Matched-pair Case-control Study Research Articles

Extremely discordant sib-pair study design to determine risk factors for neovascular age-related macular degeneration.

To search for factors that contribute to the development of neovascular age-related macular degeneration (AMD). In a matched-pair case-control study, we studied sib pairs in which the index sibling had neovascular AMD in at least 1 eye and the unaffected sibling had normal maculae (or at most only a few small drusen) and was past the age at which the index case was diagnosed. Factors studied included sex, iris color, education, alcohol consumption, body mass index, vitamin use, smoking history, hypercholesterolemia, aspirin use, hypertension, other cardiovascular disease, any autoimmune disease, and non-insulin-dependent diabetes mellitus. Conditional logistic regression was performed to identify predictors of neovascular AMD. On the basis of 73 sib pairs, multivariate regression analysis revealed a statistically significant 2% increase in risk of neovascular AMD with each pack-year of smoking (odds ratio, 1.02; 95% confidence interval, 1.01-1.04; P =.007). Suggestive but nonsignificant associations were also observed for mean lifetime alcohol consumption, adult lifetime body mass index, and hypertension in multivariate regression analyses. Using extremely discordant sib pairs to study risk factors for AMD, a novel approach in epidemiological design, we found evidence that smoking is a risk factor for neovascular AMD.

Low interleukin-12 activity in severe Plasmodium falciparum malaria.

We compared interleukin-12 (IL-12) and other cytokine activities during and after an acute clinical episode in a matched-pair case-control study of young African children who presented with either mild or severe Plasmodium falciparum malaria. The acute-phase, pretreatment plasma IL-12 and alpha interferon (IFN-alpha) levels, as well as the acute-phase mitogen-stimulated whole-blood production capacity of IL-12, were significantly lower in children with severe rather than mild malaria. IL-12 levels, in addition, showed strong inverse correlations both with parasitemia and with the numbers of circulating malaria pigment-containing neutrophils. Acute-phase plasma tumor necrosis factor (TNF) and IL-10 levels were significantly higher in those with severe malaria, and the concentrations of both of these cytokines were positively correlated both with parasitemia and with the numbers of pigment-containing phagocytes in the blood. Children with severe anemia had the highest levels of TNF in plasma. In all the children, the levels in plasma and production capacities of all cytokines normalized when they were healthy and parasite free. The results indicate that severe but not mild P. falciparum malaria in young, nonimmune African children is characterized by down-regulated IL-12 activity, contrasting markedly with the up-regulation of both TNF and IL-10 in the same children. A combination of disturbed phagocyte functions resulting from hemozoin consumption, along with reduced IFN-gamma responses, may contribute to these differential effects.

A Mexican restaurant-associated outbreak of Salmonella Enteritidis type 34 infections traced to a contaminated egg farm.

In May 1996, the Georgia Division of Public Health was notified about a cluster of persons with Salmonella Enteritidis (SE) infections in Waycross, Georgia. A matched pair case-control study to determine risk factors for illness found a statistically significant association of SE infection with a history of having eaten at Restaurant A during the 5 days before onset of illness (relative risk = 13 [95% confidence interval (CI) = 3-62, P < 0.01]). In a second case-control study, to determine specific food exposures, consumption of a deep-fried Mexican dish (chile relleno) (4 of 21 cases vs. 0 of 26 controls, odds ratio undefined, 95% CI > 1.46, P = 0.034) was found to be significantly associated with SE infection. An environmental investigation found evidence of suboptimal food storage and cooking temperatures at Restaurant A; cross contamination of foods may have contributed to the low attributable risk identified for chile rellenos. Five of 37 Restaurant A food and environment specimens yielded SE strains. All five positive specimens were from chiles rellenos. Of the seven outbreak-associated strains (six patient isolates and one food isolate from Restaurant A) for which phage typing was conducted, all were phage type 34. A FDA traceback investigation through Restaurant A's single-egg supplier identified the potential source as three interrelated farms in South Carolina. Environmental culture from one of these farms yielded SE phage type 34. As a result of this outbreak, FDA helped institute a statewide egg quality-assurance programme in South Carolina to minimize SE contamination of eggs.

Interior surface materials in the home and the development of bronchial obstruction in young children in Oslo, Norway.

