Please enable JavaScript.

Coggle requires JavaScript to display documents.

CLSA studies - Coggle Diagram

- - - - ❇️ I would add more memory tests at least: example of those would be number memory, visual memory (these were validated and are actually used in the UK Biobank). Visual memory test (lights pattern memory test): which window were lit after 10 sec. Numeric memory test (shown 2 digits, they disappear, they need to enter them in computer; then they need to recall one more and so on, up to 12 digits). In the CLSA, the memory test was also modified to take less time but it may be important to keep the original version that was validated. The first 2 plates of the stroop test could be used as measures of psychomotor speed.
    - - ❇️ (we see the mean increase as well on the CLSA portal) I believe that it is a retest effect, maybe if instructions were not the same. Participants may have remembered words or accessed the words on internet. We would typically expect a decrease in memory overtime, especially over 3 years. Usually 2 different set of words should be used and participants should be randomized to start with one/other. Another possible explanation could have been lost to follow-up that includes those who had lowest cognitive performance but the other domains performance decreased...
    - - Because they have lower cognitive reserves. sets an individual up for a lifetime of advantageous experiences, both cognitive (e.g., cognitively-demanding occupation and leisure activities) and non-cognitive (e.g., higher income, better quality neighborhoods, increased access to high-quality healthcare).
      - Linked with healthier lifestyle and higher educatin level = greeater cognitive reserve which may be protective of cognitive decline.
    - - Grip strength was significant for all cog domains. It has been extensively studied in observational studies and results have shown that strength (gait speed as well) are good predictors of cognitive impairment incidence (McGrath et al. J Alzheimer's disease 2019 - showed that a 5 kg decrease in handgrip strength was linked to 10% greater odds of having any cognitive impairment). The link may be differently explained vs the one with lean mass. Strength is reflective of the nervous system, directly linked to the brain and cognitive functioning.
      - it was observed that a decrease in hippocampal volume is associated with worsening of memory performance [21,22,23]. Conversely, an increase in hippocampal volume after a yearlong aerobic training intervention was associated with memory improvements. https://eurapa.biomedcentral.com/articles/10.1186/s11556-019-0217-2
    - - Why a specific association with executive functions?
        
        Biological answer
        
        Maybe those with greater MM engage in resistance training (there was a question on that as part of the PASE, but would have to more extensively assess this to investigate this as a mediator). Resistance training was shown to evoke functional brain changes especially in the frontal lobe which were accompanied by improvements in executive functions. https://eurapa.biomedcentral.com/articles/10.1186/s11556-019-0217-2
        
        But physical activity no longer being associated with executive functions shows that it may rather be lean mass and strength that explain/mediate the relationship between phys act and cognitive decline.
        
