In large datasets, strange anomalies can appear that may tell us something new about a condition, such as the curious association between melanoma and Parkinson’s disease.
These anomalies can also appear in small datasets, such as the idea that spring babies are more at risk of developing Parkinson’s disease. But the smaller dataset results may be a bit misleading.
In today’s post, we will look at what evidence there is supporting the idea that people born in the spring are more vulnerable to Parkinson’s disease.
Spring lambs. Source: Wenatcheemumblog
When is your birthday?
More specifically, which month were you born in? Please feel free to leave your answer in the comments section below this post.
Why do I ask?
In 1987, an interesting research report was published in a scientific research book:
Title: Season of birth in parkinsonism.
Authors: Miura, T., Shimura, M., and Kimura, T.
Book: Miura T. (ed) Seasonality of birth:Progress in biometeorology, 1987.p157-162. Hague, Netherlands.
In the report, the researchers outlined a study that they had conducted on the inhabitants of an asylum for the aged in Tokyo (Japan). They had found not only a very high rate of Parkinsonism (6.5% of the inhabitants), but also that the majority of those individuals affected by the Parkinsonism were born in the first half of the year (regardless of which year they were actually born).
Sounds crazy right? (excuse the pun)
And that was probably what everyone who read the report thought….
…except that one year later this independent group in the UK published a very similar result:
Title: Could Parkinson’s disease follow intra-uterine influenza?: a speculative hypothesis
Authors: Mattock C, Marmot M, Stern G.
Journal: J Neurol Neurosurg Psychiatry. 1988 Jun;51(6):753-6.
PMID: 3404182 (This article is OPEN ACCESS if you would like to read it)
In essence this study, looked at the birth records of 901 people with Parkinson’s disease and 2309 age-matched controls.
In the last paragraph of the results section, the authors wrote:
“When the pooled Parkinsonian data were analysed by month of birth there were excess Parkinsonian births in May (p < 0.05) and a paucity of births in July (p < 0.05). When the calendar year was divided into three periods pooled Parkinsonian births were increased from March-June, reduced for July-October, and did not differ from matched control births for November-February“
They also noted that individuals born within the years 1892, 1904, 1909, 1918, 1919 and 1929 appear to have had a higher risk of developing Parkinson’s disease in later life. For those of you who are medical history buffs, you will note that these years are close to those of the influenza pandemics of the period 1890-1930. The investigators noticed this as well. In fact, they found that 1.7 to 2.5x more people with Parkinson’s disease were born in 1918, 1919 and 1929 – dates which correspond to the years of influenza pandemics in the UK.
And they took this idea one step further: the seasonal peak in influenza activity is usually between January and March (in the northern hemisphere) they suggested which would be around 6 months gestation. In other words, they were proposing that mum having the flu, while baby was in the oven, increased the risk of Parkinson’s disease in the unborn child later in life.
Completely bonkers right?
Well guess what?
One year later (I kid you not) – and reported by a third independent group – you guessed it: here comes another research report indicating a seasonal pattern:
Title: Does idiopathic parkinsonism in Aberdeen follow intrauterine influenza?
Authors: Ebmeier KP, Mutch WJ, Calder SA, Crawford JR, Stewart L, Besson JO.
Journal: J Neurol Neurosurg Psychiatry. 1989 Jul;52(7):911-3.
PMID: 2769287 (This article is OPEN ACCESS if you would like to read it)
In this study, the investigators looked at the records of 243 people with Parkinson’s disease, and they compared them with controls selected from the same general practices around Aberdeen (Scotland). Now while they did not find a significant seasonal effect, there was a trend towards higher proportion of births between January and March (30% compared to 25% of controls).
In addition, they found an excess of people with Parkinson’s disease born in 1902 compared to surrounding years, but none in 1918, 1919, or 1929.
That third report was not so convincing. Is this seasonal effect really a thing?
You are right, and no it’s not (for Parkinson’s disease at least).
This is where the tide started to turn on the “seasonal birth” story for Parkinson’s disease. And after a brief interlude, a series of research reports appeared that questioned the previous findings, starting with this one:
Title: Is there seasonal variation in risk of Parkinson’s disease?
Authors: Postuma RB, Wolfson C, Rajput A, Stoessl AJ, Martin WR, Suchowersky O, Chouinard S, Panisset M, Jog MS, Grimes DA, Marras C, Lang AE.
Journal: Mov Disord. 2007 Jun 15;22(8):1097-101.
Using the clinical records of 10 movement disorder clinics across Canada,the researchers examined the birth dates of 8,168 people with Parkinson’s disease for any evidence of seasonal variation or clustering of birth dates. They also sought to determine any associations with the major world influenza pandemics of 1918 to 1919 and 1957 to 1958.
They found no evidence of seasonal variation in Parkinson’s incidence, nor did they find any evidence of clustering of birth dates during influenza pandemic years in people with Parkinson’s disease.
Maybe it’s just Canadians?
Well, a large analysis of English health records found the same thing:
Title: Seasonal distribution of psychiatric births in England.
