Intended for healthcare professionals


Association between mid-life marital status and cognitive function in later life: population based cohort study

BMJ 2009; 339 doi: (Published 02 July 2009) Cite this as: BMJ 2009;339:b2462

Re: On mid-life marital status and cognitive function in later life: notes inspired by a journal club discussion

Adrian Loerbroks and Nasir Omar present some very interesting ideas
from a discussion on our article about marital status and cognitive
impairment.(1) First we would like to point out that the outcome variables
we used were diagnoses of mild cognitive impairment and dementia,
including results specifically for Alzheimer’s disease. In other words, we
did not use the level of non-pathological cognitive performance as outcome
in this study. In addition the participants were relatively young,
averagely 50.4 years at baseline. At the follow-up averagely 21 years
later we used a random sample of survivors from the baseline measurements.
These methodological features have important implications for the first
issue raised in the comment: that we should have scanned participants for
possible dementia already at baseline. We have not found reliable
estimations for prevalence of dementia before the age of 60, but in the
age group 60-64, the probably most reliable estimation is 0.9%.(2) Based
on these estimations we feel safe in concluding that the prevalence of
pathological cognitive impairment in our baseline sample must have been
very close to 0. Another fact to consider is the median survival time
after a diagnosis of either vascular dementia, possible Alzheimer’s
disease or probable Alzheimer’s disease has been estimated to be 5.7 (4.54
to 6.86) years for the youngest group (65-74 years). For older persons,
the corresponding expected survival period was even shorter.(3) So even if
one or two of our subjects would have suffered from cognitive impairment
already at baseline, how likely is it that any of them would have survived
to be included in the follow-up, and thus in our study, averagely 21 years

Another issue brought up by Loerbrok and Umar concerns the
preclinical period and how it relates to reverse causation. We regret if
we in some instances in the article used “sub-clinical” and “preclinical”
synonymously, while the more adequate term should be sub-clinical in this
case. We also think it is important to distinguish between what is sub-
clinical and what is a risk factor in discussing reverse causation: A
factor that increases the risk of triggering a disease process should be
referred to as a risk factor and the changes that occur as a result of the
disease process once it has started, but before the diagnostic criteria
are fulfilled, should be referred to as a sub-clinical. Applying this
distinction to Snowdons findings,(4) referred to by Loerbrok and Umar,
means that certain cognitive/linguistic characteristics early in life
should be regarded as possible risk factors for the development of
dementia later in life, not as a sign of an underlying disease process
that has already started. We think that careful reading of the referred
article by Elias et al,(5) and especially the editorial comment to the
same article,(6) supports the distinction we like to make. When we say
that reverse causation is unlikely with a 21-year follow-up, this refers
to changes during the sub-clinical phase (the latency phase) that might
have affected behaviour already at baseline. We agree with the authors(5)
and the editorial comment(6) that the findings by Elias et al(5) cannot be
taken as evidence for sub-clinical changes 22 years before a diagnosis of

How the quality of a couple relationship contributes to the
potentially protective effect of living with someone is another
interesting idea brought up in this comment. We are currently working on
trying to identify such factors and hope to be able to report some
interesting results on this issue in the near future.

As for the last two points, we agree that the emotional consequences
of losing a partner may be one of the critical and common dimensions for
divorced and widowed in relation to singles. Another factor may be that
singles form long-term social networks and activity patterns based on a
single life style, while divorced and widowed might need to partly
restructure their networks and activity patterns after having lost their

1. Hakansson K, Rovio S, Helkala E-L, Vilska A-R, Winblad B, Soininen
H, Nissinen A, Mohammed AH, Kivipelto M. Association between mid-life
marital status and cognitive function in later life: population based
cohort study. BMJ. 2009;339(jul02_2):b2462-.

2. Ferri CP, Prince M, Brayne C, Brodaty H, Fratiglioni L, Ganguli M,
Hall K, Hasegawa K, Hendrie H, Huang Y, Jorm A, Mathers C, Menezes PR,
Rimmer E, Scazufca M. Global prevalence of dementia: a Delphi consensus
study. Lancet (British edition). 2005;366(9503):2112-2117.

3. Wolfson C, Wolfson DB, Asgharian M, M'Lan CE, Ostbye T, Rockwood
K, Hogan DB. A reevaluation of the duration of survival after the onset of
dementia. New England Journal of Medicine. 2001;344(15):1111-1116.

4. Snowdon DA, Kemper SJ, Mortimer JA, Greiner LH, Wekstein DR,
Markesbery WR. Linguistic ability in early life and cognitive function and
Alzheimer's disease in late life. Findings from the Nun Study. JAMA.

5. Elias MF, Beiser A, Wolf PA, Au R, White RF, D'Agostino RB. The
preclinical phase of Alzheimer disease - A 22-year prospective study of
the Framingham cohort. Archives of Neurology. 2000;57(6):808-813.

6. Mayeux R, Small SA. Finding the beginning or predicting the
future? Archives of Neurology. 2000;57(6):783-784.

Competing interests:
None declared

Competing interests: No competing interests

10 August 2009
Krister Håkansson
research fellow
Miia Kivipelto
Växjö University, 35195 Växjö and the Karolinska Institutet, 14186 Stockholm, Sweden