Skip to main content
. 2015 Sep 25;10(9):e0138036. doi: 10.1371/journal.pone.0138036

Table 2. Linear regression of cognitive function in midlife (~50 y) on inflammatory markers at age ~30 y (baseline) and 13 y change in GlycA: the Jerusalem LRC longitudinal study.

Inflammatory marker f Global a (n = 507) Attention a (n = 507) Information processing speed (n = 477) Executive (n = 506) Visual spatial (n = 505) Memory a (n = 499)
Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3 Model 1 Model 2 Model 3
β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p) β b (p)
1. CRP (nmol/L) a c at baseline -.053 .012 .010 -.038 .008 .019 -.097* -.061 -.044 -.080 -.033 -.019 -.036 .014 -.009 -.004 .044 .037
(.243) (.770) (.809) (.404) (.865) (.676) (.036) (.180) (.342) (.076) (.473) (.688) (.426) (.744) (.833) (.925) (.343) (.430)
2. WBC (cells/μL) c at baseline -.007 .018 .014 .008 .032 .032 -.067 -.051 -.049 -.028 -.014 -.016 -.023 -.012 -.016 .004 .027 .021
(.881) (.659) (.729) (.862) (.463) (.466) (.151) (.255) (.280) (.528) (.758) (.720) (.611) (.784) (.706) (.922) (.566) (.650)
3. Fibrinogen (mg/dL) c at baseline -.125** -.036 -.031 -.053 .014 .022 -.048 .002 .011 -.138** -.074 -.062 -.089 .007 .005 .091 -.065 .061
(.007) (.383) (.467) (.235) (.760) (.636) (.317) (.967) (.821) (.003) (.107) (.181) (.055) (.883) (.918) (.052) (.169) (.198)
4. GlycA (μmol/L) c at baseline -.082 .006 .008 -.071 -.013 -.006 -.065 -.005 .008 -.104* -.046 -.036 -.058 .006 -.003 .003 .082 .080
(.072) (.890) (.857) (.119) (.768) (.901) (.163) (.908) (860) (.022) (.323) (.448) (.204) (.886) (.950) (.948) (.083) (.095)
5. GlycA (μmol/L) change c d -.118** -.086 -.076 -.079 -.059 -.067 -.109* -.084 -.098 -.064 -.055 -.056 -.076 -.055 -.033 -.093 -.062 -.048
(.023) (.051) (.112) (.127) (.218) (.195) (.042) (.089) (.063) (.216) (.263) (.289) (.142) (.245) (.511) (.077) (.214) (.370)
6. GlycA (μmol/L) change (quintiles, 4df) d
    1 ≤ 49.88 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
    2 49.89–83.21 -.050 -.050 -.048 -.078 -.083 -.083 -.017 -.030 -.030 -.049 -.058 -.057 -.101 -.108* -.110* -.007 -.003 -.002
(.395) (.323) (.347) (.183) (.128) (.132) (.772) (.587) (.601) (.401) (.294) (.313) (.084) (.045) (.043) (.901) (.955) (.967)
    3 83.22–108.81 -.081 -.043 -.042 -.102 -.068 -.069 -.012 .015 .015 -.066 -.039 -.038 -.159** -.134* -.140* .038 .066 .068
(.171) (.399) (.418) (.089) (.217) (.220) (.842) (.793) (.798) (.270) (.489) (.509) (.008) (.014) (.011) (.524) (.249) (.243)
    4 108.82–139.19 -.034 -.032 -.027 -.032 -.034 -.040 .005 -.004 -.016 -.029 -.031 -.034 -.101 -.101 -.090 .018 .027 .034
(.575) (.543) (.611) (.602) (.547) (.499) (.935) (.943) (.787) (.637) (.585) (.568) (.098) (.069) (.117) (.765) (.646) (.579)
    5 ≥ 139.20 -.186** -.148** -.143* -.131 -.106 -.115 -.146 -.117 -.132* -.117 -.111 -.116 -.145* -.122* -.111 -.134* -.098 -.087
(.003) (.006) (.015) (.039) (.072) (.070) (.025) (.052) (.040) (.063) (.065) (.074) (.022) (.036) (.074) (.035) (.109) (.182)
    p for trend d (.010) (.018) (.041) (.181) (.297) (.277) (.036) (.077) (.051) (.141) (.157) (.167) (.106) (.180) (.362) (.038) (.111) (.205)
7. High GlycA (μmol/L) change (highest quintile, 1df) d e -.141** -.114** -.109* -.073 -.056 -.060 -.140** -.113* -.124** -.078 -.075 -.078 -.044 -.025 -.010 -.148** -.123** -.117*
(.003) (.005) (.011) (.122) (.207) (.194) (.004) (.012) (.008) (.099) (.093) (.098) (.353) (.561) (.824) (.002) (.007) (.014)

LRC, Lipid Research Clinic; Beta = standardized regression coefficient; CRP, C-reactive protein.

a Box-Cox transformed z-scores (λ = - 0.5), C-reactive protein (λ = 0).

b Standardized regression coefficient.

c Per 1 SD increment.

d Additionally adjusted for baseline measurement of GlycA and time elapsed between measurements.

e Compared to the 4 lowest quintiles grouped.

Model 1: Unadjusted

Model 2: Adjusted for age, sex, educational level, origin, childhood SES (ICBS ranking) and adult SES (ICBS ranking), depression, and baseline leisure-time vigorous activity, smoking status and BMI measured at ages 28–32.

Model 3: Model 2 plus change in smoking status, change in BMI, and duration between measurements. Reference level of dummy variables: sex, females; educational level, elementary (<9 yrs.); origin, Israel; leisure-time vigorous activity, none; smoking status, never; change in smoking status, no change. Age, childhood SES, adult SES, depression, BMI, change in BMI; all introduced as continuous/ interval variables.

f Missing data: CRP (n = 14), WBC (n = 12), fibrinogen (n = 43), GlycA at baseline (n = 21), GlycA change (n = 23), adult SES (n = 5), depression (n = 4), leisure-time vigorous activity (n = 1), change in BMI (n = 1). Missing values of adult SES, depression, leisure-time vigorous activity and change in BMI were replaced with non-missing median or mean values.

P-values in parentheses.

* p < .05.

** p < .01