This study assessed the role of polyvinyl chloride (PVC) plastics and textile materials in the home in the development of bronchial obstruction during the first 2 years of life. The study was a matched pair case-control study based on a cohort of 3754 newborns in Oslo in 1992 and 1993 who were followed up for 2 years. The case group consisted of 251 children with bronchial obstruction; the control group was matched one-to-one for date of birth. In conditional logistic regression analysis, the risk of bronchial obstruction was related to the presence of PVC flooring (adjusted odds ratio [OR] = 1.89; 95% confidence interval [CI] = 1.14, 3.14) and textile wall materials (adjusted OR = 1.58; 95% CI = 0.98, 2.54). The reference category was wood or parquet flooring and painted walls and ceiling. Further analysis revealed an exposure-response relationship between the assessed amount of PVC and other plasticizer-containing surface materials and the risk of bronchial obstruction. This study provides new evidence of the role of PVC and textile wall materials in the development of bronchial obstruction in young children.

Cigarette Smoking as a Triggering Factor of Hidradenitis suppurativa

Background: Hidradenitis suppurativa is a chronic inflammatory skin disease involving the axillary, inguinal and anogenital regions and sometimes, in addition, the submammary or sacral areas. The etiology of this condition is unknown. Objective: A matched-pair case-control study was performed to evaluate the influence of smoking habits on the manifestation of this disease. Methods: Patients who had received surgical treatment for hidradenitis suppurativa in two dermatological centers completed a questionnaire dealing with family history, course of the disease and smoking habits. To form a randomized matched-pair control group, an equal number of patients admitted for various other skin diseases such as atopic dermatitis, varicose veins, skin tattoos, alopecia areata or melanoma was matched for sex and age and evaluated for smoking habits. Statistical analysis was performed by use of several χ² tests in a cross-table setting. Moreover, a comparison to the expected smoking prevalence in Germany based on national statistics was performed. Results: Out of 84 patients treated for hidradenitis suppurativa, 63 subjects (27 men, 36 women) completed the questionnaire. The rate of active cigarette smokers was 88.9% (56 patients), whereas 4 subjects (6.4%) had never smoked. 3 patients (4.8%) stated to be ex-smokers, but 2 of these had quit smoking only recently and after onset of the disease. The rate of smokers in the matched-pair control group was 46%. The significantly higher proportion of active smokers among patients with hidradenitis suppurativa can be expressed by an odds ratio of 9.4, the calculated 95% confidence interval was 3.7–23.7 (p < 0.001). The expected smoking prevalence in Germany was 26.7% according to national statistics. 73% of our patients had no family history of hidradenitis suppurativa whereas 27% reported at least one affected first-degree relative. Conclusion: From the exceedingly high rate of smokers among patients with this condition we conclude that cigarette smoking is a major triggering factor of hidradenitis suppurativa. Remarkably, the disease can be categorized as a smoking sequel that is neither of vascular nor neoplastic nature. Because familial occurrence was rather rarely reported, and because an environmental factor in the form of cigarette smoking appears to be of crucial importance to trigger the disease, we assume that the genetic basis of hidradenitis suppurativa is polygenic rather than mendelian. Smoking cessation should be encouraged particularly in patients with hidradenitis suppurativa although it is unknown whether this improves the course of the disease.

Maternal fever in term labour in relation to fetal tachycardia, cord artery acidaemia and neonatal infection.

To determine 1. whether maternal fever during term labour is associated with acidaemia at birth and neonatal infection and 2. whether fetal tachycardia precedes maternal fever and is associated with neonatal infection. Retrospective matched-pair case-control study. Two hundred and forty-eight newborn infants whose mothers developed fever during term labour (cases) and 248 control infants. The women were matched for parity and duration of labour. Cord artery acid-base status, Apgar scores, neonatal infections, and fetal heart rate before maternal fever. Mean pH, as well as the rate of cord artery acidaemia at birth (pH < 7.10) was equal in cases and controls (in both groups 7.24% and 6%, respectively). Signs of septicaemia and/or pneumonia were identified in 17 case newborns (7%) and in one control (0.4%; OR 17.0, P < 0.001). Of 212 pairs with complete heart rate data, fetal tachycardia preceded maternal fever in 39 cases (18%) and in 16 controls (8%) (OR 2.6, P = 0.003). Tachycardia before maternal fever was not associated with increased neonatal infectious morbidity. Maternal fever during term labour was associated with perinatal infection, but not with acidaemia at birth. Elevated fetal heart rate preceded maternal fever in a minority of cases and was not associated with perinatal infection.

