Review: Metabolic Syndrome and Risk for Vascular Dementia

The featured paper is ‘Metabolic Syndrome and Risk for Incident Alzheimer’s Disease or Vascular Dementia. The Three-City Study’ by Raffaitin and colleagues and freely available here.

As the title suggests, the aim of this study was to investigate the relationship between the metabolic syndrome and two types of dementia. Participants were obtained from an epidemiological study – ‘Three City Study’. Those in institutions are excluded from the study (although the characteristics of the institutions is not characterised). This means that the sample population may have been slightly ‘healthier’ than the general population. A number of further exclusions were then made on the basis of people already having dementia or insufficient information being available for baseline metabolic syndrome identification. Again the characteristics of this excluded sample are not given in this section and so it is difficult to work out the impact this might have had on the sample population other than to speculate that those excluded may already have had a vascular dementia which could have deteriorated over the course of the study (and therefore added potentially useful information about the relation of the risk factors to the evolving pathology) or may have had another type of dementia which then developed into a mixed picture again providing potentially useful information. Given the nature of the study question and the scale of the study this is probably necessary anyway. We dont have the characteristics of the subjects who dropped out of the study for various reasons and whilst this might represent a random sample of the population there may equally be a defining feature of this sample which interacts with the main study variables e.g. were they too unwell to participate?

While the MMSE was used and the number of measures is unclear, DSM-IV criteria are used with a consensus on diagnosis being obtained by a group of neurologists with the usual work-up for dementia and then further subclassification according to established criteria. Metabolic syndrome was diagnosed using established criteria and the investigations used are detailed. There is a description of confounding variables that are also identified including APOE4 status and sociodemographic data. The authors then identify the t-test and Pearson’s test for the different data types. However we are not given the justification for using these tests – for instance is the continuous data normally distributed – it most probably is but it would be interesting to see the justification. Proportional hazards models were then used to identify the relationships between variables. Similarly in the method section there is no justification for supposing that the proportional hazards assumption holds with the study sample.

In the results section it is stated that 651 people did not have follow-up. This is considered as a homogenous group although as above we see that there are various subpopulations within the sample. However we are informed there was no significant difference between those who were not followed-up and the sample population. 208 cases of incident dementia developed at follow-up, a relatively small number compared to the overall sample size. There was no significant relationship between metabolic syndrome and dementia although perhaps a larger sample size was needed to identify a relationship. An analysis (presumably secondary) between the components of the metabolic syndrome and dementia revealed high triglyceride levels as being significantly associated although this must be assessed in terms of the absent primary outcome. There wasn’t a significant relationship between metabolic syndrome and Alzheimer’s Disease. However a positive association was found between Vascular Dementia and Metabolic syndrome after adjusting for APOE4 status. High triglyceride levels were also significantly associated with VaD. Diabetes was also significantly associated with VaD and all-cause dementia.

The authors concluded that the metabolic syndrome didn’t have significant value in stratifying risk. In effect they are suggesting that the individual components of the metabolic syndrome are more useful. They have highlighted the relationship between the triglycerides and VaD and all-cause dementia as worth pursuing. However, they have identified a significant relationship between the metabolic syndrome and vascular dementia so it could be argued that there is some utility in this concept for the purposes of identifying risk.

While it would be useful to have further information about those excluded from the study in order to translate these findings to other populations, this is a large study with rigorous methods for identifying caseness and with some useful results which can be explored further.

Steps to Treatment = 3

Steps To Treatment (STT)

STT = Steps To Treatment. An estimate of the number of steps between the results and translation into practice i.e. treatment. This is an opinion.


If you have any comments, you can leave them below or alternatively e-mail


The comments made here represent the opinions of the author and do not represent the profession or any body/organisation. The comments made here are not meant as a source of medical advice and those seeking medical advice are advised to consult with their own doctor. The author is not responsible for the contents of any external sites that are linked to in this blog.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s