Exploratory multigroup structural equation modeling approach to analyzing H. pylori infection, serological pepsinogen levels, and life style factors in an elderly population in Costa Rica

• Published in: Población y salud en mesoamérica Vol. 19; no. 2
• Main Authors: Molina-Castro, Gabriel, Matarrita-Muñoz, Stefany, Molina-Castro, Silvia Elena, Méndez-Chacón, Ericka, Une, Clas, Ramírez-Mayorga, Vanessa
• Format: Journal Article
• Language: English
• Published: San Jose Universidad de Costa Rica, Centro Centroamericano de Poblacion 07.12.2021
• Subjects: Age differences; Body mass index; Caloric intake; Habits; High risk; Lifestyles; Obesity; Older people; Costa Rica
• ISSN: 1659-0201; 1659-0201
• DOI: 10.15517/psm.v19i2.48113

The aim of this work is to describe and analyze the association of PGI/PGII ratio (indicator of gastric atrophy) with H. pylori-CagA and life style factors such as caloric intake, obesity, and harmful habits amongst H. pylori-positive elderly people infected in Costa Rica using an exploratory multigroup structural equations model (SEM). Using a sample of 1748 H. pylori-positive elderly people from CRELES first wave study, a SEM was employed analyze if the relationships between PGI/PGII ratio with levels of H. pylori-CagA, caloric intake, obesity, and harmful habits, differs by sex, age and risk areas subgroups. The proposed SEMs exhibited a good fit in males (RMSEA = 0.039), females (RMSEA = 0.000), low-risk area (RMSEA = 0.038), middle-risk area (RMSEA = 0.042), individuals under 80 years (RMSEA = 0.038) and individuals aged 80 and over (RMSEA = 0.042), while an acceptable fit was observed for the high-risk area (RMSEA = 0.061). Fitted SEMs showed that CagA predicted PG-ratio as expected, with effects increasing with the risk area, but similar between sex and age groups. All indicators measuring obesity (BMI, arms, and waist) showed significant standardized coefficients, with similar effects between sex, age and risk area groups. No other significant effects or differences between groups were identified. We propose a good-fitted SEM model for the possible relationships between CagA and PG ratio and the geographical risk area level for elderly people. No differences were observed on measured parameters between male and female population, or between under 80 years and older individuals.