A matched case-control study of risk factors associated with multiple sclerosis in Kuwait

A matched case-control design was implemented to address the objective of this study. Cases were enrolled from neurology clinics at Ibn Sina Hospital and MS clinic at Dasman Diabetes Institute. These are the only two neurology clinics which deal with confirmatory diagnosis and treatment of MS cases in Kuwait. Cases suspected for MS seen in any other clinics/ hospitals in Kuwait are referred to one of these two clinics (our study sites) for confirmatory diagnosis and treatment. Subsequently confirmed MS cases are registered in Kuwait National MS Registry. Controls were selected from among the staff and students of the Kuwait University. University staff included academic members, administrative and secretarial staff. This is the only public-sector university in the country, offering free education to all enrolled students. Therefore, it was assumed that faculty, staff and enrolled students represent the typical Kuwaiti population.

A case was defined as a patient of any age, gender, nationality and resident of Kuwait with a confirmed MS, diagnosed by a neurologist using the revised 2010 McDonald Criteria [19]. Individuals with cognitive impairment or any neurological disease other than MS were excluded. Resident status in Kuwait indicates that the person is not a short-term visitor. Resident status defines that the person is either a Kuwaiti citizen or a migrant worker (with work permit) including dependent(s) of a migrant worker.

A control was defined as a resident of Kuwait without prior history of confirmed MS diagnosis. Controls were individually matched with each case for age (± 5 years), gender and nationality. Each control was given a date of pseudo-onset of MS which matched with the date of onset of corresponding MS case. Individuals with poor cognitive ability or any past or present neurological disease were excluded.

A face-to-face interview of each case and his/ her pair matched control was conducted using a structured questionnaire to collect the data on socio-demographic, potential genetic and environmental factors identified through literature review [3, 9]. Enrollment of cases and controls was carried out from July 1, 2016 to September 30, 2016. Date of MS onset was labelled as index year and history of potential exposures were assessed prior to the end of that year. Similarly, in the matched control a date of pseudo-onset of MS was constructed, which is the date of MS onset of the case. Control subjects were asked to report their exposures prior to the end of the index year (i.e. year of their pseudo-onset of MS). Questions on socio-demographic factors included age (completed years), gender, nationality, place of birth, birth order, number of siblings and relationship of parents. Participants were asked about their education level, occupation, family income, marital status and governorate of residence at the time of MS onset, and history of all countries lived in prior to disease. Age (completed years), self-reported height and weight at time of MS onset were recorded. History of vaccination, presence of other disease(s), head trauma, and self-reported vitamin D status was taken. Assessment of genetic and environmental exposures included family history of MS and other diseases including autoimmune diseases (inflammatory bowel disease, systemic lupus erythematosus, rheumatoid arthritis, Graves’ disease), presence in Kuwait during first Gulf war of 1990, history of exposure to solvents and vaccinations’ history. History of exposure to tobacco smoke and secondhand smoking was obtained from cases and controls, by asking about previous smoking status [20, 21]. Exposure to tobacco smoke for the period prior to MS onset/pseudo-onset was assessed both in cases and controls respectively. Data were gathered on average number of cigarettes/waterpipes smoked, frequency and duration of tobacco smoking. Subsequently the amount of tobacco smoked over a period of interest was calculated as pack-years. Data on exposure to secondhand smoking was obtained, by asking participants if they have been exposed to tobacco smoke at home or elsewhere. Approximate duration (years) of secondhand smoking was calculated. The questionnaire was pretested and modified if required before actual use in the study. The final study questionnaire can be found as a Additional files 1 and 2.

By taking into account 10% potential non-response and to achieve a statistical power (1-β) of 80%, this matched case-control study was designed to enroll 100 patients with MS and 100 MS-free matched controls (1:1 ratio) to relate most of the potential exposures (having a prevalence 0.25 or higher in the general population) and MS status with a matched odds ratio (mOR) 2.5 or higher using conditional logistic regression at a 5% significance level (α) [22].

Written consent was obtained from each study participant after explaining the aims and potential benefits of the study, length of questionnaire and time required to answer all the questions. Ethical approval of the study protocol was granted by Health Sciences Center Ethical Committee, Kuwait University and The Standing Committee for Coordination of Health and Medical Research, Ministry of Health, Kuwait.

Descriptive statistics including means and standard deviations of quantitative variables and frequencies (%) of qualitative variables were computed. Variables significantly (p ≤ 0.15) related with MS status on univariable conditional logistic regression analysis were considered for their possible inclusion in the final multivariable conditional logistic regression model. The variables significantly (p < 0.05) and independently associated with MS status were retained in the final model. Adjusted mOR (mOR_adj) and their 95% confidence intervals (CI), were calculated from the parameter estimates of the final model. Statistical analyses were performed using Stata 14.2. (College Station, TX: StataCorp LP).

In this case-control study, 112 MS cases were invited to participate and 110 (98.2%) consented for enrollment. One hundred and ten MS cases and 110 control subjects, individually matched with cases (1:1) on age (± 5 years), gender and nationality were enrolled (Fig. 1). The mean (SD) age (years) was 34.8 (10.3) and 34.9 (10.6) for cases and controls respectively. The mean (SD) age was 27.3 (9.0) years was nearly the same at the time of MS onset in cases or pseudo-onset in controls. Both in cases and controls, 82.7% were Kuwaiti nationals and 76.4% were females. The distributions of demographic characteristics such as marital status, monthly income (KD), governorate of residence and relative distributions of potential risk factors in cases and controls are shown in Tables 1 and 2 respectively.

Univariable conditional logistic analyses showed that cases were significantly more likely to have positive family history of MS (p = 0.002) or less likely to receive influenza virus A and B vaccine (p = 0.030). However, cases compared to their matched controls were not statistically significantly different in terms of active or passive tobacco smoking, number of siblings, parent’s relationship, presence in Kuwait during first Gulf war of 1990, history of tonsillectomy and appendectomy. Also, distributions of self-reported vitamin D deficiency, area of residency, history of chickenpox, mumps and measles infections, regular exposure to solvents, vaccination status against hepatitis B virus and measles-mumps-rubella viruses were not statistically significantly different between MS cases and their matched controls (Table 3).

Final multivariable conditional logistic regression model showed that after adjusting for the effect of educational level, the cases compared to controls were more likely to have had a positive family history of MS (mOR_adj = 5.1; 95% CI: 2.1–12.4; p < 0.001). Additionally, after taking into account the effect of educational level, cases compared to control were significantly less likely to have had received influenza virus A and B vaccine before MS onset (mOR_adj = 0.4; 95% CI: 0.2–0.8; p = 0.010) (Table 4).