Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
The Conception of Conception
- Arthur Leibovitz (20 October 2001)
-
better knowledge in chronobiology
- Joseph Watine (22 October 2001)
-
Sunlight, Vitamin D, and Crohn's
- Cory Mermer (23 October 2001)
-
Observations seen
- Kathryn Snook (23 October 2001)
-
Medical Epidemiology or Medical Astrology?
- J Murugan (31 October 2001)
-
Are month or season of birth relevant risk factors for Crohn’s Disease?
- Danielle Morris (6 December 2001)
-
No Evidence of Month of Birth or Seasonality as a Risk Factor for Crohn’s Disease in Children
- Lucy B Taylor, Mary Jane Platt, Chris West (20 December 2001)
-
Research pointer to a blind alley?
- Timothy R Card, Andrew Sawczenko, Miguel Goncalves, Bhupinder K. Sandhu, Richard F.A. Logan (24 December 2001)
-
Drawing seasonal inferences from biomodal data
- R Moineddin, Ross E.G. Upshur (18 June 2002)
|
|
|||
|
Arthur Leibovitz Shmuel Harofe Hospital
Send response to journal:
|
Congratulation. However you should consider the time of conception: Birth date - 9 months as a possible influencing factor. |
|||
|
|
|||
|
Joseph Watine, consultant, laboratory medicine Hôpital de Rodez, France
Send response to journal:
|
It is quite likely that a better knowledge in chronobiology could help us to better understand Henrik Toft Sørensen et al.’s observation that month of birth affect risk of Crohn's disease in childhood and adolescence [1]. The temporal organisation and variation of living organisms in general, and of the mammalian immune and gastrointestinal systems, as well as of the inflammatory reactions in particular, have indeed been largely demonstrated, with a likely effect on human physio-pathological balance [2, 3]. For example, the proportions of bone marrow, gut, skin and oral mucosa cells engaged in DNA synthesis (the so-called S-phase of the cell division cycle) vary by 50% or more along the circadian time-scale in healthy human subjects [2]. Similar circadian variations have been described in the functioning of the gastrointestinal systems, as well as in that of inflammatory reactions [3]. Quite logically, variations along the seasonal time-scale may have an even larger amplitude because, as logically assumed by chronobiologists, the necessity of shifting from one metabolic strategy to another, in order to adapt to environmental changes and to trigger growth and reproduction only when favourable season arrives, is led by selective advantage to utilise some mode of measuring the daily duration of darkness or light as the only reliable indicator of the course of the season [3]. In gastrointestinal and inflammatory diseases, much less in known about seasonal rhythms than about circadian ones. Thence perhaps the difficulty for Henrik Toft Sørensen et al.’s to propose convincing mechanisms that could clearly explain their observation [1]. This stresses, once again, the need for more systematic research in chronobiology in general and seasonal biological rhythms in particular. References: [1] Henrik Toft Sørensen, Lars Pedersen, Bente Nørgård, Kirsten Fonager, Kenneth J Rothman. Does month of birth affect risk of Crohn's disease in childhood and adolescence? BMJ 2001;323:907. [2] Haus E, Smolensky MH. Biologic rhythms in the immune system. Chronobiol Int. 1999;16: 581-622. [3] Touitou Y, Haus E (eds). Biologic rhythms in clinical and laboratory medicine. Springer-Verlag, 2nd printing, Heidelberg, 1994, 730 pages. |
|||
|
|
|||
|
Cory Mermer, Medical Researcher
Send response to journal:
|
Is it possible that the seasonal variation in terms of Crohn's Disease and month of birth is due to the biochemical changes that occur due to sunlight exposure, such as vitamin D synthesis? The peak occurred in August (end of summer) births, and therefore these infants will be exposed to less sunlight in their first 6 months of life than those born in March, at the trough of Crohn's incidence. An association between Crohn's and vitamin d receptor genes has already been proposed (1) and Crohn's patients are known to be at higher risk of bone mineralization disorders. Reference: 1. Simmons JD, Mullighan C, Welsh KI, Jewell DP. Vitamin D receptor gene polymorphism: association with Crohn's disease susceptibility. Gut 2000 ;47:211-4 . |
|||
|
|
|||
|
Kathryn Snook, Hospital Pharmacist Burwood Hospital
Send response to journal:
|
There is no need to publish this reply but I would like to point out my observations of my husband who has had crohns diagnosed at 24yr at the time of a bowel resection at the terminal ileum due to obstruction. He is now 40yr and we have noticed every spring the same time each year he has a major bout of crohns symptoms tired, pain and diarrhoea. Would this not suggest an allergen triggering an allergic or autoimmune response. I have also noticed these bouts a preceeded by low mood and depression. Prior to the bowel resection he suffered from severe allergy since the operation this has settled considerably. Do you think there is any strong connection with allergens? |
