporation, Landover, Maryland). SAS stware (SAS Institute, Cary, North Carolina) was used to analyze the risk trachoma in women associated with caretaker status, trachoma in the children, and other explanatory factors. Direct age adjustment techniques for calculating age-adjusted prevalences were used, with the entire population women as the standard. Tests for trend in odds ratio were calculated by using the Mantel-Haenzel chi-square test for trend. For the analyses clean faces in the children and the risk trachoma in the mother, each child was characterized by trachoma status and facial cleanliness status. If at least one child had both trachoma and an unclean face, the mother was considered to be exposed to both factors. Otherwise, if the mother was exposed to at least one child who had trachoma, but had a clean face, the mother was considered exposed to trachoma alone. The remaining risk groups were comprised women exposed to children with unclean faces, but no trachoma, and women whose children had no trachoma and had clean faces. Noncaretakers were defined as those who lived in a house with preschool children, but did not report taking care any children. Caretakers were defined as those who reported taking care preschool children. RESULTS Household data and ocular examinations were obtained for 46 the 5 sampled women (89 percent) (table 1). Noncaretakers were significantly less likely to come in for ocular examinations. There was no statistically significant difference between noncaretakers with and those without ocular examination by age or by any the other variables in the interview. Thus, those with missing ocular examinations were unlikely to have had differential trachoma rates. The women who did not have complete ocular examinations were absent from the village during the week in which the examinations were held. Of the 5 women within the sample frame, 1 (22 percent) lived in houses without young children, 11 (2 percent) were noncaretakers, and the majority, 299 (58 percent), were caretakers preschool-age children at the time the study. The number children aged 1-7 years in the house for caretakers and noncaretakers was not significantly different (p >.3). Of 277 caretakers with complete ocular and household information, had children with missing trachoma grades. Complete facial and ocular observations were obtained for 472 (9 percent) 527 children aged 1-7 years in the sample clusters. Women who were noncaretakers preschool children were either relatively young or relatively old, creating a bimodal distribution age (table 2). The median age caretakers and women with no preschool children in the home was identical (age 3 years). The prevalence chronic manifestations trachoma (scarring, trichiasis, and TABLE 1. Number and percentage women and children in sample clutters with completed examinations, Tanzania, 1989 Group Preschool children Women No preschool children in household Noncaretakers Caretakers in census Participants with completed examinations %

4 Exposure to Children and Risk Active Trachoma in Women 369 TABLE 2. Ages by caretaker status women from whom household data obtained, Tanzania, 1989 and ocular examinations were Age group (years) Caretaker status No preschool children in house 37 Noncaretaker 35 Caretaker 69 % % % Total 888 Total TABLE 3. Prevalence rates trachoma among women different age groups, Tanzania, 1969 Prevalence by age group (years) Trachoma grade (n-141) (n-218) Active follicular and/or Intense 1 Tarsal scarring 24 Trichiasls with or without comeal opacity TABLE 4. Rate ^ Rate (%) (%) Age-adjusted rates for trachoma by caretaker status,' ranzania, 1989 Disease status Active follicular and/or severe trachoma Scarring (%) Trichiasls (%) Comeal opacity (%) (n = 11) Caretaker status Caretaker Noncaretaker {n - 277) (n = 8) 9.1* Rate Total (n - 46) 19 No preschool children {n- 12) Rate *Odds ratio for active trachoma In caretakers compared with women living wtth no preschool chbdren (p =.1). 4.1 comeal opacity) rose steadily with age (table 3). The frequency active disease, however, was approximately 7-8 percent for all age groups. The age-adjusted prevalence the different grades trachoma according to caretaker status suggested an increase in active disease (follicular or severe trachoma) with increasing exposure to children (table 4). The prevalence odds ratio for active trachoma appeared to increase from women with no preschool children in the house to noncaretakers (odds ratio (OR) = 1.63) to caretakers (OR = 2.43); however, the test for trend was not strongly statistically significant (p =.8). The adjusted rates for the different chronic forms trachoma did not exhibit a clear trend with increasing child care contact (table 4). To examine the effect caring for increasing numbers young children in a house, we next confined the analysis to the 357 women who resided in a household

