Short-Term Effects of Maternal Cannabis Use on Human Milk Macronutrient Composition: The Lactation and Cannabis (LAC) Study

Abstract

Background:

Cannabis use has increased among reproductive-aged women, including those who are breastfeeding. However, aside from the appearance of cannabinoids in milk, almost nothing is known about how cannabis use might affect human milk composition. This study explored the short-term effects of maternal cannabis use on milk macronutrient content.

Methods:

Breastfeeding women who used cannabis (cases, n = 20) were matched by body mass index and time postpartum with breastfeeding women who did not use cannabis (controls, n = 19). After abstaining from cannabis use for ≥12 hours, cases collected a baseline milk sample, used cannabis as desired, and collected additional samples 30–≤40 minutes, 1–<2 hours, 2–<4 hours, 4–<6 hours, and 8–<12 hours after use. Controls collected milk at matched time points. Delta-9-tetrahydrocannabinol (Δ⁹-THC) and lipid concentrations were quantified in all samples, and fatty acids, lactose, and protein concentrations were quantified in baseline, 1–2 hours, and 8–12 hours samples.

Results:

There were no differences between controls and cases in the concentration of any macronutrient at baseline. After cannabis use, concentrations of lipids and 10 of the 39 identified fatty acids were lower in milk from cases compared to controls. Milk lactose levels increased over time in controls but not in cases. Protein levels were not different between groups at any timepoint. In milk produced by cases, Δ⁹-THC levels were positively correlated with lipids and negatively correlated with lactose.

Conclusion:

Cannabis use may transiently influence lipid, fatty acid, and lactose concentrations, highlighting the need for further research to understand the physiology of these alterations.

Keywords

breastfeeding breastmilk cannabinoids cannabis human milk macronutrients

Introduction

Cannabis use has increased in United States adults including reproductive-aged women with nearly 6% and 5% of pregnant and breastfeeding women, respectively, now reporting use.^1,2 While current evidence is limited, studies have reported associations between prenatal cannabis exposure and greater risks of adverse neonatal outcomes.³ Even less is known about the impacts of cannabis use during breastfeeding. Nonetheless, several studies have demonstrated the presence of cannabinoids in milk produced by women who use cannabis.^4–12

Cannabinoids that reach the mammary gland may alter the function of the mammary epithelial cell (MEC), potentially altering the nutrient composition of milk.¹³ Recently, an in vitro study suggested that cannabinoid exposure modifies expression of genes related to lipid and protein metabolism in murine MEC,¹⁴ but only two studies—both observational—have examined possible effects of maternal cannabis use on nutrient composition of milk. Josan et al.⁷ reported that cannabis use was associated with lower milk lactose concentrations, while Narayanan et al.¹⁵ reported that cannabis use was associated with lower milk fat and greater milk protein concentrations.

The objectives of this study were to explore (1) potential short-term effects of maternal cannabis use on the macronutrient composition of milk, (2) associations between Δ⁹-tetrahydrocannabinol (Δ⁹-THC) and macronutrient concentrations in milk, and (3) associations between previous cannabis use patterns and the macronutrient composition of milk. We hypothesized that maternal cannabis use is associated with alterations in the macronutrient composition of milk, particularly total lipids and fatty acids.

Methods

Participants

20 breastfeeding women residing in Washington and Oregon who reported using cannabis at least weekly were enrolled as cases in the Lactation and Cannabis study.^6,16 After collecting milk and data from cases, 20 breastfeeding women residing in Washington, Oregon, and Idaho who reported not using cannabis in the last 2 years (controls) were enrolled and matched with cases based on body mass index and time postpartum. Sample and data collection took place in 2022.

