Mortality Seasonality In Ischemic Heart Disease and Cerebrovascular Disease

                                      

Ischemic heart disease and cerebrovascular diseases are major causes of mortality in the United States.  Therefore factors associated with these diseases are important to identify.  Previous studies have suggested seasonal influences in mortality data in certain populations (Seretakis et al, Douglas et al).  This study examines mortality data for the US from 1995-1998 for seasonality of cerebrovascular and ischemic heart disease.  Knowing if and what seasonal patterns exist will bring to attention certain risk factors for ischemic heart disease and cerebrovascular diseases.  Additionally, knowing of seasonal trends will promote proper resource allocation of people and money.

          This study examined death certificate data of the entire United States.  The data was obtained from the Centers for Disease Control and Prevention.  All data sets from 1995-1998 were examined.  For the purposes of this study, deaths due to ischemic heart disease consisted of those death certificates noted with ICD9 codes 410.0-414.9.  Deaths due to cerebrovascular disease were noted with ICD9 codes 430.0-439.9. Additionally, acute myocardial infarction (ICD9 code of 410.0) and angina pectoris (413.0-413.9) were examined individually for seasonality.  This study specifically examined seasonality in days of the week and months of the year.  A c2 test for equal proportions was used to determine if the proportion of deaths across days and months were equal.

          The results of the c2 tests and frequency distributions are presented below:

 

Ischemic Heart Disease over Days of Week

Cumulative    Cumulative

day    Frequency     Percent     Frequency      Percent

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

1      267897       14.09        267897        14.09

2      279901       14.73        547798        28.82

3      272124       14.32        819922        43.13

4      269914       14.20       1089836        57.33

5      269276       14.17       1359112        71.50

6      271105       14.26       1630217        85.76

7      270620       14.24       1900837       100.00

 

Chi-Square Test

for Equal Proportions

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

Chi-Square   339.9897

DF                  6

Pr > ChiSq     <.0001

 

Sample Size = 1900837

The c2 test is significant, indicating that the days of the week do not have the same incidence of deaths due to IHD.  Monday has the greatest incidence of IHD mortality and Sunday the least.  To view pair-wise comparisons click here.

 

Ischemic Heart Disease over Months of Year

Cumulative    Cumulative

month    Frequency     Percent     Frequency      Percent

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

1      187064        9.84        187064         9.84

2      161912        8.52        348976        18.36

3      170643        8.98        519619        27.34

4      158146        8.32        677765        35.66

5      156404        8.23        834169        43.88

6      145762        7.67        979931        51.55

7      146837        7.72       1126768        59.28

8      143338        7.54       1270106        66.82

9      141369        7.44       1411475        74.26

10      154055        8.10       1565530        82.36

11      157418        8.28       1722948        90.64

12      177889        9.36       1900837       100.00

 

Chi-Square Test

for Equal Proportions

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

Chi-Square  13875.3852

DF                  11

Pr > ChiSq      <.0001

 

Sample Size = 1900837

The c2 test is significant, indicating that the months of the year do not have the same incidence of deaths due to IHD.  January has the greatest incidence of IHD mortality and September the least.  To view pair-wise comparisons click here.

 

Acute Mycardial Infarction over Days of Week

Cumulative    Cumulative

day    Frequency     Percent     Frequency      Percent

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

1      119650       14.05        119650        14.05

2      126104       14.81        245754        28.87

3      122077       14.34        367831        43.21

4      121339       14.25        489170        57.46

5      120824       14.19        609994        71.65

6      120879       14.20        730873        85.85

7      120435       14.15        851308       100.00

 

Chi-Square Test

for Equal Proportions

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

Chi-Square   220.8745

DF                  6

Pr > ChiSq     <.0001

 

Sample Size = 851308

The c2 test is significant, indicating that the days of the week do not have the same incidence of deaths due to MI.  Monday has the greatest incidence of MI mortality and Sunday the least.  To view pair-wise comparisons click here.

 

Acute Mycardial Infarction over Months of Year

Cumulative    Cumulative

month    Frequency     Percent     Frequency      Percent

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

1       83489        9.81         83489         9.81

2       72161        8.48        155650        18.28

3       75822        8.91        231472        27.19

4       70923        8.33        302395        35.52

5       70712        8.31        373107        43.83

6       64844        7.62        437951        51.44

7       65738        7.72        503689        59.17

8       63988        7.52        567677        66.68

9       63317        7.44        630994        74.12

10       69347        8.15        700341        82.27

11       71012        8.34        771353        90.61

12       79955        9.39        851308       100.00

 

Chi-Square Test

for Equal Proportions

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

Chi-Square  6164.5828

DF                 11

Pr > ChiSq     <.0001

 

Sample Size = 851308

The c2 test is significant, indicating that the months of the year do not have the same incidence of deaths due to MI.  January has the greatest incidence of MI mortality and September the least.  To view pair-wise comparisons click here.