5 37 Congdon et al. where children aged 1-7 years were present (table 5). A total 8 women were noncaretakers in such houses. A significantly increased risk active trachoma associated with the total number young children cared for was observed among women with children in the house. The age-adjusted odds ratio for active disease in women rose with the number preschool children, reaching 3.38 (95 percent confidence interval (CI) ; p =.4) among women caring for three or more preschool children compared with noncaretakers. However, the most striking increased risk was observed when caretaker status was evaluated accord- TABLE 5. Active trachoma status In women by number children cared (or and number children cared for who had active disease, Tanzania, 1989 children cared for Total With active trachoma 1-2 ;>3 Without active trachoma Total no. women 8* Women with active trachoma % These are noncaretakers In houses with chidren. Ageadjusted odds ratio for three or more children compared with (p».4). ing to numbers children with active trachoma (table 5). The prevalence active trachoma rose from ( 88) among caretakers with no infected children to 9.5 percent ( 8) among those with one to two infected children. Caretakers caring for three or more infected children had a prevalence rate 4. percent (6 ) (p <.1, Fisher's exact test). Conversely, the rate was zero among caretakers who had three or more children, none whom had trachoma ( ). The only other factor associated with active trachoma in women was religious beliefs. Adjustment for ancestor worship and other potential confounding factors did not change the findings. Caring for children with active trachoma was also associated with some the chronic trachomatous sequelae among caretakers. All eight caretakers with trichiasis cared for at least one child with active trachoma {p =.5, Fisher's exact test), as did all four caretakers with trachomatous corneal opacity (table 6). The age-adjusted odds ratio for conjunctival scarring was 1.55 for caretakers who cared for children with active disease compared with caretakers whose children did not have disease, although the increase was not statistically significant. Individuals who cared for three or more children with active disease experienced an age-adjusted odds ratio for scarring 3. (95 percent CI ; p =.4) relative to those who did not care for such children. Among caretakers, those whose children had unclean faces and trachoma were at no increased risk compared with caretakers TABLE 6. Trachomatous scarring, trictilasis, and corneal opacity among caretakers by active disease status children, Tanzania, 1989 Caretakers with signs caring for Chronic trachoma signs No chid with active trachoma 1 or more chidren with active trachoma (n-174) No %~ Scarring Trichiasis 8 5 Trachomatous comeal opacity Age-adjusted odds ratio t Fisher's exact test p value.16*.5t.3t

6 Exposure to Children and Risk Active Trachoma in Women 371 TABLE 7. Prevalence active trachoma In caretakers according to trachoma status and facial signs their preschool children, Tanzania, 1989 Face status Trachoma status child* crtfdt caretakers Unclean Unclean Clean Clean None Active None Active % caretakers with active trachoma * Unclean, at least one chid the caretaker had an unclean face; clean, ad chbdren the caretaker had clean faces. t None, no child the caretaker had active disease; active, at least one chid the caretaker had active dsease. whose children had trachoma alone (table 7). In particular, neither flies on the face nor nasaj discharge, signs found in our previous work to be significance in intersibling spread trachoma in this population (1), appeared to increase the risk trachoma for caretakers. The increased risk for active disease in women appeared to be greatest for caretakers whose children had trachoma, regardless the children's facial characteristics. DISCUSSION The results this study provide evidence that close, physical contact with preschool children, especially preschool children with trachoma, was associated with an increased odds ratio for active trachoma in women. While there is a trend in the risk active disease with increasing exposure to children, with caretakers experiencing the highest rate active disease, there is a significantly enhanced risk to caretakers from exposure to greater numbers children with active trachoma. The importance exposure to preschool children with disease as a risk factor for active disease in women in this population is underscored by the fact that only a small minority women (22 percent) were without at least household contact with preschool children at the time the study. The question could be raised as to whether the mothers were the source infection for their children, rather than the converse. As this study was cross-sectional in design, the data do not permit a definitive answer. Nevertheless, other epidemiologic data suggest that preschool children are the reservoir active trachoma in these villages (8, 1). Active disease declines with age in both sexes; however, women childbearing age tend to have more active disease than do men, which provides evidence that women are the recipients infection within the family, rather than the source (2). Moreover, the increase in risk for women associated with an increase in the number infected children also suggests that transmission is from child to mother. Because the cross-sectional nature the data, we cannot rule out the theoretical possibility relapse a chronic infection as the source active disease in women. Relapse chlamydia ocular infection is not felt to be an important cause infection in trachoma (13). Rather, reinfection is the more