Background:
Helicobacter pylori (H. pylori) is the common etiology of gastric tumors. This study aimed to evaluate the risk factors associated with H. pylori infection in the eastern part of the Democratic Republic of the Congo (DR Congo), where these tumors seem to be more frequent than in its western part.
Patients and Methods:
Between January and December 2021, the authors conducted a multicenter case–control study in three hospitals in Bukavu City involving 90 individuals with dyspeptic complaints. Risk factors for H. pylori infection were assessed in a participant interview and H. pylori status from stool antigen detection.
Results:
Among the risk factors assessed, only history of H. pylori in the family and the habit of adding salt to already-seasoned food were found positively associated with the risk of H. pylori infection (adjusted odds ratio: 7, 95 CI: 2.742–17.867; P<0.0001 and 2.911, 95% CI: 1.010–8.526; P=0.048, respectively). On the other hand, low-temperature food storage seems to be protective with a negative association (adjusted odds ratio: 0.044, 95% CI: 0.009–0.206; P=0.0001).
Conclusion:
This study demonstrated again the importance of lifestyle-related factors on the risk of acquisition of H. pylori. These findings call for preventive interventions for this group of individuals.
Keywords: Helicobacter pylori, risk factors, DR of the Congo
Introduction
Highlights
Helicobacter pylori infection remains a public health concern in the DR Congo.
A family history of H. pylori infection influence positively the presence of this infection.
The addition of salt to already-seasoned foods influences positively the H. pylori infection.
Low-temperature food storage seems to protect against H. pylori infection.
Helicobacter pylori (H. pylori) is a spiral, motile, and microaerophilic Gram-negative bacterium 3–4-μm long. This bacterium particularly colonizes the gastric mucosa and was first isolated in 1982 from gastric biopsies by two Australian researchers: J. Robin Warren and Barry J. Marshall1,2.
Transmission is essentially human-to-human within the family (parent–child or sibling), and the infection is acquired during childhood by oral or fecal-oral contamination due to unfavorable hygiene and promiscuity conditions3.
H. pylori infection is the most common chronic bacterial infection worldwide and the leading cause of gastric cancer4. In many studies, risk factors reported include low economic status, low education, crowded housing, poor access to safe water, and young age5. These factors, which are frequent in developing countries, explain the high prevalence of this disease. Indeed, while the rate of H. pylori infection decreases in developed countries due to improved living conditions, developing countries still suffer from this disease with a high prevalence that can reach 80 or even 90%6,7.
Since 1997, the European Consensus Conferences of Maastricht allowed standardization of clinical practices for H. pylori infection worldwide7. Unfortunately, the management of this infection is confronted nowadays by the antibiotic resistance of the bacteria. This latter result leads to treatment failure and favors the evolution of the disease toward complications such as peptic ulcer, Mucosa-Associated Lymphoid Tissue lymphoma, gastric atrophy, and even gastric adenocarcinoma8. The diagnosis and management of these complications remain very expensive for developing countries, where, unfortunately, the infection is the most frequent. That is why identification and better control of risk factors seem to be the most effective measures in the fight against this disease.
In the DRC and especially in South Kivu Province, studies on H. pylori infection are rare. In 1991, a prospective series in a rural region of South Kivu showed a 90% prevalence in symptomatic patients according to histological and bacteriological analysis9. Between 2001 and 2011, a multicenter study analyzing precancerous and cancerous gastric lesions noted a higher frequency of cancer in the eastern provinces (including South Kivu Province) of the DRC than elsewhere in the country10. Thus, the prevalence and, especially, the consequences of this infection could be more frequent in our region. Hence the importance of research in identifying its predisposing factors in our environment.
This study aims to prevent H. pylori infection in Bukavu City by determining its risk factors.
Patients and methods
Study framework
This was a multicenter case–control study conducted from January to December 2021 in the three main hospitals of Bukavu City, South Kivu Province, in Democratic Republic of the Congo. This work has been reported in line with the STROCSS criteria11.
