alexa Identification of the Risk Factors Associated with Helicobacter pylori Infection in Lahore, Pakistan | Open Access Journals
ISSN: 2155-6180
Journal of Biometrics & Biostatistics
Like us on:
Make the best use of Scientific Research and information from our 700+ peer reviewed, Open Access Journals that operates with the help of 50,000+ Editorial Board Members and esteemed reviewers and 1000+ Scientific associations in Medical, Clinical, Pharmaceutical, Engineering, Technology and Management Fields.
Meet Inspiring Speakers and Experts at our 3000+ Global Conferenceseries Events with over 600+ Conferences, 1200+ Symposiums and 1200+ Workshops on
Medical, Pharma, Engineering, Science, Technology and Business

Identification of the Risk Factors Associated with Helicobacter pylori Infection in Lahore, Pakistan

Ashfaq Ali1, Muhammad Riaz Ahmad2, Zafar Iqbal3 and Abdul Basit4*

1Department of Statistics, Govt. M.A.O. College, Lahore, Pakistan

2Govt. Postgraduate College, Jaranwal, Pakistan

3Department of Statistics, Govt. Post Graduate College, Gujranwala, Pakistan

4State Bank of Pakistan, Karachi, Pakistan

*Corresponding Author:
Abdul Basit
State Bank of Pakistan, Karachi, Pakistan
Tel: 0092-333-4859346
E-mail: [email protected]

Received Date: January 03, 2017; Accepted Date: May 27, 2017; Published Date: May 31, 2017

Citation: Ali A, Ahmad MR, Iqbal Z, Basit A (2017) Identification of the Risk Factors Associated with Helicobacter pylori Infection in Lahore, Pakistan. J Biom Biostat 8: 348. doi: 10.4172/2155-6180.1000348

Copyright: © 2017 Ali A, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Visit for more related articles at Journal of Biometrics & Biostatistics

Abstract

In this study, significant risk factors of Helicobacter pylori infection in Lahore are investigated through a case-control study by using descriptive and analytical approaches. A sample of 362 subjects was selected from the Gastroenterology Departments and OPDs of different hospitals of Lahore. About 25 risk factors with sub categories were included in the study. For bivariate analysis, the chi-square, phi/v statistics and Kendall’s tau-b are used. From descriptive analysis, it was found that the persons who eat from restaurants have more risk of infection as compare to persons who eat homemade food. By the descriptive analysis, it was also observed that risk of Helicobacter pylori infection increases with an increase in the number of family members per house and in the number of persons living per room. Furthermore, similar results were observed in the bivariate analysis. In the analysis, the five risk factors including age, food eat, food liked, dental complains and number of persons living per room are found to be positively significant having the odds ratios and 95% confidence intervals of odds ratios (1.025; 1.003-1.047), (9.596; 4.767-19.314), (3.500; 1.509-8.119), (3.204; 1.685-6.094) and (2.772; 1.496-5.139), respectively. While the three risk factors including usage of tea, educational level and sewerage system are found to be negatively significant having odds ratios and 95% confidence intervals for the odds ratios (0.221; 0.119-0.411), (0.216; 0.115-0.404) and (0.401; 0.218-0.738), respectively, which indicates that these three risk factors are protective factors against Helicobacter pylori infection. According to this study, the subjects who eat from restaurants have higher risk of Helicobacter pylori infection as compared to all other risk factors.

Keywords

Helicobacter pylori infection; Risk factors; Logistic regression; Odds ratio; Controls; Significance; Retrospective

Introduction

During infections due to bacteria, factors related with bacteria, host and environment, all three are cause of infection results and infection appearance. For the development of a stubborn infection, the bacteria basically have the ability of adapting to a number of different ecological variations that happen during the infection. An example of the bacterium is the Helicobacter pylori (H. pylori) which is the cause of the stubborn infections in the stomach of human.

H. pylori infect more than 30% population of the world and in few countries this infects above than 50% people from the total population [1]. H. pylori is one of the most common bacterial infections in human beings. In 1994, Helicobacter pylori was categorized as a class I human carcinogen by the WHO International Agency for Research on Cancer (IARC) due to its epidemiological connection to gastric cancer.

In Pakistan, acid peptic disease due to the prevalence of Helicobacter pylori Bacterium infection is very high among population and the numbers of patients are continuously increasing mainly due to non-availability of ideal diagnostic and treatment facilities in public sector healthcare services. As H. pylori is a rising issue in Pakistan so according to experts of health H. pylori infection may become chronic with the chances of gastric carcinoma and stomach cancer, if it is not treated fully and properly [2].

