Skip to main content
PMC home page
Sage Choice logoLink to Sage Choice
. 2019 Jul 9;58(10):1085–1123. doi: 10.1177/0009922819861281

Cannabis Teratology Explains Current Patterns of Coloradan Congenital Defects: The Contribution of Increased Cannabinoid Exposure to Rising Teratological Trends

Albert Stuart Reece 1,2,✉, Gary Kenneth Hulse 1,2
PMCID: PMC6691650  PMID: 31288542

Abstract

Rising Δ9-tetrahydrocannabinol concentrations in modern cannabis invites investigation of the teratological implications of prenatal cannabis exposure. Data from Colorado Responds to Children with Special Needs (CRCSN), National Survey of Drug Use and Health, and Drug Enforcement Agency was analyzed. Seven, 40, and 2 defects were rising, flat, and falling, respectively, and 10/12 summary indices rose. Atrial septal defect, spina bifida, microcephalus, Down’s syndrome, ventricular septal defect, and patent ductus arteriosus rose, and along with central nervous system, cardiovascular, genitourinary, respiratory, chromosomal, and musculoskeletal defects rose 5 to 37 times faster than the birth rate (3.3%) to generate an excess of 11 753 (22%) major anomalies. Cannabis was the only drug whose use grew from 2000 to 2014 while pain relievers, cocaine, alcohol, and tobacco did not. The correlation of cannabis use with major defects in 2014 (2019 dataset) was R = .77, P = .0011. Multiple cannabinoids were linked with summary measures of congenital anomalies and were robust to multivariate adjustment.

Keywords: delta9-tetrahydrocannabininol, cannabidiol, epigenetic genotoxicity, congenital teratogenicity, congenital, cardiovascular malformations

Introduction

While the teratogenic activities of cannabis have been investigated since the 1960s,1,2 substantially higher levels of Δ9-tetrahydrocannabinol of currently used cannabis3 suggests that the neonatal epidemiology of former years requires reexamination.4,5

Urgency for epidemiological reassessment achieves particular currency in view of recent US data indicating that 24% of pregnant Californian teenagers test positive for cannabinoids,6 that 69% of pregnant Coloradan mothers have cannabis recommended to them by cannabis dispensaries,7 and that 161 000 pregnant women across the United States admitted to cannabis use during their pregnancy.8

In such a context, experience from flagship states such as Colorado, which has been a pioneer in US cannabis use and also supports a detailed and public database of congenital defects, is invaluable to ascertain current trends and likely future directions. Cannabis was permitted for medicinal use from November 2000 and was decreed legal in November 2011 with full effect from 2014.

Colorado also has one other considerable advantage that greatly simplifies the statistical analysis of its data, as during the period 2000 to 2014, nationally representative datasets indicate that the use of other drugs was static or falling. In this sense, therefore, the Coloradan context is ideal from a statistical and public health perspective to ascertain current teratological trends while statistically isolating the effect of rising cannabinoid exposure to facilitate the study of prenatal cannabis exposure (PCE).

This study explores the presence of any overall trends in the pattern of Coloradan congenital anomalies data and investigates the extent to which ecologically documented drug use trends explained some of this variance.

Methods

Data

Data on birth defects in Colorado were taken from the Colorado Responds to Children with Special Needs (CRCSN) online database as single data points in January 2019.9 Total 2013 defect data were taken from the April 2018 CRCSN dataset. Data on drug use were taken from the National Survey of Drug Use and Health (NSDUH) conducted annually by the Substance Abuse and Mental Health Administration (SAMHSA).8 Data on cannabinoid concentration were taken from the National Drug Enforcement Agency seizures10,11 and multiplied by annual cannabis use to derive state-wide cannabinoid exposure.

Relationship to Cannabis

Defects were classified as cannabis-related if strong published evidence had previously identified a relationship to cannabis exposure. Papers from Centers for Disease Control and Prevention (CDC) and National Birth Defects Prevention Network (NBDPN) have established that anencephaly,12,13 diaphragmatic hernia, esophageal atresia with or without tracheoesophageal fistula, and gastroschisis are cannabis-related.12 A joint statement by the American Academy of Pediatrics and the American Heart Association linked Ebstein’s anomaly and ventricular septal defect (VSD) with cannabis use.14 A large 2007 epidemiological study from Hawaii also linked encephalocele, hypoplastic left heart, syndactyly, reduction deformity of the upper limbs, hydrocephaly, cleft lip and cleft palate both separately and together, anotia/microtia, tetralogy of Fallot, pyloric stenosis, microcephaly, pulmonary valve atresia and/or stenosis, large bowel or rectal atresias or stenosis, obstructive genitourinary defect, polydactyly, atrial septal defect (ASD), and trisomy 21 with PCE.15 Although this study is an outlier in terms of the literature, this list of defects was accepted as being cannabis-related in view of its high predictive value and pointed real-world applicability particularly in the United States (see Results and Discussion sections).

Statistics

Data were processed in “R” v3.5.2 and “R Studio” v1.1.463 from the Central “R” Archive Network. Model reduction was conducted by the classical method with progressive removal of the least significant term. Models were compared by analysis of variance (ANOVA). Model parameters were compared with the “purrr” and “broom” packages. Regression line slope change was assessed with the “segmented” package. Differing quantitative scales were adjusted using the “scales” package. The “nlme” package was used for mixed-effects regression. Principal components analysis was conducted using the “psych” package. P < .05 was considered significant.

Ethics

The study was approved by the Human Research Ethics Committees of South City Medical Centre and the University of Western Australia.

Results

The January 2019 CRCSN dataset consists of annual numbers and rates on 49 defects for each of the years 2000 through 2014 and comprises 746 data points together with 180 data points relating to 13 summary indices by major organ system. These defects are graphed by time in Figures 1 and 2.

Figure 1.

Figure 1.

Open in a new tab

Colorado congenital defects A-E by time, regression lines fitted.

Figure 2.

Figure 2.

Open in a new tab

Colorado congenital defects H-V by time, regression lines fitted.

Table 1 lists the slope and confidence intervals of these time-dependent changes. Seven defects are noted to be significantly rising and 2 significantly falling. Table 2 repeats this exercise for the major defect summary groups. Nine of 11 slopes are noted to be rising. Supplementary Figures 1 and 2 (available online) present loess curves for these data.

Table 1.

Time-Dependent Trends of CRCSN Defects.

