Table 1

Study characteristics

Publication and study designParticipants, country, setting and length of follow-upSedentary behaviour (exposure assessment)Cognitive function (outcome assessment)Results
Cohort designs
Hamer and Stamatakis23
Cohort design
6359 men and women from the English Longitudinal Study of Ageing
2-year follow-up
Self-reported TV viewing considered sedentary behaviour (SB)Immediate word recall, delayed word recall and verbal fluency.31 All three used to create a global cognitive function score (primary outcome)Linear inverse relationship between TV time and cognitive function. Decreased cognition from baseline (EMM=0.39, 95% CI [0.33 to 0.45)) to follow-up (EMM=0.25, 95% CI (0.19 to 0.31)), but no association between baseline SB and changes in cognitive function
Kesse-Guyot et al24
Cohort design
2430 participants from the Supplémentation en Vitamines et Minéraux Antioxydants Study
13-year follow-up
Self-administered French version of the Modifiable Activity Questionnaire (MAQ).32 Participants reported average time spent at home watching TV (min/day)Digit span forward and backward (primary outcome),33 Delis-Kaplan Trail Making Test,34 RI-48 cued recall test,35 semantic fluency and phonemic fluency36SB associated with decreased global cognitive function (β= −1.28; 95% CI (−2.46 to −0.11)) and decreased verbal memory (β= −1.38; 95% CI (−2.58 to −0.18)) over time
Kesse-Guyot et al25
Cohort design
2579 participants who agreed to participate in the follow-up period of the Supplémentation en Vitamines et Minéraux Antioxydants Study
8-year follow-up
Self-administered French Modifiable Activity Questionnaire (MAQ).32 Participants asked about average daily time spent with SB (min/day)Phonemic and semantic fluency (primary outcome),36 RI-48 test,35 digit span forward and backward,33 Delis-Kaplan Trail Making Test34Negative association observed between TV viewing and executive function cross-sectionally (β=−0.98; 95% CI (−1.93 to −0.04)), no association between executive function and SB over time
Case-control designs
Kivipelto et al26
Nested case-control design
1449 participants from the Cardiovascular Risk Factors, Aging and Dementia Study (65–79 years)
Mean follow-up time of 21 years
Self-reported leisure-time physical activity (PA) dichotomised into categories: active and sedentary (persons who participated in leisure-time PA less than two times per week)Cognitive status determined via scores on the Mini-Mental State Examination (MMSE),37 and all-cause dementia diagnosis (primary outcome) confirmed according to the Diagnostic and Statistical Manual of Mental Disorders38The odds of developing all-cause dementia were 2.07 times greater for participants who were sedentary (95% CI 1.12 to 3.86) as compared to physically active when controlling for age, sex, follow-up time, education, body mass index (BMI), cholesterol, blood pressure, heart attack, stroke and diabetes
Lindstrom et al27
Case-control design
Participants born between 1915 and 1944. 135 cases of Alzheimer's disease
331 controls recruited from clinical settings and from the community.
Participants self-reported daily hours of television viewingDiagnosed case of Alzheimer's disease (primary outcome)Cases watched significantly more television than controls (F (1, 464)=35.37). The odds of developing Alzheimer's disease increased 1.32 times for every hour of daily television viewing (95% CI 1.08 to 1.62)
Cross-sectional designs
Rosenberg et al28
Cross-sectional design
307 older adults (67–100 years) from 11 retirement communities
Self-reported SB assessed using a modified version of the Sedentary Behaviour Questionnaire.39 Objective sedentary time measured using ActiGraph GT3X+ accelerometer40Trail Making Test41Self-reported sedentary time was associated with improved performance on Trails A (β= −0.01±0.01) but was not associated with improved executive performance. Objectively measured sedentary time was not associated with Trail Making Test performance
Steinberg et al29
Cross-sectional design
125 healthy adults 65 or older with no clinical evidence of cognitive impairment
Hours spent in SB according to the Community Health Activities Model Program for Seniors (CHAMPS) questionnaire42CogState computerised battery measured multiple domains of cognition including: psychomotor speed, visual attention, visual recognition and memory (primary outcome)43Lower scores on executive function measures associated with increased SB (β=0.006±0.003; R2=0.2323). Memory scores and processing speed were not associated with increased SB.
Vance et al30
Cross-sectional design
158 participants with a mean age of 75.05 years were recruited from the Accelerate study
The total amount of time spent sitting, sleeping or lying down was used as an indicator of SB44Benton Visual Retention Test,45 Trail Making Test41 and the Rey-Osterrieth Complex Figure Copy and Recall Tests.46 A composite score for cognitive function was then created (primary outcome)Structural equation modelling predicted SB was associated with increased cognitive function (β=0.34)