To be eligible, participants had to self-report being ≥21 years, have an infant <180 days old at enrollment who had been born full-term (≥37 weeks gestational age), be breastfeeding and/or providing pumped milk to their infant ≥5 times/day, and have provided pumped milk to their infant previously. Exclusion criteria included self-reported use of illicit drugs or opioid-use disorder treatments or opioids in the past 5 years (excluding any administered/prescribed opioids relating to delivery or cesarean-sections), participant or infant having symptoms of illness in the last 7 days, and/or indications of a breast infection (e.g., fever, red streaks or hard red portions of the breast, breast pain, discomfort, or lumps on the breast). The Washington State University Institutional Review Board approved all study procedures (protocol number #19087). A certificate of confidentiality was obtained from the National Institutes of Health.

Study design

In this prospective, case–control, quasi-experimental study, cases were instructed to observe at least a 12-hour abstention from cannabis and then collect a baseline milk sample (sample 1, S1). After S1 collection, cases used cannabis at least once, although the timing and frequency of use after abstention was at their discretion. Then they collected five additional milk samples (S2–S6) within timed intervals (30–≤40 minutes, 1–<2 hours, 2–<4 hours, 4–<6 hours, and 8–<12 hours, respectively) after initial cannabis use. Controls collected milk samples at the same times of the day (±75 minutes) as their matched counterparts.

Milk collection

Participants were provided with instructions and necessary supplies for milk collection. At each timepoint, participants thoroughly cleaned and dried their personal breast pump components. Using gloves, participants twice cleaned the breast using individually wrapped castile wipes before pumping a full breast expression. Subjects were instructed to collect all samples from the same breast throughout the study if possible. Participants thoroughly mixed and aliquoted each sample into five separate tubes using a single-use sterile pipette. Aliquots were stored in participants’ freezers until the end of the study period, when project personnel transported them on ice and stored them at −80°C.

Quantification of cannabinoids

Milk samples from both case and control groups were analyzed for Δ⁹-THC concentrations using previously described methods.⁶

Quantification of macronutrients

One aliquot of each milk sample was thawed on ice and divided into two tubes: one for lipid and fatty acids analyses, and one for protein and lactose analyses, that were again stored at −80°C. Quantification of lipids was performed in duplicate on 250 µL of milk collected at each timepoint following a modified Folch method.^17,18 The Percentage of lipid was determined gravimetrically. Fatty acid, protein, and lactose contents were quantified in samples collected at baseline and at 1–<2 hours and 8–<12 hours after first cannabis use by cases (matched timepoints for controls).

For fatty acid analysis, extracted lipids were methylated using a base-catalyzed transesterification¹⁹ and analyzed on a gas chromatography (Hewlett-Packard 6890 Series with auto injector) fitted with a flame ionization detector. Detailed method specifications are provided in Supplementary Data S1. Relative fatty acid composition was expressed as weight-percent of total fatty acid methyl esters, and absolute concentrations of each fatty acid were calculated as total lipid concentration multiplied by the relative amount of each fatty acid.

Milk protein concentrations were quantified using the Pierce bicinchoninic acid assay (Thermo Scientific, Waltham, MA) using human serum albumin in the standard curve. Milk lactose concentrations were determined through an enzymatic assay based on glucose oxidation.²⁰

Metadata, surveys, and cannabis use

Participants completed online surveys using REDCap^21,22 that assessed demographics, health, prenatal and postpartum substance use, and previous and current cannabis use (Daily Sessions, Frequency, Age of Onset, and Quantity of Cannabis Use Inventory).²³ On the day of milk collection, the timing of cannabis abstention and use was recorded. Project personnel were available via phone throughout the study period to answer questions and assist.

Statistical analyses

All analyses were performed using R v4.4.1,²⁴ with p ≤ 0.05 considered significant.

Numerical variables with normal distribution (Shapiro–Wilk test, p value >0.05) are reported as mean ± standard deviation (SD), while non-normally distributed variables are reported as median and interquartile range (IQR). Group differences in normally distributed variables were assessed using t-test; otherwise, Wilcoxon rank-sum test was used. p Values were adjusted for multiple tests using the Benjamin–Hochberg method. Categorical variables were compared using Fisher’s Exact test when any expected cell count was <5; otherwise, Chi-squared test was used.