Study population
In total, 90 patients of both sexes who consulted at these three hospitals with dyspeptic symptoms were enrolled in this study. Recruitment was continuous throughout the study period to achieve a 1 : 1 ratio of cases to controls (one infected for one uninfected). At the end of the recruitment period, we had a total of 50 cases and 45 controls. We, therefore, decided to eliminate the last five patients recruited from the case group to achieve the set proportions (45 cases to 45 controls). A case was defined as a patient with a positive antigenic test. In the same period, we also enrolled controls among dyspeptic patients with negative tests using the same method. We have excluded from this study any participant, case, or control with unclear results. The majority of cases were recruited at the Hopital Général de Panzi (44.4%), while two-thirds of control patients came from the Hopital Général de Bukavu (66.7%).
Sample collection and H. pylori infection status
The H. pylori infections were diagnosed based on the combined stool analysis antigen test. For this purpose, 50 mg of solid stools (or two drops of liquid stool) were transferred to a tube containing an extraction buffer. The tube was then shaken and left to settle for at least 2 min. Two drops of the extracted specimen were then transferred to the test device, and results were read after 10 min, according to the manufacturer’s instructions.
Data collection
Epidemiological data from questionnaires included sociodemographic data (age, sex, marital status, occupation, educational level, and socioeconomic level), socioeconomic factors (socioeconomic status, household population, source of drinking water), and factors influencing the onset of the disease (history of H. pylori infection, dietary, and eating behavior, family factors, alcohol, and tobacco consumption).
Statistical analysis
All data were computerized, and statistical analysis was performed using Epi-info 7.1 software. χ 2 and t-tests were used for categorical and continuous data, respectively. The strength of association between variables was assessed by simple logistic regression to estimate the adjusted odds ratio and 95% CI. P less than 0.05 was considered statistically significant.
Ethical considerations
The participants were aware of the purpose and methodology of this study and voluntarily agreed to answer our questions. Written informed consent was obtained from each participant, and the ethical committee of each hospital approved the study. The confidentiality of the collected data was scrupulously respected.
Results
Table 1 shows the characteristics of the 90 participants. Patients under 50 years of age were the most represented in both groups, at 77.7 and 97.7% in cases and controls, respectively. The male sex remained the most represented in both groups with 55.6 and 53.4% of patients, respectively (P=0.832). However, cases appear to be older than controls (37.4±18 vs. 29.1±10 years) (P=0.0082). According to the socioeconomic data, neither the level of education nor the economic level seemed to influence the H. pylori infection status. Concerning family history, H. pylori infection in a first-degree relative and the presence of gastric cancer in the family were associated with H. pylori-positive infection. Other factors, such as the presence of dyspeptic symptoms in the family or the history of gastric surgery, did not show a statistically significant difference.
Table 1.
Characteristics of study participants and H. pylori positivity.
Cases [n (%)] | Control [n (%)] | P value | |
---|---|---|---|
Age group (years) | |||
<50 | 35 (77.7) | 44 (97.7) | |
>50 | 10 (22.2) | 1 (2.2) | 0.01 |
Sex | |||
Female | 20 (44.4) | 21 (46.6) | |
Male | 25 (55.6) | 24 (53.4) | 0.832 |
Education | |||
Primary or lesser | 7 (15.5) | 4 (8.8) | 0.51 |
Secondary or above | 38 (84.4) | 41 (91.1) | |
Socioeconomic level | |||
Low | 17 (37.8) | 12 (26.7) | 0.143 |
Medium | 12 (26.7) | 21 (46.6) | |
High | 16 (35.6) | 12 (26.7) | |
Family history | |||
Family history of H. pylori infection | 35 (77.8) | 15 (33.3) | <0.0001 |
Family history of dyspepsia | 20 (44.4) | 12 (26.7) | 0.078 |
Family history of gastric cancer | 7 (15.6) | 1 (2.2) | 0.026 |
Family history of gastric surgery | 2 (4.4) | 1 (2.2) | 0.557 |
This table shows the principal characteristics of the study population. Family history of H. pylori infection and family history of gastric cancer seem to influence the H. pylori infection.
H. pylori, Helicobacter pylori.
According to lifestyle, no factors related to life in promiscuity (i.e. household crowding, community toilets, community meals, etc.); access to drinking water, alcohol consumption, or smoking seem to be associated with the presence of H. pylori infection (Table 2).
Table 2.
