41 - From Card to Cradle: Examining Medical Cannabis Purchasing Among Pregnant Women in Arkansas

Publication Number: 3038.41

Nahed O. ElHassan, University of Arkansas for Medical Sciences College of Medicine, Little Rock, AR, United States; Cain Farnam, ACHI, Little Rock, AR, United States; Joseph W. Thompson, ACHI, University of Arkansas for Medical Sciences, Little Rock, AR, United States; Nichole Stanley, ACHI, Little Rock, AR, United States; Corey J. Hayes, University of Arkansas for Medical Sciences College of Pharmacy, Little Rock, AR, United States; Chenghui Li, University of Arkansas for Medical Sciences College of Medicine, Little Rock, AR, United States; Teresa Hudson, UAMS, Little Rock, AR, United States; Robert E. Mcgehee, University of Arkansas for Medical Sciences College of Medicine, Little Rock, AR, United States; Peter Mourani, University of Michigan Medical School, Ann Arbor, MI, United States; Bradley C. Martin, University, Little Rock, AR, United States

Background: Medical cannabis (MC) use has increased among pregnant women, yet little is known about the quantity, product types, and timing of use.

Objective: To conduct a population-level analysis of MC purchasing patterns among pregnant women in Arkansas (AR).

Design/Methods: This descriptive exploratory analysis examined MC purchasing among pregnant women in AR, 5/11/2019-8/31/2022, after MC availability. Data from three statewide sources, the MC Card Registry (patient and qualifying condition data), MC Dispensary Database (transaction and product details), and Birth Certificate Records (maternal demographics and obstetric diagnoses), were linked with the National Social Determinants of Health database, providing neighborhood-level socioeconomic indicators. Descriptive statistics summarized study characteristics and factors associated with the amount of purchased delta-9-tetrahydrocannabinol (THC), cannabis main psychoactive ingredient.

Results: Of the 72,992 pregnancies, 1,185 (1.6%) purchased MC during pregnancy. Among these, 774 were continuers, defined as those with MC purchases within 90 days before pregnancy (preconception) that continued during pregnancy, and 411 initiated MC during pregnancy (initiators). Ninety-four pregnancies purchased MC in the preconception period but discontinued by pregnancy onset (discontinuers). Among in-pregnancy purchasers, 964, 871, and 690 purchased in first, second, and third trimester. Compared to non-purchasers, MC purchasers were more likely ≥30 years old and tobacco smokers and less likely non-Hispanic Black and privately insured (Table 1). Post-traumatic stress disorder was the most common qualifying condition for MC purchase (67%), while severe nausea accounted for only 6%. Plant cannabis was the most commonly purchased product (52%). The mean (SD) daily THC purchased during pregnancy was 137 (170) mg/day. Continuers purchased more daily THC during pregnancy than initiators (P < 0.001) and more pre-pregnancy THC than discontinuers (P < 0.001) (Figure 1). The adjusted mean ratio of daily THC purchased in-pregnancy was higher among continuers compared to initiators and lower for women < 20 years, privately insured, receiving prenatal care, living in micropolitan/small-town areas or in neighborhoods with higher household food-stamp receipt (Table 2).

Conclusion(s): The mean daily THC purchased during pregnancy exceeded therapeutic dosing ranges (2-60 mg/day) established for FDA-approved cannabinoid formulations in non-pregnant adults and was high among women initiating MC purchase pre-pregnancy. These findings highlight the need for research on MC safety during pregnancy.

Table 1. Adjusted Odds of Maternal and Neighborhood Characteristics of Being a Medical Cannabis (MC) Purchaser during Pregnancy (N= 72,992)
Pregnancies that purchased exclusively MC during the pre-pregnancy period (n=94) were excluded from the non-purchaser group to ensure an unbiased comparison with in-pregnancy MC purchasers.
Multiple variables did not have significant adjusted odds ratios (aORs) or P-values and were therefore not shown in the table. These variables included:
Maternal demographic variables: Geographic location of maternal residence.
Maternal clinical variables: Prenatal care, gestational diabetes, and pregnancy-induced hypertension.
SDOH economic stability domain: Household income, percentage of households receiving food stamps, and percentage of housing units without a vehicle.
SDOH health care access and quality domain: Percentage of population with Medicaid or other means-tested public health coverage, percentage of uninsured population, and distance to the nearest clinic in miles (calculated using population-weighted ZIP centroids).
SDOH education access and quality domain: Percentage of population with less than high school education among adults aged ≥25 years and percentage of population with high school education among adults aged ≥25 years
SDOH neighborhood and built environment domain: Percentage of housing units lacking complete plumbing facilities and percentage of housing units built before 1979.
SDOH social and community context domain: Percentage of limited English-speaking households and percentage of households with White-alone householders.

Figure 1. Average Daily Delta-9-Tetrahydrocannabinol Purchased per Different Time Periods (N=1,279)
THC, Delta-9-Tetrahydrocannabinol

Table 2. Multivariable Analysis of Maternal and Neighborhood Characteristics Associated with Average Daily THC Purchased During Pregnancy Among Medical Cannabis (MC) Purchasers (N=1,185)
Abbreviations by alphabetical order: CI, confidence interval; PTSD, post-traumatic stress disorder; THC, Delta-9-Tetrahydrocannibol.
Average daily THC was calculated by dividing the total calculated amount of THC purchased during pregnancy by the number of days of pregnancy.
Qualifying medical conditions were categorized based on the most common combination of qualifying medical conditions for medical cannabis purchase per pregnancy.
Similar to Table 1, many variables were assessed but not reported due to non-significant findings. The maternal and neighborhood characteristics assessed were identical in Tables 1 and 2.