To evaluate the transient effect of cannabis use during the study on the nutrient composition of milk, linear mixed-effects (LME) model analyses were performed using lme4 v1.1.34 and lmerTest v3.1–3.^25,26 Each LME model included participant as a random effect, group (case or control), timepoint (S2–S6), and group-by-timepoint interaction as fixed effects. Model assumptions were evaluated through Q-Q plots, Shapiro–Wilk test, and simulated residual diagnosis (DHARMa v0.4.7).²⁷ Post-hoc pairwise comparisons were performed using emmeans v1.8.7²⁸ with Tukey’s correction. Effect sizes (Cohen’s f²) were calculated for group, timepoint, and group-by-timepoint interaction by comparing R² from full and reduced LMEmodels using MuMIn v1.48.11.²⁹

To explore relationship between milk Δ⁹-THC concentration and milk macronutrient concentrations over the study period, repeated measures correlations were performed using rmcorr v0.7.0.³⁰ The associations between the number of times participants used cannabis during the study period (once vs multiple times) with the milk macronutrients content were tested using LME models following the same specifications detailed previously. The area under the curve (AUC) for lipid concentration over time was calculated using the trapezoidal rule.

To explore possible associations between participants’ previous patterns of cannabis use and milk macronutrients at baseline, Spearman’s correlations were used for numerical variables, and t-test or Wilcoxon tests for categorical variables.

Results

Participant characteristics

Milk samples were collected from 20 cases and 20 controls. Two cases did not provide samples for one timepoint (S3), and one case did not provide samples for three timepoints (S3–S6). One control was excluded from the analysis due to measurable levels of Δ⁹-THC in her milk.

Characteristics of the participants are summarized in Table 1. Age, body mass index, parity, postpartum time, exclusive breastfeeding, and infant sex were similar between cases and controls. However, infants of controls had longer gestational ages than those of cases (mean ± SD: 40.0 ± 1.0 vs. 38.8 ± 1.2 weeks; Student’s t-test: p = 0.001). Cases also had a longer lifetime history of cannabis use compared to controls [median (IQR): 9 (5.75–12.25) vs. 1 (1–3.5) yearr; Wilcoxon rank-sum test: p = 0.007], although it is notable that seven controls reported cannabis use at some point in their lives. Only one participant reported using tobacco during pregnancy and postpartum periods.

Table 1.

Selected Demographic and Anthropometric Characteristics and Cannabis Use Patterns of the Women Who Participated in This Study

Characteristic	Cases(n = 20)	Controls(n = 19)	p value^a
Age (years)	26.6 ± 3.5	29.2 ± 4.3	0.051
Body mass index (kg/m²)	29.3 ± 5.6	28.7 ± 6.4	0.767
How many times have you given birth in your life?	2 (2–2)	2 (2–3.5)	0.259
Gestational age at parturition (weeks)	38.8 ± 1.2	40.0 ± 1.0	0.001
Time postpartum (days)	107 ± 51	99 ± 49	0.645
Education level
High school graduate/GED	6 (30)	1 (5)	0.019
Some college but not degree	4 (20)	3 (16)
Technical/Vocation degree or associate degree	5 (25)	3 (16)
College graduate (bachelor’s degree)	5 (25)	5 (26)
Post-graduate education	0 (0)	7 (37)
Are you exclusively breastfeeding?
No	3 (15)	2 (11)	>0.999
Yes	17 (85)	17 (89)	>0.999
Infant sex
Female	12 (60)	10 (53)	0.593
Male	6 (30)	9 (47)
Not reported	2 (10)	0 (0)
Time since last breastfeeding before collecting S1 sample (minutes)	90 (60–277.5)	120 (80–195)	0.921
How many years have you used cannabis in your entire life?	9 (5.7–12.2)	1 (1–3.5)	0.007
While pregnant, did you use tobacco?
Yes	1 (5)	0 (0)	>0.999
No	19 (95)	19 (100)	>0.999
Since giving birth, have you used tobacco?
Yes	1 (5)	0 (0)	>0.999
No	19 (95)	19 (100)	>0.999

Cases reported using cannabis, whereas controls did not. Numerical variables expressed as means ± SD or medians (IQR); categorical variables expressed as counts (percentage).

