The meaning of distal sensory loss and absent ankle reflexes in relation to age: a meta-analysis.
ABSTRACT Polyneuropathy is a common disease and is more prevalent (at least 3 %) in elderly people. However, routine neurological examination of healthy elderly people may show distal sensory loss and absent tendon reflexes, which can obscure the distinction from polyneuropathy.
To investigate the relation between age and the prevalence of distal sensory loss, absent tendon reflexes, or muscle weakness, and to ascertain above which age these neurological signs could be considered as normal in ageing.
PubMed, Embase, the Cochrane Library, and Current Contents from 1960 until 2004. Reference lists of relevant studies were searched for additional studies, reviews or textbooks.
Studies reporting on neurological signs upon routine neurological examination in generally healthy adult persons were considered for inclusion. Two reviewers independently assessed study eligibility and performed study inclusion. Of 629 studies initially identified, 50 (8 %) met the inclusion criteria.
Two reviewers independently performed data extraction and assessed study quality based on study design and the rigour by which confounding co-morbidity was excluded.
The 50 included studies comprised a total of 9,996 adult persons. Assuming heterogeneity between studies, the prevalence data from different studies were pooled for separate age groups with a random-effects model. In healthy persons older than 60 years the prevalence of absent vibration sense at the big toes (29 % [95 % CI 18 % to 38%]) or ankles (15 % [95 % CI 11 % to 20%]), and absent ankle reflexes (23 % [95 % CI 16 % to 30 %]) was increased.
Self-declared healthy adult persons younger than 60 years do not have neurological signs. After the age of 60 absent vibration sense at the big toes or ankles, and absent ankle reflexes are more prevalent, although the majority does not have these neurological signs. It seems more appropriate to apply different diagnostic criteria for polyneuropathy in adult persons younger and older than 60 years.
Article: The ankle jerk in older people.[show abstract] [hide abstract]
ABSTRACT: Fewer people both male and female have ankle jerks as age increases. Significantly fewer women than men have an ankle jerk in the age group 63–69.Copyright © 1972 S. Karger AG, BaselGerontologia clinica 02/1972; 14(2):86-8.
- [show abstract] [hide abstract]
ABSTRACT: Reports on the incidence of alcoholic polyneuropathies are variable depending on diagnostic tools. In this study, 50 chronic alcoholics with positive MALT (Munich Alcoholism Test) and greater than seven years history of excessive alcohol abuse were examined neurologically. Tibial and peroneal motor and sural nerve conduction velocities (NCV) were studied. Warm and cold perception was evaluated in the area behind the internal malleolus using a Somedic-Thermotest. Thresholds were determined by the method of limits. The effect of a slow, medium and fast temperature change rate on thermal perception was tested. Thirty-eight patients (76%) showed signs of neuropathy. Thermal perception was more often abnormal (62%) than NCV (42%) and clinical examination (56%). A medium temperature change rate of 2.0-2.5 degrees C/s was the most sensitive index of small fiber neuropathy. Thermal threshold measurement proved to be a reliable, sensitive and easy to perform method that should become standard in the examination of polyneuropathies.Acta Neurologica Scandinavica 06/1995; 91(5):389-93. · 2.47 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Meta-analysis may be used to estimate an overall effect across a number of similar studies. A number of statistical techniques are currently used to combine individual study results. The simplest of these is based on a fixed effects model, which assumes the true effect is the same for all studies. A random effects model, however, allows the true effect to vary across studies, with the mean true effect the parameter of interest. We consider three methods currently used for estimation within the framework of a random effects model, and illustrate them by applying each method to a collection of six studies on the effect of aspirin after myocardial infarction. These methods are compared using estimated coverage probabilities of confidence intervals for the overall effect. The techniques considered all generally have coverages below the nominal level, and in particular it is shown that the commonly used DerSimonian and Laird method does not adequately reflect the error associated with parameter estimation, especially when the number of studies is small.Statistics in Medicine 04/2001; 20(6):825-40. · 2.04 Impact Factor
THE MEANING OF DISTAL SENSORY
LOSS AND ABSENT ANKLE REFLEXES
IN RELATION TO AGE:
Journal of Neurology 2006;253:578-589
Background: Polyneuropathy is a common disease and is more prevalent (at least 3%) in
elderly people. However, routine neurological examination of healthy elderly people may
show distal sensory loss and absent tendon reflexes, which can obscure the distinction
from polyneuropathy. The aim of the study was to investigate the relation between age
and the prevalence of distal sensory loss, absent tendon reflexes, or muscle weakness, and
to ascertain above which age these neurological signs could be considered as normal in
Methods: PubMed, Embase, the Cochrane Library, and Current Contents from 1960 until
2004 were searched electronically. Reference lists of relevant studies were searched
manually for additional studies, reviews or textbooks. Studies reporting on neurological
signs upon routine neurological examination in generally healthy adult persons were
considered for inclusion. Two reviewers independently assessed study eligibility,
performed study inclusion and data extraction, and assessed study quality based on study
design and the rigour by which confounding co-morbidity was excluded. Assuming
heterogeneity between studies, the prevalence data from different studies were pooled for
separate age groups with a random-effects model.
