Controlling for continuous confounding factors: non- and semiparametric approaches

Confounding is one of the major types of bias encountered in observational epidemiologic surveys designed to study the relation between an exposure factor and a health event. A common way to remove confounding bias during the statistical analysis phase is to adjust for the confounders in a regression model. If a confounding factor is assessed as a continuous variable, it is necessary to define how the variable is entered into the regression model. In the case of logistic regression, we illustrate through simulation that coding by a binary variable or a categorical variable with broad categories may lead to substantial residual confounding. Specific approaches can be used to define a coding method that limits residual confounding. Among these, we briefly present nonparametric approaches and describe in detail several semiparametric approaches (generalised partial linear models, spline regression and fractional polynomials). These can be used to estimate the relation between a continuous factor and the health event of interest by a smooth non pre-specified function. In semiparametric models, the effect of certain covariates is coded by a parametric function, whereas the coding of one or two continuous variables is represented by a nonparametric function. These models can be used in exploratory analyses to describe dose-effect relations between the confounder and the health event, and thus help to define a relevant coding for the confounder.

Selection bias in semen studies due to self-selection of volunteers

BACKGROUND

Reports of a secular decrease in semen quality remain controversial, particularly due to the possibility of selection bias. We aimed to describe the potential bias due to self-selection of volunteers in semen studies involving fecund men.

METHODS

Using data from the French multicentre study REPRHOM, we compared the characteristics of the partners of pregnant women for three levels of participation: completion of a refusal questionnaire (n = 698), agreement to complete the study questionnaires only (n = 676) and agreement to complete the study questionnaires and give a semen sample (n = 331, 13% of the subjects approached).

RESULTS

Poorly educated men refused more often to participate than highly educated men. Semen providers were more likely to have experienced unfavourable pregnancy outcomes (odds ratio 1.68, 95% confidence interval 1.14-2.49) compared with participants completing the questionnaires only. Time to pregnancy was similar for all participants.

CONCLUSIONS

This study demonstrates the existence of selection bias in semen studies associated with fertility and socio-demographic characteristics of men. The results of semen analysis for this population sample cannot be extrapolated to the whole population from which the volunteers originate. More information is required on who participates, and participation rates should be reported in semen studies to make it possible to interpret the results correctly.

Epidemiology of male reproductive function: a field searching for tools

BACKGROUND

An increase in the frequency of disorders of the male human reproductive organs has been described over recent decades. Neither its causes nor its consequences on fecundity, the ability of the couples to produce a live child, are clearly known. This lack of knowledge may partly be due to methodological difficulties specific to reproductive epidemiology.

METHODS

We discuss the relevance and limits of some markers of the male aspects of human reproduction, focusing on semen parameters and fecundability, a measure of the probability of pregnancy.

RESULTS

Semen parameters are associated with fecundability, although they have a relatively low sensitivity to detect couples with low fecundability. The study of semen parameters proved central to describe the influence of environmental factors on the male side of reproductive function. The main limitation of semen studies is low participation rates and the possible selection biases ensuing. Fecundability can be estimated by collecting waiting time to pregnancy. Its assessment in retrospective studies often excludes the least fecund, those couples remaining childless, which entails a bias and a decrease in statistical power. The prospective approach and an approach relying on the enrollment of a cross-sectional sample of the couples currently trying to obtain a pregnancy (current duration approach) do not have these limitations. Although it has never been used, the assessment of fecundability using the current duration approach is promising both for aetiologic research and monitoring.

CONCLUSION

Most of the potential markers of male reproductive function are not assessed in the general population of France, with the exception of the incidence rate of testis cancer, which is currently increasing. We present some alternatives for a monitoring system of reproductive function.

[Use of national French health insurance register to identify the current address of a cohort]

BACKGROUND

The aim of the study was to assess the efficiency of the Registre National Inter-régimes des bénéficaires de l'Assurance-Maladie (RNIAM), which is the French register of health insurance, in order to identify the present address of subjects who have lived in the Beaumont-Hague county, France, between 1978 and 1998, when they were less than 25 years old.

METHODS

A cohort of 4,118 persons was defined by consulting school and civil status registers from three villages. We drew at random 824 subjects and between October 2000 and August 2001, we asked the RNIAM to locate them. For each subject, the usual identification parameters (first name, last name, birth date and eventually birth place) were provided. In case of a doubt concerning these parameters, a second request was undertaken with a slight modification.