Defect Term β-Estimate Standard Error t P Lower CI Upper CI
Atrial septal defect secundum Year 6.4518 0.7943 8.1229 .0000 4.7359 8.1677
Ventricular septal defect Year 1.1825 0.1623 7.2866 .0000 0.8319 1.5331
Patent ductus arteriosus Year 0.9925 0.2382 4.1660 .0011 0.4778 1.5072
Chromosomal anomalies Year 0.6543 0.1545 4.2357 .0010 0.3206 0.9880
Anomalies pulmonary artery Year 0.4621 0.3210 1.4396 .1736 −0.2314 1.1556
Microcephalus Year 0.3046 0.0812 3.7519 .0024 0.1292 0.4801
Trisomy 21 Year 0.1850 0.0673 2.7480 .0166 0.0396 0.3304
Renal agenesis Year 0.0961 0.0394 2.4378 .0299 0.0109 0.1812
Total anencephalus and spina bifida Year 0.0843 0.0467 1.8052 .0942 −0.0166 0.1852
Hirschsprung’s Year 0.0754 0.0457 1.6485 .1232 −0.0234 0.1741
Spina bifida without anencephalus Year 0.0693 0.0405 1.7094 .1111 −0.0183 0.1568
Anomalies abdominal wall Year 0.0507 0.0528 0.9610 .3541 −0.0633 0.1647
Choanal atresia Year 0.0489 0.0405 1.2086 .2483 −0.0385 0.1364
Microphthalmos Year 0.0296 0.0305 0.9723 .3486 −0.0362 0.0955
Endocardial cushion defects Year 0.0221 0.0434 0.5106 .6182 −0.0716 0.1158
Anencephalus Year 0.0154 0.0224 0.6847 .5056 −0.0331 0.0638
Trisomy 18 Year 0.0150 0.0267 0.5619 .5838 −0.0427 0.0727
Anophthalmos Year 0.0139 0.0218 0.6402 .5332 −0.0331 0.0609
Encephalocele Year 0.0129 0.0110 1.1654 .2648 −0.0110 0.0367
Transposition great vessels Year 0.0107 0.0540 0.1983 .8458 −0.1060 0.1274
Congenital biliary atresia Year 0.0096 0.0273 0.3538 .7291 −0.0492 0.0685
Exstrophy urinary bladder Year 0.0072 0.0166 0.4304 .6770 −0.0305 0.0448
Common ventricle Year 0.0064 0.0284 0.2266 .8242 −0.0548 0.0677
Coarctation aorta Year 0.0032 0.0768 0.0418 .9673 −0.1628 0.1692
Congenital scoliosis Year 0.0029 0.0234 0.1222 .9046 −0.0477 0.0534
Polydactyly syndactyly Year −0.0014 0.1724 −0.0083 .9935 −0.3738 0.3710
Leg reduction Year −0.0018 0.0324 −0.0551 .9569 −0.0717 0.0682
Congenital buphthalmos Year −0.0032 0.0157 −0.2043 .8413 −0.0372 0.0308
Common truncus Year −0.0032 0.0267 −0.1205 .9059 −0.0608 0.0544
Orofacial anomalies Year −0.0046 0.0921 −0.0504 .9606 −0.2036 0.1943
Hypoplastic left heart Year −0.0096 0.0367 −0.2631 .7966 −0.0888 0.0695
Cleft Lip with/without cleft palate Year −0.0114 0.0828 −0.1381 .8923 −0.1903 0.1674
Limb reduction Year −0.0125 0.0330 −0.3785 .7112 −0.0838 0.0588
Trisomy 13 Year −0.0125 0.0235 −0.5311 .6043 −0.0633 0.0383
Tracheoesophageal fistula esophageal atresia stenosis Year −0.0146 0.0475 −0.3086 .7625 −0.1172 0.0879
Anomalies diaphragm Year −0.0146 0.0543 −0.2697 .7917 −0.1320 0.1027
Total anomalous pulmonary venous connection Year −0.0204 0.0262 −0.7768 .4512 −0.0770 0.0363
Cleft palate without cleft lip Year −0.0214 0.0916 −0.2340 .8186 −0.2193 0.1764
Atresia stenosis bladder neck Year −0.0304 0.0391 −0.7759 .4517 −0.1149 0.0542
Congenital hydrocephalus without spina bifida Year −0.0318 0.0584 −0.5443 .5954 −0.1579 0.0944
Tetralogy Fallot Year −0.0389 0.0524 −0.7425 .4710 −0.1522 0.0743
Arm reduction Year −0.0414 0.0381 −1.0881 .2963 −0.1237 0.0408
Cong stenosis aortic valve Year −0.0568 0.0337 −1.6866 .1155 −0.1295 0.0160
Hip dysplasia Year −0.0639 0.1674 −0.3819 .7087 −0.4256 0.2977
Congenital cataract Year −0.0689 0.0346 −1.9903 .0680 −0.1437 0.0059
Atresia stenosis large intestine Year −0.0936 0.0641 −1.4594 .1682 −0.2321 0.0449
Infantile cerebral palsy Year −0.1325 0.0730 −1.8158 .0925 −0.2901 0.0251
Pyloric stenosis Year −0.2529 0.1057 −2.3912 .0326 −0.4813 −0.0244
Pulmonary valve stenosis atresia Year −0.3271 0.1009 −3.2417 .0064 −0.5452 −0.1091
Open in a new tab

Abbreviations: CRCSN, Colorado Responds to Children with Special Needs; CI, confidence interval.

Table 2.

Time-Dependent Trends of CRCSN Major Defect Classes.

Defect Term β-Estimate Standard Error t P Lower CI Upper CI
Major Defects Number 2013 Year 228.4791 17.7906 12.8427 .000000 189.7167 267.2415
Major Defects Number 2014 Year 92.9179 11.5577 8.0395 .000002 67.9489 117.8868
Major Defects Rate 2014 Year 15.6757 2.2823 6.8684 .000011 10.7451 20.6063
Major Genitourinary Defects Year 6.1111 0.6297 9.7052 .000000 4.7508 7.4714
Major Cardiovascular Defects Year 6.0657 0.8369 7.2476 .000006 4.2576 7.8738
Major Musculoskeletal Anomalies Year 3.6582 0.5886 6.2149 .000031 2.3866 4.9298
Major Musculoskeletal Defects Year 3.6329 0.5912 6.1449 .000035 2.3556 4.9101
Respiratory Anomalies Year 1.9304 0.2758 6.9991 .000009 1.3345 2.5262
Chromosomal Anomalies Year 0.6543 0.1545 4.2360 .000973 0.3210 0.9880
Major Gastrointestinal Defects Year 0.2061 0.3224 0.6393 .533760 −0.4903 0.9025
Major Eyes Defects Year 0.0289 0.0807 0.3585 .725688 −0.1454 0.2032
Open in a new tab

Abbreviations: CRCSN, Colorado Responds to Children with Special Needs; CI, confidence interval.

Since the data are rather difficult to mentally digest en masse, Figures 3 to 8 present data grouped by organ system. Figure 9 illustrates the summary data by organ system.

Figure 3.

Figure 3.

Open in a new tab

Central nervous system (CNS) defects by time.

Figure 8.

Figure 8.

Open in a new tab

Limb defects by time.

Figure 9.

Figure 9.

Open in a new tab

Major defects by time.

Figure 4.

Figure 4.

Open in a new tab

Neural tube defects by time.

Figure 5.

Figure 5.

Open in a new tab

Cardiovascular defects by time.

Figure 6.

Figure 6.

Open in a new tab

Chromosomal defects by time.

Figure 7.

Figure 7.

Open in a new tab

Face defects by time.

Figure 10 shows the numbers of defects as a total number and as a percentage of live born babies. The total figure in the April 2018 CRCSN dataset is noted to be substantially higher than that in the January 2019 CRCSN dataset. Figure 11 shows the relative rise from baseline of the various categories with the origin of each dataset forming the baseline comparator for that group.

Figure 10.

Figure 10.

Open in a new tab

Total defects by time.

Figure 11.

Figure 11.

Open in a new tab

Relative rises in selected defects compared with baseline by time.

Supplementary Table 1 (available online) shows the summaries of regression models for these major defects and defect classes. Table 3 lists the number of cases in each group by year, sums the total, compares it with the calculated total based on 15 times (2000:2014) the lowest rate in either 2000 or 2001, calculates the absolute and relative case excess, and compares it with the rise in births from 2000 to 2014 of 3.3069%. These relative case excesses are then graphed in order in Figure 12.

Table 3.

Case Excess.

Population
 Specific Defects
 Major Systems
 Total
Year Births Microcephalus Spina_Bifida Atrial_Septal_Defect_Secundum Patent_Ductus_Arteriosus Ventricular_Septal_Defect Trisomy_21 Major_CNS Major_CVS Resp_Anom Major_GUT2 Major_MS Chromosomal_Anomalies Majors_N_2014 Majors_N_2013
2000 65 429 30 18 299 244 272 78 159 898 184 1101 860 157 4026 4830
2001 67 006 33 10 382 239 261 74 168 999 199 1176 800 183 3514 4942
2002 68 420 38 18 518 268 254 92 179 1196 203 1240 856 193 3795 5406
2003 69 325 38 21 548 263 285 104 174 1194 207 1297 814 200 3797 5311
2004 68 491 46 25 568 263 316 88 213 1295 221 1322 814 223 3909 5482
2005 68 929 47 21 571 277 303 95 197 1280 230 1355 929 205 4080 5978
2006 70 732 63 15 578 247 314 99 212 1303 223 1430 927 204 4168 6325
2007 69 580 55 23 553 266 296 97 212 1218 254 1434 927 235 4111 6213
2008 70 024 71 22 770 312 337 81 253 1436 303 1543 1004 195 4592 7010
2009 68 603 67 30 922 352 317 106 246 1478 293 1550 1025 238 4637 6826
2010 66 339 64 17 887 331 328 96 250 1434 296 1444 1112 257 4666 7171
2011 65 026 65 26 866 313 314 88 253 1483 348 1515 1008 227 4728 7174
2012 65 173 46 19 912 324 342 95 231 1498 370 1744 966 220 4619 6939
2013 64 996 56 18 880 324 380 100 264 1585 345 1668 1164 213 5117 8165
2014 65 817 51 24 785 277 359 97 224 1346 282 1581 1058 225 4704
Total 1 013 890 770 307 10039 4300 4678 1390 3235 19643 3958 21400 14264 3175 64463 87772
Calculated total 981 435 450 150 4485 3660 3915 1110 2385 13470 2760 16515 12000 2355 52710 67620
Case excess 32 455 320 157 5554 640 763 280 850 6173 1198 4885 2264 820 11753 20152
% Excess 3.3069% 71.11% 104.67% 123.84% 17.49% 19.49% 25.23% 35.64% 45.83% 43.41% 29.58% 18.87% 34.82% 22.30% 29.80%
Excess relative to births 1.00 21.50 31.65 37.45 5.29 5.89 7.63 10.78 13.86 13.13 8.94 5.71 10.53 6.74 9.01
Open in a new tab

