From the Comprehensive Rural Health Services Project, Ballabgarh, All India Institute of Medical Sciences, New Delhi 110 029, India.
Reprint requests: Dr. K. Anand, Assistant Professor, Center for Community Medicine, All India Institute of Medical Sciences, New Delhi 110 029, India.

Manuscript received: May 24, 1999;
Initial review completed: July 26, 1999;
Revision accepted: August 27, 1999

Seasonality of health events, especially infections is well known. The seasons of high occurrence is related to the mode of spread of the disease concerned. Gastro-intestinal diseases, for example, are more common in summer and rainy seasons where as respiratory infections are common in winter season. Human births have also been reported to follow a seasonal pattern. Cowgill, in 1966 reported a seasonal pattern in birth of newborns in humans(1). The reason(s) for this seasonality of births is not known and many hypotheses have been postulated(2-8). Briefly, these can be classified as seasonality of marriages, changes in frequency of sexual intercourse due to differences in availability of leisure time, changes in weather conditions and seasonality of biological chances of conception. This study was done to document the seasonality of births in a rural area of Haryana in north India, and to assess possible reason(s) for the observed seasonality.

Subjects and Methods

The data for this study has been collected from the twenty eight villages of Ballabgarh Block in District Fardiabad, Haryana, India. These villages fall under the Comprehensive Rural Health Services Project of All India Institute of Medical Sciences, New Delhi. Health care to all the residents of these villages is provided by two Primary Health Centers (PHCs). In 1972, the estimated population of these villages was approximately 40,000 with a birth rate of 45 per thousand, while in 1997, the population was 70,079 and the birth rate 28.4 per thousand.

The health care in these villages is provided by male and female multipurpose workers (M and F MPWs), a pattern similar to the rest of country. The data for marriages, births and deaths are collected by male and female MPWS during their domiciliary visits. The antenatal registration in the study area was around 95% and immunization coverage more than 90% for the last ten years. Initially these data were entered into a family demographic register which had separate pages for each family. This data of the community was computerized in 1987. At present all births and deaths are entered into the database on a monthly basis. Children, whose exact date of birth was not known or children who immigrated and had no written record for the exact date of birth were excluded from the present analysis.

In addition to routine continuous collection of demographic information, yearly census is also conducted in the months of May and June. This information is cross-checked by the health assistant and the medical officer of the two PHCs for completeness and accuracy. To adjust for different number of days per month in a year, the monthly data was standardized by considering a month of 30.44 days.

Results

The total number of births during these 19 years (1972-1990) was 35,720. Of these, 476 (1.3%) births were excluded from analysis as their exact date of birth was not known. Remaining 35,244 births were therefore included in the present anlaysis. The maximum number of deliveries occurred in the month of August and September. The least number of deliveries occurred in the month of April (Fig. 1). Extent of variation around the mean was 34.5%.

It was noticed that the marriages peaked in May when almost two fifth of all marriages took place. The relationship between "lagged" marriage (marriage plus nine months) and child birth was analyzed for first born. The peak of the two curves did not coincide. The median interval between marriage and first birth showed a decreasing trend for marriages taking place in the quartiles of February-April (1.80 years) to November-January (1.62 years). The difference was however not statistically significant (p = 0.063).

The seasonality was also analyzed in blocks of four year periods, except for the period 1976-78 when following forced vasectomies during 1976-78, there was a drop in the number of births (Table I). The seasonal trend in birth remained unchanged. Even during the period 1976-78 when the number of births were much less, the overall pattern of seasonality remained same.

The seasonality curves of upper and lower castes (proxy for socio-economic status) were similar in their months of peak and trough. However, the peak was slightly higher for the lower castes. The effect of use of spacing method of contraception by a couple on the seasonality of births was assessed by comparing the curves for ever users of spacing method and never users. The two seasonality curves were similar.

[ Fig. 1. Distribution of births of by month in the study area (1972-90) ]

[ Fig. 2. Proposed bio-social model for seasonality of births in humans ]

Discussion

A definite seasonality in the births was observed in the study area with peak occurring during the months of August and September and the trough during March and April. Seasonality of births have been reported from different parts of the world including, Nigeria(2), Australia(3), Canada(4), United States(5), Japan(6), Bangla-desh(7) and India(8). The amplitude of variation (35%) was higher in our study compared to that reported in literature. Lower values of 7% for Nigeria(2), 5% Australia(3) and around 10-15% for Japan(6) have been reported. A study from Bangladesh, which is similar to India in its geo-cultural characteristic also reported a high amplitude of variation of 40%(7).

Matsuda et al.(6) reported a difference in seasonality of the first born and subsequent born in Japan and attributed it to seasonality of marriages. The comparison of seasonality of marriages with first borns is valid only if the practice of contraception does not exist among the newly wed couples. In our area, the use of contraceptive method was less than 5% among newly wed couples. Our study did not include abortions and still births, estimated at 17% of all births(9). This might influence the observed seasonality. Another factor which could influence this relationship is the practice of Gauna (delayed cohabitation in the event of an early marriage). This practice was near universal in the early seventies and around 25% in the late eighties.

The peak in births in the months of August and September implied that there is increased conception during December and January. It has been postulated that high summer temperatures reduce conception directly (affecting ovulation or spermatogenesis or fetal loss) or indirectly (reduced sexual intercourse due to physical discomfort).(2,3,5,7,8). Most of the houses, in the study area consisted of one or two rooms. The average number of family members ranged between four and five. There is a lack of privacy in these houses and winter season may actually reduce the opportunity for sexual activity since all family members may sleep inside.