RESULTS

Altogether, 94.5% of the included people were identified by the RNIAM. Identification was better for people born in France than for those born abroad (97% versus 52%) and 84.5% of people were linked to a health insurance regimen. The RNIAM was able to identify a correct address for 68.7% of the subjects.

CONCLUSION

The RNIAM seems promising for further epidemiological investigations. Nonetheless, it still remains insufficient by itself to identify addresses. Other means (tax records) should be evaluated and associated with the register data.

Population mixing and leukaemia in young people around the La Hague nuclear waste reprocessing plant

In order to investigate for an association between population mixing and the occurrence of leukaemia in young people (less than 25 years), a geographical study was conducted, for the years 1979 to 1998, in Nord Cotentin (France). This area experienced between the years 1978 and 1992 a major influx of workers for the construction of a nuclear power station and a new nuclear waste reprocessing unit. A population mixing index was defined on the basis of the number of workers born outside the French department of 'La Manche' and living in each 'commune', the basic geographical unit under study. The analyses were done with indirect standardisation and Poisson regression model allowing or not for extra-Poisson variation. Urban 'communes' were considered as the reference population. The Incidence Rate Ratio was 2.7 in rural 'communes' belonging to the highest tertile of population mixing (95% Bayesian credible interval, 95%BCI=1.2-5.9). A positive trend was observed among rural strata with increasing population mixing index (IRR for trend=1.4, 95%BCI=1.1-1.8). The risk became stronger for Acute Lymphoblastic Leukaemia in children 1-6 years old in the highest tertile of population mixing (IRR=5.5, 95%BCI=1.4-23.3). These findings provide further support for a possible infective basis of childhood leukaemia.

Time to pregnancy and semen parameters: a cross-sectional study among fertile couples from four European cities

BACKGROUND

In fertile populations, little is known about the association between semen parameters and time to pregnancy (TTP).

METHODS

Pregnant women from Copenhagen, Edinburgh, Paris and Turku who conceived without medical intervention were asked for their TTP (942 couples), and their partners provided a semen sample. The proportion of morphologically normal sperm and the multiple anomalies index (MAI, ratio of the total number of anomalies to the number of abnormal sperm) were centrally estimated. We estimated rate ratios for the occurrence of a pregnancy by a discrete survival model, adjusted for sexual activity and female factors affecting fecundity.

RESULTS

Increasing sperm concentration influenced TTP up to 55 x 10(6)/ml. The proportion of morphologically normal sperm influenced TTP up to 39% according to David's criteria, and this association held among the subjects with a sperm concentration >55 x 10(6)/ml. For strict criteria, the threshold value was 19% normal sperm. An increase of 0.5 in MAI was associated with an adjusted rate ratio for the occurrence of a pregnancy of 0.68 (95% confidence interval: 0.54-0.85).

CONCLUSIONS

These results highlight the importance of sperm morphology parameters and indicate that the effect of proportion of normal sperm on TTP may be independent of sperm concentration.

[Ionizing radiation and health: data and objectives]

Three dimensions must be considered to analyze the consequences of exposure to ionozing radiation on public health: assessment of exposure, measure of the public health risk, and monitoring the perception of risk by the population. We reviewed these three dimension for several situations of exposure to ionizing radiation: exposure related to the Chernobyl accident, occupational exposure, military-related exposure and medical exposure. For each situation, we discuss the pertinence of available data an monitoring systems used in France, and propose when necessary new systems. An extension of the cancer registries to the entire country would be a significant progress in terms of research and surveillance. This must be discussed in terms of potential utility in other domains of public health.

Regional differences in waiting time to pregnancy among fertile couples from four European cities

BACKGROUND

A previous European study found a longer time to pregnancy (TTP) among fertile women from Paris compared with women from other Western European countries. A co-ordinated, cross-sectional study of pregnant couples from Denmark (Copenhagen), France (Paris), Scotland (Edinburgh) and Finland (Turku) was therefore undertaken to assess differences in waiting TTP among couples from these cities.

METHODS

Pregnant women were invited to participate when they showed up for their first antenatal visit in one of the four centres. Inclusion criteria included that their partner was 20-45 years of age and born in the country in which he was currently living and that the pregnancy was achieved without fertility treatment. Both partners filled in a questionnaire and the man underwent a physical examination and delivered a semen sample (Turku: n = 237, Copenhagen: n = 302, Edinburgh: n = 212, Paris: n = 191).