Abbreviations: CNS, central nervous system; CVS, cardiovascular system.

Figure 12.

Figure 12.

Open in a new tab

Rise in selected defects relative to rise in births by time.

En passant one notes that the rate of rise of the 2 common cardiac defects ASD (secundum type) and patent ductus arteriosus (PDA) appears to rise sigmoidally across this time period of the cannabis legalization process (Figure 13). One notes that the quartic model accounts for the time-dependent variance significantly better than the linear model for both ASD (ANOVA F = 6.6319, degrees of freedom [df] = 3, P = .0096) and PDA (ANOVA F = 5.413, df = 3, P = .018).

Figure 13.

Figure 13.

Open in a new tab

Atrial septal defect, ventricular septal defect, and patent ductus arteriosus—Loess curves by time.

Since both the American Academy of Pediatrics and the American College of Obstetricians and Gynecologists concur that drug use in the peripartum period is harmful to the fetus,16,17 it is reasonable to consider the potential role of drug use by the parents in a possible epidemiological association with this overall increasing defect profile.

Drug use in Colorado is presented from the SAMHSA NSDUH data as least squares regression lines in Figure 14, and the slopes of these lines are summarized in Table 4. Only the slopes of the cannabis curves are seen to be rising; the slopes of the tobacco, cigarette, cocaine, and pain reliever curves are falling significantly.

Figure 14.

Figure 14.

Open in a new tab

Drug use in Colorado from National Survey of Drug Use and Health (NSDUH) dataset by time with regression lines fitted.

Table 4.

Regression Slope Trend Estimates for Drug Use—NSDUH.

Drugs Term β-Estimate Standard Error t P Lower CI Upper CI
Cannabis—Annual Year 0.6509 0.1107 5.8808 .0001 0.4097 0.8921
Cannabis—Monthly Year 0.5822 0.0960 6.0671 .0001 0.3731 0.7913
Alcohol Monthly Year 0.1498 0.0825 1.8159 .0925 −0.0284 0.3281
Binge Alcohol Year 0.0703 0.0896 0.7842 .4481 −0.1250 0.2656
Cocaine Annual Year −0.0592 0.0260 −2.2795 .0417 −0.1158 −0.0026
Pain Relievers Year −0.0849 0.0358 −2.3698 .0419 −0.1660 −0.0039
Tobacco Monthly Year −0.2859 0.0933 −3.0651 .0098 −0.4892 −0.0827
Cigarettes Monthly Year −0.3743 0.0817 −4.5838 .0005 −0.5507 −0.1979
Open in a new tab

Abbreviations: NSDUH, National Survey of Drug Use and Health; CI, confidence interval.

Figure 15 presents these drug use data with loess curves. Formal testing for change of regression slope for monthly cannabis use showed a significant change in 2007 from .0293 to .11917 (Davies test, k = 3, P = .0002).

Figure 15.

Figure 15.

Open in a new tab

Drug use in Colorado from National Survey of Drug Use and Health (NSDUH) dataset by time with loess curves fitted.

Monthly cannabinoid exposure was calculated by multiplying the concentration of Federal cannabis seizures by within-state monthly cannabis use. These data are presented as regression lines and loess curves in Figures 16 and 17.

Figure 16.

Figure 16.

Open in a new tab

Cannabinoid exposure in Colorado from National Survey of Drug Use and Health (NSDUH) dataset by time with regression lines fitted.

Figure 17.

Figure 17.

Open in a new tab

Cannabinoid exposure in Colorado from National Survey of Drug Use and Health (NSDUH) dataset by time with loess curves fitted.

Because many of the 49 defects had different quantitative rates, they were scaled to mean of 0 and standard deviation of 1 using the “scales” package. The time-dependent plots shown in Figure 18 were obtained.

Figure 18.

Figure 18.

Open in a new tab

Scaled drug use in Colorado from National Survey of Drug Use and Health dataset by time with loess curves fitted.

A similar exercise was conducted, illustrated in Figure 19, which charts the scaled defect rate as a linear temporal function of the various drug exposures. Increasing levels of binge alcohol, cocaine, cannabis, and pain relievers are all noted to be linked to higher rates of congenital defects. These relationships are demonstrated in Table 5. One notes that the quartic model for cannabis has a higher F value and lower model P value than that for opioid pain relievers (7.83 vs 4.422 and 3.5 × 10−7 vs 3.4 × 10−5).

Figure 19.

Figure 19.

Open in a new tab

Scaled drug use in Colorado from National Survey of Drug Use and Health dataset by time with regression lines fitted.

Table 5.

Regression Slopes for All Scaled Defects by Drug Classes.

Parameter
 Model
Parameter Estimate Standard Error t Pr(>|t|) Adjusted R2 F df P
Linear models
Cannabis
Cannabis_Monthly −9.4175 4.3409 −2.169 0.0304 0.005634 2.932 2680 .05395
Year:Cannabis_Monthly 0.0047 0.0021 2.173 0.0301
Opioids
Year 0.3486 0.1756 1.985 0.0477 0.009133 2.644 3532 .04856
Pain_Relevers 112.5612 62.2827 1.807 0.0713
Year:Pain_Relevers −0.0559 0.0310 −1.804 0.0718
Quartic models
Tobacco
(Year)^3: Tobacco 5.1567 1.7701 2.913 0.0037 0.04503 5.02 8674 4.55E-06
(Year)^3 −145.182 51.2001 −2.836 0.0047
Alcohol
NS
Cannabis
(Year)^2 −13.6307 3.5007 −3.894 0.00011 0.0477 7.833 5677 3.51E-07
(Year)^3 4.8938 1.5169 3.226 0.00132
(Year)^4 −9.6683 1.6810 −5.751 1.3E-08
Cannabis_Monthly 0.2002 0.0710 2.822 0.00492
Opioids
Year −610.237 228.352 −2.672 0.0078 0.04869 4.422 8527 3.39E-05
(Year)^4 −309.336 103.395 −2.992 0.0029
(Year)^2: Pain_Relevers 83.360 35.953 2.319 0.0208
(Year)^4: Pain_Relevers 69.235 22.935 3.019 0.0027
Cocaine
(Year)^2 32.2730 14.3249 2.253 0.0246 0.04574 5.087 8674 3.67E-06
(Year)^2: Cocaine −13.2694 5.0409 −2.632 0.0087
Open in a new tab

Abbreviation: df, degrees of freedom.

Table 6 compares the defect rates against multiple drug exposure in additive models and increasingly complex interactive mixed-effects models with defect as the random variable. Terms including cannabis exposure persisted in final models.

Table 6.

Regression Slopes for All Scaled Defects Against Various Drugs—Mixed-Effects Models.