Availability of leisure time due to slackness in agricultural work may result in increased sexual activity. During the peak agricultural work especially during the harvest (April to May), most of the villagers work till late night and often sleep in the field resulting in decreased sexual activity. Over the years there has been a change in the agricultural practice in the study area. Though, wheat continues to be the major crop, cultivation of other cash crops like sunflower, maize, sugar cane is gradually increasing. However, there has been no change in the seasonality of births during these years.

In United States, a seasonality similar to this study (peak in August and September) was explained by festival season of Christmas and New Year(1). In Malaysia, the seasonality of births in sixties varied among the indigenous Malayan (peak in October) and the Chinese (peak in January) immigrants(10). Authors attributed this difference to religious obser-vances/holidays. The amplitude of the variation was also lesser among the Malayans probably because they had changed their traditional practices(10).

The availability of contraception may influence the seasonality of births as couples can now choose the time of conception. The couple protection rate (CPR) in the study area increased from less than five per cent in 1972 to around 35% in 1990. A major proportion to CPR is contributed by females undergoing tubal ligation after completing their families.

Our study area is a predominantly agri-cultural community with even those working in urban areas returning to the villages during harvesting and sowing seasons. The staple diet in these areas is wheat which is harvested in April and May. The seasonality in the availability of food may result in seasonality of malnutrition in the population(11-13). There may, however, be a latent period of some months between the change in nutritional status and its influence on reproductive biology. Many studies have reported a shift in seasonality over the years. This has been reported from Japan(6), United States(1), Malaysia(10), and Canada(4). The important reasons suggested for this are increase in use of birth control methods in Canada(4), industrialization and americani-zation of way of life in Japan(6), changes in way of life and decreased religious practice among Malayans(10).

Based on the review of literature, our study findings and our experience of working in the community, we propose a model which tries to explain the seasonality of births in different areas of the world (Fig. 2). We believe that there is a baseline seasonality of births. This baseline seasonality is subjected to push and pull from two groups of factors: potentiating (push) factors and inhibiting (pull) factors. The potentiating factors push the curve away from baseline. Thus, they could be the factors which accentuate the peak or which depress the trough further. This includes holidays/festivals, condu-cive weather, agricultural cycle, seasonality of marriage, use of contraception. The inhibiting factors are the factors which mask the biological seasonality of conception. They, thus, pull the curve towards the baseline. These include air conditioning of homes, contraception, religious proscription, food availability; all features of socio-economic development. The overall seasonality depends upon the combined influ-ence of potentiating factors and inhibiting factors on the baseline seasonality. In less developed communities, the potentiating (push) factors may be more important and as the development progresses the inhibiting (pull) factors may assume greater importance. A very high seasonality seen in this study could be because of the presence of potentiating factors like temperature differential, agricultural leisure time. But, more importantly there is an almost complete lack of any inhibiting factors.

We suggest further studies to address issues related to seasonality of births. Firstly, studies should be done to find out factors causing baseline seasonality and quantify it. Our hunch is that it will be around 15%. Future studies should look into the questions of seasonality of ovulation, survival of ovum, sperm, zygote, etc. Secondly, there is a need to test the validity of the proposed model by comparing studies from different geographical areas of the world on the factors listed above. It may be possible to have a mathematical model based on the bio-social model suggested above. We hope all subsequent studies will provide data on the factors listed in our model so that a comprehensive evaluation of the proposed model can be carried out.

Contributors: SK conceived the idea of the study and took part in supervision of data collection, and helped in analysis and drafting. He is the guarantor of the paper. GK was responsible for handling the data from collection, computer entry and analysis. KA supervised data collection, conceived the pattern of analysis and wrote the draft. SK was involved in data analysis and revising the draft.

Funding: None.
Competing interests: None.

1. Cowgill UM. Season of birth in Man. Ecology 1966; 47: 614-623.

2. Ayani O. Seasonal variation of birth in rural Southwestern Nigeria. Int J Epidemiology 1986; 15: 91-96.

3. Mathes CD, Harris RS. Seasonal distribution of births in Australia. Int J Epidemiol 1983; 12: 326-331.

4. Halil SS. The seasonality of births in Canada. J Biosoc Sci 1989; 21: 321-327.

5. Erhardat CL, Nelson FG, Pakter J. Seasonal patterns of conception in New York City. Am J Public Health 1971; 61: 2246-2258.

6. Matsuda S, Kahyo H. Geographical differences and time trends in the seasonality of birth in Japan. Int J Epidemiology 1994, 23: 107-118.

7. Stoeckel J, Choudhury AKM. Seasonal variation in births in rural east Pakistan. J Biosoc Sci 1972; 4: 107-116.

8. Bernard RP, Bhatt RV, Potts DM, Rav AP. Seasonality of births in India. J Biosoc Sci 1972; 4: 107-116.

9. Murthy GVS, Reddaiah VP, Kapoor SK, Kailash S. Dynamics of induced abortions in rural community of Haryana. Hlth Popn Persp Iss 1987; 10: 93-104.

10. Holland B. Seasonality of births: Stability and change in a developing country. Hum Biol 1989; 61: 591-598.

11. Anand K., Sundaram KR, Lobo J, Kapoor SK. Are diarrheal incidence and malnutrition related in under five children - A longitudinal study in an area of poor sanitary conditions in India. Indian Pediatr 1994; 31: 943-948.

12. George SM, Latham LC, Abel R, Ethirajan N, Frongillo EA. Evaluation of effectiveness of good growth monitoring in south Indian villages. Lancet 1993; 342: 348-351.

13. Mosher SW. Birth seasonality among peasant cultivators: The inter realtionship of workload, diet and fertility. Hum Ecology 1979; 7: 151-181.