Sample description and missing data
Of the 10,308 participants at Phase 1 (1985–1988), 7,830 participated in at least one part of Phase 5 (1997–1999), 2,204 were non responders and 274 dead (). At Phase 5, data on cognitive function, smoking history and all covariates were available for 5,388 respondents. Compared to baseline, this group was younger (55.5 years versus 56.1 years) and composed of fewer women (27.6% versus 33.1%) and fewer low socioeconomic position participants (14.6% versus 22.7%) (p < 0.001). From this population, calculation of cognitive decline, implying participation in cognitive tests at Phase 7, was possible for 4,659 participants (). Here again, missing data were similarly influenced by age, gender and socioeconomic position compared to data available for analysis on cognitive deficit (N=5,388).
In order to assess whether the smoking-cognition association is underestimated due to premature mortality among smokers, we examined the association between smoking status at Phase 1 and mortality during the 17.1 (standard deviation=2.3) years of follow-up till Phase 7 (). “Current smokers” at Phase 1 had a higher risk of dying during follow-up compared to “never smokers” after adjustment for age, socioeconomic position and marital status among men (Hazard Ratio (HR), 2.00 (95% confidence interval (CI), 1.58–2.52)) and women (HR, 2.46 (1.80–3.37)). Ex-smokers at Phase 1 did not have a higher risk of death during the period of follow-up examined (HR, 1.09 (0.84–1.41) among men, HR, 1.23 (0.84–1.79) among women). Among survivors at Phase 7 (N=9,625), we examined the association between smoking status at Phase 1 and non-participation in the cognitive tests at Phase 7. In analyses adjusted for age, socioeconomic position and marital status, “current smokers” at Phase 1 were more likely to be non-participants among men (Odds Ratio (OR), 1.32 (1.16–1.51)) and women (OR, 1.69 (1.41–2.02)). In order to examine the persistence of this association, we repeated the analysis with smoking history at Phase 5 and participation in cognitive tests at Phase 7 (N=7,221). Greater numbers of both male (OR, 1.47 (1.20–1.81)) and female smokers (OR, 1.81 (1.35–2.43)) did not undertake the cognitive tests. “Long-term ex-smokers” and “recent ex-smokers” at Phase 5 were not different from “never smokers”.
Association between smoking and mortality and non-participation (2002–2004).
Characteristics of individuals included in the analyses on smoking and cognitive deficit at Phase 5 are shown in . The test for trend shows that smoking status was associated with socioeconomic position, education, alcohol and fruit and vegetable consumption (p < 0.0001). Prevalence of CHD, stroke and diabetes was not associated with smoking history. Among the vascular risk factors, smoking history was associated only with cholesterol (p < 0.0001). Cognitive scores at Phase 5 as a function of health measures are presented in .
Characteristics of the study population at Phase 5 (1997–1999).*
Cognitive function (mean (standard deviation)) as a function of health measures at Phase 5.*
Smoking history and cognitive function at Phase 5
The fully-adjusted mixed-effects model showed that smoking history was associated with memory (p=0.01), reasoning (p=0.0004), vocabulary (p<0.0001), phonemic (p<0.0001), and semantic fluency (p=0.0009). presents results of the logistic regression using binary cognitive outcomes; the sex-specific cut-offs used are also shown. In age- and sex-adjusted models, “current smokers” were more likely to have cognitive deficits on all tests: memory (OR, 1.54 (1.25–1.90)), AH4-I (OR, 1.53 (1.27–1.85)), Mill Hill (OR, 1.42 (1.18–1.70)), phonemic (OR, 1.32 (1.09–1.60)) and semantic fluency (OR, 1.30 (1.08–1.57)). In fully adjusted models, the association remained for memory (OR, 1.37 (1.10–1.73)). Compared to “never smokers”, the “long-term ex-smokers” were less likely to have deficits in memory (OR, 0.79 (0.65–0.96)), the Mill Hill (OR, 0.73 (0.60–0.87)), phonemic (OR, 0.73 (0.61–0.87)) and semantic fluency (OR, 0.75 (0.63–0.89)) in fully adjusted models. “Recent ex-smokers” also had a reduced risk of poor vocabulary score (OR, 0.65 (0.49–0.85)) and semantic fluency (OR, 0.72 (0.55–0.94)).
Odds ratio of being in the worst quintile of cognitive function at Phase 5 as a function of smoking status (1997–1999), N=5,388.†
Among current smokers at phase 5, in fully adjusted models, there was no evidence of a dose-response association between pack-years of smoking and cognitive deficit (memory, p=0.97; AH4-I, p=0.13; Mill Hill, p=0.33; phonemic, p=0.25; semantic fluency, p=0.97).
Smoking history and cognitive decline between Phases 5 & 7
The interaction term between smoking history and time in the fully-adjusted mixed-effects model showed that smoking history was associated with cognitive decline in reasoning (p=0.0004), but not with memory (p=0.64), vocabulary (p=0.68), phonemic (p=0.63), and semantic fluency (p=0.61); detailed results shown in the Appendix. Further analysis on decline () uses the worst quintile of change, implying decrease greater than 1 point for memory and the Mill Hill, 7 points for the AH4-I and 3 points for the fluency measures. In fully adjusted models, both “current smokers” (OR, 1.40 (1.11–1.75)) and “recent ex-smokers” (OR, 1.38 (1.07–1.77)) were more likely to decline on the AH4-I. No other association was evident. Further adjustment for health behaviours at Phase 7 did not much change these results.
Odds ratio of being in the worst quintile of change in cognitive function between Phase 5 (1997–1999) and Phase 7 (2002–2004), N=4,659.
Among current smokers at Phase 5, in fully adjusted models, there was no dose-response association between pack-years of smoking and cognitive decline (memory, p=0.22; AH4-I, p=0.88; Mill Hill, p=0.54; phonemic, p=0.30; semantic fluency, p=0.94).
This analysis was aimed at the exploration of changes in other health behaviours along with change in smoking status (giving up smoking) over the follow-up period. Those who stopped smoking between Phases 1 and 5 (“recent ex-smokers”) had the smallest increase in consumption of alcohol between Phases 1 and 7 (0.82 g of alcohol per week) compared to the other groups (1.46 g among “never smokers”). In terms of healthy eating, the percentage of participants consuming at least one fruit or vegetable per day increased more among “recent ex-smokers” than among “never smokers”. shows that “recent ex-smokers” were at the same level of fruit and vegetable consumption as “current smokers” at Phase 1 but by Phase 7 they had reached the same level as “long-term ex-smokers” and “never smokers”.
Participants (%) consuming at least one fruit or vegetable per day as a function of smoking history at Phase 5.