        Stats answer
        
        more tests leading to bias. Less tests and more variability in other cog domains. This could be addressed with more tests in other domains. Maybe then considering a weighted composite score would be worth to attribute appropriate weight to each test of the construct cognitive domains.
    - - ❇️ Study design is important here to answer this question. Majority of studies were cross-sectional limiting conclusions. It is not much possible to inflict low muscle mass to participants (maybe through bed rest but it comes with important ethical concerns) so the most appropriate design remains a longitudinal prospective cohort study which provides direction to the associations. A nested case-control could be applied with sarcopenic individuals matched to non-sarcopenic and we follow them and measure their cognitive performance over time. More measures would be worth adding to such study including PET and/or functional MRI, serum growth factors such as IGF-1, BDNF (CSF easures although invasive). I would think other studies could be done in animal, maybe knockout mice to induce low muscle mass and measure cognition and other factors.
    - - I had tested lean mass as continuous and it was not associated with cognitive decline. It is possible that the relationship is not linear. Here the objective was really to see if it was associated with our specifc threshold of low muscle mass vs normal mass.
    - - DSM-5, published by the American Psychiatric Association in 2013. The term dementia was changed to Major neurocognitive disorders as: Another significant change is that the term “dementia” has been eliminated and replaced with major or minor neurocognitive disorder. It was believed that the word dementia was stigmatizing toward older individuals and not well accepted by younger individuals with HIV dementia. The new term focuses on the decline from a previous level of functioning as opposed to a deficit.
  - - - Some tests were shown to be affected by rater of the test (from ICC).
      - You calculate the MDC95 from standard error of measurement (SEM)
      - Is change above the SEM? and MDC95?
  - - - Advanced age and low educational level are associated with decreased BDNF serum levels. Decreased BDNF serum levels correspond to the severity of cognitive impairment.
    - - MMSE, greater or equal to 3 points. MoCA greater or equal to 4 points.
- - - - For sure, having age as a factor will help better predict impaired PP and low strength. It is highly correlated. Might in fact better predict than low ALM and low strength themselves. Having some sort of score with different variables would help estimate the risk of impaired PP. Ideally you woul
    - - Although the objective of this study was not to find causal relationship but rather tools to identify persons most likely to have impaired PP. As shown from R squared and beta estimated from path analyses, strength and ALM do not fully explain PP. Other factors do. Now it will be interesting to explore other outcomes low ALM may be linked with. A longitudinal design will allow to understand whether low ALM is linked with greater incident of diseases, with incident impaired PP, it will confirm direction to the association vs when cross-sectional.
    - - We cannot conclude whether someone with low ALM has greater risk to develop impaired PP vs someone with normal ALM. A longitudinal design will provide us with this information.
  - - - ICD-10 codes by the WHO are established for clinical and epidemiological classification of conditions. Health care professionals and researchers use it to identify conditions and report it for public health tracking (incidence and death; statistical use or economic). Independent of language and culture. Ex. COVID-19, the coding has been released for countries to start tracking although it is continually being updated.
      - It is indeed problematic if not standard and different people apply the coding differently. But it is a way to initiate tracking I believe. Physicians and other health care providers should follow the international guidelines.
      - Classification is a logical algorithm/rules. mutually exclusive groups. Does not include guidelines as to HOW to dx it. If muscle dystrophy due to X, attribute Y code. If muscle dystrophy due to age, attribute W code.
      - It may have been established to stimulate research in that area. Not long after the ICD-10, the FDA was conducting focused groups with person with sarcopenia and were asking about the disease burden, willingness to take medication for their condition, etc. In fact, participants mentioned that the condition was not well recognized by their physician and wished they had more info and intervention for it.
      - New classification: the network receives application, the committee evaluate. If they judge it should be included they will derive coding for it.
    - - We wanted to have a comprehensive evaluation of physical performance. In the CLSA, we had not only gait speed but also the TUG (a measure of mobility), balance, 4-meter walk test (endurance), chair rise (balance/strength). I also identified the weight of each in the construct "physical performance" using factor analysis.
    - - T-scores are developed based on younger population to represent peak of muscle mass in that case (O value-Y mean/Y SD). That would be around the age of 30 y and we did not have such data. Yet our approach is more robust given it was empirically derived to predict clinical outcomes (weakness and low physical performance). T-score remains arbitrary - where is the cut-off below the distribution of young where it is abnormal? It would be another valuable approach and would be interesting to compare results to current proposed approach.
    - - These results were found in other populations (AWGS for example in East China). It also seems that muscle loss is slower in women vs men. Muscle metabolism seem to differ between men and women. ...(There is also a greater shift toward smaller fibres in women vs men.)
      - Prevalence for low strength was similar between both. But prevalence of low ALM was significantly lower in women. This means that the best cut-off to find those with low strength was lower and more conservative in women. In men that ideal cut-off was higher giving greater %. The relationship between ALM and strength was weaker in women maybe explaining the lower cut-point. Also different factors seem to explain this relationship (fat mass more important in women than man vs age is the opposite). It may be explained by statistics or there may be a social or biological explanation.
      - *When looking at path analyses: fat mass is a stronger determinant of ALM in women. While in men age is a stronger determinant of ALM.
      - More studies on trajectories in men and women, interactions in observation studies and studies in vivo
      - Men and women seem to have similar mTOR activation. But processes may differ.
    - - Physical performance is a crucial factor for older adults autonomy. Parameters are of high importance for clinicians to assess disability, risk of disability, falls, autonomy, etc. Limited or impaired physical performance should be termed as is simply, not sarcopenia. Adding it to the definition of sarcopenia surely helped in predicting impaired phys perf but as Bulow et al. said it, it is a tautology.
      - It also distracts from researching the etiology and consequences of low muscle mass itself. Also adding multiple measurements and parameters to the definition of a condition complexifies it (more tests needed, algorithm to follow), denatures it (no longer assesses same thing), difficult to study treatment because of multiple outcomes.
    - - We cannot expect someone short to have very high ALM compared to the distribution and alternatively, we cannot expect some very tall to have very low ALM or at least it would indicate they have very low ALM and vice versa. Like BMI is you weight divided by your height squared, an index of ALM/ht squared indicates your muscle mass relative to your height. Your height barely changes also over the years (vs BMI).
      - BMI is inclusive of ALM. height is not inclusive of ALM.
      - height correlates strongly with muscle mass and square of height is the best common power for minimizing the correlation of the index with height across all sex, ethnic, age groups and study pop.
    - - Prevalence differ, dx differ, difficult to identify proper intervention as outcome differs... a good example of this is clinical trials that have been conducted and reached phase 2-3 and were discontinued because they were successful at increase MM but did not significantly increase MS.
    - - We wanted to have a measure of mobility that is reflective of MM and muscle function and not excess weight.
      - Why not adjust ALM directly for weight/BMI/fat mass?
        