Parameter Parameter
 Model
Value Standard Error df t P AIC BIC LogLik
Additive model
Rate~Year+Cannabis_Monthly+Opioids+Tobacco+Cocaine+BingeAlc
Opioids 0.3479 0.1560 278 2.2311 .0265 848.7998 867.4013 −419.3999
Year 0.0448 0.0214 278 2.0954 .0370
Increasing levels of interactive models
Rate~Year*Cannabis_Monthly+Opioids+Tobacco+Cocaine+BingeAlc
Opioids 0.4003 0.1756 278 2.2796 .0234 863.8158 882.4173 −426.9079
Year:Cannabis_Monthly 0.0000 0.0000 278 2.0248 .0438
Rate~Year*Cannabis_Monthly*Opioids+Tobacco+Cocaine+BingeAlc
Year 6.3170 1.6760 273 3.7689 .0002 861.0326 898.0704 −420.5163
Opioids 2489.7840 680.8430 273 3.6569 .0003
Year: Opioids −1.2360 0.3390 273 −3.6495 .0003
Cannabis_Monthly: Opioids −392.6320 114.0330 273 −3.4431 .0007
Year: Cannabis_Monthly: Opioids 0.1950 0.0570 273 3.4470 .0007
Cannabis_Monthly 2101.5210 617.6850 273 3.4023 .0008
Year: Cannabis_Monthly −1.0440 0.3060 273 −3.4071 .0008
Rate~Year*Cannabis_Monthly*Tobacco+Opioids+Cocaine+BingeAlc
Year: Cannabis_Monthly −0.0030 0.0009 275 −4.0238 .0001 875.2932 904.9767 −429.6466
Cannabis_Monthly: Tobacco 0.2550 0.0637 275 4.0089 .0001
Year 5.5130 1.4606 275 3.7741 .0002
Tobacco 396.3240 105.2453 275 3.7657 .0002
Year: Tobacco −0.1990 0.0527 275 −3.7677 .0002
Rate~Year+Cannabis_Monthly*Opioids*Tobacco+Cocaine+BingeAlc
Opioids: Tobacco −0.4067 0.1071 272 −3.7971 .0002 866.9679 907.6728 −422.4839
Cannabis_Monthly: Opioids: Tobacco 0.1857 0.0530 272 3.5015 .0005
Cannabis_Monthly: Opioids −4.4878 1.2866 272 −3.4882 .0006
Cannabis_Monthly 18.7962 5.6135 272 3.3484 .0009
Cannabis_Monthly: Tobacco −0.7761 0.2343 272 −3.3130 .0010
Cocaine −1.5894 0.5995 272 −2.6510 .0085
Opioids 6.2896 3.0387 272 2.0698 .0394
Open in a new tab

Abbreviations: df, degrees of freedom; AIC, Akaike information criterion; BIC, Bayesian information criterion; LogLik, log likelihood.

As described in Methods, defects were assigned to be either cannabis-related or not based on reports in the published literature. However, as the Hawaiian report of pyloric stenosis being cannabis-linked15 has not been confirmed elsewhere, this condition was removed from the cannabis-associated group. Moreover, 2 reports from CDC/NBDPN indicate that PCE is linked with anencephaly.12,13 Several drugs linked with anencephaly are similarly linked with spina bifida, which is accepted to be a prototypical neural tube closure defect so that it seems likely that cannabis may also be linked with spina bifida with or without anencephaly. Graphs showing the effect of these 2 adjustments are included as Supplementary Figures 3 to 6 (available online).

Figure 20 shows the time relationship of the 49 scaled defects by the above-described relationship to cannabis. These data are shown on single plots with both loess curves and linear regression lines in Figure 21.

Figure 20.

Figure 20.

Open in a new tab

Scaled defects rate as a function of drug use exposure with regression lines fitted in facetted plot by relationship to cannabis use, after omission of pyloric stenosis and inclusion of spina bifida.

Figure 21.

Figure 21.

Open in a new tab

Scaled defects rate as a function of drug use exposure with (A) loess curves and (B) regression lines fitted.

A model quartic-in-time was superior to a linear-only model (ANOVA F = 4.6099, df = 5, P = .0004).

Table 7 shows that the results of both linear and quartic models are significant with cannabis terms remaining in final models both as a factor and in interaction with time and time-squared.

Table 7.

Comparisons of Cannabinoid Models Linear and Quartic in Time for All Scaled Defects.

Parameter
 Model
Parameter Estimate Standard Error t Pr(>|t|) Adjusted R2 F df P
Linear models
Defect_Rate ~ Year * Cannabis_Related
Year: Cannabis_Related 0.0402 0.0108 3.712 0.0002 0.01523 4.763 3727 .002697
Cannabis_Related −98.8011 33.3927 −2.959 0.0032
Quartic-in-time models
Defect_Rate ~ I(poly(Year, n=4)) * Cannabis_Related
(Year)^4 −4.7042 1.4591 −3.224 0.0013 0.03908 4.711 8722 1.20E-05
Year: Cannabis_Related 5.7531 1.9252 2.988 0.0029
(Year)^2: Cannabis_Related −4.8258 1.9287 −2.502 0.0126
Open in a new tab

Abbreviation: df, degrees of freedom.

Figure 22 shows the time relationship of exposure to various cannabinoids with regression lines, and loess curves are shown in Supplementary Figure 7 (available online).

Figure 22.

Figure 22.

Open in a new tab

Scaled defects rate as a function of cannabinoid exposure with loess curves fitted. Facetted plot by cannabinoid.

Figure 23 shows the defects charted against cannabinoid exposure. These relationships are formalized in Table 8.

Figure 23.

Figure 23.

Open in a new tab

Scaled defects rate as a function of cannabinoid exposure with regression lines fitted. Facetted plot by cannabinoid.

Table 8.

Rises in Defects by Cannabinoid Exposure—Models Linear and Quartic in Time.

Parameter
 Model
Parameter Estimate Standard Error t Pr(>|t|) Adjusted R2 F df P
Linear models
d8_THC
Year −0.2870 0.1131 −2.537 .0117 0.0169 3.496 2289 .03162
Year:d8_THCMON 0.0005 0.0002 2.209 .0280
Quartic models
d9_THC
Year: d9_THC 0.8892 0.3821 2.327 .0203 0.0431 4.84 8674 8.15E-06
(Year)^4 −20.9952 9.1100 −2.305 .0215
Year −99.5298 49.4149 −2.014 .0444
CBD
(Year)^3 55.9750 22.1400 2.528 .0117 0.0489 5.378 8674 1.42E-06
(Year)^3: CBD −14.8342 5.9146 −2.508 .0124
CBN
(Year)^4 −6.6144 1.2950 −5.108 4.3E-07 0.0477 7.832 5577 3.52E-07
(Year)^2 −8.8041 2.0599 −4.274 2.2E-05
(Year)^3 4.8154 1.5064 3.197 .0015
CBN 0.1645 0.0583 2.821 .0049
CBG
NS
THCV
(Year)^4 −8.0989 1.4039 −5.769 1.2E-08 0.04671 7.683 5677 4.88E-07
(Year)^2 −8.4791 2.0250 −4.187 3.2E-05
THCV 0.9907 0.3681 2.691 .0073
(Year)^3 3.7194 1.4095 2.639 .0085
Open in a new tab

Abbreviation: df, degrees of freedom.

Figure 24 illustrates the complex relationship between monthly cannabis use, falling cannabidiol concentration, and the population exposure to cannabidiol.

Figure 24.

Figure 24.

Open in a new tab

(A) Monthly cannabis use by year. (B) Cannabidiol concentration by year. (C) Cannabidiol exposure by year as the product of (A) and (B).

Figure 25 is a point and box plot graph of the movement of cannabis-related versus nonrelated defects for each year to address the complex relationship of cannabidiol exposure.

Figure 25.

Figure 25.

Open in a new tab

Relationship of scaled defects by year to described relationship to cannabis consumption from the published literature (see Discussion).

Figure 26A shows these 2 rates side by side from 2000 to 2014. The difference between the 2 groups is plotted in Figure 26B, and their adjusted ratio (adjusted by adding unity [1] to numerator and denominator) appears in Figure 26C. Figure 26D shows the ratio of the absolute values of the cannabis-related and non–cannabis-related values, which correlates broadly with cannabidiol exposure (Figure 24C, R = 0.4857, P = .0783). These measures clearly peaked in 2009-2010 when cannabidiol exposure also peaked.