likely explanation for ocular infection. Moreover, if relapse were the primary explanation, we would expect to see more equal rates among women, regardless their caretaker status. The finding that rates active disease in women were higher if the women were caretakers children with active disease suggests that relapse is not an important explanation. Trachomatous trichiasis and the other potentially blinding sequelae active infection are thought to be the result constant reinfection over the course many years (12). Presumably, women who continue to be exposed to reinfection from their children should be an important risk group for chronic and potentially blinding disease. We cannot assess risk trichiasis from chronic reinfection in this population women because our study was cross-sectional in design. Given this limitation, it is noteworthy that we did find some evidence that women with scarring, trichiasis, and corneal opacity associated with trichiasis were more likely to be currently caring for children with active trachoma. The association active disease in children with chronic sequelae in their caretakers may reflect a prile trachomatous families in the area, where chronic and acute disease cluster together. In this population, 6 percent children exhibit active trachoma, and, therefore, a

7 372 Congdon et al. large proportion caretakers are exposed to risk infection. This raises a central question: If the majority caretakers are at risk exposure to active trachoma from their children, why do fewer than one in 1 show evidence active infection themselves at the time the survey? We postulated that transmission from child to mother may be heightened in children with unclean faces. Our previous work in this population has shown that both flies and nasal discharge on a child's face increase the risk trachoma to the sibling (1). Overall, only 1 percent preschool children had clean faces, which is consistent with our findings in other villages in this area (8, 1). Among the few children with clean faces, there was little evidence that the risk trachoma in the caretaker was decreased. The extreme lack face washing in this area may also have led to some misclassification the children with clean faces, although that misclassification is unlikely to be differential with respect to maternal trachoma status. Women have been observed to have a higher prevalence all signs trachoma than do men the same age, with caretakers at greater risk than noncaretakers (2). These differences have been thought to be due to more immediate physical contact with preschool children, as opposed to more general household contact, as would be true for men or noncaretakers. We addressed this issue by posing the following question: How did the risk caretakers exposed to infected children compare with that noncaretakers, whose exposure is more similar to that men? As noted above, the rate for active trachoma increased from for caretakers with no infected children to nearly 4 percent among those with three or more children with active trachoma. This is in contrast to no increased risk for a noncaretaker in a house with a child with active trachoma, compared with a noncaretaker in a house with children free active trachoma (5.4 vs. 5.6 percent). The intimacy caretaking exposure to an infected child, rather than the more general household contact that a male or other noncaretaker would have, appears to be necessary to increase the risk trachoma. These data underscore the need for effective intervention strategies that interrupt transmission at the household level. Any program whole-scale antibiotic therapy should be aimed not only at children, but at the caretakers as well. REFERENCES 1. Dawson CR, Jones BR, Tarizzu ML. Guide to trachoma control. Geneva, Switzerland: World Health Organization, West SK, Munoz B, Turner V, et al. The epidemiology trachoma in Central Tanzania. Int J Epidemiol 1991,2: Ballard RC, Sutter EE, Fotheringham P. Trachoma in a rural South African community. Am J Trop MedHyg 1978;27: Tielsch JM, West KP, Katz J, et al. The epidemiology trachoma in Southern Malawi. Am J Trop Med Hyg 1988;38: Reinhards J, Weber A, Nizetic B, et al. Studies in the epidemiology and control seasonal conjunctivitis and trachoma in Southern Morocco. Bull World Health Organ 1968;39: Taylor HR, Millan-Velasco F, Sommer A. The ecology trachoma; the epidemiology trachoma in Southern Mexico. Bull World Health Organ 1985;63: Jones BR. The prevention blindness from trachoma. Trans Ophthalmol Soc UK 1968;95: Taylor HR, West SK, Mmbaga BBO, et al. Hygiene factors and increased risk trachoma in Central Tanzania. Arch Ophthalmol 1989; 17: Lane SD. A bicultural study trachoma in an Egyptian hamlet. Doctoral Thesis. University California, San Francisco, CA, West SK, Congdon NG, Katala S, et al. Facial cleanliness and risk trachoma in families. Arch Ophthalmol 1991; 19: Thylefors B, Dawson CR, Jones BR, et al. A simple system for the assessment trachoma and its complications. Bull World Health Organ 1987;65: Taylor HR, West SK, Katala S, et al. Trachoma: evaluation a new grading system in the United Republic Tanzania. Bull World Health Organ 1987;65: Grayston JT, Wang SP, Yeh LJ, et al. Importance reinfectionin the pathogenesis trachoma. Rev Infect Dis 1985;7: