1. C-Reactive Protein & Cognitive Function
Ben Tappin
Undergraduate at University of Surrey

2. Overview
C-Reactive Protein (CRP) is a bodily protein that is related to infection, tissue damage, and inflammation. Chronic elevated CRP levels have great clinical utility in predicting biological conditions associated with systemic bodily inflammation e.g. cardiovascular disease.
Recent research has identified a link between elevated CRP and cognitive impairment, however it is unclear whether this relationship is direct or whether elevated CRP acts as a secondary marker of other disorders only.
The broad aim of the present study was to investigate the CRP-cognitive impairment relationship further.

3. Background The present study was (sort of) a secondary data analysis.
Data were obtained from the Institute of Psychiatry, King’s College, London.
The data were collected as part of a longitudinal study of children in and around Greater London that began in 2006 and is still on-going.
I was part of the research team there for a year and collected the majority of the data during this time.

4. Literature review Biology & elevated CRP
CVD/CHD/atherosclerosis + association with BMI
Alzheimer’s disease, dementia

5. Literature review Cognition & elevated CRP
A multitude of cognitive tests are used to measure cognitive performance, covering predominantly:
-Executive Function (EF)
-Memory
-Visuospatial ability
Most consistent associations are found between EF and CRP e.g.
Moderate correlation coefficients (r = .20, .25)
Small standardized beta values (β = .09 to .16)

6. Problems & gaps in research
A major problem within the literature is the over-use of older-age adult samples. General age-related decline may be confounding the CRP-cognition relationship.
Research using child populations is scarce: there are only a handful of studies investigating the CRP-cognition link in children (some positive findings e.g. Gozal et al. 2007)
Why is this?
Difficulty/cost in obtaining blood samples from nonclinical paediatric populations; &
Lack of perceived importance in identifying CRP-cognition link in those not at immediate risk for CRP-correlated disorders e.g. AD.

7. Aims/hypotheses of the present study
To investigate the CRP-cognition relationship in a child population, free of older-age confounds;
Hypothesis 1) CRP level will predict cognitive performance, associated primarily with executive function, in a child population.
To investigate the validity of a novel form of CRP measurement – requiring saliva samples, rather than blood;
Hypothesis 2) The relationship between salivary CRP levels and cognitive performance in the present study will be similar to the observed relationship between blood CRP levels and cognitive performance in the extant literature.

8. Method Cognitive measures: 89 participants, aged between 9-16 years.
All participants provided saliva using the passive drool method, and samples were subsequently analysed using the salivary CRP assay (analysis method) provided by Salimetrics™.
Cognitive measures:
Stroop Test – test of EF with consistent links to CRP level in adult literature
Tower Test – another test of EF including a visuospatial component
Story Memory – an exploratory immediate-recall memory test

9. Method
Design:
Three linear regression models (one for each cognitive test) were used to determine if there was a predictive relationship between CRP and cognitive function.
Documented covariates were also investigated using t-tests and bivariate correlational analyses:
-Age
-Gender
-BMI
-Dental health

10. Results Primary findings
1) Stroop Test performance was significantly predicted by CRP levels: with standardized beta value = -.25 (a small-medium sized relationship).
This equates to approx. 6% of the variance in Stroop performance being explained by CRP levels.
2) Tower Test and Story Memory performance was not associated with CRP level.

11. Results Secondary findings
1) Dental health significantly influenced the salivary CRP measurements: those who reported no dental health problems had significantly lower salivary CRP concentrations than those who reported one or more.
This effect was moderate (Cohen’s d = 0.63).
2) There was no effect of any tested covariates. CRP was not positively related to its most consistent correlate, BMI: in fact, the relationship tended toward a negative one (ns correlation coefficient of -.21). Somewhat surprising.

12. Critical reflections & Implications
Hypothesis 1 was partially supported: the Stroop-CRP association was replicated in children, though Tower Test and Story Memory performance was not related to CRP.
Hypothesis 2 was supported: the effect size found for the Stroop-salivary CRP relationship (β = -.25) is similar to the current blood CRP-cognition literature.

13. Critical reflections & Implications
Hypothesis 1: replication of Stroop-CRP relationship in children contributes to growing evidence that suggests CRP is a causal agent in cognitive impairment, and not simply a secondary marker of general age-related decline.
Caveats:
-Lack of an association for Tower and Story variables means there is a lack of generalizability to all cognitive function, and the findings must be interpreted as such.
Why?
-Tower Test performance is rarely reported as directly related to CRP; it was included here as an EF test with a spatial component. It may be that other spatial tests that are related to CRP have different cognitive demands that the present Tower Test does not.
-Story Memory was included as an exploratory variable as the findings are fairly conflicted in supporting the CRP-memory link.

14. Critical reflections & Implications
Hypothesis 2: the acceptance of hypothesis 2 indicates that the present study has assisted in validation of a novel method of CRP analysis which may help stimulate future larger-scale CRP-cognition research on children.
Caveats:
-Dental health problems served to increase concentration of salivary CRP, suggesting that poor dental health can confound salivary CRP measurement. It is recommended that future research using salivary CRP controls for dental health.
-Lack of a significant association with BMI.

15. Acknowledgements + Q & A
University of Surrey:
Dr. Henriette Hogh (dissertation supervisor)
Institute of Psychiatry:
Dr. Kristin Laurens (longitudinal project supervisor)
Alexis Cullen (PhD student)
Contact me:
Ben Tappin