Material and Methods

This study is hospital-based, an epidemiological and analytical case and control which is conducted to find out the risk factors of the Helicobacter pylori infection in patients of all age groups. For case group, the patients from the Gastroenterology Departments and Outdoor Patient Departments (OPD) of different hospitals of Lahore, was the target population while the attendants of the patients or all the other patients having any kind of stomach disease other than H. pylori infection who came to the above mentioned departments was the target population for the control group.

This study consists of 362 respondents including 181 cases and 181 controls. There are almost 25 risk factors including age, gender, marital status, smoking, quantity of cigarettes smoked on each day, usage of tea, quantity of cups of tea taken on each day, usage of coffee, quantity of cups of coffee taken on each day, food’s type to eat, type of food to like, source of drinking water, dental complains, educational level, socioeconomic level, garbage collection system, type of accommodation, number of rooms in accommodation, number of persons per room, evaluation of sewerage system, use of drugs, use of antibiotics, contacts with animals and travelling outside the native district.

From above factors seven are treated as quantitative including age, quantity of cigarettes smoked on each day, quantity of cups of tea taken on each day, quantity of cups of coffee taken on each day, income, number of rooms in accommodation, number of persons per room. All the other variables are qualitative from which most of the variables are in dichotomous form. To present the complete and comprehensive analysis, this part of study is divided into two parts; analytical and model building.

In this study, response variable is binary and independent variables are both categorical and continuous. Also in this situation, there exists nonlinear relationship between response variable and independent variables, so the best and most appropriate choice for modelling is the binary logistic regression model in order to predict the occurrence of H. pylori infection [3].

Results

The sample consists of 181 (50.0%) cases and 181 (50.0%) controls. The age was treated as continuous variable. The minimum and maximum ages of the subjects were 15 years and 80 years, respectively. The average age of patients was 39.89 ± 13.972 years.

Association analysis shows that the risk factors, smoking, number of cigarettes smoked per day, type of food, type of liked food, dental complains, number of family members per house and number of persons living per room are found to be positively significantly associated with Helicobacter pylori infection, while the risk factors use of tea, source of drinking water, educational level, socio economic status, garbage collection system and sewerage system are found to be negatively associated with Helicobacter pylori infection [4-9]. On the other hand, Gender, marital status, use of antibiotics in last month, contact with animals and travelled outside the native city/town have not significant association infection. The value of V/Phi statistic for the type of food is 0.421, which is the highest value, so it is highly associated nominal categorical variable with Helicobacter pylori infection among all the significant nominal risk factors [10-15]. The value of Kendall’s Tau-b statistic for the number of persons living per room is 0.338, which is the highest value, so it is highly associated ordinal categorical variable with Helicobacter pylori infection among all the significant ordinal risk factors.

Binary logistic regression model is fitted. In this model, regression coefficients, odds ratios, p-values and 95% confidence intervals of odds ratios are calculated for patients. Omnibus and Hosmer-Lemeshow tests are used to check the overall appropriateness of the model (Table 1).

