李津1,罗国春2,武革3,林少达4,袁琳5,李路6,许美伦7,曾龙驿1,李绍清8,陈定宇9,吴建能10,方懿珊11,欧文新12,林健才13,张秀薇14,梁干雄15,严励16,吴永新17,柯涓18,郎江明19,蔡德鸿20,陈妍21,钟可荣22,杨华章23,谢乃强24,蔡悦25,谭国据26,翁建平1(通信作者)
1中山大学附属第三医院;2深圳大学第一附属医院;3广东医学院附属医院;4汕头大学医学院第一附属医院;5珠海市人民医院;6深圳市人民医院;7惠州市中心人民医院;8清远市人民医院;9广州市第一人民医院;10粤北人民医院;11揭阳市人民医院;12茂名市人民医院;13江门市中心医院;14东莞市人民医院;15中山市人民医院;16中山大学附属第二医院;17梅州市人民医院;18潮州市中心医院;19佛山市中医院;20珠江医院;21云浮市人民医院;22紫金县人民医院;23广东省人民医院;24肇庆市人民医院;25汕尾市人民医院;26阳江市人民医院
背景和目的:SEARCH研究(美国的一项针对1型糖尿病患儿的大型研究)中发现,5月份出生的儿童发生1型糖尿病的机率最大,而1月份出生儿童的机率最小。其它在高加索人群中进行的研究也得到了类似的结果。然而在中国,很少有研究关注这一领域。所以本研究的目的即调查中国广东省出生1型糖尿病发病是否与出生季节有关。
材料与方法:我们通过覆盖全省的糖尿病注册登记系统,在广东省全部21个地级市的105家医院开展了这项研究。我们收集了自2000年1月至2011年3月在这些医院住院的1型糖尿病患者的出生月份信息;然后采用核平滑法计算每月出生的1型糖尿病患者的病例数期望值。
结果:总共纳入了2,227例1型糖尿病患者(其中1,062例为女性)。在这些患者中,在3、4和5月份出生的病例数较多(分别为191例、210例和198例),而在12月、1月、2月和4月份出生的病例数较少(分别为165例、139例、167例和156例)。经核平滑法处理数据后,我们发现在在6月至10月出生的1型糖尿病患者较多,而在12月至4月出生的患者较少。其中以8月份出生患者的风险最高,1月份出生患者的风险最低(RR=1.26,P<0.001)。按照性别分层后并未发现类似的结果(P均>0.05)。
结论:8月份出生者日后发生1型糖尿病的风险最高;其原因我们推测可能在于:出生前4-5月母体维生素D水平对胎儿胰岛β细胞发育非常关键;8月份出生的患儿在出生前4-5月正好处于3月至4月份,而这段时间正好又是广东地区阳光照射最少的时期,从而导致这段时期母体维生素D水平较低。
In the SEARCH for diabetes in youth study, type 1 diabetes had been found to be born the most common in May and the least common in January. Other studies in Caucasian came to similar conclusions. Few studies, however, focused on this field among Chinese population. Hence, the aim of our survey was to investigate whether birth seasonality influenced type 1 diabetes incidence in later life among people born in Guangdong Province, China.
We collected birth month data of type 1 diabetic patients (medical records documented from January 2000 to March 2011) in 105 hospitals from all 21 cities in Guangdong Province by a provincial-wide type 1 diabetes registry. Kernel smoother technique was conducted to calculate expected incident number in each month.
A total of 2,227 (1,062 females) type 1 diabetic patients were included. Unsmoothed data showed more cases were born in March (191), June (210), July (198), August (217), September (212) and October (201), as compared with those in December (165), January (139), February (167) and April (156). After data smoothing, we found a deficit in December to April and an excess in June to October, among which the maximum risk was August birth and the minimum risk was January birth (RR = 1.26, P < 0.001). However, similar effect was not found after sex stratification (P > 0.05).
In conclusion, August birth was associated with increased risk of type 1 diabetes incidence in later life, possibly due to decreased sunlight in March and April (4~5 months before birth, which is critical to prenatal development of pancreatic β cells).