Scarring trachoma is associated with polymorphism in the tumor necrosis factor alpha (TNF-alpha) gene promoter and with elevated TNF-alpha levels in tear fluid.

Tumor necrosis factor alpha (TNF-alpha) may play a central role in the disease pathogenesis which occurs as a consequence of chlamydial infection. To investigate the importance of TNF-alpha gene promoter polymorphisms and TNF-alpha levels in tear fluid in scarring trachoma, a large matched-pair case-control study was performed in The Gambia. The -308A allele was present in a higher proportion of patients (28.4%) than controls (18.4%), with an increasing association for homozygotes (chi2 for trend, P = 0.032; allele frequency, 0.163 in patients and 0.099 in controls; chi2, P = 0.025). For the -238A allele, the association was similar but not significant. The disease association was highly significant when the number of either -308A or -238A sites in an individual was considered (P = 0.003). TNF-alpha promoter alleles are tightly linked to some HLA class I and II alleles, but multivariate analysis confirmed that the disease associations were independent of HLA, although a class I allele, A*6802, is also associated with disease. TNF-alpha was more frequently detected in tear samples from patients (27.6%) than from controls (15.9%), increasingly so for higher levels of detectable TNF-alpha (P = 0.015). Among patients, detectable TNF-alpha in tears was highly associated with the presence of ocular chlamydial infection (P < 0.001). The results indicate that TNF-alpha plays a major role in the tissue damage and scarring which occurs as a consequence of Chlamydia trachomatis infection.

Socioeconomic and behavioural risk factors for mortality of individuals with IDDM in Japan: population-based case-control study. Diabetes Epidemiology Research International (DERI) US-Japan Mortality Study Group.

The aim of this population-based, one-to-one matched-pair case-control study was to evaluate the factors concerning the markedly increased risk for dying among Japanese subjects with insulin-dependent diabetes mellitus (IDDM) from a social and behavioural perspective. The study was based on the population-based cohort of IDDM subjects in the Diabetes Epidemiology Research International Mortality Study. We studied 90 cases who died and 90 living control subjects, selected from the rest of the cohort, who were matched for sex, birth year, year of diagnosis and duration of diabetes. Socioeconomic and behavioural status were surveyed through a questionnaire. Conditional logistic regression analyses based on 55 respondent pairs revealed that the better educated patients (year of completing education: odds ratio = 0.66) who kept the same physician (number of times a patient changed physician: odds ratio = 2.77) and who attended a clinic specializing in diabetes (attendance at university hospital clinic: odds ratio = 0.18) injecting insulin several times a day (number of injections, odds ratio = 0.31) and more frequently attending the clinic (> or = 12 times per year, odds ratio = 0.23) were at substantially lower risk of death. The results begin to profile the patients with the highest risk of dying who could be identified earlier and undergo intervention treatment.

Matched-pair case-control studies when risk factors are correlated within the pairs.

If pair members are independent, simple matched-pair case-control studies are known to yield consistent estimates of the population odds ratio. If pair members are not independent, this is not necessarily true. It has been shown previously that the usual matched-pair estimate remains consistent if the exposure of interest is correlated within the pairs. However, the effect of correlation of unmeasured risk factors within the pairs has not been studied. We examine the effect of within-pair correlation of unmeasured risk factors independent of the measured exposure. This is done within the context of a simple matched-pair case-control study. We compare the large-sample expectation of the usual matched-pair estimate to the population odds ratio. We show that the usual estimate may be inconsistent in the presence of this correlation. However, if the disease is rare, the magnitude of the bias will be negligible. Correlation of unmeasured risk factors independent of the measured exposure is not a practical problem in this setting.

Risk factors for fever in labor

To identify risk factors for fever in labor. A retrospective case-control study was conducted. Maternal sublingual temperature was measured every 2-4 hours during labor in 3109 of 3860 consecutive term parturients presenting from September 1992 through December 1993. Women who had fever (at least one recorded temperature of 38C or more, n = 72) during labor were compared with those who remained afebrile (n = 3037). Furthermore, a matched-pair case-control study was conducted, involving 250 women at term who developed fever in labor and 250 controls matched for parity and duration of labor; all delivered between January 1989 and December 1993. A conditional multiple logistic regression analysis was used to identify independent risk factors for fever during labor. In the case-control study, fever was associated with epidural analgesia, nulliparity, and a long duration of labor. These three variables were also related among themselves. However, multiple regression analysis showed that all three variables were independently associated with maternal temperature. In the matched-pair study, epidural analgesia, rupture of membranes longer than 24 hours, latency phase exceeding 8 hours, and a temperature in the upper normal range (37.5-37.9C) at admission were independent risk factors for developing fever in labor. Epidural analgesia, duration of labor, and a long interval from rupture of membranes to delivery were independent risk factors for maternal fever in labor.