Student’s t-test or Wilcoxon rank-sum test for numerical variables; chi-square test or Fisher test for categorical variables.

IQR, interquartile range; SD, standard deviation. Values in bold denote statistically significant differences.

Cannabis use patterns among cases are summarized in Table 2. Cases reported using cannabis a median of 28 days in the past 30 days. During pregnancy, most cases reported using cannabis either daily (45%) or weekly (20%). The primary forms of cannabis use were marijuana (e.g., flower, bud, herb; 60%) and concentrates (40%), primarily via inhalation. Additional details regarding participants’ cannabis use have been published previously.¹⁶

Table 2.

Cannabis Use Patterns of the Women Who Reported the Use of Cannabis Participating in This Study (Cases; n = 20)

Cannabis use patterns	Median (IQR); n (%)
How many days have you use cannabis in the past 30 days?	28 (14–30)
How often have you used cannabis since giving birth?
Daily	14 (70)
Weekly	6 (30)
How often did you use cannabis during pregnancy?
Prefer not to answer	1 (5)
Daily	9 (45)
Weekly	4 (20)
Once or twice during pregnancy	2 (10)
Not applicable	4 (20)
What is the primary form of cannabis you use?
Marijuana (e.g., flower, bud, herb)	12 (60)
Concentrates (e.g., oil, wax, shatter, butane hash oil, dabs)	8 (40)
What is the primary method by which you use cannabis?
Joints	7 (35)
Hand pipe	1 (5)
Bong (water pipe)	8 (40)
Vaporizer (e.g., volcano, vape pen)	3 (15)
Other	1 (5)

Differences in human milk macronutrients at baseline

At baseline, concentrations of total lipids (mean ± SD: 43.81 ± 16.11 vs. 40.13 ± 17.22 mg/mL), lactose [median (IQR): 72.86 (70.2–75.46) vs. 71.61 (66.86–73.61) g/L], and protein [median (IQR): 1.33 (1.21–1.54) vs. 1.49 (1.26–1.65) g/dL] in milk produced by cases and controls did not differ (Table 3). Similarly, no differences were detected in absolute concentrations or relative abundances of individual and grouped milk fatty acids between cases and controls at baseline (Table 3 and Supplementary Table S1).

Table 3.

Milk Macronutrients Composition at Baseline in Women Who Used Cannabis (Cases) and Those Who Did Not Use Cannabis (Controls)