Figure 26.

Figure 26.

Open in a new tab

(A) Box plot of relationship of scaled defects to time by described relationship to cannabis use. (B) Difference between cannabis-related and non–cannabis-related rates of scaled scores. (C) Ratio of cannabis-related and non–cannabis-related scaled scores after adjustment by adding unity (1) to both scores. (D) The ratio of the absolute value of the cannabis-related defects to that of the absolute value of cannabis-unrelated defects.

Figure 9 and Table 2 showed that defects in 5 major organ systems are rising: central nervous system, cardiovascular, genitourinary, musculoskeletal, and respiratory systems. These 5 may then be combined by principal component analysis. A scree plot (Supplementary Figure 8, available online) shows that 1 principal component—PC1—was sufficient to combine these data and accounted for 90% of the variance. Together with total rates from the CRCSN dataset, this produces 3 summary statistics, the totals for 2013, 2014, and PC1.

Figure 27 charts these parameters against each other along with the monthly cannabis exposure. A close visual relationship is immediately apparent. These correlations are presented formally in Table 9.

Figure 27.

Figure 27.

Open in a new tab

Line plots showing relationship between scaled (A) major defect rates 2013 and last month cannabis use; (B) major defect rates 2014 and last month cannabis use; (C) principal component 1 and last month cannabis use; (D) major defects 2014 and principal component 1 for 2014.

Table 9.

Correlation Coefficients—Major Summary Indices With Cannabis Use (see Figure 27).

Group 1 Group 2 t df P R Lower CI Upper CI
Major Defects 2014 Cannabis_Monthly 4.2597 12 .0011 0.7758 0.4169 0.9255
Major Defects 2013 Cannabis_Monthly 5.1534 11 .0003 0.8409 0.5402 0.9512
PC1 Cannabis_Monthly 4.2722 12 .0011 0.7767 0.4187 0.9258
Majors Defects 2014 PC1 11.035 13 5.7E-08 0.9505 0.8542 0.9838
Open in a new tab

Abbreviations: df, degrees of freedom; CI, confidence interval.

Table 10 summarizes the regression of all scaled defects against various drug combinations.

Table 10.

Linear Regression of Major Summary Indices by NSDUH Drug Exposure.

Parameter Model
Parameter Estimate Standard Error T Pr(>|t|) Adjusted R2 F df P
Linear models
Majors2014~Year+Tobacco*Cannabis_Monthly+Opioids+Binge_Alcohol+Cocaine
Cannabis_Monthly 1.2139 0.1678 7.234 .0002 0.8656 22.48 3,7 .0005718
Tobacco: Cannabis_Monthly 0.4423 0.1116 3.964 .0054
Opioids 0.3683 0.1373 2.683 .0314
Majors2014~Year*Δ9-THC+CBDM+Tobacco+Opioids
Δ9-THC 253.3969 94.7077 2.676 .0281 0.7856 19.33 2,8 .0095
Year: Δ9-THC −0.1257 0.0471 −2.668 .0284
PC1~Year+Tobacco*Cannabis_Monthly+Opioids+Binge_Alcohol+Cocaine
Cannabis_Monthly 1.2404 0.1962 6.322 .0004 0.8221 16.4 3,7 .001508
Tobacco: Cannabis_Monthly 0.4829 0.1305 3.701 .0076
Opioids 0.3957 0.1605 2.465 .0431
Major_CNS~Year+Tobacco*Cannabis_Monthly+Opioids+Binge_Alcohol+Cocaine
Cannabis_Monthly 1.1631 0.2332 4.988 .0016 0.6956 8.619 3,7 .00949
Opioids 0.6071 0.1908 3.182 .0154
Tobacco: Cannabis_Monthly 0.4133 0.1551 2.665 .0322
Major_CVS~Year+Tobacco*Cannabis_Monthly+Opioids+Binge_Alcohol+Cocaine
Cannabis_Monthly 1.1303 0.2331 4.85 .0019 0.6951 8.599 3,7 .009549
Tobacco: Cannabis_Monthly 0.4770 0.1550 3.077 .0179
Opioids 0.4469 0.1907 2.344 .0516
Quartic-in-time models
PC1~I(poly(Year, n=4))*Δ9-THC+CBDM+Tobacco+Opioids
Δ9-THC 0.6793 0.1625 4.181 .0024 0.6223 17.48 1,9 .0095
Majors2014~(poly(Year, n=4))*Cannabis_Monthly+Tobacco+Opioids+Binge_Alcohol+Cocaine
(Year)^4: Cannabis_Monthly −20.4192 0.9427 −21.66 .0294 0.9982 609.9 9,1 .03141
Year: Cannabis_Monthly 302.1905 14.3060 21.12 .0301
Cannabis_Monthly −58.0400 2.8018 −20.71 .0307
(Year)^3: Cannabis_Monthly 99.9858 4.8684 20.54 .0310
(Year)^2: Cannabis_Monthly −209.6725 10.3174 −20.32 .0313
Year 340.6153 16.8783 20.18 .0315
(Year)^2 −270.0904 13.4573 −20.07 .0317
(Year)^3 119.3061 6.4290 18.56 .0343
(Year)^4 −40.3087 2.4539 −16.43 .0387
PC1~I(poly(Year, n=4))*Cannabis_Monthly+Tobacco+Opioids+Binge_Alcohol+Cocaine
Year: Cannabis_Monthly −5.7912 1.1506 −5.033 .0024 0.8062 11.4 4,6 .005744
(Year)^2: Cannabis_Monthly 5.0229 1.0371 4.843 .0029
(Year)^3: Cannabis_Monthly 2.9675 0.8060 3.682 .0103
(Year)^4: Cannabis_Monthly −1.6296 0.6479 −2.515 .0456
Major_CNS~I(poly(Year, n=4))*Cannabis_Monthly+Tobacco+Opioids+Binge_Alcohol+Cocaine
Year: Cannabis_Monthly −5.6711 1.5086 −3.759 .0094 0.5964 4.694 4,6 .04649
(Year)^2: Cannabis_Monthly 4.2187 1.3598 3.102 .0211
(Year)^3: Cannabis_Monthly 2.6681 1.0568 2.525 .0450
Major_CVS~I(poly(Year, n=4))*Cannabis_Monthly+Tobacco+Opioids+Binge_Alcohol+Cocaine
Cannabis_Monthly −54.5060 3.3030 −16.5 .0385 0.997 373 9,1 .04016
Year: Cannabis_Monthly 274.4250 16.8670 16.27 .0391
(Year)^2: Cannabis_Monthly −192.5180 12.1640 −15.827 .0402
(Year)^4: Cannabis_Monthly −17.3420 1.1110 −15.604 .0407
Year 309.6860 19.8990 15.562 .0409
(Year)^3: Cannabis_Monthly 84.3810 5.7400 14.701 .0432
(Year)^2 −228.2030 15.8660 −14.383 .0442
(Year)^3 103.9890 7.5800 13.719 .0463
(Year)^4 −28.4810 2.8930 −9.844 .0644
Open in a new tab

Abbreviations: NSDUH, National Survey of Drug Use and Health; df, degrees of freedom.

Table 11 is a regression summary for all scaled defects against various cannabinoids.

Table 11.

Linear Regression of Major Summary Indices Against Selected Cannabinoids.