Variables Classification H. pylori infection Total
No Yes
No. %age No. %age No. %age
Gender Female 65 48.9 68 51.1 133 36.7
Male 116 50.7 113 49.3 229 63.4
Marital Status Unmarried 67 51.9 62 48.1 129 35.6
Married 114 48.9 119 51.1 233 64.4
Smoking No 155 54.6 129 45.4 284 78.5
Yes 26 33.3 52 66.7 78 21.5
No. of cigarettes smoked per day 0 155 54.6 129 45.4 284 78.5
1–10 18 34.0 35 66.0 53 14.6
11–20 6 26.1 17 73.9 23 6.4
Above 20 2 100 0 0 2 0.6
Use of tea No 52 31.7 112 68.3 164 45.3
Yes 129 65.2 69 34.8 198 54.7
Use of coffee No 181 50 181 50 362 100
Yes 0 0 0 0 0 0
Type of food Homemade 153 65.7 80 34.3 233 64.4
Restaurant 28 21.7 101 78.3 129 35.6
Food liked Less spicy 59 76.6 18 23.4 77 21.3
Spicy 95 47.5 105 52.5 200 55.2
Over spicy 27 31.8 58 68.2 85 23.5
Source of drinking water Municipality/water pump 98 38.4 157 61.6 255 70.4
Filtered 83 77.6 24 22.4 107 29.6
Dental complains No 142 64.0 80 36.0 222 61.3
Yes 39 27.9 101 72.1 140 38.7
Educational Level Low 53 30.1 123 69.9 176 48.6
High 128 68.8 58 31.2 186 51.4
Socio economic status Below 20000 60 41.1 86 58.9 146 40.3
20000–40000 72 51.4 68 48.6 140 38.7
Above 40000 49 64.5 27 35.5 76 21.0
Garbage collection system No 81 44.5 101 55.5 182 50.3
Yes 100 55.6 80 44.4 180 49.7
No. of family members per house ≤ 5 82 70.1 35 29.9 117 32.3
>5 99 40.4 146 59.6 245 67.7
No. of persons living per room ≤ 3 129 65.5 68 34.5 197 54.4
>3 52 31.5 113 68.5 165 45.6
Evaluation of sewerage system Poor 16 16.8 79 83.2 95 26.2
Good 89 54.9 73 45.1 162 44.8
Excellent 76 72.4 29 27.6 105 29.0
Use of drugs No 178 49.7 180 50.2 358 98.9
Yes 3 75.0 1 25.0 4 1.1
Use of antibiotics in last month No 97 51.1 93 48.9 190 52.5
Yes 84 48.8 88 51.2 172 47.5
Contacts with animal No 113 48.3 121 51.7 234 64.6
Yes 68 53.1 60 46.9 128 35.4
Travelled outside native city/district No 50 56.2 39 43.8 89 24.6
Yes 131 48.0 142 52.0 273 75.4

Table 1: Counts and percentages of cases and controls according to different risk factors.

It is examined that model is significant, because in Omnibus test the value of p=0.000 is less than the level of significance 0.05 (5%) for model, so at least one of the predictors is playing a significant role in predicting the response variable and the fitted model is adequate [16-20]. In Hosmer-Lemeshow test, the value of p=.996, so Hosmer- Lemeshow test is insignificant, which shows that the fitted model is adequate.

For checking the overall goodness of fit of the model, Cox & Snell and Nagelkerke R2 are used for model.

The values of Cox & Snell and Nagelkerke R2 for male model are 0.464 and 0.618, respectively, which shows that strength of association is high between the response and independent [21].

From Table 2, it is examined that 147 (81.2%) uninfected (controls) and 153 (84.5%) infected (cases) are rightly predicted, while 28 (15.5%) infected (cases) and 34 (18.8%) uninfected (controls) are misclassified 28 (15.5%) as uninfected (controls) and 34 (18.8%) as infected (cases). On the whole, the numbers (percentages) of rightly classified and misclassified respondents are 300 (82.9%) and 62 (17.1%), respectively. The value of accurate classification is sufficiently high, so the fitted model is satisfactory.

Observed Predicted Percentage correct
H. pylori infection
No Yes
H. pylori infection No 147 34 81.2
Yes 28 153 84.5
Overall percentage 82.9

Table 2: Correct and incorrect classification for the model.

From Omnibus test, Hosmer-Lemeshow test and correct classification, it is examined that the logistic regression model satisfactorily fit the data, so the 95% CI are furthermore applicable for inferences.

From Table 3, eight risk factors consisting of Age, Tea, Food eat (FE), Food liked (FL), Dental complains (DC), Educational level (Edu.), Number of persons living per room (NPPR) and Sewerage system (SS) are significant.

FACTORS Β S.E. Wald d.f. Sig. Exp(B) 95% CI forExp(B)
Lower Upper
Age 0.024 0.011 4.837 1 .028 1.025 1.003 1.047
Tea -1.509 0.317 22.722 1 .000 0.221 0.119 0.411
Food eat 2.261 0.357 40.143 1 .000 9.596 4.767 19.314
Food liked 1.253 0.429 8.519 1 .004 3.500 1.509 8.119
Dental comp. 1.164 0.328 12.600 1 .000 3.204 1.685 6.094
Edu. level -1.534 0.320 22.921 1 .000 0.216 0.115 0.404
NPPR 1.020 0.315 10.489 1 .001 2.772 1.496 5.139
Sewerage sys. -0.913 0.311 8.618 1 .003 0.401 0.218 0.738
Constant -1.694 0.661 6.566 1 .010 0.184    

Table 3: Model coefficients, P-values, OR and 95% C.I. for OR.