Lifestyle and dietary/eating habits are associated risk factors for H. pylori infection.
Cases [n (%)] | Controls [n (%)] | P value | |
---|---|---|---|
Lifestyle | |||
Household crowding (mean±SD) | 7.8±3 (2–21) | 6.8±3 (3–12) | 0.117 |
Drinking water from public water distribution system | 31 (68.9) | 36 (80) | 0.226 |
Drinking water from wells | 6 (13.3) | 2 (4.4) | 0.138 |
Community toilets | 7 (15.5) | 6 (13.3) | 0.764 |
Community meals | 40 (88.9) | 36 (80) | 0.244 |
Frequent alcohol driking | 17 (37.8) | 15 (33.3) | 0.824 |
Smoking | 4 (8.9) | 3 (6.6) | 0.693 |
Method of food preservation | |||
Refrigeration | 22 (48.8) | 43 (88.8) | <0.001 |
Smoking | 7 (15.5) | 1 (2.2) | 0.026 |
Salting | 5 (11.1) | 1 (2.2) | 0.091 |
Fruit and vegetable consumption | |||
Low fruit consumption | 18 (40) | 12 (26.7) | 0.179 |
Low vegetable consumption | 11 (24.4) | 5 (19.94) | 0.098 |
Salt consumption | |||
Adding salt at the table | 14 (31.1) | 6 (13.3) | 0.042 |
Low salt diet | 16 (35.5) | 15 (33.3) | 0.824 |
This table shows lifestyle and eating habits that may influence the H. pylori infection. Food smoking storage and adding salt on the table to already-seasoned food are associated with the risk of H. pylori Infection. Refrigeration food storage seems to have a protective effect.
H. pylori, Helicobacter pylori.
However, concerning the dietary habits, preserving food by smoking was a risk factor for developing H. pylori infection, as well as the addition of salt at the table. On the other hand, some factors, such as the cold storage of food seem protective.
The logistic analysis (Table 3) using factors with statistical significance in univariate analysis showed that family history of H. pylori infection [odds ratio (OR): 7; 95% CI: 2.742–17.867; P<0.0001] and addition of salt at the table (OR: 2.911, 95% CI: 1.010–8.526; P=0.048) were positively associated with H. pylori infection. In contrast, low-temperature storage food seems to be a protective factor (OR: 0.044, 95% CI: 0.009–0.206; P=0.0001). However, factors such as family history of gastric cancer or smoking and food preservation seem to show an association but are not statistically significant (P=0.055, respectively).
Table 3.
Risk factors for H. pylori infection as evaluated by a multivariate analysis.
Adjusted OR | 95% CI | P | |
---|---|---|---|
Age >50 years | 1.75 | 0.746–4.106 | 0.198 |
Family history of H. pylori infection | 7.00 | 2.742–17.867 | <0.0001 |
Family history of gastric cancer | 8.10 | 0.953–68.877 | 0.055 |
Cold storage of food | 0.044 | 0.009–0.206 | 0.0001 |
Preservation of food by smoking | 8.10 | 0.953–68.877 | 0.055 |
Adding salt at the table | 2.911 | 1.010–8.526 | 0.048 |
This table shows a logistic regression of factors influencing the H. pylori infection in univariate analysis. The family history of H. pylori infection and the adding salt on the table keep their positive association. Cold storage food shows a protective effect.
H. pylori, Helicobacter pylori; OR, odds ratio.
Discussion
The present study aimed to evaluate the risk factors associated with the acquisition of H. pylori infection in the eastern part of the DR of the Congo. Several studies have been conducted on risk factors for H. pylori infection, but some findings are still controversial. Generally, the infection is frequent among people with low socioeconomic and hygiene state. Our study appears to show the importance of lifestyle factors in the transmission of this infection.
H. pylori infection usually occurs in childhood, and the prevalence is higher in the young adult population12. In our study, the participants in both groups were predominantly young, which is largely related to the current age pyramid of African countries. As in several other studies, sex seemed not to be associated with H. pylori infection7,13–15.
