|Year : 1998 | Volume
| Issue : 4 | Page : 211-215
Cataract, glaucoma and season of birth amongst patients born on the Indian Subcontinent
Robert A Weale
Moorfields Eye Hospital and Age Concern Institute of Gerontology, King's College London, United Kingdom
Robert A Weale
Age Concern Institute of Gerontology, King's College London, Cornwall House, Waterloo Road, London SE1 8WA
Source of Support: None, Conflict of Interest: None
The relation between the prevalence of open-angle glaucoma and of different types of cataract on the one hand and the patient's season of birth on the other was studied with special reference to natives of the Indian Subcontinent. Partial comparisons were made with results obtained on British-born patients. Both primary open-angle glaucoma (POAG) and the mixed nuclear/posterior-subcapsular type (NP) showed a statistically significant variation with the month of birth for the Asian patients, but only NP showed significant variation for the European patients in a month-by-month analysis. The Asian data are examined from the point of view of seasonal variations in maternal nutrition and of low birth-weight babies.
Keywords: Cataract, open-angle glaucoma, season of birth, risk factors, ethnicity
|How to cite this article:
Weale RA. Cataract, glaucoma and season of birth amongst patients born on the Indian Subcontinent. Indian J Ophthalmol 1998;46:211-5
|C IS CORTIAL CATARACT, N IS NUCLEAR CATARACT, P IS POSTERIOR SUBCAPSULAR CATARACT, S IS NUMBER OF PATIENTS.
Click here to view
Several items of information suggest that a number of so-called age-related conditions can be traced to an individual's early life if not, indeed, to ante-natal influences. This is true of both animals and man. For example, the prevalence of congenital nuclear cataract in cattle varies systematically with the season of birth. Numerous studies have shown in the case of man that seasonal influences can play a role in early development, often resulting from low birth weight. This, in turn, is significant because it is associated with ocular abnormalities such as a tendency to develop keratoconus and myopia, especially in girls. Birth weight is subject to many environmental influences. There are social causes, which may reflect poor physiological status during pregnancy, with accompanying dietary insufficiency. Again, a study of lactation, seasonality and mother's postpartum weight change in Bangladesh revealed marked seasonal patterns with losses in the winter and the rainy season.
|C IS CORTIAL CATARACT, N IS NUCLEAR CATARACT, P IS POSTERIOR SUBCAPSULAR CATARACT, S IS NUMBER OF PATIENTS.
Click here to view
It is known that the development of the eye starts early during pregnancy (when the woman may not even be aware of her state). Also, the shape of the eye makes a spurt during the last few weeks of term. Studies on individuals born in the Netherlands toward the end of World War II, when that country was subjected to a debilitating famine, have demonstrated the importance of an adequate maternal diet during the last three months of term if growth is not to be inhibited. This may be linked to the above-mentioned ocular involvement in low birth-weight individuals.
It may be postulated that the normal physiological anlage of the eye is liable to be inhibited in unfavourable developmental circumstances; there then arises the question of possible long-term effects related to seasonal environmental influences on the foetus. It must be appreciated that these notions are recent, and that prospective studies have yet to be undertaken, ethical considerations permitting.
In some preliminary studies it was found that primary open-angle glaucoma varies with the season of a patient's birth, both in patients born in the British Isles and in those born in the Indian Subcontinent. As indicated above, tentative explanatory hypotheses rest, for example, on seasonal variations in maternal antenatal nutrition, seasonal sensitivity to pathogens, and seasonal variations in the relative number of low birth-weight babies, which may also be related to maternal nutritional levels.
The influence of the season of birth on the occurence of human eye diseases does not seem to have been considered even though a large number of risk factors have been identified. A seasonal variation in the prevalence of a given condition can be considered meaningful only if the precondition of a statistically significant difference from the birth-index (the relative number of births per month) is met.
| Materials and Methods
Patients presenting in outpatient clinics of Moorfields Eye Hospital, London, UK, were included if diagnosed as suffering from age-related lens opacities or open-angle glaucoma. Place and date of birth were verified in every case in order to correct typing and other errors, if any. It is appreciated that in the earlier part of this century, births were not always registered in all parts of the Indian Subcontinent. Moreover, there was a tendency for the authorities to list births as having occurred in January in case of doubt. In the present study it was found that far more patients claimed to have been born on the first of January than was statistically probable: with very few exceptions those making this claim were not counted. No reliance was placed on information based on passports. In general, patients were confident about the month, rather than the day, of birth. It will be recalled that this study is concerned with the relation, if any, of the season of birth and the probability of presenting with glaucoma or cataract of a special type: it is most improbable that any significant uncertainty in the information offered by patients should lead to such a bias in the results as to create a significant effect where there actually is none. More specifically, if, for example, a seasonal effect were to be observed in fewer than all types of cataract, the problem would seem to be controlled.
