Correlation Between Alcohol Consumption and Life Expectancy

The study of factors influencing life expectancy is important not only for scientist and researchers, but also for the general public in the bid to live long in this world. For instance, some scientist have tried to come up with demands that eating certain food stuffs either increases or decreases the life span of an individual. Moreover, other factors like the income of people, a country’s GDP and level of education have also been studied regarding the longevity of an individual (Lin, Chen, Chien, & Chan, 2012).. However, this study tries to determine if there is a correlation between the level of alcoholism and the life expectancy of individuals in the world. In this regard, if in deed the two variables to be investigated will have a relationship, then the nature and strength of the relationship will be discussed. It is worth noting, that there has been an increase in general trend of life expectancy from the year 1800 to date.

Data Collection

The data to be analyzed in this study was obtained from gapminder.org website. The 2008 raw data for the two variables was used in developing a sample of 31 countries in Europe ("Gapminder Tools," 2018). Since the countries were selected from one region, we expect the variability to be minimal unlike when we would have considered taking a sample that encompasses countries in the whole world. The distribution of the data points between the levels of alcohol consumption and life expectancy will be shown using scatter plot diagram for visual representation. Also, regression analysis will used to determine the strength of the model and the correlation between the explanatory and the response variable (Seber & Lee, 2012). In this regard, the former represents the level of alcohol consumption and the latter the life expectancy.

Sampling Technique

Since the population of interest is known, the lottery also known as the simple random sampling approach was employed. The countries in Europe were designated specific numbers from 1 to 50 and mixed in a box, after which a random sample of 31 nations was selected. Apparently, the sample size represents 62% of the total population which implies that the results in this study will be significant in making inference to the entire population.

Analysis

Scatter Plot Diagram

Regression Line

The least square method is used to determine the line of best fit. This is because it is more accurate than the other methods. The linear model is represented by,

 

Where, is the y-intercept, and

 is the slope.

S/No.

7.29

76.8

559.872

53.1441

5898.24

-5.0432

-0.8742

25.43387

4.40877

10.17

84.6

860.382

103.4289

7157.16

-2.1632

6.9258

4.67943

-14.98189

13.66

72.3

987.618

186.5956

5227.29

1.3268

-5.3742

1.76040

-7.13049

12.4

80.4

996.96

153.76

6464.16

0.0668

2.7258

0.00446

0.18208

10.41

79.6

828.636

108.3681

6336.16

-1.9232

1.9258

3.69870

-3.70370

9.6

77.5

744

92.16

6006.25

-2.7332

-0.1742

7.47038

0.47612

11.4

73.2

834.48

129.96

5358.24

-0.9332

-4.4742

0.87086

4.17532

15

76.2

1143

225

5806.44

2.6668

-1.4742

7.11182

-3.93140

8.84

80

707.2

78.1456

6400

-3.4932

2.3258

12.20245

-8.12448

12.02

78.9

948.378

144.4804

6225.21

-0.3132

1.2258

0.09809

-0.38392

13.1

79.6

1042.76

171.61

6336.16

0.7668

1.9258

0.58798

1.47670

12.48

81.1

1012.128

155.7504

6577.21

0.1468

3.4258

0.02155

0.50291

12.14

80

971.2

147.3796

6400

-0.1932

2.3258

0.03733

-0.44934

11.01

80.2

883.002

121.2201

6432.04

-1.3232

2.5258

1.75086

-3.34214

16.12

73.9

1191.268

259.8544

5461.21

3.7868

-3.7742

14.33985

-14.29214

7.38

82.4

608.112

54.4644

6789.76

-4.9532

4.7258

24.53419

-23.40783

9.72

81.5

792.18

94.4784

6642.25

-2.6132

3.8258

6.82881

-9.99758

16.3

72.1

1175.23

265.69

5198.41

3.9668

-5.5742

15.73550

-22.11174

8.94

74.5

666.03

79.9236

5550.25

-3.3932

-3.1742

11.51381

10.77070

23.01

70.4

1619.904

529.4601

4956.16

10.6768

-7.2742

113.99406

-77.66518

9.75

80.3

782.925

95.0625

6448.09

-2.5832

2.6258

6.67292

-6.78297

8.35

80.8

674.68

69.7225

6528.64

-3.9832

3.1258

15.86588

-12.45069

14.43

75.4

1088.022

208.2249

5685.16

2.0968

-2.2742

4.39657

-4.76854

13.89

79.4

1102.866

192.9321

6304.36

1.5568

1.7258

2.42363

2.68673

16.15

73.2

1182.18

260.8225

5358.24

3.8168

-4.4742

14.56796

-17.07713

12.21

74.3

907.203

149.0841

5520.49

-0.1232

-3.3742

0.01518

0.41570

14.94

78.7

1175.778

223.2036

6193.69

2.6068

1.0258

6.79541

2.67406

9.5

81.1

770.45

90.25

6577.21

-2.8332

3.4258

8.02702

-9.70598

11.41

82

935.62

130.1881

6724

-0.9232

4.3258

0.85230

-3.99358

17.47

67.8

1184.466

305.2009

4596.84

5.1368

-9.8742

26.38671

-50.72179

13.24

79.7

1055.228

175.2976

6352.09

0.9068

2.0258

0.82229

1.83700

Total

382.33

2407.9

29431.758

5054.8625

187511.41

339.50028

-265.41642

Mean

12.3332

77.67419

But we know that,

On the other hand, the y-intercept is obtained from the value of the slope as follows,

