Discussion

In the present study, we identified a similar mortality rate (1.225, 95% CI 0.944–1.563) for patients with AD compared to the general population over a 10-year follow-up period and also found that lower MMSE score and comorbid DM are predictors for poor survival. Although some previous studies indicated differential mortality of patients with dementia compared to the general population,^{Reference Li, Shen and Chen1,Reference Cereda, Pedrolli and Zagami5} our present study supports a nominally normal survival outcome despite an AD diagnosis during prior decade in Shanghai.

Several reasons can potentially explain why our current study demonstrates no elevated mortality for AD patients compared to various prior studies. (1) First, the diagnostic criteria used and rigorous clinical diagnosis in the present study exclude other types of dementia, such as frontotemporal dementia and vascular dementia. Indeed, according to a previous population-based study,^{Reference Garciaptacek, Farahmand and Religa6} non-AD dementias are associated with shorter survival and a higher risk of death than associated with AD. (2) Second, patients in this study were urban residents, a majority having a high level of education, and such persons tend to be sensitive to memory decline and to seek medical attention actively. Recently, a study by Zhou et al.^{Reference Zhou, Wang and Zeng7} also found that age-standardized years of life lost were significantly lower than the national mean for major causes including AD in Shanghai. Meanwhile, participants in our study are mostly from urban Shanghai. It is reasonable that patients from the urban, coastal region in eastern China generally have a better health outcome than individuals from other regions. In addition, in that population-based study, age-standardized disability-adjusted life years (DALYs) for AD and other dementias were significantly lower than expected in Shanghai in 2017, in which the ratio of observed to expected difference in DALYs was only 0.69.^{Reference Zhou, Wang and Zeng7} (3) Furthermore, most previous studies were conducted in the 1990s. Improved quality of care in clinical practice and better awareness of disease might play a role in the observed reduced mortality in our cohort. As practitioners, the authors indeed note that there are relatively higher levels of diagnosis, treatment, and improved nursing standards for AD patients in Shanghai during the most recent two decades, compared with the two prior decades.

The present study confirms previous studies that found shorter survival time for individuals with AD and comorbid type 2 DM.^{Reference Helzner, Scarmeas and Cosentino8–Reference Imfeld, Bodmer and Jick11} A previous study determined that the duration of diabetes was an important modifier of dementia onset and survival.^{Reference Zilkens, Davis and Spilsbury9} Moreover, a history of diabetes may be associated with a faster annual rate of cognitive decline in AD patients.^{Reference Helzner, Luchsinger and Scarmeas12} DM seems to promote specific neuropathologic processes that contribute to dementia, concurrent with cerebral atrophy, lacunar infarcts, and white matter lesions.^{Reference Jongen, Grond and Kappelle13,Reference van Harten, de Leeuw and Weinstein14} However, the evidence for an association between DM and cognitive decline in AD patients is contradicted in other studies.^{Reference Regan, Katona and Walker15,Reference Mielke, Rosenberg and Tschanz16} One hypothesis is that diabetic patients may have more frequently received cardiovascular risk-modifying medications or insulin-sensitizing agents, which might contribute to a slower progression of cognitive decline in patients with AD.^{Reference Mielke, Rosenberg and Tschanz16} Notably, both AD and DM are chronic diseases that may share common pathologic features,^{Reference Knowles, Vendruscolo and Dobson17} where there is mounting evidence of brain glucose dysregulation in AD.^{Reference An, Varma and Varma18} In our view, insulin resistance and abnormal glucose metabolism of pre-existing type 2 DM patients may influence insulin sensitivity in brain and aggravate β-amyloid accumulation^{Reference Han and Li19} and, consequently, accelerate the progression of AD.

Consistent with previous studies, lower MMSE score indicated a higher risk of death.^{Reference Go, Lee and Seo10,Reference Zhao, Zhou and Ding20} Notably, patients with faster cognitive decline tend to have significantly lower MMSE score at the time of diagnosis of AD.^{Reference Carcaillon, Pérès and Péré21} Another cohort study on mortality risk after diagnosis of various types of dementia demonstrated an association between worse baseline cognition (as represented by MMSE) and higher risk of death.^{Reference Garciaptacek, Farahmand and Religa6} Interestingly, both age at onset and age at diagnosis are not associated with mortality in AD in our study.

In the present study, sex and education levels were not significant risk factors impacting AD patient survival. Although we presently found that AD increased mortality risk more significantly among women (SMRs of female AD patients are 1.593 (95% CI 1.129–2.183)), this controversial result could be related to the fact that males in general already have an elevated mortality risk and a reduced life expectancy compared to females. For this reason, significant sex differences of survival have not been observed, although our results are consistent with a 15-year epidemiological study.^{Reference Ganguli, Dodge and Shen22}