Parameter
 Model
Parameter Estimate Standard Error T Pr(>|t|) Adusted R2 F df P
Quartic-in-time models
Additive models
Majors2014~poly(Year, n=4)+Δ9-THC+CBD+CBN+THCV
Δ9-THC 0.8746 0.1535 5.699 .0001 0.7077 32.48 1,12 9.9E-05
PC1~poly(Year, n=4)+Δ9-THC+CBD+CBN+THCV
Year 2.5605 0.5875 4.358 .0024 0.9582 60.67 5,8 3.8E-06
(Year)^4 −0.8531 0.2711 −3.147 .0137
(Year)^2 −1.0638 0.4303 −2.472 .0386
CBN 0.4179 0.1789 2.336 .0477
Major_CVS~poly(Year, n=4)+Δ9-THC+CBD+CBN+THCV
(Year)^4 −1.5260 0.3572 −4.273 .0027 0.9149 28.97 5,8 6.4E-05
(Year)^2 −1.9670 0.5670 −3.469 .0085
Year 2.3018 0.7741 2.974 .0178
CBN 0.5051 0.2357 2.143 .0645
Majors2013~poly(Year, n=4)+Δ9-THC+CBD+CBN+THCV
CBD −1.3975 0.3085 −4.530 .0062 0.9718 60.04 7,5 .0002
THCV 1.1978 0.2796 4.284 .0078
(Year)^4 1.9973 0.4900 4.076 .0096
(Year)^3 −2.4039 0.6190 −3.884 .0116
(Year)^2 −3.5165 1.0913 −3.222 .0234
CBN 0.8580 0.2758 3.111 .0265
Year −8.0678 2.6757 −3.015 .0296
Interactive models
PC1~poly(Year, n=4)*Δ9-THC+CBD+CBN+THCV
Year: Δ9-THC −2.1082 0.3168 −6.655 .0006 75.94 75.94 7,6 2.0E-05
CBD −0.3900 0.0713 −5.469 .0016
CBN 0.5980 0.1536 3.894 .0080
THCV 0.4206 0.1444 2.914 .0268
Majors2013~poly(Year, n=4)*LMCann+Tob+Opioids+Binge_Alcohol+Cocaine
(Year)^3 4.5310 0.9640 4.700 .0053 0.8206 11.29 4,5 .0102
Year 3.0367 0.9432 3.219 .0235
(Year)^2 −2.9994 1.1589 −2.588 .0490
Majors2013~poly(Year, n=4)*Δ9-THC+CBD+CBN+THCV
Δ9-THC 3.7695 0.0983 38.354 .0166 0.9999 14750.0 11,1 .0064
CBD −1.2900 0.0356 −36.203 .0176
(Year)^4: Δ9-THC 9.4066 0.2628 35.801 .0178
Year: Δ9-THC −31.1351 0.9101 −34.211 .0186
(Year)^4 23.1233 0.7273 31.791 .0200
(Year)^2 40.8303 1.3204 30.923 .0206
(Year)^3: Δ9-THC −23.2475 0.8131 −28.59 .0223
CBN 1.0334 0.0472 21.908 .0290
Year −4.7978 0.3662 −13.103 .0485
Open in a new tab

Abbreviation: df, degrees of freedom.

Table 12 presents final regression models of various key summary parameters against the indicated combinations of drugs and cannabinoids in linear and/or time-quartic models.

Table 12.

Linear Regression of Major Defect Indices Against Drugs and Cannabinoids Together.

Parameter
 Model
Parameter Estimate Standard Error t Pr(>|t|) Adjusted R2 F df P
Linear models
Majors2014~Year*Δ9-THC+CBDM+Tobacco+Opioids
Δ9-THC 253.3969 94.7077 2.676 .0281 0.7856 19.33 2,8 .0008648
Year:Δ9-THC −0.1257 0.0471 −2.668 .0284
Quartic-in-time models
Majors2014~poly(Year, n=4)*Δ9-THC+CBDM+Tobacco+Opioids
(Year)^3: Δ9-THC 3.8734 1.1020 3.515 .0126 0.7258 7.619 4,6 .01561
(Year)^2: Δ9-THC −5.4212 1.5780 −3.435 .0139
Year: Δ9-THC 4.0698 1.3505 3.013 .0236
(Year)^4: Δ9-THC −1.7240 0.8807 −1.958 .0980
PC1~poly(Year, n=4)*Δ9-THC+CBDM+Tobacco+Opioids
Δ9-THC 0.6793 0.1625 4.181 .0024 0.6223 17.48 1,9 .002374
Major_CNS~poly(Year, n=4)*Δ9-THC+CBDM+Tobacco+Opioids
(Year)^2: Δ9-THC −6.4699 1.7370 −3.725 .0098 0.6103 4.915 4,6 .04217
Year: Δ9-THC 4.8429 1.4866 3.258 .0173
(Year)^3: Δ9-THC 2.7165 1.2130 2.239 .0664
Major_CVS~poly(Year, n=4)*Δ9-THC+CBDM+Tobacco+Opioids
Δ9-THC 0.5267 0.1820 2.894 .0178 0.4244 8.374 1,9 .01777
Majors2013~poly(Year, n=4)*Δ9-THC+CBDM+Tobacco+Opioids
(Year)^3: Δ9-THC 4.7555 1.0784 4.410 .0070 0.7843 9.179 4,5 .01592
Year: Δ9-THC 3.3902 1.2164 2.787 .0386
(Year)^2: Δ9-THC −3.6253 1.3525 −2.680 .0438
Open in a new tab

Abbreviation: df, degrees of freedom.

Discussion

This study portrays a detailed picture of congenital defects in the state of Colorado based on the latest intrastate defect registry data from CRCSN and provides compelling evidence that the generally rising pattern both of individual defects and of systems levels summary and total measures closely parallels the rise in cannabis use in Colorado in the context of static or falling levels of other drug use.

While there is substantial heterogeneity in the trend of birth defects in Colorado, the overall trend of the CRCSN dataset is upward, a trend that closely parallels cannabis use during the progression of that state toward cannabis legalization. This is reflected in some of the most common birth defects such as ASD, PDA, VSD, and Down’s syndrome and also in summary measures such as central nervous, cardiovascular, respiratory, chromosomal, and genitourinary defects, the overall total defects in both 2013 and 2014 and on principal component analysis. Indeed, ASD and PDA showed an uptick temporally associated with rising cannabis use. Cannabis use showed a statistically significant rise about 2007 related to the movement toward cannabis legalization. Moreover, the relationship to cannabis use was robust to multivariate adjustment with all other drug use. Data implicated several cannabinoids including Δ9-tetrahydrocannabinol, Δ8-tetrahydrocannabinol, tetrahydrocannabivarin, cannabinol, and cannabidiol. Although the relationship with cannabidiol is temporally complex, data show that the relative elevation of cannabis-related defects compared with non–cannabis-related defects peaked in 2009 to 2010 when cannabidiol exposure was peaking.

It should be underscored again that the reported changes are all at the associational level only: such a study cannot by itself establish or interrogate causal pathways.

Moreover, as has been described elsewhere, numerous published mechanistic reports link PCE with molecular pathways to teratogenesis and form a critical backdrop and highly pertinent context to the present report.18-23 This confluence of strong mechanistic links together with the present compelling teratological profile in the situation where the use of other drugs is uniformly static or falling strengthens the argument that causal pathways may be operating in clinical populations.

Space precludes detailed consideration of possible teratogenic mechanisms, but these have been addressed elsewhere.18-23 Neurotoxic mechanisms include withdrawal of glutamate receptors from synapses,24 misconstruction of synapses from disruption of neurexin-neuroligin synaptic scaffolding,25 excessive dendritic and spine pruning,26 mitochondrial impairment,27 stem cell inhibition,28 CB1R-mediated neuraxis inflammation,29 and cytoskeletal impairment and motility disruption.30 Cardiovascular toxic mechanisms include inflammatory vasculitis and CB1R signaling to CB1R-rich endovascular and endocardial tissues.31,32 Importantly, cannabis has been described as blocking both notch33,34 and robo-slit receptor-ligand35 signaling, which are important as both neuronal and vascular guidance cues,36 and critically involved in heart and brain morphogenesis.36 Cannabis induces severe epigenetic disruption22,37-39 and has long been known to stimulate micronucleus formation and genetic anomalies secondary to chromosomal missegregation.22,40

The present work did not have access to Coloradan early termination of pregnancy for anomaly data. Since many of the defects mentioned are known to be carefully sought by prenatal screening programs and have high applicable termination rates, the present results represent underestimates and set a lower bound for effect, which is likely to be greatly exacerbated by incorporation of the complete dataset.