In this regard, he regression line is given by,

From the liner regression line, there exists a negative relationship between the variables in this stud from the negative nature of the slope. The y- intercept value tells us that mean period of people from the sample population can live for a period of 87.3156 years if and only if they are not taking alcohol. However, their life expectancy decreases as soon as one starts to engage in drinking alcohol.

Correlation Coefficient

The correlation coefficient will be used to predict the relationship between the explanatory and the response variables.

In statistical analysis, the correlation value is used to determine if the explanatory and response variables have a relationship. It should be noted that the relationship can either be positive or negative (Zou, Tuncali, & Silverman, 2003). In this case, the correlation has a negative value which implies than an increase in one variable will cause a significant decrease in the other variable under study. However, the value of 0.6577 means that the two variables, that is life expectancy moderately depends on the level of alcohol consumption. In the bid to determine if the regression model better accounts for the variability we proceed to compute the R squared value (Montgomery, Peck, & Vining, 2012).

Coefficient of Determination (R Squared)

 

From the R squared value, it is seen that the model only accounts for 43.26% of the variability in the variables. Also, the value reveals that the data points are not close to the best fi line.

Data Grouping For Alcohol Consumption Level

Row Labels

Alcohol Consumption Level

CF

Sec. Angle

Mid-Point

Upper Class Limit

7.29-9.29

5

5

16%

8.29

9.29

9.29-11.29

7

12

23%

10.29

11.29

11.29-13.29

9

21

29%

12.29

13.29

13.29-15.29

5

26

16%

14.29

15.29

15.29-17.29

3

29

10%

16.29

17.29

17.29-19.29

1

30

3%

18.29

19.29

21.29-23.29

1

31

3%

20.29

21.29

Grand Total

31

23.29

Table 1: Grouped Frequency for alcohol consumption level.

The data in Table 1 was used in the drawing of histograms, frequency polygons, ogive curves, and pie charts to aid in the visual interpretation of data. Indeed, this is fundamental particularly for decision makers who have little or no knowledge in statistical analysis.

Histogram

The data from the sample is not normal since the histogram is skewed to the right. This implies that most of the countries have a mean value less than the mean of the sample in the study.

Frequency Polygon and Ogive Curve

Pie Chart

From the pie chart, a high percentage of the countries had alcohol levels ranging from 11.29 to 13.29 with 23%, followed by 27% for those countries ranging from 9.29 to 11.29 mean alcohol level.

Class

f

CF

x

fx

(x-Mean)^2

Upper Class Limit

7.29-9.29

5

5

8.29

41.45

16.5202

9.29

9.29-11.29

7

12

10.29

72.03

4.2622

11.29

11.29-13.29

9

21

12.29

110.61

0.0042

13.29

13.29-15.29

5

26

14.29

71.45

3.7462

15.29

15.29-17.29

3

29

16.29

48.87

15.4882

17.29

17.29-19.29

1

30

18.29

18.29

35.2302

19.29

21.29-23.29

1

31

20.29

20.29

62.9722

23.29

Grand Total

31

382.99

138.2231

Mean

 

Mode

Median

 

Standard Deviation

Inter-Quartile Range

Therefore, the inter-quartile range is given by,

Data Grouping For Life Expectancy

Row Labels

Count of Life expectancy

Mid-point

C.F

Sector angles

Upper Class Limit

67.8-70.8

2

69.3

2

6%

70.8

70.8-73.8

4

72.3

6

13%

73.8

73.8-76.8

5

75.3

11

16%

76.8

76.8-79.8

8

78.3

19

26%

79.8

79.8-82.8

11

81.3

30

35%

82.8

82.8-85.8

1

84.3

31

3%

85.8

Grand Total

31

Table 2: Grouped data for life expectancy

Histogram

The histogram above is skewed to the left meaning that it has a long left tail as compared to the right tail. Besides, this also implies that most of the values in the sample are on the highest side, the life expectancy in most countries is high.

The frequency polygon depicts the same results as the histogram but it uses lines at the midpoints of the range. A significant number of countries had a life expectancy of 81.3 years in accordance with the sharp peak in the frequency polygon. From the data grouping table, very few countries had a life expectancy ranging from 82.8 to 85.8 as evident in the frequency polygon and histogram above.

It is evident that the mean life expectancy of the countries in Europe significantly increases from 70.85 to 82.85. However, there is a small change from 82.85 to 85.85 as per the small nature of the slope at that point.

