Impact of Converging Sociocultural and Cannabinoid-Related Trends on US Autism Rates Dataset: Combined Geospatiotemporal and Causal Inferential Analysis

Background: Whilst cannabis is known to be toxic to brain function and brain development in many respects it is not known if its increasing availability is associated with the rising US autism rates, whether this contribution is sufficient to effect overall trends and if its effects persist after controlling for other major covariates. Methods: Longitudinal epidemiological study using national autism census data from the US Department of Education Individuals with Disabilities Act (IDEA) 1991-2011 and nationally representative drug exposure (cigarettes, alcohol, analgesic, and cocaine abuse, and cannabis use monthly, daily and in pregnancy) datasets from National Survey of Drug Use and Health and US Census (income and ethnicity) and CDC Wonder population and birth data. Geotemporospatial and causal inference analysis conducted in R. Results: 266,950 autistic of a population of 40,119,464 eight year olds 1994-2011. At the national level after adjustment daily cannabis use was significantly related (β-estimate=4.37 (95%C.I. 4.06-4.68), P<2.2x10-16) as was cannabis exposure in the first trimester of pregnancy (β-estimate=0.12 (0.08-0.16), P=1.7x10-12). At the state level following adjustment cannabis use was significant (from β-estimate=8.41 (3.08-13.74), P=0.002); after adjustment for varying cannabis exposure by ethnicity and other covariates (from β-estimate=10.88 (5.97-15.79), P=1.4x10-5). Cannabigerol (from β-estimate=-13.77 (-19.41—8.13), P = 1.8x10-6) and Δ9-tetrahydrocannabinol (from β-estimate=1.96 (0.88-3.04), P=4x10-4) were also significant. Geospatial state-level modelling showed an exponential relationship between ASMR and both Δ9-tetrahydrocannabinol and cannabigerol exposure; effect size calculations reflected this exponentiation. Exponential coefficients for the relationship between modelled ASMR and THC- and cannabigerol- exposure were 7.053 (6.39-7.71) and 185.334 (167.88-202.79; both P<2.0x10-7). In inverse probability-weighted robust generalized linear models ethnic cannabis exposure (from β-estimate=3.64 (2.94-4.34), P=5.9x10-13) and cannabis independently (β-estimate=1.08 (0.63-1.54), P=2.9x10-5) were significant. High eValues in geospatial models indicated that uncontrolled confounding did not explain these findings. Therefore the demonstrated relationship satified the criteria of causal inference. Dichotomized legal status was geospatiotemporally linked with elevated ASMR. Conclusions: Data show cannabis use is associated with ASMR, is powerful enough to affect overall trends, and persists after controlling for other major drug, socioeconomic, and ethnic-related covariates. Selected cannabinoids are exponentially associated with ASMR. The cannabis-autism relationship satisfies criteria of causal inference.研究背景：尽管已知大麻（cannabis）在诸多方面对大脑功能与大脑发育具有毒性作用，但目前尚不明确其可及性的不断提升是否与美国自闭症患病率的上升存在关联，也不清楚该关联的贡献度是否足以影响整体流行趋势，以及在控制其他主要协变量后，大麻的相关影响是否仍会持续存在。 研究方法：本研究为纵向流行病学研究，使用了1991-2011年美国教育部《残疾人教育法案》（Individuals with Disabilities Act, IDEA）中的全国自闭症普查数据，以及来自全国药物使用与健康调查、美国人口普查局（收入与种族数据）和疾控中心（CDC）Wonder数据库的人口与出生数据构成的具有全国代表性的药物暴露数据集（涵盖香烟、酒精、镇痛剂、可卡因滥用情况，以及月度、每日使用大麻和孕期大麻暴露情况）。研究采用R语言开展时空地理分析与因果推断分析。 研究结果：1994-2011年间，在40119464名8岁儿童的研究队列中，共确诊266950名自闭症患儿。 在国家层面校正后，每日大麻使用与自闭症患病率（ASMR）显著相关（β估计值=4.37，95%置信区间：4.06~4.68，P<2.2×10^-16），孕期妊娠首三个月的大麻暴露亦呈现显著关联（β估计值=0.12，95%置信区间：0.08~0.16，P=1.7×10^-12）。 在州级层面校正后，大麻使用与ASMR的关联仍具有统计学意义（β估计值=8.41，95%置信区间：3.08~13.74，P=0.002）；在按种族与其他协变量校正不同水平的大麻暴露情况后，该关联依然显著（β估计值=10.88，95%置信区间：5.97~15.79，P=1.4×10^-5）。 大麻萜酚（Cannabigerol, CBG，β估计值=-13.77，95%置信区间：-19.41~-8.13，P=1.8×10^-6）与Δ9-四氢大麻酚（Δ9-tetrahydrocannabinol, THC，β估计值=1.96，95%置信区间：0.88~3.04，P=4×10^-4）同样呈现显著关联。 州级时空地理模型分析显示，ASMR与Δ9-四氢大麻酚、大麻萜酚暴露均呈指数关联；效应量计算结果也印证了这一指数关系。建模所得ASMR与Δ9-四氢大麻酚、大麻萜酚暴露的指数系数分别为7.053（95%置信区间：6.39~7.71）与185.334（95%置信区间：167.88~202.79），二者P值均<2.0×10^-7。 在逆概率加权稳健广义线性模型中，种族相关大麻暴露（β估计值=3.64，95%置信区间：2.94~4.34，P=5.9×10^-13）与独立的大麻使用（β估计值=1.08，95%置信区间：0.63~1.54，P=2.9×10^-5）均具有统计学意义。时空地理模型中的高E值表明，未被控制的混杂因素无法解释上述研究结果，因此所证实的关联符合因果推断的标准。将大麻法律状态二分后，其时空分布与升高的ASMR存在关联。 研究结论：本研究数据显示，大麻使用与ASMR存在显著关联，其影响强度足以改变整体流行趋势，且在控制其他主要药物、社会经济与种族相关协变量后，该关联依然存在。部分大麻素与ASMR呈指数关联。大麻与自闭症的关联符合因果推断的判定标准。