MMPI and Critical Flicker Frequency (CFF) Analysis in Headache Patients With and Without Drug Abuse

Sixty-three headache patients (migraine: n = 28; tension-type headache: n = 35) who fulfilled the IHS criteria of 'drug abuse' were investigated by means of the Minnesota Multiphasic Personality Inventory (MMPI) and the Critical Flicker Frequency (CFF) analysis. The results were compared to those of 63 headache patients without drug abuse (matched-pair case-control study). With respect to the MMPI results, no statistically significant differences between patients with drug abuse and patients without drug abuse were found. However, patients with drug abuse showed significantly decreased CFF values compared to patients without drug abuse. This was true both for patients with migraine and for patients with tension-type headache. Thus, CFF analysis may serve as a useful method to differentiate between headache patients with drug abuse and those without drug abuse.

Risk Factors for Nosocomial Urinary Tract and Postoperative Wound Infections in Renal Transplant Patients: A Matched-Pair Case-Control Study

A matched-pair case-control study was done on 4 groups of renal transplant patients who acquired nosocomial infections: 1) urinary tract infection, 2) postoperative wound infection, 3) urinary tract plus postoperative wound infection and 4) the entire group of patients. For urinary tract infection patients a prolonged period of hemodialysis before hospitalization was considered a risk factor. Renal transplantation with an HLA-1 (identical) donor graft was a characteristic related to the control group. High levels of plasma creatinine and prolonged vesical catheterization were risk factors for acquiring postoperative wound infection. The latter was also considered to be a risk factor for both infections, as well as the inadequate use of antibiotic prophylaxis and the number of antibiotics used. For the entire group of patients surgical wall hematoma was a risk factor. In this group the independent risk factors analyzed by multivariate logistic regression were renal transplantation with a cadaver donor graft, prolonged vesical catheterization and prolonged use of antibiotics. Careful management of the cadaver donor allograft, decreasing the chances of contamination, decreasing the interval of urinary catheter maintenance and use of antibiotics in the postoperative period are measures that can contribute to lessen the incidence of these nosocomial infections in renal transplant recipients.

Vibration-induced White Finger and Auditory Susceptibility to Noise Exposure

It is well known that there are wide individual variations in the susceptibility of human hearing to noise exposure. The hearing of male forest workers were examined with vibration-induced white finger (VWF) as an indicator of noise susceptibility of hearing. The results were as follows: 1. Hearing losses at 4 and 8 kHz in subjects with VWF were greater than those of unaffected men in a matched pair case-control study allowing for confounding due to differences of age and noise exposure. 2. The hearing loss at 4 kHz on the ipsilateral side of the hand with VWF was greater than the loss on the contralateral side in subjects with VWF in one hand. 3. The hearing loss at 4 kHz in subjects with VWF progressed more rapidly during a five-year follow-up period than the loss in men with no history of VWF. These results indicate that the hearing of subjects with VWF was more vulnerable to noise than the hearing of subjects without VWF. It was suggested that a pathological change causing VWF, such as enhanced vasoconstriction due to elevated sympathetic nervous tone, could also cause this additional auditory vulnerability to noise exposure.

Non-01 Vibrio cholerae infections in Cancun, Mexico.

To determine the role of Vibrio cholerae as a cause of diarrheal illness in Cancun, Mexico, an investigation was conducted in July and August 1983. Although toxigenic V. cholerae 01 were not found, non-01 V. cholerae were isolated from 22 (16%) of 134 stools from persons with diarrheal illness and none of 22 stools from well persons; 58 (92%) of 63 sewage samples; 12 (86%) of 14 untreated well water samples; a home storage tank for treated water; and 5 (21%) of 24 samples of raw seafood. None of the V. cholerae isolates from patients were toxigenic. The illness occurred mainly in small children, and were characterized principally by diarrhea and abdominal pain. No patient was seriously ill, and all recovered without sequelae. Seven different serotypes of non-01 V. cholerae were isolated from the stool specimens, and Smith serotype 12 accounted for 10 (46%) of the 22 isolates. A matched-pair case-control study found that cases were more likely than controls to have eaten home prepared gelatin (P = 0.03, OR = 5/0) and seafood (P = 0.06, OR = 4/0).

