Demographic characteristics, smoking and alcohol consumption status, immune responses, levels of inflammatory markers, and Ig levels in the entire study population and in the different treatment groups are summarised in the Table .
Smoking,alcohol drinkingand antibody response following pneumococcal vaccination in entire study population/different groups
At vaccination, RA patients on anti-TNF+MTX had significantly lower CRP compared to the other two RA treatment groups (P = 0.019 and P = 0.009, compared to group I and group II, respectively). IgG, IgA and IgM levels were significantly lower in RA patients on MTX (group I) compared to RA patients on anti-TNF as monotherapy or RA on anti-TNF +MTX (P between 0.032 and <0.001; independent sample T-test).
SpA patients on anti-TNF+MTX had lower CRP compared to the other SpA treatment groups (P = 0.001 and <0.001, compared to group IV and VI, respectively). No statistically significant differences of Ig levels were seen among groups of patients with SpA.
The effect of cigarette smoking in the total study population
Of 505 patients enrolled in the study, 290 (57.4%) smoked for some period of their life or still smoke (ever smokers). The total smoking load expressed in number of smoking pack-years (mean; SD; range) was 19.3 (14; 0 to 65). In total, 212 patients (42%) had never smoked and no data on smoking status were available in 3 patients (0.6%). Two hundred, two patients (40%) had previously smoked but quit smoking before the study entry (ex-smokers). Mean (SD; range) of smoking pack-years in this group was 16.6 (13; 1 to 65).
In all, 88 patients (17.4%) of the population studied were current smokers at the time of vaccination. Mean (SD; range) number of cigarettes per day was 10.1 (6.4; 0 to 30). Mean (SD; range) number of smoking pack-years was 24.5 (13; 0 to 57). Of all current smokers, 64 individuals (72.7%) were females and 24 (27.3%) males (P = 0.029).
Differences in measured markers of inflammation, IgG levels and GMC of immune responses for each serotype between non-smokers, ex-smokers and current smokers are given in Table .
Markers of inflammation, Ig and immune response according to smoking status in total study population
Current smokers had significantly higher CRP and ESR compared to non-smokers (Table ). Number of pack-years of smoking was significantly correlated with both CRP (Spearman's rho = 0.152; P = 0.001) and ESR (Spearman's rho = 0.159, P = 0.001). The findings persisted in univariate ANCOVA with age, gender, number of tender and swollen joints in 28-joint count index, MTX, anti-TNF treatment, prednisolone and alcohol consumption included in the analysis. Smoking had a significant impact on both CRP (P = 0.008) and ESR (P = 0.037).
At the time of vaccination, current smokers had significantly lower total IgG levels compared to non-smokers but IgA and IgM levels did not differ significantly. Both daily cigarette number at the time of vaccination and number of pack-years were inversely correlated with IgG levels (Spearman's rho -0.132 (P = 0.003) and -0.130 (P = 0.006), respectively). These findings were confirmed by ANCOVA with age, gender, DAS, MTX, anti-TNF prednisolone treatment and alcohol consumption as covariates.
GMC of antibody responses, that is, the ratio between post- and pre-vaccination antibody levels, were significantly higher in non-smokers for both serotypes (Table ). This was not corroborated by univariate ANCOVA after adjustment for demographic, disease and treatment characteristics. Neither did smoking status predict a posIR for each serotype separately nor both 23F and 6B together. Number of pack-years did not correlate with antibody levels or IR and did not predict a posIR for either serotype or both serotypes together.
Effect of cigarette smoking within different diagnostic and treatment groups
The percentage of current smokers in different treatments groups were: 19, 18, 19, 12, 24 and 12% (Table ). The proportion of current smokers did not differ significantly between the groups of patients with RA. Among patients with SpA more patients on anti-TNF+MTX were smokers (P = 0.05 and 0.04, compared to group 4 and 6, respectively).
Group I (RA on MTX)
In RA patients on MTX, current smokers had significantly higher CRP and ESR. This was also demonstrated using univariate ANCOVA, where age, gender, number of tender and swollen joints in 28-joint count index were included in the analysis (P = 0.023 for CRP and P = 0.023 for ESR). However, total IgG, IgA and IgM levels did not differ between smokers and non-smokers in this treatment group. Ever-smokers among RA patients on MTX had a lower proportion of responders to vaccination for both serotypes (P = 0.025; OR 0.21; 95% CI 0.05 to 0.81) compared to never-smokers (Figure ). Also, higher pack-years were associated with lower posIR (P = 0.046; OR 0.93 95% CI 0.86 to 0.99) after adjustment for age, sex, CRP, counts of affected joints (tender joint count (TJC) and swollen joint count (SJC)) and pre-vaccination antibody levels (Table ).
Figure 1 Responders(%) to 23F, 6B and both pneumococcal serotypes in methotrexate treated RA patients according to smoking status. Responders i.e. patients with positive immune response defined as post-vaccination antibody levels ≥2 times pre-vaccination (more ...)
Impact of smoking on positive immune response forpneumococcal serotypes(23F and 6B) in RA patients on methotrexate
Groups II to VI
Smoking habits had no effect on acute phase reactants, Ig levels, pre- and post-vaccination antibody levels, IR or posIR against pneumococcal serotypes 6B or 23F in any of the other treatment groups.
Effect of alcohol drinking in the total study population
Of the participating 505 patients, 148 (29.3%) reported no or sporadic alcohol consumption in very small amounts, while 333 (65.9%) patients reported alcohol consumption on a regular basis. Median (range) consumption among patients using alcoholic beverages was 70.8 (5.3 to 758) g/week. The majority of patients participating in the study reported low to moderate alcohol use corresponding to ≤30 g 100% ethanol daily. Alcohol consumption was more common in men (P <0.001).
Compared to non-drinkers, patients regularly consuming alcohol showed lower levels of CRP and ESR (P = 0.007 and P <0.001; T-test) also remaining after univariate ANCOVA, including age, gender, number of tender and swollen joints in 28-joint count index, and after adjustment for multiple comparisons (P = 0.05 and P = 0.003, for CRP and ESR, respectively).
However, serum Ig levels, GMC of antibody responses or posIR were not influenced by the reported alcohol consumption.
Effect of alcohol drinking within different diagnostic and treatment groups
Within each treatment group, no significant differences in CRP, ESR, Ig levels, GMC of antibody responses or pos IR were detected among persons not consuming alcoholic beverages compared to those drinking alcohol.