|||
|
|
|||
|
J Murugan, Senior House Officer Department of Obstetrics and Gynaecology. Royal Gwent Hospital, Cardiff Rd. Newport
Send response to journal:
|
Editor - Sorensen's et al paper (1). This appears to be another recent study investigating the month of birth as a risk factor for future pathology. Recent studies have included the month of birth as a risk factor for breast carcinoma, obesity, cataracts and even glaucoma(2,3,4). Unfortunately due to mulitiple confounding variables it is extremely difficult to interpret these observations. Sorensen's suggestion that the seasonal variations of Crohns Disease is due to the seasoanl variations of an infectious agent must be taken cautiously. Other confounding variables associated with seasonal varibility could also be a possible hypothesis. These include the known seasonal variability of birth weight, climate and even pre-natal sunlight exposure (5)! Although these studies are very interesting we must question is this valued epidemiology or simply "medical astrology". (1) Sorenson H, Pedersen L et al Does month of birth affect risk of Crohn's disease. BMJ 2001;323:907 (2) Yuen J, Ekbom A, et al Season of birth and breast cancer risk in Sweden. Bri. J. of Cancer 1994;70(3):564-8 (3) Phillips D I, Young J B. Birth weight, climate at brith and the risk of obesity in adult life. International J. of Obesity & Related Metabolic Disorders. 2000;24(3):281-7 (4)Weale R A. Cataract, gaucoma and season of birth amongst patients born on the Indian subcontinent. Indian J. of Opthalmology 1998;46(4):211-5 (5) Waldie K E, Poulton R. et al The effect of pre and post-natal sunlight exposure on human growth:evidence from the Southern Hemisphere. Early Human Development. 2000;60(1):35-42 J Murugan
Anthony N Griffiths
|
|||
|
|
|||
|
Danielle Morris, Wellcome training fellow in Epidemiology Royal Free Hospital
Send response to journal:
|
Are month or season of birth relevant risk factors for Crohn’s Disease? EDITOR- Further to the recent article by Sorensen et al showing a seasonal variation in birth of subjects with Crohn's disease1 and other studies2, 3,4,5, we would like to report on a similar finding from our study in Northeast Scotland. We hypothesised that perinatal or early postnatal infection as implicated by seasonal variations in birth, may be associated with subsequent Crohn's disease in susceptible individuals. In addition material circumstances in childhood may also be important in determining exposure and outcome from such infections. We examined date of birth of 856 patients with Crohn's disease diagnosed by a single practitioner in Northeast Scotland between 1955 and 1988(420 born before 1949). Months of birth statistics were obtained for all births in Scotland from 1889 to 1981. Season of birth was defined as winter (December to February), spring (March to May), summer (June to August) and autumn (September to November). Infant mortality, a proxy measure of childhood material circumstances, showed a sharp decline in Scotland in the post World War II period and so births before 1949 were analysed separately from those during and after 1949. The Chi-squared test and odds ratios with 95% confidence intervals for later Crohn's disease were calculated for each season in both time periods. There was no overall statistical association between season of birth and subsequent Crohn's disease but analysis of birth pre-1949 showed a significant increase in Crohn's disease in those born in Autumn (Pearson chi squared =11.97, p=0.007, OR 1.53, 95% CI 1.17 to 2.01), compared with winter births. Those born in November in the pre-war period were significantly more likely to develop Crohn's disease than those born in January. (Pearson chi squared = 29.56,p=0.002, OR 1.63, 95% CI 1.06 to 2.50). Such clustering supports the hypothesis that perinatal events, possibly infectious epidemics, may be associated with later Crohn's disease when conditions in childhood are poor, but that improvements in childhood circumstances may have masked these changes in recent decades, with other risk factors becoming more important. DL Morris, Wellcome Research Fellow in Epidemiology
1. Sorenson HT, Pederson L, Norgard B. Does month of birth affect risk of Crohn's disease in childhood and adolescence? BMJ 2001; 323:907 2. Ekbom A, Adami H, Helmick C, et al. Perinatal risk factors for inflammatory bowel disease: A case-control study. Am J Epidemiol 1990;132:1111-1119. 3. Gilat T, Hacohen D, Lilos P, Langman MJS. Childhood factors in ulcerative colitis and Crohn’s disease. Scand J Gastrenterol 1987; 22 : 1009-1024. 4. Haslam N, Mayberry JF, Hawthorne AB, Newcombe RG, Holmes GKT, Probert CSJ. Measles, month of birth, and Crohn's disease. Gut 2000;47:801-3. 5. Miller DS, Keighley A, Smith PG, Hughes AO, Langman MJ. Crohn’s disease in Nottingham: a search for time-space clustering. Gut 1975;16:454 -7. |