Low socioeconomic status is generally recognized as a risk factor for H. pylori-related gastritis5,12. However, many authors from developing countries whose studies involved adult participants did not find the influence of socioeconomic level7,14. This was also found in our study, where neither the level of education (P=0.533) nor a low economic level (P=0.143) seemed to influence the presence of this bacterium. The explanation provided is that people with high living standards in developing countries have mostly been subjected to bad living conditions during their childhood, which is the common transmission period of the disease16.
The risk of H. pylori transmission between family members or cohabiting individuals is suggested by the presence of the same strain in people living together15. This is mainly related to oral or fecal transmission. In our study, factors suggesting family contamination were also assessed. Our study showed that the presence of H. pylori infection in the family increases the risk of acquiring the infection by seven times. Several other studies have reported an increased risk of intrafamilial H. pylori infection, and this is obvious in the available literature17,18.
Some behaviors, such as promiscuity (assessed by household size), unsafe drinking water, use of shared toilets, and consuming food in community, did not influence the presence of H. pylori although these factors are known to promote the risk of infection5,16. This may be explained by the small sample size of our study population but also by the fact that the current socioeconomic status of individuals does not necessarily reflect the status during childhood, the usual period of contamination.
Studies on the association between H. pylori and smoking or alcohol consumption are participant of contradictions in the literature. Even though one study recently incriminated smoking as a risk factor for H. pylori eradication treatment failure19, most studies have not found the influence of smoking on susceptibility to infection20–23. However, in one study, Wu et al.24 curiously found that smoking was a risk factor in males but protective in females. Moreover, a Turkish study has incriminated regular smoking as a risk factor for H. pylori infection25. Our study reported no statistically significant effect of smoking and alcohol consumption on H. pylori infection. In the same way, tobacco consumption also did not influence the presence of H. pylori. It is known that alcohol has antimicrobial activity and stimulates gastric acid secretion, which could interfere with the growth of the bacterium26. Some other studies have also found no association between alcohol consumption and H. pylori infection27,28.
The role of salt consumption in the susceptibility to developing H. pylori infection is not well studied. However, some studies have suggested an increase in gastric disease associated with H. pylori infection in the population with high sodium chloride consumption29,30. Our study showed that H. pylori infection was higher among individuals with the habit of adding salt at the table to already-seasoned food. In this group, the risk of acquiring the infection is majored (ORa: 2.911, 95% CI: 1.010–8.526; P=0.048).
The relationship between H. pylori infection and cold storage is little discussed in the literature. The present study revealed that low-temperature storage of food seems to be protective. Although H. pylori have demonstrated resistance to low-temperature conditions31, this latter could retain several substances that interfere with the growth of this bacterium. These include vitamins that act as natural antioxidants32. But also, the preservation of food by cold could discourage other, more harmful forms of storage (such as preservation by salt), which is positively associated with the risk of developing an H. pylori infection. Finally, the widespread use of refrigerators in Western countries in the recent past may partly explain the continuing decline in the incidence of gastric cancer33
In conclusion, this study is the only one so far carried out in our setting that has investigated the susceptibility factors of H. pylori infection in the DRC, where the prevalence of gastric tumors is highest. Our study has found that classically favoring factors such as a low socioeconomic level and some unfavorable living conditions, although recognized in the literature as risk factors, did not influence the presence of H. pylori infection. However, the history of H. pylori infection in the family and some dietary behaviors seem to be positively or negatively associated with the risk of H. pylori infection.
The limitation of this study is its small sample size, which does not allow generalize the results. Finally, the difficulty of objectively quantifying some parameters, such as the quantity of salt consumed, should be taken into account when considering our findings.
Ethical approval
020/2021/UOB/FM/EC/MKS.
Consent
Yes.
Sources of funding
None.
Authors’ contributions
Y.C.B. drafted the manuscript and assisted in data analysis. A.M.B. drafted the manuscript and assisted in sample collection. D.K.M. reviewed the manuscript. T.A.S. conceived and designed the study, analyzed the data, and reviewed the manuscript. All authors read and approved the final manuscript.
Conflicts of interest disclosure
None.
Provenance and peer review
Not commissioned, externally peer reviewed.
Acknowledgments
The authors thank nurses and laboratory technicians of the three hospitals, for their technical support.
Footnotes
Sponsorships or competing interests that may be relevant to content are disclosed at the end of this article.