RESULTS

French couples had a decreased probability of conception compared with couples from the other three countries, although only after adjustment for confounders. No significant differences between couples from the three other countries were found.

CONCLUSION

The observed geographical differences in TTP remain unexplained and were not due to differences in semen quality, but may be caused by varying exposures to environmental factors or psychological distress. In addition, selection bias due to the low participation rates cannot be ruled out. Future studies examining the causes for geographical differences in TTP are needed.

The incidence of childhood leukaemia around the La Hague nuclear waste reprocessing plant (France): a survey for the years 1978-1998

BACKGROUND

A previous study has suggested an increased incidence rate of leukaemia from 1978 to 1992 in people aged 0 to 24 years and living in the vicinity of the La Hague nuclear waste reprocessing plant without considering age and cytological type.

SETTING

The Nord Cotentin region (France) and the island of Alderney (United Kingdom).

STUDY OBJECTIVE

To describe the occurrence of leukaemia for each age group and cytological type from 1978 to 1998 in the same area, using accurate reference incidence rates and adequate estimation of the at risk population.

DESIGN

A geographical study of incidence using three zones defined according to their distance from the site (0 to 10 km: Beaumont-Hague electoral ward, 10 to 20 km and 20 to 35 km) has been conducted. The risk of leukaemia was estimated from the standardised incidence ratio (SIR) of the number of cases observed to the number expected. Exact 95% confidence intervals (CI) have been computed.

PARTICIPANTS

All people under the age of 25 years living in the study region between 1978 and 1998.

MAIN RESULTS

The observed number of cases of leukaemia in the study region as a whole was consistent with the expected value (SIR=1.03; 95%CI: 0.73, 1.41). No cases were observed on Alderney. The SIR in the Beaumont-Hague electoral ward was 2.17 (95%CI: 0.71, 5.07). The highest SIR was observed in the 5 to 9 years age group (SIR=6.38; 95%CI: 1.32, 18.65). This consists in acute lymphoblastic leukaemia cases.

CONCLUSION

This study indicates an increased incidence of leukaemia in the area situated at less than 10 km from the plant. Monitoring and further investigations should be targeted at acute lymphoblastic leukaemia occurring during the childhood incidence peak (before 10 years) in children living near the La Hague site and may be other nuclear reprocessing plants.

[Value of maximum expiratory volume per second and peak expiratory flow monitoring in asthma. 77 patients]

A retrospective study was conducted on a sample of 77 subjects to assess the interest of computerized flowmeter "one-flow" in the diagnosis and follow up of asthma. 77 patients and a control group matched for age and sex, measured four times by day, the forced expiratory volume in one second as well as the peak expiratory flow during 15 to 30 days. The monitoring of these two variables allowed to participate to the exclusion from the study of fifteen patients whose asthma diagnosis was informed. It also allowed the confirmation of four occupational asthma. Concerning the adherence of the treatment by patients, the conclusion of the study showed that 47% of patients modified their treatment at home without consulting their doctors, 38% resorted to their doctor and only 3% resorted to the emergencies.

[Medical treatment of atrial fibrillation in the healthy heart]

Whatever the cause and clinical features, establishing the strategy for the medical treatment of atrial fibrillation, requires three basic steps. First sinus rhythm must be restored, followed by the prescription of antiarrhythmic drugs to prevent recurrence and finally the use of anticoagulants. Obviously the underlying heart disease has a major effect on the patient's tolerance of atrial fibrillation, the risks of peripheral embolism and overall mortality. In patients with "solitary atrial fibrillation" and an apparently normal heart, each step of the treatment must be examined in light of the individual patients clinical situation and the risks of haemorrhage with anticoagulants as well as the risk of lethal ventricular arrhythmia due to the arrhythmogenic effect of class la or 1c antiarrhythmic drugs.

[Functional contribution of inhalation spacers in the treatment of asthma]

The aim of this study is to evaluate the ventilatory gain obtained by using metered dose inhaler (MDI) plus spacer versus MDI alone in 30 asthmatic patients (19 men and 11 women); aged 30 to 70 years old. Initial spirometry showed air flow obstruction. A reversibility test was performed with beta 2 agonists: first with MDI and then, later, with MDI plus spacer. In 27 cases (90%) the improvement of FEVI, referring to its initial value, was significantly better with spacer. This improvement was equal or superior to 20% in 19 patients with spacer versus only 9 patients with MDI. The improvement of FEVI was always better with spacer which ever the ways of expressing the bronchodilating response (referring to initial, predicted or absolute value). in conclusion, since the treatment of asthma is now based on local administration of medications, it is recommended to use spacers not only for children and patients who have coordination problems but more widely specially in severe asthma.