        Because you will not identify the same persons as demonstrated in the CLSA. These people have normal ALM, they are not the ones you find at the tip of the distribution with low physiological reserves.
      - Mobility correlated very well with BMI. BMI is also a measure of weight relative to height vs absolute fat mass. Gait speed is slower in shorter individuals so it accounts for all these parameters.
      - It is expected that older adults have lower strength and lower ALM. Adjusting for age would mean that someone older with low strength from younger distribution may be considered to have normal strength for their specific age. We prefered to have higher threshold. You have than higher prevalence among older adults. which is normal because as we age we lose muscle mass and strength.
    - - Type 2 diabetes: do not report taking diabetes medications but for whom fasting blood glucose levels are ≥ 7.0 mmol/l
      - Ischemic Heart Disease: self-reported dx and symptoms, medical procedure, medication use. electrocardiogram (ECG) and the Rose Questionnaire
      - They are algorithm based on self-reprted, medication use and clinical measurements. it is not validated by a physician. This approach has been validated in the literature.
    - - Second step. We had out binary variable and needed to find an ideal cut-off that would classify low ALM vs normal ALM. Typically the number of muscle fibers correlates with the number of motor neuron (nerve cells) (one for 10-100 fibers). It is logical to think the skeletal muscle mass predicts strength although the opposite may be true as well - strength appears to be more dynamic than MM such that when submitted to exercise it increases more vs MM.
      - The outcome of this study was one that is very important to clinicians as it reflects independence or risk of losing independence, capacity to pursue activities of daily living, etc. That was physical performance. ALM was not a predictor of PP but strength was and so naturally strength was our independent variable and we wanted to find the ideal cut-point to predict limited PP as a 1st step.
    - - impaired physical performance = 7.4% is reflective of what has been observed in other studies, but tends toward low. Our 1.5 SD was conservative. A. Mayhew at McMaster also studies sarcopenia, physical performance in the CLSA. She demonstrated that all phys performance in lowest perf quintile (20%, vs. 1.5 SD = 19%) were associated with ADL and IADL limitations, individually and combined. She also reported a prevalence of 5-10% limitation in ADL or IADL (using self reported and phys perf test)
      - Low ALM = similar to prevalence observed in European countries and applying the european cut-points. Namely because of the same index used although a different approach was used.
      - Low handgrip strength=
    - - our findings represent the most recent cut-points for sarcopenia and dynapenia in a contemporary database (recruitment between 2011-2015) and the first established in a single large national-representative population addressing generalizability and methodological issues encountered with previously derived cut-points.
    - - p interaction was not significant so it predicted similarly across groups. Only for those at nutritional risk (better prediction of HGS in men with ALM cutoffs) and it is the opposite in women...
      - In women, sig better prediction of impaired PP in those younger than older. Not sig diff in men.
      - In women, those without kidney disease = better prediction of low HGS.
    - - Predictive of physical limitations, ADL, prediction of sarcopenia was better WITHOUT muscle mass component.....
    - - Ultrasound is interesting for bedside assessment. It has been validated to estimate muscle thickness in patient populations and especially ICU. But very difficult to use in standardized manner. High variability among administrators.
      - DXA - is 3 compartments, provides lean soft tissue, fat tissue and bone tissue.
      - BIA, measures resistance to a current (greater resistance in fat tissue vs lean). Provides FFM and FM (only 2 compartments). FFM includes bones.
    - - There is high concordance for FM but less so for FFM. It is heavily relying on assumptions as it is a 2 compartment model vs DXA which is a 3 compartment model. It allows to differentiate bone from lean soft tissue and fat tissue.
      - DXA scan for BMD is recommended especially for women above the age of 65y who have been through menopause. Or in younger who have at least one risk factor (e.g. smoking, low weight, etc) which are already risk factors for low muscle mass. I think that if clinical screening points toward possible sarcopenia, full body scan should be requested.
- - - - Because:
      - How to address:
        
        Other databases that exists adn whats now available in the CLSA
        
        Which FU is now available? Only first follow up data is available. Physical assessment as well. https://www.clsa-elcv.ca/doc/3162 gender identity is only available in FU 1 not baseline. Subjective cognitive decline now available in FU1 which is usually part of dementia dx. Chemistry report not yet available for FU1. Genotype baseline is now available
        
        were older adults in CLSA free of dementia at baselne? what were the inclusion/excusion criteria?
  - - - There is higher prevalence of dementia in women vs men (2:1). This may be explained by greater longevity of women. Typically, dementia is diagnosed when symptoms are full blown and this is around age 80-85. Life expectancy of men is 80 in Canada. Also brain cell apoptosis seems to be greater in women vs men. It is not well understood.
      - Gender differences have been reported: men tend to perform better in spatial memory tasks and women in verbal memory and word/object memory. Usually no difference in working memory. Limited data on sex differences in executive functions. Women perform better on inhibition tasks.
      - Among biological factors, sex hormones (estrogen and testosterone) may play a role in these difference between men and women. Also, genetic traits that are inherited.
    - - Nutrients important in brain health, e.g. low B12, associated with high homocysteine levels and link with cognitive dysfunction.
      - dietary patterns and food modifying microbiome modulating learning and memory via gut-brain axis
    - - Emphasis was put on executive function as they are most strongly related to those capabilities that enable people to engage in independent, purposive, and self-regulatory behaviors (Wecker et al. 2000)
        Executive functions are involved in many higher order mental processes and complex behaviors.
    - - Potentially because of adjustment for intake in other nutrient dense/key nutrients/matrix relevant foods sucha s meants (B12, zinc), legumes (zinc, fibres), egg (choline, B12), salmon (omega-3, B12).
      - Choline is a precursor to the neurotransmitter acetylcholine. Loss of cholinergic neurons is associated with impaired cognitive function (memory and AD in particular)