Published online ■ ■
Contributor Information
Yannick C. Birato, Email: ychibinda@gmail.com.
Bahati Armand Masimango, Email: armandobahati1@gmail.com.
Delphin M. Katabana, Email: delphinmurhula@gmail.com.
Tony A. Shindano, Email: tonyshinda@gmail.com.
References
- 1. Appelman HD. Gastritis: terminology, etiology, and clinicopathological correlations: another biased view. Hum Pathol 1994;25:1006–1019. [DOI] [PubMed] [Google Scholar]
- 2. Mignon M. Helicobacter pylori awarded [Helicobacter pylori couronné]. Méd Sci 2005;21:993–994. [DOI] [PubMed] [Google Scholar]
- 3. Mladenova I, Durazzo M. Transmission of Helicobacter pylori . Minerva Gastroenterol Dietol 2018;64:251–254. [DOI] [PubMed] [Google Scholar]
- 4. de Korwin JD. Epidemiology of Helicobacter pylori infection and gastric cancer. Rev Prat 2014;64:189–193. [PubMed] [Google Scholar]
- 5. Bommelaer G, Stef A. Gastroduodenal ulcer before and after Helicobacter pylori [Ulcère gastroduodénal: avant et après Helicobacter pylori]. Gastroentérologie Clin Biol 2009;33(8–9):626–634. [DOI] [PubMed] [Google Scholar]
- 6. Heluwaert F. PYLORI HEBDO*: résultat d’une enquête de pratique sur la prise en charge d’Helicobacter pylori dans 31 centres ANGH en 2014 | SNFGE.org – Société savante médicale française d’hépato-gastroentérologie et d’oncologie digestive. Accessed 15 October 2021. https://www.snfge.org/content/pylori-hebdo-resultat-dune-enquete-de-pratique-sur-la-prise-en-charge-dhelicobacter-pylori
- 7. Andoulo FA, Noah Noah D, Tagni-Sartre M, et al. Epidemiology of infection Helicobacter pylori in Yaoundé:specificity of the African enigma [Epidémiologie de l’infection à Helicobacter pylori à Yaoundé: de la particularité à l’énigme Africaine]. Pan Afr Med J 2013;16:115. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Graham DY. Benefits from elimination of Helicobacter pylori infection include major reduction in the incidence of peptic ulcer disease, gastric cancer, and primary gastric lymphoma. Prev Med 1994;23:712–716. [DOI] [PubMed] [Google Scholar]
- 9. Glupczinski Y, Bourdeaux L, De Prez C, et al. Prevalence of Helicobacter pylori in rural Kivu, esatern Zaire: a prospective endoscopic study. Eur J Gastroenterol Hepatol 1991;3:449–455. [Google Scholar]
- 10. Bomba E, Nkondi NJ, Kabongo MJ, et al. Cancer gastrique et infection à Helicobacter pylori en RD Congo. Aspects épidémiologiques. Ann Afr Médecine 2013;6(4):1–7. [Google Scholar]
- 11. Mathew G, Agha R, Albrecht J, et al. Strocss 2021: strengthening the reporting of cohort, cross-sectional and case-control studies in surgery. Int J Surg Open 2021;37:100430. [DOI] [PubMed] [Google Scholar]
- 12. Peleteiro B, Bastos A, Ferro A, et al. Prevalence of Helicobacter pylori infection worldwide: a systematic review of studies with national coverage. Dig Dis Sci 2014;59:1698–1709. [DOI] [PubMed] [Google Scholar]
- 13. Jaiswal S, Tiwari BR, Sharma DC. Prevalence and factor assessment of Helicobacter pylori infection in a rural setting. Int J Life Sci Pharma Res 2021;11:L194–L199. [Google Scholar]
- 14. Bougouma A, Diomande I, Iboudo D, et al. Epidemiological and clinical aspects of helicobacter pylori infection in the tropics: about 150 patients at the national hospital of Ouagadougou (Burkina Faso) [Aspects epidemiologiques et cliniques de l’infection a Helicobacter pylori en zone tropicale: a propos de 150 patients a l’hopital national de Ouagadougou]. Med Afr Noire En Ligne 1997;44:24–28. [Google Scholar]
- 15. Bamford KB, Bickley J, Collins JS, et al. Helicobacter pylori: comparison of DNA fingerprints provides evidence for intrafamilial infection. Gut 1993;34:1348–1350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Klein PD, Opekun AR, Smith EO, et al. Water source as risk factor for Helicobacter pylori infection in Peruvian children. Lancet 1991;337:1503–1506. [DOI] [PubMed] [Google Scholar]