Estimation of exposure-specific rates from sparse case-control data

Several methods for estimation of exposure-specific rates from case-control data have been developed, but none are practical for application to sparse stratifications, such as individually matched case-control data. This paper presents a method for estimating standardized exposurespecific rates that can be applied to sparse data, as well as other (non-sparse) stratifications. An application to a matched-pair case-control study of prostate cancer is given.

An Association Between Raynaud's Phenomenon and Hearing Loss in Forestry Workers

Forestry workers using chain saws run the risk of both vibration disease and noise-induced hearing loss. It was reported that people with Raynaud's phenomenon of occupational origin had more severe hearing loss than those without it. A possibility of confounding due to age and duration of noise exposure from the tools remained unsolved, however. A matched pair case-control study was conducted to eliminate such confounding. Forestry workers operating chain saws, bush cleaners and/or winches were investigated. For 37 men with Raynaud's phenomenon a control was chosen from those unaffected whose age and operating hours for the tools were almost equal to those of the cases. Good matching was achieved for age and for total hours of work with the tools. The cases had higher median hearing threshold than the controls at every frequency. The difference was significant (p less than 0.05) at 4 and 8 kHz and almost significant at 2 kHz (p not equal to 0.06). According to our classification of the audiogram, the cases had more advanced types of noise-induced hearing loss than did the controls. Thus, an association between Raynaud's phenomenon and hearing impairment in forestry workers handling chain saws and/or other tools was found to persist even after the effect of age and exposure time was eliminated.

Risk Factors for Fireworks-Related Injury in Washington State

To determine the frequency and effects of and risk factors for fireworks -related injury, we identified all 146 persons who were injured by fireworks and sought emergency care during the 1983 July 4 holiday in the Seattle area. The mean charge for medical care for the injuries received was +562; 7.1% of those injured required hospitalization. In a matched-pair case-control study, use of either of two fireworks types-- firecrackers or aerial devices--was significantly associated with injury (odds ratios [ORs], 3.3 and 2.9, respectively; 95% confidence intervals [CI], 1.2, 8.5, and 1.2, 6.6, respectively). Also associated with injury were several fireworks misuse behaviors, including lack of adult supervision of children (OR, 11.5; CI, 2.8, 100.6). We conclude that fireworks cause serious injuries that theoretically could be prevented by behavioral changes or decreased availability of high-risk fireworks devices.

FOODBORNE TRANSMISSION OF CHOLERA IN MICRONESIAN HOUSEHOLDS

During a cholera epidemic in Truk (Micronesia), in a survey of 1 village, transmission of Vibrio cholerae 01 in the 28 households with illness appeared to be through food contaminated in the home. Households in which the index case was a foodhandler had significantly higher attack rates than households in which the index case was not a foodhandler. Members of households with illness were significantly more likely to become ill if they had eaten food prepared by a foodhandler who had recently been ill. A matched-pair case-control study on several adjacent islands confirmed this relation. Those who had tended an ill person were not at increased risk of cholera in either study. This study suggests that in this outbreak, foodborne transmission of V cholerae 01 was more important than contact spread.

Case-control studies: a sequential approach.

A sequential approach to the design of a matched pair case-control study is proposed as an alternative to fixed sample plans when information on case-control pairs is acquired sequentially in time. The method described is that of Wald using the Sequential Probability Ratio Test (SPRT) for comparing two binomial populations. The test is an open, non-truncated procudure. Several tables are presented for the average sample numbers needed under the sequential plan and fixed sample plan for selected Type 1 and 2 errors, alternative relative risks of interest and exposure probabilities in case and controls. A hypothetical example is presented to illustrate the use of the method and discussion is given as to the possible advantages and disadvantages of the sequential approach to case-control studies.

Cholera in Portugal, 1974. II. Transmission by bottled mineral water.

During a cholera epidemic, Vibrio cholerae was isolated from two springs which supplied mineral water to a spa and to a commercial water bottling plant. Epidemiologic investigation found that cholera attack rates were 10-fold greater among visitors to the spa than among non-visitors. A subsequent matched-pair case-control study which excluded persons who had visted the spa showed that a history of consumption of the bottled non-carbonated water was significantly more common among bacteriologically confirmed cholera cases than among paired controls.