Skip to main content

Table 9 Geospatiotemporal regression models

From: Cannabinoid exposure as a major driver of pediatric acute lymphoid Leukaemia rates across the USA: combined geospatial, multiple imputation and causal inference study

Parameter Model
Parameter Estimate (C.I.) P-Value LogLik S.D. Model Parameter Estimate P-Value
Cannabis Alone
spreml (ALL_Rate ~ Cannabis)
  Cannabis 0.22 (0.09, 0.36) 0.0015 − 193.8962 0.4040 phi 0.5060 0.0005
Additive Model
spreml (ALL_Rate ~ Cigarettes + Cannabis + AUD + Analgesics + Cocaine)
  Cannabis 0.16 (0.06, 0.27) 0.0032 N/A 0.3869 phi 0.4604 0.0005
      rho −0.3274 0.0229
Interactive Models - Drugs
spreml (ALL_Rate ~ Cigarettes * Cannabis * AUD + Analgesics + Cocaine)
  Cigarettes: Cannabis: AUD 22.02 (10.13, 33.91) 0.0003 −189.4824 0.3760 phi 0.3669 0.0022
  Cigarettes: AUD 45.76 (17.83, 73.69) 0.0013    rho −0.3098 0.0349
  Cannabis: AUD −2.39 (−4.75, −0.02) 0.0482    lambda 0.2865 0.0162
Interactive Models - Drugs, Income, Ethnicity
spreml (ALL_Rate ~ Cigarettes * Cannabis * AUD + Analgesics + Cocaine + Income + 5_Races)
  Hispanic 0.21 (0.15, 0.26) 8.3E-13 −170.6536 N/A rho 0.3314 0.0007
  Asian 0.12 (0.04, 0.2) 0.0035    lambda −0.3523 0.0011
  Income −0.29 (− 0.57, − 0.01) 0.0404      
  African-Am. − 0.16 (− 0.21, − 0.12) 6.2E-14      
Interactive Cannabinoid Models - Drugs, Income, Ethnicity
spreml (ALL_Rate ~ Cigarettes * THC * CBG + AUD + Analgesics + Cocaine + Income + 5_Races)
  Hispanic 0.23 (0.17, 0.28) 9.9E-15 − 163.7442 0.3325 phi 0.0495 0.1403
  Cannabigerol 0.26 (0.01, 0.52) 0.0444    psi 0.0573 0.2640
  Cigarettes: THC: Cannabigerol −1.18 (−1.87, −0.49) 0.0008    rho 0.1791 0.2751
  Cigarettes: THC −4.96 (−7.8, −2.12) 0.0006    lambda −0.2047 0.2159
  African-Am. −0.13 (− 0.18, − 0.09) 3.4E-09      
  1. Technical Notes:
  2. phi: - Idiosyncratic component of the spatial error term
  3. psi: - Individual time-invariant component of the spatial error term
  4. rho: - Spatial autoregressive parameter
  5. lambda: - Spatial autocorrelation coefficient