- 17. Escobar ML, Kawakami E. Evidence of mother-child transmission of Helicobacter pylori infection. Arq Gastroenterol 2004;41:239–244. [DOI] [PubMed] [Google Scholar]
- 18. Tindberg Y, Blennow M, Bengtsson C, et al. Helicobacter pylori infection in Swedish school children: lack of evidence of child-to-child transmission outside the family. Gastroenterology 2001;121:310–316. [DOI] [PubMed] [Google Scholar]
- 19. Itskoviz D, Boltin D, Leibovitzh H, et al. Smoking increases the likelihood of Helicobacter pylori treatment failure. Dig Liver Dis 2017;49:764–768. [DOI] [PubMed] [Google Scholar]
- 20. Ferro A, Morais S, Pelucchi C, et al. Smoking and Helicobacter pylori infection: an individual participant pooled analysis (Stomach Cancer Pooling-StoP Project. Eur J Cancer Prev 2019;28:390–396. [DOI] [PubMed] [Google Scholar]
- 21. Amaral O, Fernandes I, Veiga N, et al. Living conditions and Helicobacter pylori in adults. BioMed Res Int 2017;2017:1–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Brenner H, Rothenbacher D, Bode G, et al. Relation of smoking and alcohol and coffee consumption to active Helicobacter pylori infection: cross sectional study. BMJ 1997;315:1489–1492. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Bateson MC. Cigarette smoking and Helicobacter pylori infection. Postgrad Med J 1993;69:41–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Wu W, Leja M, Tsukanov V, et al. Sex differences in the relationship among alcohol, smoking, and Helicobacter pylori infection in asymptomatic individuals. J Int Med Res 2020;48:0300060520926036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Ozaydin N, Turkyilmaz SA, Cali S. Prevalence and risk factors of Helicobacter pylori in Turkey: a nationally-representative, cross-sectional, screening with the 13C-urea breath test. BMC Public Health 2013;13:1215. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Singer MV, Leffmann C, Eysselein VE, et al. Action of ethanol and some alcoholic beverages on gastric acid secretion and release of gastrin in humans. Gastroenterology 1987;93:1247–1254. [DOI] [PubMed] [Google Scholar]
- 27. den Hollander WJ, Holster IL, den Hoed CM, et al. Ethnicity is a strong predictor for Helicobacter pylori infection in young women in a multi-ethnic European city: Urban Helicobacter pylori colonization. J Gastroenterol Hepatol 2013;28:1705–1711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Zhu Y, Zhou X, Wu J, et al. Risk factors and prevalence of Helicobacter pylori infection in persistent high incidence area of gastric carcinoma in Yangzhong City. Gastroenterol Res Pract 2014;2014:1–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Loh JT, Beckett AC, Scholz MB, et al. High-salt conditions alter transcription of Helicobacter pylori genes encoding outer membrane proteins. Infect Immun 2018;86:e00626–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Cover TL, Peek RM, Jr. Diet, microbial virulence, and Helicobacter pylori-induced gastric cancer. Gut Microbes 2013;4:482–493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Jiang X, Doyle MP. Effect of environmental and substrate factors on survival and growth of Helicobacter pylori . J Food Prot 1998;61:929–933. [DOI] [PubMed] [Google Scholar]
- 32. Fei SJ, Xiao SD. Diet and gastric cancer: a case-control study in Shanghai urban districts. Chin J Dig Dis 2006;7:83–88. [DOI] [PubMed] [Google Scholar]
- 33. Yan S, Gan Y, Song X, et al. Association between refrigerator use and the risk of gastric cancer: a systematic review and meta-analysis of observational studies. PLoS One 2018;13:e0203120. [DOI] [PMC free article] [PubMed] [Google Scholar]