Only highly experienced clinical specialists were involved in the diagnoses of glaucoma and of cataract type, but not lenticular location. The following 7 classes of cataract were identified: cortical (C), nuclear (N), posterior subcapsular (P), and their four combinations (CN, NP, PC, CNP). Diabetic patients and those who had taken steroids at any time were excluded, as were those having traumatic, radiation or mature cataracts. That is, only age-related cases capable of being classified under the C, N, P, scheme and its 4 combinations, were included. The existence of different etiologies for the different types of cataract did not affect the counting procedure: it was based on patients, not cataracts, also in the case of mixed types. Therefore, someone with a mixed cortical and nuclear cataract would be recorded as a CN type, and not separately in the classes of cortical (C) and nuclear (N) cataracts. First cataracts were recorded; in the case of unilaterally aphakic patients information on the first eye was obtained from available notes.
Ethnicity was defined with reference to the place of birth, name, and appearance, care being taken to avoid misjudgement in cases of inter-ethnic marriages, or first-generation descendants of immigrants.
Finally, only three of the ethnic groups, namely natives of the British Isles (BI), of the Indian Subcontinent (IS), and of the Caribbean Region (CR) were found to provide the necessary number of glaucomas and of cataracts (see under analysis).
There were 119 IS, 447 BI, and 393 CR glaucoma patients. This met the statistical requirements in each of these groups. Patients from Punjab and Gujarat made up almost 50% of the sample population. Previous studies have indicated that, with POAG, both IS and BI showed a statistically significant seasonal variation in the prevalence of glaucoma when the data were amalgamated on a quarterly basis. In the present study we are applying a more stringent criterion, namely a month-by-month analysis. It is easily seen that if the seasonal variation were sinusoidal, the level of significance would depend on the starting month: in principle there is no reason to start with January. This partly fortuitous influence persists, of course, when fewer months are amalgamated into cells, but is eliminated when the analysis is done month by month as in this paper.
As described in the Introduction, the significance of seasonal variations was tested by comparing the observed monthly prevalence with the mean prevalence multiplied by the monthly birth index.
The birth index shows a marked variation on the Indian Subcontinent, possibly due to the influence of the monsoon (nocturnal temperature?) on the frequency of conception, but in the British Isles the index is almost constant. It is realized that there is an east-west variation in the Indian index perhaps because of the annual westward slide of the monsoon across the Subcontinent, but this was considered to be a second-order effect: an average value for the birth index was used. The results for men and women were amalgamated for all parameters.
The criterion of significance was based on the Chi-square test, the product of the mean monthly frequency times the monthly birth index providing the projected figure with which to compare the observed data. For Chi-square to be significant at a 5% level its value must not be less than 19.675 for 12 months; if at all, this might be expected to be achieved with no fewer than 80-90 patients per condition and ethnic group. The test is exacting, and preferable, for example, to multivariate analysis, as it is reasonable to assume that confounding factors will be randomly distributed over the months of the year. Admittedly, all statistical procedures involve some element of arbitrariness.
[Table - 1] shows the number and average age of both cataract and glaucoma patients (by gender) who took part in this study.
Primary open-angle glaucoma
[Figure - 1] shows the monthly variations in the prevalence of POAG for 119 immigrants born on the Indian Subcontinent. Values expected on the basis of demographic data for the monthly birth index are also shown. The differences between the observed and the expected data are significant with p=0.0003.
The results for the other ethnic groups were not significant on a month-for-month basis, and are, therefore, not presented.
[Figure - 2] shows the prevalence of all cataract data for the IS patients, the other overall data not being statistically significant (but see below). The difference between observation and expectation is significant (p=0.00032), and raises the question of whether one type of cataract is specially sensitive to seasonal influences. [Figure - 3] and [Figure - 4] show results for nuclear cataract (p=0.786) and all types of posterior subcapsular combinations (p=0.014) respectively. [Figure - 5]a gives the results for NP (p=0.0167) which, as explained below, accounts for all the observations (also see [Table - 2]).
This positive result for NP prompted a renewed examination of the data obtained for other ethnic groups. It is noteworthy that the only set to show a significant seasonal variation on a month-to-month basis was NP in the British Isles with p=0.002 [Figure - 5]b. Where a significant effect appeared with other cataract types it involved semesters, trimesters or quarters, but never a monthly effect. Semestrial effects may, of course, depend on where the cycle begins. A monthly variation is independent of this and hence noteworthy.
Dates of birth
The figures show that, notwithstanding the precautions reported above, high values appear for birthdays reported in January. The possibility has to be faced that they may be partly artefactual.
[Figure - 1] shows a clear seasonal variation with the observations for IS differing from the expectation (p=0.0003). The excess in the spring (corresponding to conceptions in the summer) is followed 6 months later by an equal deficit in the autumn (corresponding to conceptions in the winter). If the results are valid, they suggest that the season of birth may constitute a risk factor for primary open-angle glaucoma. Demographically speaking, conceptions in the course of the monsoon season appear to highlight the risk, which may reflect, for example, on a "low" in a woman's physiological state at times of heat stress; dietary problems may play a role alternatively or additionally.