|||
|
|
|||
|
Lucy B Taylor, Research Assistant Royal Liverpool Children's Hospital, Eaton Road, Liverpool, L12 2AP, Mary Jane Platt, Chris West
Send response to journal:
|
Editor – The recent article by Sorenson et al (1) found a seasonal pattern in a cohort of young people with Crohn’s disease in Denmark. The paediatric inflammatory bowel disease database (funded by CICRA)has been prospectively collecting data on new cases of inflammatory bowel disease (IBD) in children aged under 16 years of age, from 53 tertiary centres and district general hospitals within the UK. We have used this population to examine whether these cases demonstrate an association between month of birth (seasonality) and the incidence of Crohn’s disease. A total of 1411 cases of IBD have been registered between 1st May 1997 and 23rd October 2001. Of these cases, 837 are Crohn’s disease. The median age of this cohort at diagnosis was 12.4 years (interquartile range = 3.7 years). Using a chi squared test to examine the association of month of birth and a periodic regression model to examine seasonality (in a similar manner to that used by Sorenson et al), no significant association between month of birth and Crohn’s disease was found (Chi Square= 13.02, df=11, p=0.29). Similarly, no significant association was seen between seasonality and the incidence of the disease. These results differ from those in the article by Sorenson et al (1). Although this may be explained within this study group by either age acting as a confounding factor, or because monthly fluctuation in birth rate has not been specifically adjusted for in the analysis, it would be expected that, if month of birth or seasonality was a significant risk factor in the development of Crohn’s disease in young patients, it would also be evident within this patient population. 1 Lucy Taylor Register Co-ordinator, 2 Mary Jane Platt Senior
Lecturer, 2 Chris West Lecturer.
1. Soreson HT, Pederson L, Norgard B. Does month of birth affect risk of Crohn’s disease in childhood and adolescence? BMJ 2001;323:907 |
|||
|
|
|||
|
Timothy R Card, Wellcome Training Fellow in Clinical Epidemiology Division of Public Health Sciences, University Hospital, Nottingham, Andrew Sawczenko, Miguel Goncalves, Bhupinder K. Sandhu, Richard F.A. Logan
Send response to journal:
|
Editor We read the recent report from Sorensen and co-workers [1] suggesting a possible (causal) relationship between month of birth and risk of developing Crohn’s disease with interest. The initial suggestion of such a relationship was made by Ekbom and co-workers in 1990[2] with a report of a weak and inconsistent association for birth in the first half of the year. A similarly weak association has also been recently reported in a multi-centre report from the UK [3]. Given that the report of Sorensen et al is at variance with these initial two reports we feel that it is important to examine why its findings are so different. We believe that this may at least in part be explained by an examination of the statistical treatment of the Danish data, which has in our opinion potential to be misleading. The analysis employed by Sorensen and colleagues [1] fits the data to a smooth sinusoidal curve with a single maximum, and enforces a periodicity of one year. This curve is then tested by means of a chi- square test for goodness of fit. We believe that the results of this analysis are misleading for two reasons. Firstly no account was taken of the error in the measurement of the population rate between years, and hence no standard error for the means was presented. (These errors may of course themselves be misleading since not all annual birth cohorts included have experienced equal numbers of years at risk, and hence in years when the risk is low this may be artefactual.) Secondly, no attempt was made to assess how well the data were described by a constant (the average across the months): i.e. no risk variation with month of birth. Hence no test was presented to assess whether the use of the periodic model significantly improves the fit with data. The hypothesis upon which this and the previous studies are based, is that seasonal variation in the incidence of a causative (infectious) agent around birth may be manifested at a latter date as Crohn's disease. The finding of an association in such studies would require to some extent consistent seasonal variation in any aetiological agent(s) both