[Right accessory pathways with slow and decremential anterograde conduction. Electrophysiological changes during ablation of the proximal atrial poles]

Right sided accessory pathways with slow and decremential anterograde conduction differ from the anatomical substrates of the nodo-ventricular connections described by Mahaim. The connection may also be atrio-ventricular or atrio-fascicular. The authors report a series of 6 patients (2 men, average age 31 +/- 9 years) with antidromic reciprocating tachycardia implicating accessory pathways with decremential conduction (wide QRS complexes with left bundle branch block) the radiofrequency ablation of which was centered on the proximal atrial pole. The increment of pre-excitation was 75 +/- 18 ms. No nodo-ventricular or nodo-fascicular connections were observed. One patient had an atrio-ventricular connection, the mapping and ablation (5 applications) of which were performed on the annulus, as if it were a bundle of Kent. Five patients had a nodo-fascicular connection which was masked during sinus rhythm. The localisation of the proximal atrial insertion on the tricuspid annulus was initially estimated by studying the superior junction of the circuit and by endocavitary stimulation, and then by recording the M potential (activation of the accessory pathway). This was recorded in 4 patients, either at a specific point on the annulus (lateral or postero-lateral) which was the target of ablation (3 patients, 5 +/- 3 applications) or over a wider region (2 cm) at which ablation failed (1 patient: 13 applications). The site of the increment was determined in 2 patients and was proximal to the M potential. The M potential could not be recorded on the annulus in one patient. In 2 patients, the M potential was also recorded along the side of the pathway on the antero-lateral wall of the right ventricle of the tricuspid valve at its distal pole. In 2 patients, ablation was centered on the distal pole in second intention, near to the terminal part of the right bundle where it was localised in one of the two (18 applications). Finally, ablation was obtained in 5 out of 6 patients. They remain asymptomatic after 12 +/- 7 months follow-up. These cases show that accessory pathways with slow decremential conduction usually behave like a complete conduction pathway with an accessory laterally situated atrioventricular node on the tricuspid, an accessory bundle and a distal insertion which are all accessible to ablation.

[Catheter ablation of atrioventricular nodal reentrant tachycardia]

Electrophysiological data of atrioventricular nodal reentrant tachycardia recensed over the last 40 years in the animal and in man has not resolved the question as to the exact site of the reentry circuit: an exclusively intranodal pathway or a pathway involving part of the atrium? The remarkable efficacy of modern radical therapy of this arrhythmia with preservation of atrioventricular conduction reinforces the concept of reentry involving not only the atrioventricular node but also the juxta nodal atrium and the superior and inferior atrionodal connections. Radical treatment was initially surgical and then by catheter ablation. The technique of specific ablation of the rapid anterior pathway was the first to be described. Its limitation is the relatively high risk (about 10%) of complete atrioventricular block. Very quickly, radiofrequency ablation of the slow posterior pathway became the method of reference. Most patients do not have retrograde conduction in the slow pathway. The pathway is located in sinus rhythm by recording its specific potentials: either the rapid potential described by Jackman et al or the fragmented potential described by Haïssaguerre and Warin. The former is recorded from the posterior septal position anterior to the orifice of the coronary sinus; the second is recorded at the same level but slightly above in the mid septal position. Ablation of the slow pathway can be performed on these purely anatomical criteria. Using these approaches, an immediate success rate of over 90% may be obtained. The recurrence rate is 0 to 5%; that of complete atrioventricular block ranges from 0 to 4%.(ABSTRACT TRUNCATED AT 250 WORDS)

[Automatic implantable defibrillators: long-term results]

The automatic implantable defibrillator is an essential component of the preventive management of sudden death of patients with malignant ventricular arrhythmias. Though its efficacy in this indication is not contested, the data concerning its influence on global cardiac mortality is more controversial. The elements of this controversy are reviewed. Several prognostic factors are implicated, the principal of which being the patient's haemodynamic status as assessed by objective evaluation of left ventricular function. Prospective randomised studies comparing medical therapy with the automatic implantable defibrillator are under way and should provide a better understanding of its indications in the future.