Some discussion of the attribution of cannabis association to the listed defects is appropriate. Many of the defects listed as cannabis-associated have been attributed as such based on the large population survey of Forrester and Merz from Hawaii in 2007.15 While this article is an outlier in the clinical cannabis-related teratogenesis literature, albeit highly concordant with previous animal studies,1,2 its very uniqueness places it in a signal position to face the most stringent test of predictive theories, namely, the test of prediction of future trends. By this test, the Forrester-Merz article towers above the remainder of the literature. It alone predicts the increased incidence of ASD, Downs’ syndrome, microcephaly, and chromosomal defects found in the present study. Moreover, this is the only study that explains the current pattern of cannabis-related defects such as ASD, Down’s syndrome, VSD, encephalocele, limb reductions, anotia, and gastroschisis across the high cannabis-using states of the United States41 and recently reported elevated rates of limb defects in France in hemp-fed cattle and babies.42,43 As noted above, pyloric stenosis was omitted from the cannabis-related group as it has not been independently verified by other studies, and spina bifida is believed to share much in common with other neural tube closure defects such as anencephalus so this has been included.

Four of 4 longitudinal studies of cortical executive functioning following PCE indicate serious deficits in cerebral associational function.44-48 Data on these deficits are not included within the CRCSN dataset, which therefore forms an additional disease burden to that described above. However, one notes that there has been a movement in Colorado for several years to declare a state of medical emergency related to a rapidly accelerating renaissance of autistic spectrum disorders in that community.49 Importantly, rapid growth of autism in Colorado may shortly overshadow the classical anomalies described in the present report, which again suggests that this work describes a lower bound of cannabis teratogenesis.

Taken together, these various data imply that the full spectrum of cannabis-associated defects is potentially much broader than has previously been delineated. It may still be expanding.

A major finding of this statistical study was that models quartic in time outperformed strictly linear models. This suggests a feed-forward–type positive-feedback process.

In October 2018, the CRCSN revised their total database from 2000 to 2014 without explanation in a manner that mainly affected the total congenital anomalies. The previous historical totals from 2000 to 2013 appear as indicated.

This study has several strengths. Colorado is unusual among the United States in that it makes extracts from its birth defects register publicly available. Colorado is also unusual as it is one of the only states with legal cannabis to do so. This study also utilizes the very large nationally representative NSDUH dataset to assess intrastate drug exposure. Limitations include the lack of individual-level drug use data, which might be available to a case-control study. Due to the uncertainties involved with self-report studies,6 we would suggest that future studies employ objective evidence of drug exposure such as hair analysis.50

Conclusion

An excess of 11 753 to 20 152 birth defects occurred in Colorado from 2000 to 2014, which represents a 6.7- to 9.4-fold excess of growth in defects compared with growth in births. Defects in 6 of 8 major organ systems increased significantly in frequency. While other drug use was falling over this period, cannabis use alone rose. Cannabis and many cannabinoids were shown to be associationally linked with this rise with correlation coefficients up to 0.78, were confirmed on bivariate analysis, and were robust to multivariate adjustment. In the context of multiple mechanistic pathways, causality is strongly implied. Longitudinal case-control series denominated by an objective measures of drug use are indicated.

Author Contributions

ASR did the background research, analyzed data and wrote the first draft. GKH provided supervision, provided administrative, professional and academic support of several types, provided meaningful intellectual input and co-wrote the final draft of this paper.

Supplemental Material

Supplementary_Table_and_figures – Supplemental material for Cannabis Teratology Explains Current Patterns of Coloradan Congenital Defects: The Contribution of Increased Cannabinoid Exposure to Rising Teratological Trends
Click here for additional data file. (1.7MB, pdf)

Supplemental material, Supplementary_Table_and_figures for Cannabis Teratology Explains Current Patterns of Coloradan Congenital Defects: The Contribution of Increased Cannabinoid Exposure to Rising Teratological Trends by Albert Stuart Reece and Gary Kenneth Hulse in Clinical Pediatrics

Footnotes

Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD: Albert Stuart Reece Inline graphic https://orcid.org/0000-0002-3256-720X

Supplemental Material: Supplemental material for this article is available online.

