Table 1.

Quality assessment of cohort studies on obesity and dementia risk

Study	1	2	3	4	5	6	7	8	9	10
Yoshitake et al. [47]	★	★	★	★	★			★	★	★
Borenstein et al. [44]	★	★	★	★	★		★	★	★	★
Gustafson et al. [43]	★	★	★	★	★	★	★	★	★	★
Nourhashemi et al. [29]	★	★		★	★		★
Buchman et al. [50]	★	★	★	★	★		★	★	★	★
Hayden et al. [41]	★	★		★	★		★	★	★	★
Lunchsinger et al. [33]	★	★	★	★	★		★	★	★	★
Atti et al. [30]	★	★	★	★	★		★	★	★	★
Dahl et al. [31]	★	★	★	★	★	★	★	★	★	★
Hughes et al. [32]	★	★	★	★	★	★	★	★	★	★
Fitzpatrick et al. [35]	★	★	★	★		★	★	★	★	★
Scarmeas et al. [46]	★	★	★	★	★	★	★	★	★	★
Power et al. [34]	★	★	★	★		★	★	★	★	★
Lucca et al. [28]	★	★		★	★	★	★	★
Tolppanen et al. [52]	★	★	★	★	★	★	★	★		★
Neergaard et al. [42]	★	★	★	★		★	★	★	★	★

(1) Cohort truly representative; (2) Controls from the same cohort; (3) Clear measurement of obesity at baseline; (4) Adequacy of Follow-up duration (≥24 months); (5) Reliable methods of dementia and AD diagnosis (i.e., Quality of outcome); (6) Data analysis controlled for smoking and medical co-morbidities; (7) Data analysis controlled for any other three confounders (Age, social class/education, alcohol, ApoE4 carrier status, medical therapies and ethnicity etc.); (8) Findings interpreted well; (9) Weakness mentioned and explained clearly; (10) Paper written well.