Structural analysis and modelling of alcoholism as a non-communicable disease

Abstract

Alcohol use has been part of many cultures globally for many years. However, alcoholism or alcohol use disorders was recognized in Kenya in 2014 as a disease as classified by World Health Organization. Alcoholism is one of the highest causes of global disease burden resulting from liver cirrhosis, road traffic accidents and several types of cancers. In Kenya alcoholism is a persistent health problem with harmful alcohol consumption, especially among young people (18-35 years), being on the rise in spite of stringent laws governing alcohol use. Existing models consider transmission of alcoholism through social interaction in addition prevalence and incidences of alcohol use and alcoholism in Kenya have mainly been determined using surveys. Models that address the progression stages of alcoholism, from susceptible through social drinking to alcoholism are lacking. This study sought to analyse structural relationship between risk factors and alcoholism and model alcoholism as a non communicable disease. The specific objectives were to: Analyse structural relationship between risk factors and alcoholism; model incubation period of alcoholism; evaluate parametric and non-parametric hazard of alcoholism and predict incidences of alcoholism. Method: Secondary data was sourced from the Ministry of Health. Structural Equation Modelling (SEM) was used to show structural relationship between alcoholism and the latent variables; to model incubation period of alcoholism Birnbaum-Saunders (B-S) distribution hinged on biological process of cumulative cell damage caused by excessive alcohol consumption was used; hazard of becoming alcoholic was determined non-parametrically using discrete-time hazard model and incidences of alcoholism were found using logistic regression and back-projection. Results: There was a significant relationship between alcoholism and the risk factors (gender, peer influence, age at onset, social-cultural, economic status, environmental settings, drinking habits and pattern, family and family attention and personality) (RMSEA =0.06; CFI =0.80; SRMR=0.06); B-S distribution (𝛼 =0.77 [CI: 0.68, 0.85] and 𝛽 = 6.13 [CI: 5.44, 6.83], R 2 =0.94, was appropriate for modelling incubation period of alcoholism; the probability of becoming alcoholic increased from 0.31% when drinking once per week to 57% when drinking seven sessions a week, in addition the hazard of becoming alcoholic was higher for females than for males; The model was used to predict incidence of alcoholism between 2014 and September 2019. The predicted number of alcoholics (6632) do not differ significantly from the reported cases alcoholics (6631). In conclusion the analyses of the relationship between risk factors and alcoholism showed that risk of becoming alcoholic was affected differently by different risk factors with gender having the largest impact. Biophysical process of fatigue failure caused by cumulative cell damage yielded a model of alcoholism as a non-communicable disease. Recommendation: The study recommends initiatives for sensitization on impacts of alcohol use and early diagnosis of alcoholism to help initiate prevention policies.