 

Angina Pectoris over Days of Week

Cumulative    Cumulative

day    Frequency     Percent     Frequency      Percent

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

1         448       13.48           448        13.48

2         499       15.02           947        28.50

3         475       14.29          1422        42.79

4         485       14.60          1907        57.39

5         443       13.33          2350        70.72

6         515       15.50          2865        86.22

7         458       13.78          3323       100.00

 

Chi-Square Test

for Equal Proportions

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

Chi-Square     9.0948

DF                  6

Pr > ChiSq     0.1683

 

Sample Size = 3323

The c2 test is insignificant, indicating that the days of the week have the same incidence of deaths due to angina.  However, note the small sample size and hence low power of this subgroup. 

 

Angina Pectoris over Months of Year

   Cumulative    Cumulative

month    Frequency     Percent     Frequency      Percent

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

1         309        9.30           309         9.30

2         271        8.16           580        17.45

3         291        8.76           871        26.21

4         266        8.00          1137        34.22

5         283        8.52          1420        42.73

6         272        8.19          1692        50.92

7         253        7.61          1945        58.53

8         274        8.25          2219        66.78

9         269        8.10          2488        74.87

10         266        8.00          2754        82.88

11         273        8.22          3027        91.09

12         296        8.91          3323       100.00

 

Chi-Square Test

for Equal Proportions

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

Chi-Square     9.3346

DF                 11

Pr > ChiSq     0.5910

 

Sample Size = 3323

The c2 test is insignificant, indicating that the months of the year have the same incidence of deaths due to angina.  However, note the small sample size and hence low power of this subgroup. 

 

Cerbrovascular Disease over Days of Week

 Cumulative    Cumulative

day    Frequency     Percent     Frequency      Percent

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

1       89935       13.98         89935        13.98

2       92492       14.38        182427        28.35

3       93112       14.47        275539        42.82

4       91556       14.23        367095        57.05

5       92420       14.36        459515        71.42

6       92154       14.32        551669        85.74

7       91751       14.26        643420       100.00

 

Chi-Square Test

for Equal Proportions

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

Chi-Square    66.9519

DF                  6

Pr > ChiSq     <.0001

 

Sample Size = 643420

The c2 test is significant, indicating that the days of the week do not have the same incidence of deaths due to stroke.  Tuesday has the greatest incidence of stroke mortality and Sunday the least.  To view pair-wise comparisons click here.

 

Cerbrovascular Disease over Months of Year

   Cumulative    Cumulative

month    Frequency     Percent     Frequency      Percent

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

1       62733        9.75         62733         9.75

2       54986        8.55        117719        18.30

3       57592        8.95        175311        27.25

4       53166        8.26        228477        35.51

5       52260        8.12        280737        43.63

6       48598        7.55        329335        51.19

7       48979        7.61        378314        58.80

8       49092        7.63        427406        66.43

9       48193        7.49        475599        73.92

10       53397        8.30        528996        82.22

11       54263        8.43        583259        90.65

12       60161        9.35        643420       100.00

 

Chi-Square Test

for Equal Proportions

ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ

Chi-Square  4526.5773

DF                 11

Pr > ChiSq     <.0001

 

Sample Size = 643420

The c2 test is significant, indicating that the months of the year do not have the same incidence of deaths due to stroke.  January has the greatest incidence of MI mortality and September the least.  To view pair-wise comparisons click here.

 

 

The results tend to support the hypothesis that adjusting to the stresses of the work week and the colder temperatures of winter increase the incidence of mortality among cerebrovascular and ischemic heart diseases. 

Take this evidence of seasonality with a grain of salt, note that deaths due to all causes also had a significant equal proportions chi-square test.  So, deaths are not uniform across days of week and months of the year.  More deaths occur on Saturdays and in January.  Perhaps the grim reaper doesn't work as hard every day...

 

References:

 

Seretakis D. Lagiou P. Lipworth L. Signorello LB. Rothman KJ. Trichopoulos D., Changing seasonality of mortality from coronary heart disease, JAMA. 278(12):1012-4, 1997 Sep 24.

 

Dougla AS. Dunnigan MG. Allan TM. Rawles JM., Seasonal variation in coronary heart disease in Scotland, Journal of Epidemiology and Community Health 49(6):575-82, 1995 Dec.

 

Analysis of Dr. Mcgees framingham dataset.