References

  • 1. Geber WF, Schramm LC. Teratogenicity of marihuana extract as influenced by plant origin and seasonal variation. Arch Int Pharmacodyn Ther. 1969;177:224-230. [PubMed] [Google Scholar]
  • 2. Graham JDP. Cannabis and health. In: Graham JDP, ed. Cannabis and Health. 1st ed London, England: Academic Press; 1976:271-320. [Google Scholar]
  • 3. Volkow ND, Compton WM, Weiss SR. Adverse health effects of marijuana use. N Engl J Med. 2014;371:879. [DOI] [PubMed] [Google Scholar]
  • 4. Volkow ND, Compton WM, Wargo EM. The risks of marijuana use during pregnancy. JAMA. 2017;317:129-130. [DOI] [PubMed] [Google Scholar]
  • 5. Volkow ND, Han B, Compton WM, Blanco C. Marijuana use during stages of pregnancy in the United States. Ann Intern Med. 2017;166:763-764. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Young-Wolff KC, Tucker L, Alexeeff S, et al. Trends in self-reported and biochemically tested marijuana use among pregnant females in California from 2009-2016. JAMA. 2017;318:2490-2491. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Dickson B, Mansfield C, Guiahi M, et al. Recommendations from cannabis dispensaries about first-trimester cannabis use. Obstet Gynecol. 2018;131:1031-1038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Substance Abuse and Mental Health Services Administration. National Survey on Drug Use and Health. https://www.samhsa.gov/data/data-we-collect/nsduh-national-survey-drug-use-and-health. Accessed July 15, 2018.
  • 9. Department of Public Health and the Environment. Colorado Responds to Children With Special Needs—Birth Defect Data, Colorado. Denver, CO: Department of Public Health and the Environment; 2018. [Google Scholar]
  • 10. ElSohly MA, Mehmedic Z, Foster S, Gon C, Chandra S, Church JC. Changes in cannabis potency over the last 2 decades (1995-2014): analysis of current data in the United States. Biol Psychiatry. 2016;79:613-619. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. ElSohly MA, Ross SA, Mehmedic Z, Arafat R, Yi B, Banahan BF., 3rd Potency trends of delta9-THC and other cannabinoids in confiscated marijuana from 1980-1997. J Forensic Sci. 2000;45:24-30. [PubMed] [Google Scholar]
  • 12. van Gelder MMHJ, Donders ART, Devine O, Roeleveld N, Reefhuis J; National Birth Defects Prevention Study. Using Bayesian models to assess the effects of under-reporting of cannabis use on the association with birth defects, national birth defects prevention study, 1997-2005. Paediatr Perinat Epidemiol. 2014;28:424-433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. van Gelder MMHJ, Reefhuis J, Caton AR, Werler MM, Druschel CM, Roeleveld N; National Birth Defects Prevention Study. Maternal periconceptional illicit drug use and the risk of congenital malformations. Epidemiology. 2009;20:60-66. [DOI] [PubMed] [Google Scholar]
  • 14. Jenkins KJ, Correa A, Feinstein JA, et al. ; American Heart Association Council on Cardiovascular Disease in the Young. Noninherited risk factors and congenital cardiovascular defects: current knowledge: a scientific statement from the American Heart Association Council on Cardiovascular Disease in the Young: endorsed by the American Academy of Pediatrics. Circulation. 2007;115:2995-3014. [DOI] [PubMed] [Google Scholar]
  • 15. Forrester MB, Merz RD. Risk of selected birth defects with prenatal illicit drug use, Hawaii, 1986-2002. J Toxicol Environ Health A. 2007;70:7-18. [DOI] [PubMed] [Google Scholar]
  • 16. American College of Obstetricians and Gynecologists. Substance abuse and pregnancy. https://www.acog.org/About-ACOG/ACOG-Departments/State-Legislative-Activities/Substance-Abuse-and-Pregnancy?IsMobileSet=false. Accessed February 4, 2019.
  • 17. Behnke M, Smith VC; Committee on Substance Abuse; Committee on Fetus and Newborn. Prenatal substance abuse: short- and long-term effects on the exposed fetus. Pediatrics. 2013;131:e1009-e1024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Reece AS, Hulse GK. Explaining contemporary patterns of cannabis teratology. Clin Pediatr OA. 2019;4:1000146. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Reece AS. Cannabis problematics include but are not limited to pain management. JAMA. https://jamanetwork.com/journals/jama/article-abstract/2723649. Accessed June 25, 2019.
  • 20. Reece AS, Hulse GK. Gastroschisis and autism—dual canaries in the Californian coal mine. JAMA Surg. 2019;154:366-367. [DOI] [PubMed] [Google Scholar]
  • 21. Reece AS, Norman A, Hulse GK. Cannabis exposure as an interactive cardiovascular risk factor and accelerant of organismal ageing: a longitudinal study. BMJ Open. 2016;6:e011891. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22. Reece AS, Hulse GK. Chromothripsis and epigenomics complete causality criteria for cannabis- and addiction-connected carcinogenicity, congenital toxicity and heritable genotoxicity. Mutat Res. 2016;789:15-25. [DOI] [PubMed] [Google Scholar]
  • 23. Richardson KA, Hester AK, McLemore GL. Prenatal cannabis exposure—the “first hit” to the endocannabinoid system. Neurotoxicol Teratol. 2016;58:5-14. [DOI] [PubMed] [Google Scholar]
  • 24. Martel G, Uchida S, Hevi C, et al. Genetic Demonstration of a role for stathmin in adult hippocampal neurogenesis, spinogenesis, and NMDA receptor-dependent memory. J Neurosci. 2016;36:1185-1202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Anderson GR, Aoto J, Tabuchi K, et al. β-Neurexins control neural circuits by regulating synaptic endocannabinoid signaling. Cell. 2015;162:593-606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Miller ML, Chadwick B, Dickstein DL, et al. Adolescent exposure to Δ9-tetrahydrocannabinol alters the transcriptional trajectory and dendritic architecture of prefrontal pyramidal neurons. Mol Psychiatry. 2019;24:588-600. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Hebert-Chatelain E, Desprez T, Serrat R, et al. A cannabinoid link between mitochondria and memory. Nature. 2016;539:555-559. [DOI] [PubMed] [Google Scholar]
  • 28. Noonan MA, Eisch AJ. Regulation of adult neurogenesis by cannabinoids. Chem Today. 2006;24:84-88. [Google Scholar]
  • 29. Cutando L, Maldonado R, Ozaita A. Microglial activation and cannabis exposure. In: Preedy VR, ed. Handbook of Cannabis and Related Pathologies: Biology, Pharmacology, Diagnosis and Treatment. 1st ed New York, NY: Academic Press; 2017:401-412. [Google Scholar]
  • 30. Wang J, Yuan W, Li MD. Genes and pathways co-associated with the exposure to multiple drugs of abuse, including alcohol, amphetamine/methamphetamine, cocaine, marijuana, morphine, and/or nicotine: a review of proteomics analyses. Mol Neurobiol. 2011;44:269-286. [DOI] [PubMed] [Google Scholar]
  • 31. Barber PA. Cannabis and stroke. In: Preedy VR, ed. Handbook of Cannabis and Related Pathologies: Biology, Pharmacology and Treatment. 1st ed New York, NY: Academic Press; 2017:486-493. [Google Scholar]
  • 32. Menahem S. Cardiovascular effects of cannabis usage. In: Preedy VR, ed. Handbook of Cannabis and Related Pathologies: Biology, Pharmacology and Treatment. 1st ed New York, NY: Academic Press; 2017:481-485. [Google Scholar]
  • 33. Kim D, Lim S, Park M, et al. Ubiquitination-dependent CARM1 degradation facilitates Notch1-mediated podocyte apoptosis in diabetic nephropathy. Cell Signal. 2014;26:1774-1782. [DOI] [PubMed] [Google Scholar]
  • 34. Tanveer R, Gowran A, Noonan J, Keating SE, Bowie AG, Campbell VA. The endocannabinoid, anandamide, augments Notch-1 signaling in cultured cortical neurons exposed to amyloid-beta and in the cortex of aged rats. J Biol Chem. 2012;287:34709-34721. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Alpar A, Tortoriello G, Calvigioni D, et al. Endocannabinoids modulate cortical development by configuring Slit2/Robo1 signalling. Nat Commun. 2014;5:4421. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Carlson BM. Human Embryology and Developmental Biology. Philadelphia, PA: Elsevier; 2014. [Google Scholar]
  • 37. Szutorisz H, Hurd YL. High times for cannabis: epigenetic imprint and its legacy on brain and behavior. Neurosci Biobehav Rev. 2018;85:93-101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Zumbrun EE, Sido JM, Nagarkatti PS, Nagarkatti M. Epigenetic regulation of immunological alterations following prenatal exposure to marijuana cannabinoids and its long term consequences in offspring. J Neuroimmune Pharmacol. 2015;10:245-254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. DiNieri JA, Wang X, Szutorisz H, et al. Maternal cannabis use alters ventral striatal dopamine D2 gene regulation in the offspring. Biol Psychiatry. 2011;70:763-769. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Russo C, Ferk F, Mišík M, et al. Low doses of widely consumed cannabinoids (cannabidiol and cannabidivarin) cause DNA damage and chromosomal aberrations in human-derived cells. Arch Toxicol. 2019;93:179-188. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. National Birth Defects Prevention Network. Annual reports. https://www.nbdpn.org/ar.php. Accessed July 15, 2018.
  • 42. Agence France-Presse in Paris. France to investigate cause of upper limb defects in babies. The Guardian. https://www.theguardian.com/world/2018/oct/21/france-to-investigate-cause-of-upper-limb-defects-in-babies. Published 2018. Accessed June 24, 2019.
  • 43. Willsher K. Baby arm defects prompt nationwide investigation in France. The Guardian. https://www.theguardian.com/world/2018/oct/31/baby-arm-defects-prompt-nationwide-investigation-france. Published 2018. Accessed June 24, 2019.
  • 44. Meier MH, Caspi A, Ambler A, et al. Persistent cannabis users show neuropsychological decline from childhood to midlife. Proc Natl Acad Sci U S A. 2012;109:E2657-E2664. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45. Brents L. Correlates and consequences of prenatal cannabis exposure (PCE): identifying and characterizing vulnerable maternal populations and determining outcomes in exposed offspring. In: Preedy VR, ed. Handbook of Cannabis and Related Pathologies: Biology, Pharmacology, Diagnosis and Treatment. 1st ed New York, NY: Academic Press; 2017:160-170. [Google Scholar]
  • 46. Smith AM, Fried PA, Hogan MJ, Cameron I. Effects of prenatal marijuana on visuospatial working memory: an fMRI study in young adults. Neurotoxicol Teratol. 2006;28:286-295. [DOI] [PubMed] [Google Scholar]
  • 47. Smith AM, Longo CA, Fried PA, Hogan MJ, Cameron I. Effects of marijuana on visuospatial working memory: an fMRI study in young adults. Psychopharmacology (Berl). 2010;210:429-438. [DOI] [PubMed] [Google Scholar]
  • 48. Smith AM, Mioduszewski O, Hatchard T, Byron-Alhassan A, Fall C, Fried PA. Prenatal marijuana exposure impacts executive functioning into young adulthood: an fMRI study. Neurotoxicol Teratol. 2016;58:53-59. [DOI] [PubMed] [Google Scholar]
  • 49. Garcia M. Colorado lawmakers push to declare autism an epidemic. https://denver.cbslocal.com/2018/04/02/lawmakers-autism-epidemic/. Published April 2, 2018. Accessed February 4, 2019.
  • 50. David AL, Holloway A, Thomasson L, et al. A case-control study of maternal periconceptual and pregnancy recreational drug use and fetal malformation using hair analysis. PLoS One. 2014;9:e111038. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary_Table_and_figures – Supplemental material for Cannabis Teratology Explains Current Patterns of Coloradan Congenital Defects: The Contribution of Increased Cannabinoid Exposure to Rising Teratological Trends
Click here for additional data file. (1.7MB, pdf)

Supplemental material, Supplementary_Table_and_figures for Cannabis Teratology Explains Current Patterns of Coloradan Congenital Defects: The Contribution of Increased Cannabinoid Exposure to Rising Teratological Trends by Albert Stuart Reece and Gary Kenneth Hulse in Clinical Pediatrics

Articles from Clinical Pediatrics are provided here courtesy of SAGE Publications

RESOURCES