The fitted logit model is given as;

equation

It is observed that the five factors including Age, Food eat, Food liked, Dental complain and number of person living per room are positively significant, which means that H. pylori infection and these risk factors are directly related and the three factors including Tea, Educational level and Sewerage System are observed to be negatively significant which means that the H. pylori [22,23] infection and these risk factors are inversely related.

Explanation for the model coefficients, odds ratios and its 95% confidence intervals are discussed here.

As the value of odds ratio for the risk factor age is 1.025, which investigated that the subjects with the older age have about 1.025 times more risk of Helicobacter pylori infection than the subjects with young age [24-26]. From this study, it is examined that the effect of age is directly associated with the Helicobacter pylori infection, because in the model the value of logit coefficient is 0.024 with p=0.028. The 95% confidence interval for odds ratio of the age is (1.003, 1.047), which indicates that age has significant association with H. pylori infection, because confidence interval does not contain the value of one.

In this study, it is examined that the effect of tea is inversely associated with the Helicobacter pylori infection. As the value of odds ratio for tea in the model is 0.221, which explained that the subjects who take tea have (1 – 0.221=0.779) 77.9% protection against infection as compared to subjects who do not take tea. The 95% confidence interval for odds ratio of the tea is (0.119, 0.411), which indicates that tea has significant association with H. pylori infection, because confidence interval does not contain the value of one.

In this study, subjects were divided into two categories, one who eat homemade food and second who eat from restaurants and examined that the effect of food is directly associated with the Helicobacter pylori infection. Because the value of odds ratio for food in the model is 9.596, so the subjects who eat from restaurants have risk of infection 9.596 times more than the subjects who eat homemade food. The 95% CI for OR of the food is (4.767, 19.314), which indicates that food has significant association with H. pylori infection, because confidence interval does not contain the value of one.

As the value of odds ratio for types of liked foods is 3.500, which investigated that the subjects who liked spicy food have about 3.500 times more risk of Helicobacter pylori infection than the subjects who liked less spicy food. From this study, it is also examined that the effect of types of liked foods is directly associated with the Helicobacter pylori infection, because the value of logit coefficient is 1.253 with p=0.004. The 95% confidence interval for odds ratio of the types of liked foods is (1.509, 8.119), which indicates that types of liked foods have significant association with H. pylori infection, because confidence interval does not contain the value of one.

As the value of odds ratio for the dental complains is 3.204, which stated that the subjects who have dental diseases have risk of Helicobacter pylori infection 3.204 times more than the subjects who have not dental diseases. The 95% CI for OR of dental complain is (1.685, 6.094), which indicates that dental complain has significant association with H. pylori infection, because confidence interval does not contain the value of one.

The educational level was categorized as low and high. The educational level is examined as significant, because the values of odds ratio and its 95% confidence interval are 0.216 and (0.115, 0.404), respectively. The respondents with high education have (1 – 0.216=0.784) 78.4% protection against the subjects who have low education. The 95% confidence interval for odds ratios for education is (0.115, 0.404), which indicate that educational level has significant association with H. pylori infection, because confidence interval does not contain the value of one. From this study it is examined that the effect of educational level is inversely associated with the Helicobacter pylori infection because the value of logit coefficient is -1.534 with p=0.000.

The number of persons per room was distributed into two mutually exclusive groups as 3 or less persons per room and above 3 persons per room and examined that the effect of number of persons per room is directly associated with the Helicobacter pylori infection, because value of logit coefficient is 1.020 with p=0.001. Because the value of odds ratio for the number of person living per room in the model is 2.772, so the subjects who live more than 3 persons per room have 2.772 times more risk of Helicobacter pylori infection as compared to subjects who live 3 or less persons per room [27-30]. The 95% confidence interval for odds ratio of the number of persons per room is (1.496, 5.139), which indicates that number of persons per room has significant association with the H. pylori infection, because confidence interval does not contain the value of one.

From this study, it is examined that the effect of sewerage system is inversely associated with the Helicobacter pylori infection, because value of the logit coefficient is -0.913 with p=0.003. As the value of odds ratio for sewerage system in the model is 0.401, which explain that the subjects with good sewerage system have (1 – 0.401=0.599) 59.9% protection against infection as compared to subjects who have poor sewerage system in their houses. For sewerage system, the 95% confidence interval for odds ratio is (0.218, 0.738), which indicate that sewerage system has significant association with H. pylori infection, because confidence interval does not contain the value of one.