between years, and between geographical locations. Additionally exposure would need to have an effect only if it occurred within a narrow age band. The hypothesis described is clearly not the only possible explanation of any seasonality found, since month of birth may also determine the timing and nature of environmental changes around the time of entry to schooling. (This is highlighted by the growing literature to demonstrate other major effects of month of birth mediated by social and psychological factors [4].) In summary, we believe that the evidence presented does not show convincing seasonality in births. Furthermore we would caution that the psychosocial effects of birth month mean that it should not be seen solely as a marker of perinatal exposure(s), but may also be a determinant of the environment during later periods of life. For these reasons it seems to us that research effort in this area is likely to be unrewarding. 1. Sorensen, H. T., L. Pedersen, et al. (2001). "Does month of birth affect risk of Crohn's disease in childhood and adolescence?" BMJ 323(7318): 907. 2. Ekbom A, Zack M, Adami HO, Helmick C. Is there clustering of inflammatory bowel disease at birth? Am J Epidemiol 1991;134(8):876-86. 3. Haslam N, Mayberry JF, Hawthorne AB, Newcombe RG, Holmes GK, Probert CS. Measles, month of birth, and Crohn's disease. Gut 2000;47(6):801-3. 4. Menet F, Eakin J, Stuart M, Rafferty H. Month of birth and effect on literacy, behaviour and referral to psychological services. Educational Psychology in Practice 2000;16(2):225-234. |
|||
|
|
|||
|
R Moineddin, Assistant Professor Family and Community Medicine Research Unit, Suite 1100, Bay St., Toronto Ontario, Canada M5G 1N8, Ross E.G. Upshur
Send response to journal:
|
A recently published article in the British Medical Journal by Sorensen et al (1) concluded that there was a seasonal variation in month of birth of the people in Denmark with Crohn’s disease. It is evident from the reported graph that there are two clusters of rates, the first one includes the rates for months December, January, February, March, April (roughly cold months) and the second clusters includes the rest of the rates (warm months). We simply read the rates of Crohn’s disease in Denmark from the reported graph by Sorensen et al. (1). Month Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Rate 25.8 24.0 18.8 22.7 28.7 26.2 26.0 27.0 30.1 29.2 26.9 20.8 We used a simple t-test to compare the average of the rates of the two clusters namely, cold months (mean=22.42, std=2.7) and warm months (27.7, std=1.6). The test was significant at the level of 0.01. The sinusoidal shape of the reported graph by Sorensen et al (1) seems to be an artifact of the rates. By closely investigating the graphs the wavy shape of the graph is due to low rates in the beginning and the end and high rates in the middle. If the seasonality pattern existed then it is invariant under any rotation of the months. That is it is invariant with respect to selecting any month as the beginning of the year. To investigate this invariance for these rates we selected December to be first month of the year and November to be the last month then we used the SAS procedure NLIN to estimate the parameters and the 95% confidence interval for the regression curve First we set January as the first month and fit the regression model and then we set December as the first month and fit the model. The sinusoidal shape of the January graph clearly disappears which means the data is not seasonal. The level of significance for the regression equations are 0.078 and 0.0814 respectively. It is obvious that the curvature of the January graph is driven by the difference between the averages of the two clusters not by the cyclical structure of the data. Therefore, a sinusoidal regression model is not adequate for these data. This simple regression model Rate = 22.4 if month=Dec, Jan, …, April = 27.7 otherwise has the p-value 0.008. Taylor et al. in a similar study (2) didn’t find any evidence that month of birth is a risk factor for Crohn’s disease in children, however from the Sorenson et al.(1) study it is evident that the month of birth is a risk factor for Crohn’s disease in Denmark. We believe we have shown a potential pitfall in drawing seasonal inferences from bimodal data. This clearly has implications for the conclusions drawn by the authors. 1. Sorensen HT, Pedersen L, Norgard B. Does month of birth affect risk of Crohn’s disease in childhood and adolescence? BMJ 2001;323:907. 2. Taylor LB, Platt MJ, West C. No evidence of month of birth or seasonality as a risk factor for Crohn’s Disease in Children. Electronic Response to: Does month of birth affect risk of Crohn’s disease in childhood and adolescence? in BMJ 2001;323:907. |
|||