Authors: Disanto G, Morahan JM, Lacey MV, DeLuca GC, Giovannoni G, Ebers GC, Ramagopalan SV.
Journal: PLoS One. 2012;7(4):e34866.
PMID: 22496872 (This article is OPEN ACCESS of you would like to read it)
In this study, birth month data was collected for 57,971 patients from the English Hospital Episode Statistics. In total, there were 3,545 people with Parkinson’s disease. Absolutely no difference was observed in the distribution of birthdays between people suffering from Parkinson’s disease and the general population (p=0.48).
Ok, but a lot of Canadians immigrated from England, so perhaps there is a Anglo-Canuck effect occurring in the statistics?
Well, if there is, you will need to include the Italians and Sweds in that group because:
Title: Season of birth and Parkinson’s disease: possible relationship?
Authors: Palladino R, Moccia M, De Pascale T, Montella E, Torre I, Vitale C, Barone P, Triassi M.
Journal: Neurol Sci. 2015 Aug;36(8):1457-62.
These researchers looked at people with Parkinson’s disease in Naples, Italy. They collected the birth dates of 898 people with Parkinson’s were matched them with 1796 controls, and they found no “association between seasonal birth variations” and risk of developing Parkinson’s disease.
And one year later another (much larger!) report found the same result in Sweden:
Title: Early-Life Factors and Risk of Parkinson’s Disease: A Register-Based Cohort Study.
Authors: Liu B, Chen H, Fang F, Tillander A, Wirdefeldt K.
Journal: PLoS One. 2016 Apr 15;11(4):e0152841.
PMID: 27082111 (This article is OPEN ACCESS if you would like to read it)
The data used in this research report was collected from:
- The Swedish Multi-Generation Register (MGR) – which holds information about the biological and adoptive parents for all residents born in 1932 or later, and were alive or lived in Sweden in 1961. This database covers over 95% of Swedish-born residents, plus more than 22% of foreign-born residents in Sweden.
- The Swedish Patient Register – established in 1964/1965, this databases collects inpatient discharge records. It became nationwide in 1987, and since 2001, the Patient Register has recorded information on every inpatient visit and vast majority of the outpatient visits for all Swedish residents.
- They also linked their data to the Migration Register and Swedish Population and Housing Censuses from 1960, 1970, 1980, and 1990 for information on socio-economic status.
The investigators next set up a selection criteria. Within their pool of people for analysis, individuals had to:
- be born in Sweden between 1932 and 1970
- have information available regarding maternal links in the MGR
- be alive and free of Parkinson’s disease on January 1, 2002,
- 40 years or older on January 1, 2002 or turned 40 years during the study period
A total of 3.5 million people fulfilled this criteria. 8779 cases of Parkinson’s disease were identified within that population of people (a further 2658 people were identified as having Parkinson’s disease, but since they were diagnosed before 2002, they were excluded).
Monthly data on national burden of influenza-like illness during 1932–1970 were obtained from the Swedish Public Health Agency.
As expected, older age, male sex, parental occupation as farmers, and family history of PD were associated with higher PD risk. But the year, season and month of birth were not associated with the risk of developing Parkinson’s disease.
Curiously, they did find that reduction in risk of Parkinson’s among participants who have an older siblings compared to those who do not – I wrote about this is a previous post.
So how do you explain the Japanese asylum patients with parkinson’s disease all being born in the spring?
So researchers have suggested that many of the inhabitants of the asylum would have had Parkinsonisms as opposed to Parkinson’s disease. Parkinsonisms are the features of the condition without the underlying disease itself. These Parkinson’s-like features could have been brought on by long term neuroleptic drug use (eg. Anti-psychotic drugs) – this is a well documented side effect of these drugs (Click here for more on this).
And while there may not be a seasonal birth trend for Parkinson’s disease, there is for people with schizophrenia (December-March) and major depression (March-May).
More than 250 studies have been published on birth seasonality and the risk of individuals developing schizophrenia (or bipolar disorder). They cover 29 Northern and 5 Southern Hemisphere countries. And they are remarkably consistent:
There is a 5-10% winter-spring excess of births for both schizophrenia and mania/bipolar disorder.
For an OPEN ACCESS example of this research, please Click here.
Thus, it may be that people who have been committed to asylums and have had long term treatment with anti-pyschotic drugs, will exhibit Parkinsonisms. And if the reason they had been committed to the asylum was due to psychiatric problems, there would probably appear to be a trend towards many of the Parkinsonian inhabitants in the asylum being born in the first half of the year.
And what about the first English and Scottish studies?
In both of these studies the number of subjects in the studies were rather small (just 243 people with Parkinson’s disease in the Scottish study). And this results in big conclusions being made based on small samples. Such as in the English study, the apparent cluster of births in the years 1918, 1919, and 1929 was based upon only 21, 25, and 10 people with Parkinson’s born in those years. These are not large number to base big ideas on.
So what does it all mean?
Summing up: The point of this post was to dispel a myth that continues to hang around even after it has been largely disproved. There is no increased risk of developing Parkinson’s disease, based on which month, season or year one is born in.
The banner for today’s post was sourced from Mummypages