Results: Of 629 studies initially identified, 50 (8%) met the inclusion criteria. The 50
included studies comprised a total of 9,996 adults. In healthy people older than 60 years
the prevalence of absent vibration sense at the big toes (29% [95% CI 18% to 38%]) or
ankles (15% [95% CI 11% to 20%]), and absent ankle reflexes (23% [95% CI 16% to 30%])
Conclusions: Self-declared healthy adult people younger than 60 years do not have
neurological signs. After the age of 60 absent vibration sense at the big toes or ankles, and
absent ankle reflexes are more prevalent, although the majority does not have these
neurological signs. It seems more appropriate to apply different diagnostic criteria for
polyneuropathy in adults younger and adults older than 60 years.
Polyneuropathy is a prevalent disorder that mostly occurs as a complication of chronic
conditions (e.g. diabetes mellitus, alcohol abuse, chronic renal disease). Early recognition
and diagnosis of the underlying cause is important, as treatment delays progression.1 The
incidence and prevalence of polyneuropathy vary among studies, because of differences
between study populations and different criteria to define polyneuropathy. The estimated
incidence is 30 to 200 per 100.000 persons per year.2-5 The estimated prevalence is up to
3% and in people older than 50 years it may be as high as 15%, although
electrophysiological investigations were usually not performed to confirm the diagnosis.4-14
Symptoms, reduced distal sensation, decreased muscle strength or absent tendon
reflexes may also be present in generally healthy people without polyneuropathy.10,13,15-22
Consequently, the distinction between normal ageing or polyneuropathy on merely clinical
signs, especially in elderly people without known underlying diseases or risk factors, may
not always be so straightforward.
In view of the above and the ongoing ageing population there is a need for a
clinical diagnostic definition of polyneuropathy.1 In order to develop such a definition it is
necessary to establish whether findings in the routine neurological examination might be
considered as normal or abnormal in relation to age. Previous studies investigating the
relation between the prevalence of neurological signs and age were mostly small and
confounding co-morbidity or risk factors for polyneuropathy were not always adequately
excluded. In addition, the age at which these signs are more prevalent has not been clearly
We therefore performed a meta-analysis to investigate the relation between the
prevalence of distal sensory loss, absent tendon reflexes, or muscle weakness and age, and
to ascertain whether these can be considered as signs of normal ageing (i.e. not indicative
of a polyneuropathy) in healthy people above a certain age.
Medline, Embase, the Cochrane Library, and Current Contents were searched
electronically without language restriction for studies published from 1960 until 2004 and
performed in Northern America, Europe or Australia. The following keywords as MESH
terms and text words, their synonyms and derivatives were used in relevant combinations
to guide the search: neurological examination, neurological signs, healthy control subjects,
pain sense, touch sense, vibration sense, tuning fork, position sense, tendon reflex, biceps
reflex, triceps reflex, knee reflex, ankle reflex, muscle strength. In relevant studies thus
found, we manually searched the reference lists for additional studies, reviews or
Neurological Examination & Age: Meta-Analysis
textbooks. This method of cross checking was continued until no further publications
Criteria for inclusion of studies
The study populations should have consisted of generally healthy adults who were not
selected or recruited on the basis of a normal neurological examination.