[Figure - 2] shows the combined results for all cataract types obtained for patients born on the Indian Subcontinent and compares them with the expected prevalence as based on the birth index (p=0.0003). One must then address the question whether this is globally true or due more to some types of cataract rather than others. For example, the difference between the data for just nuclear cataract and the deviation from expectation [Figure - 3] is non-significant (p=0.7864). The matter is further analysed in [Table - 2]. This gives the cataract types, and shows the number of cases and the level of significance of the difference between observation and expectation for a variety of combinations. For example, line 2 shows results for all cataracts with a cortical component, Line 4 for all cataracts with a nuclear component, and line 7 for all cataracts with posterior subcapsular component. Note that, with the exception of line 8, all the lines satisfy the condition mentioned earlier, namely that the number of cases exceed 80. It may be mentioned parenthetically that the numbers under S in [Table - 1] are subsets of the total of 362 shown in Line 1. We shall return to this point below.
Again, all the combinations involving P show a significant difference between observation and expectation [Figure - 4] and [Table - 2], but it becomes clear that significant probability levels are observed only in the presence of the mixed cataract NP, as is true also for the total (line 1) and N (line 4). The fact that line 8 in [Table - 2] shows a p-value approaching statistical significance is attributable to the fact that some of the cells used in the Chi-square test contained low numbers: in one case there was only 1. The data for the mixed cataract type NP are shown in [Figure:5a] (cf. [Table - 2], line 9).
As regards the results for the British Isles [Figure:5b] there is no monsoon there, and it remains to be seen whether or not the observation is due to chance.
Number of cases
[Table - 2] illustrates the danger involved in counting cataracts rather than patients. The average number of cataracts per patient is about 1.6, and it is easily seen that statistically significant results may be secured from spuriously large numbers. It is important to state clearly which procedure is being followed. Counting patients evidently does not involve the loss of information entailed by counts of cataracts. The reason is as follows. Suppose a patient has a mixed cataract, combining a cortical (C) with a nuclear (N) lesion. If this is counted as 1 patient, classified a CN, the information is more precise than if one puts this as one cortical and one nuclear cataract. On the cataract-counting principle, another patient with a mixed cortical (C) and posterior sub-capsular (P) would be said to have 1 cortical and 1 posterior subcapsular cataract. This small population would then have 2 cortical, 1 nuclear, and 1 posterior sub-capsular cataract between them, and the original state cannot be determined. If, in contrast, they are reported as one CN and one NP all the original information is preserved. If, for some reason, one then wants to count cataracts this remains possible, whereas a reversal of the procedure is not.
More and more environmental effects on intra-uterine development are being discovered, and it has been known for a long time that prematurity/low birth weight can play an adverse role in ocular development. It is therefore not surprising that seasonal variations in the prevalence of ocular conditions should manifest in a population known to be prone to extremely adverse environmental conditions which have a seasonal element. It should not be difficult to test this tentative hypothesis in other populations. What is unexpected is that, with a type of cataract conventionally regarded to be mixed, namely NP, one has to ask whether the assumption is justified. For example, when relative risk ratios for various types of cataract are examined for two populations, statistically significant values may be obtained for cortical and mixed nuclear/posterior subcapsular cataracts. It is found almost invariably that, if the relative risk ratio is significantly greater for the cortical type, the reverse holds for the mixed type. This is noteworthy, because the result for the cortical type is age-related, so that the opposite must hold for the mixed type. [13,14] Perhaps there are two types of this mixed cataract, namely one representing the addition of its components, and another presenting as sui generis, a type of its own.
Barker DJP, editor. Fetal and Infant Origins of Adult Disease.
London: British Medical Journal; 1992.
Ashton N, Barnett KC, Clay CE, Clegg FG. Congenital cataract in cattle. Vet Rec
Fledelius H. Prematurity and the eye. Ophthalmic 10-year follow up of children of low and normal birth weight. Acta Ophthalmol
Kramer MS. Determinants of low birth weight: methodological assessment nd meta-analysis. Bull World Health Organ
Kogan MD. Social causes of low birth weight. J Roy Soc Med
Miller JA, Rodriguez, G, Pebley AR. Lactation, seasonality and mother's weight change in Bangladesh: An analysis of maternal depletion. Am J Hum Biol
Weale R. Is the season of birth a risk factor in glaucoma? Br J Ophthalmol
Golding J, Sutherland R. Month and day of delivery. In: Golding, J. editor. Social and Biological Effects on Perinatal Mortality.
Bristol, UK: University of Bristol/WHO; 1990. p 105-50.
Leske MC, Chylack Jr LT, Suh-Yuh W. The lens opacities case-control study. Arch Ophthalmol
Young RW. Age-related Cataract.
Oxford: Oxford University Press; 1991. p 33-56.
Fullard H, Darby HC. The Library Atlas.
London: George Philip & Son; 1978. p xiv-xv.
Chambers R, Longhurst R, Paley A. Seasonal Dimensions in Rural Poverty.
London: Frances Pinter; 1981.
Weale RA. Inter-ethnic risk ratios for different types of cataract. Ophthalmic
Pierscionek BK, Weale RA. Odds ratios for different types of age-related cataract: ethnicity and environment. Ophthalmic Res
[Figure - 1], [Figure - 2], [Figure - 3], [Figure - 4], [Figure - 5], [Figure - 6]
[Table - 1], [Table - 2]