It should be noted that pie charts give a good visual representation than frequency polygons and ogive curves because they give the percentage score of each grouped category in relation to the total percentage in the sample. From the pie of life expectancy, most countries in Europe have a mean life expectancy ranging from 79.8 to 82.8 with 36% followed by 76.8 to 79.8 with 26%. On the other hand, only 3% of the sample population had a mean longevity between 82.8 to 85.8 years.

Class

f

x

C.F

fx

(x-Mean)^2

Upper Class Limit

67.8-70.8

2

69.3

2

138.60

70.89

70.8

70.8-73.8

4

72.3

6

289.20

29.37

73.8

73.8-76.8

5

75.3

11

376.50

5.85

76.8

76.8-79.8

8

78.3

19

626.40

0.34

79.8

79.8-82.8

11

81.3

30

894.30

12.82

82.8

82.8-85.8

1

84.3

31

84.30

43.30

85.8

Grand Total

31

2409.30

162.57

Mean

 

Mode

Median

 

Standard Deviation

Inter-Quartile Range

Therefore, the inter-quartile range is given by,

Data Interpretation

The life expectancy of people in a particular country is affected by various factors apart from the level of alcohol drinking. The political factors cannot be ignored since if a country is politically unstable and experiences either international or domestic wars, people in that country are unlikely to live for long. Besides, the literacy level falls under the umbrella of political factors that influence life expectancy. It is worth noting that a life expectancy is high in a country with low illiteracy rate (Lin et al., 2012). This is because life expectancy index does not account for the types of deaths of people. Life expectancy is also impacted by socioeconomic inequality and access to better health care services. This is because improved health care in a country is directly proportional to increase in life expectancy (Chan & Kamala, 2015).

Conclusion

            The analysis above shows that indeed the alcohol consumption level in a country affects the longevity of individuals in that country. This has been confirmed from the correlation analysis even though the two variables have a moderate relationship. The variables in this study do not have a causal relationship because the longevity of people in the world is not caused by the amount of alcohol intake. Moreover, people take different types of drinks which have different effects to the alcohol consumers. It should be noted that the data in this study can be significantly inferred o the total population in Europe since the sample population was composed of the countries from the same continent. In this regard, researchers should advice the relevant bodies in different nation to advice the citizens on the reduction of alcohol drinking in the bid to live long. Also, the governments in the individual nations can provide a rule that limits the alcohol drinking hours to increase the life expectancy of its citizens. To improve this study, a multiple regression analysis should be encouraged since it is composed of more than one explanatory variable. The multiple regression model can then be used to determine which variables have greater effect to the life expectancy.

References

Chan, M. F., & Kamala Devi, M. (2015). Factors Affecting Life Expectancy: Evidence From 1980-2009 Data in Singapore, Malaysia, and Thailand. Asia Pacific Journal of Public Health, 27(2), 136-146.

Gapminder Tools. (2018). Retrieved from https://www.gapminder.org/tools/#_state_time_value=2008;&marker_select@_geo=usa&trailStartTime=2005;;&axis/_x_which=alcohol/_consumption/_per/_adult/_15plus/_litres&domainMin:null&domainMax:null&zoomedMin:null&zoomedMax:null&scaleType=linear;;;&data_/_lastModified:1523640284417&lastModified:1523640284417;&chart-type=bubbles

Lin, R. T., Chen, Y. M., Chien, L. C., & Chan, C. C. (2012). Political and social determinants of life expectancy in less developed countries: a longitudinal study. BMC Public Health, 12(1), 85.

Montgomery, D. C., Peck, E. A., & Vining, G. G. (2012). Introduction to linear regression analysis (Vol. 821). John Wiley & Sons.

Seber, G. A., & Lee, A. J. (2012). Linear regression analysis (Vol. 329). John Wiley & Sons.

Zou, K. H., Tuncali, K., & Silverman, S. G. (2003). Correlation and simple linear regression. Radiology, 227(3), 617-628.

Appendix

Country List

S/NO.

Country

Alcohol Consumption Level  (2008)

Life expectancy (2008)

Albania

7.29

76.8

Andorra

10.17

84.6

Armenia

13.66

72.3

Austria

12.4

80.4

Belgium

10.41

79.6

Bosnia and Herzegovina

9.6

77.5

Bulgaria

11.4

73.2

Croatia

15

76.2

Cyprus

8.84

80

Denmark

12.02

78.9

Finland

13.1

79.6

France

12.48

81.1

Germany

12.14

80

Greece

11.01

80.2

Hungary

16.12

73.9

Iceland

7.38

82.4

Italy

9.72

81.5

Lithuania

16.3

72.1

Macedonia [FYROM]

8.94

74.5

Moldova

23.01

70.4

Netherlands

9.75

80.3

Norway

8.35

80.8

Poland

14.43

75.4

Portugal

13.89

79.4

Romania

16.15

73.2

Serbia

12.21

74.3

Slovenia

14.94

78.7

Sweden

9.5

81.1

Switzerland

11.41

82

Ukraine

17.47

67.8

United Kingdom

13.24

79.7