Studies that applied a routine neurological examination were considered for
inclusion. Studies that only made use of advanced examination techniques (e.g. quantitative
sensory testing methods, muscle strength testing with handheld dynamometry, isometric
or isokinetic muscle strength testing, tendon reflex electrophysiology) were excluded.
The neurological examination should have been carried out in both legs and/or
arms, and the findings should refer to the presence of symmetrical neurological signs. The
findings for each of the neurological signs should preferably have been reported separately,
but a combination was acceptable if reported as either all normal, all abnormal/decreased,
all present, or all absent.
Although criteria for abnormality could vary between studies, the findings for pain
sense, touch sense, position sense, and manual muscle strength testing were to be
dichotomized in normal and abnormal/decreased. The findings for vibration sense and
tendon reflexes (ankle, knee, biceps, or triceps) were to be dichotomized in present and
absent. Pain sense, touch sense, and muscle strength should have been described for
testing at the feet, lower legs, hands, or forearms. Position sense should have been
described for testing at big toes, index or middle fingers. Vibration sense should have been
described for testing with a 128 Hz tuning fork at the big toes, ankles, knees, or
index/middle fingers. If only part of the neurological examination was described, studies
could still be included.
Selection of studies and data extraction
Two readers (AV and FB) assessed the eligibility of studies, independently included studies,
and independently extracted relevant data onto standard forms.
If data were presented graphically or as percentages, the number of people with
normal and abnormal findings were deduced or calculated. In case of multiple publications
describing the same population, we used either the publication with the largest number of
people or the most recent publication. If a population was studied prospectively, the
neurological signs at baseline and follow-up were used separately for meta-analysis at
different ages when possible. Authors were contacted for more detailed study information
Prior to the data extraction, the age distributions of included studies were
juxtaposed to evaluate which age groups were used most frequently. It was decided that
the following age groups formed the most optimal choice for data extraction and
subsequent meta-analysis: 18 to 39 years, 40 to 59 years, 60 to 79 years, and 80 years or
older (80+ years). In order to avoid the exclusion of studies in which age-specific findings
were not given in detail, the following age groups were defined as well: 18 to 64 years, and
65 years or older (65+ years). To investigate the prevalence of neurological signs with
increasing age more thoroughly, data were also extracted for additional age groups 60
years or older (60+ years) and 70 years or older (70+ years). A 5-year leniency below or
above the limits of the age groups just described was permitted for studies whose age
distribution did not exactly comply.
Assessment of the methodological quality of studies
The methodological quality of studies was graded similar to the OCEBM classification of
levels of evidence.23 Studies attaining a grade of class I were judged to have the lowest risk
of bias, and studies attaining a grade of class IV were judged to have the highest risk of
bias. We also decided beforehand to perform a sensitivity analysis by including only those
studies in which people without relevant co-morbidity were examined. Thus, studies were
labelled as ‘restricted’ if conditions and diseases that cause or are associated with
polyneuropathy were sufficiently excluded by medical history taking prior to the
neurological examination; otherwise studies were labelled ‘non-restricted’. Appropriate
laboratory investigations were not necessary to support the absence of confounding co-
morbidity, because we wanted to know which findings upon routine neurological
examination of a self-declared healthy person unknown to have concomitant diseases or
risk factors may be present in relation to age.
In case of disagreement on study inclusion, data extraction or study quality
assessment, the two readers reviewed the study together to reach consensus. If consensus
could not be achieved a decision was made with third party adjudication (NN or GR).
The data on the prevalence of neurological signs from the different studies were pooled
for separate age groups with a random-effects model with maximum likelihood estimation,
because we assumed heterogeneity between the studies. A random-effects model
incorporates the variation between studies. It is assumed that each study has its own
(true) exposure effect and that there is a random distribution of these exposure effects
around a central effect. The observed effects from the different studies are used to
estimate this distribution. In this model, study-specific weights are calculated as a sum of
the variance within a study and a term for the variance between the studies. The pooled
prevalence estimates and their 95%-confidence intervals were calculated for each
neurological sign in separate age groups. The meta-analysis was performed by using the
procedure Proc Mixed in the statistical computer programme SAS (version 8).24
If just one study reported on the prevalence of a neurological sign for a certain
age group, then the observed prevalence from that study was taken as an estimate of the
population prevalence and its 95% confidence interval was calculated according to the
Neurological Examination & Age: Meta-Analysis