Predicted probabilities

Because eight risk factors are involved as predictors in the above fitted logistic regression equation, so there may exists numerous combinations of these risk factors.

As shown in serial no. 1 of Table 4, when all the significant risk factors associated with Helicobacter pylori infection are present that is subjects have older age with age 50 years, do not take tea, eat food from restaurants, liked spicy food, have dental complain, have low educational level, live 6 persons per room and have poor sewerage system, then these subjects have predicted probability of the occurrence of infection is 0.999999, which means that subjects with the above mentioned risk factors have 99.9999% chance of the occurrence of infection. As shown in serial no. 2 of Table 4, when all the significant risk factors associated with Helicobacter pylori infection are present except subjects age is 30 years means that subjects are not of old age, then these subjects have the predicted probability of the occurrence of infection is 0.999998.

SN Risk Factors Z equation
Age Tea FE FL DC Edu. NPPR S.S.
1 50 No Rest. Spicy Yes Low 6 Poor 13.818 0.999999
2 30 No Rest. Spicy Yes Low 6 Poor 13.338 0.999998
3 20 No Rest. Spicy Yes Low 6 Poor 13.098 0.999998
4 50 No Rest. Spicy Yes Low 5 Poor 12.798 0.999997
5 50 No Rest. Spicy Yes Low 4 Poor 11.778 0.999992
6 50 Yes Rest. Spicy Yes Low 6 Poor 12.309 0.999995
7 50 No H.M Spicy Yes Low 6 Poor 11.557 0.999990
8 50 No H.M. Spicy Yes Low 3 Poor 8.497 0.999796
9 50 No Rest. L.S Yes Low 6 Poor 12.565 0.999996
10 50 No Rest. Spicy No Low 2 Poor 8.574 0.999811
11 50 No Rest. Spicy Yes High 6 Poor 12.284 0.999995
12 50 No Rest. Spicy Yes Low 2 Poor 9.738 0.999941
13 50 No Rest. Spicy Yes Low 6 Good 12.905 0.999997
14 20 Yes H.M L.S. No High 2 Good 0.384 0.594837

Table 4: Predicted probabilities.

As shown in serial no. 3 of Table 4, when all the significant risk factors associated with Helicobacter pylori infection are present except subjects age is now 20 years, which means that subjects are too young, then these subjects have the predicted probability of the occurrence of infection is 0.999998.

Similarly to observe the influence of some of the individual risk factor, the predicted probabilities are calculated in Table 4, which are shown in serial no. 4 to 13.

Finally as shown in serial no. 14 of Table 4, when all the significant risk factors associated with Helicobacter pylori infection are absent that is subjects are of young age with age 20 years, take tea, eat food from homes, liked less spicy food, have not dental complain, have high educational level, live 2 persons per room and have good sewerage system, then these subjects have predicted probability of the occurrence of infection is 0.594837, which is low to others.

Conclusion

The risk factors of Helicobacter pylori infection are thoroughly analyzed. Bivariate analysis revealed that the type of food is highly associated nominal categorical variable with Helicobacter pylori infection among all the significant nominal risk factors, because it has the highest value of V/Phi statistic and the number of persons living per room is highly associated ordinal categorical variable with Helicobacter pylori infection among all the significant ordinal risk factors, because it has highest value of Kendall’s Tau-b statistic.

In Bivariate Logistic regression analysis, we found that the risk factors food eat, food liked, dental complain and number of persons living per room are the most significant risk factors in developing Helicobacter pylori infection in subjects.

References

Select your language of interest to view the total content in your interested language
Post your comment

Share This Article

Recommended Conferences

Article Usage

  • Total views: 64
  • [From(publication date):
    June-2017 - Jul 22, 2017]
  • Breakdown by view type
  • HTML page views : 47
  • PDF downloads :17
 

Post your comment

captcha   Reload  Can't read the image? click here to refresh

Peer Reviewed Journals
 
Make the best use of Scientific Research and information from our 700 + peer reviewed, Open Access Journals
International Conferences 2017-18
 
Meet Inspiring Speakers and Experts at our 3000+ Global Annual Meetings

Contact Us

 
© 2008-2017 OMICS International - Open Access Publisher. Best viewed in Mozilla Firefox | Google Chrome | Above IE 7.0 version
adwords