Macronutrient	Common name	Cases (n = 20)	Controls (n = 19)	p value^a
Total lipids (mg/mL)		43.81 ± 16.11	40.13 ± 17.22	0.496
Lactose (g/L)		72.86 (70.20–75.46)	71.61 (66.86–73.61)	0.120
Protein (g/dL)		1.33 (1.21–1.54)	1.49 (1.26–1.65)	0.354
Fatty acids (mg/mL)
C6:0	Caproic	0.03 ± 0.01	0.03 ± 0.01	0.422
C8:0	Caprylic	0.07 ± 0.03	0.06 ± 0.03	0.586
C10:0	Capric	0.53 ± 0.18	0.48 ± 0.23	0.404
C11:0	Undecanoic	0.01 ± 0.00	0.00 ± 0.00	0.096
C12:0	Lauric	2.14 (1.76–2.60)	1.70 (1.56–2.77)	0.444
C13:0	Tridecanoic	0.01 ± 0.01	0.01 ± 0.01	0.326
C14:0	Myristic	2.60 ± 0.97	2.21 ± 1.16	0.261
C14:1t9		0.02 (0.01–0.02)	0.02 (0.01–0.02)	0.967
C14:1c9	Myristoleic	0.07 (0.06–0.11)	0.07 (0.05–0.08)	0.550
C15:0	Pentadecanoic	0.12 (0.09–0.14)	0.11 (0.09–0.14)	0.569
C15:1c10	Pentadecenoic	0.02 (0.02–0.03)	0.02 (0.02–0.03)	0.879
C16:0	Palmitic	9.30 ± 3.80	8.14 ± 3.40	0.323
C16:1ω7c9	Palmitoleic	0.75 (0.52–1.03)	0.74 (0.43–1.00)	0.461
C17:0	Margaric	0.12 ± 0.06	0.11 ± 0.05	0.748
C17:1ω7c10	Margaroleic	0.07 (0.05–0.09)	0.07 (0.05–0.09)	0.879
C18:0	Stearic	2.80 ± 1.22	2.86 ± 1.40	0.889
C18:1t7/8		0.02 (0.00–0.09)	0.03 (0.00–0.10)	0.848
C18:1t9	Elaidic	0.11 ± 0.05	0.10 ± 0.04	0.504
C18:1t10		0.11 ± 0.05	0.11 ± 0.05	0.805
C18:1t11		0.11 ± 0.05	0.09 ± 0.05	0.318
C18:1c9	Oleic	14.09 ± 5.71	13.31 ± 6.40	0.692
C18:1c11		0.80 ± 0.33	0.78 ± 0.41	0.875
C18:1c15		0.05 ± 0.02	0.04 ± 0.02	0.258
C19:0		0.02 ± 0.01	0.02 ± 0.01	0.376
C18:2t9t12	Linolelaidic	0.05 (0.03–0.06)	0.05 (0.04–0.05)	0.478
C18:2c9c12	Linoleic	6.69 (4.35–8.15)	5.52 (4.13–7.04)	0.336
C18:3ω6	γ-Linolenic	0.14 ± 0.07	0.14 ± 0.07	0.959
C18:3ω3	α-Linolenic	0.50 (0.39–0.73)	0.43 (0.31–0.59)	0.351
C20:1ω9	Gondoic	0.19 ± 0.08	0.18 ± 0.09	0.793
C18:2c9t11	Rumenic	0.06 (0.04–0.09)	0.06 (0.05–0.09)	0.813
C21:0	Heneicosanoic	0.01 (0.01–0.01)	0.01 (0.01–0.01)	0.967
C20:3ω6	Behenic	0.18 (0.12–0.22)	0.17 (0.12–0.23)	0.901
C20:3ω3		0.02 (0.01–0.02)	0.01 (0.01–0.02)	0.158
C20:4ω6	Arachidonic	0.19 ± 0.09	0.20 ± 0.09	0.833
C23:0	Tricosanoic	0.03 (0.02–0.04)	0.03 (0.02–0.03)	0.627
C20:5ω3	Eicosapentaenoic	0.02 (0.01–0.02)	0.02 (0.01–0.02)	0.901
C22:4ω6		0.04 (0.02–0.04)	0.03 (0.03–0.04)	0.835
C22:5ω3	Docosapentaenoic	0.05 ± 0.02	0.05 ± 0.02	0.914
C22:6ω3	Docosahexaenoic	0.05 (0.04–0.07)	0.06 (0.05–0.12)	0.322
Unidentified		1.11 ± 0.40	1.03 ± 0.43	0.541
Total saturated fatty acids		17.87 ± 6.41	16.12 ± 6.73	0.412
Total monounsaturated fatty acids		16.59 ± 6.74	15.56 ± 7.40	0.653
Total polyunsaturated fatty acids		7.98 (5.45–9.75)	6.49 (5.00–8.42)	0.411
Total ω3 fatty acids		0.60 (0.47–0.84)	0.54 (0.39–0.68)	0.380
Total ω6 fatty acids		7.27 (4.82–8.74)	6.07 (4.39–7.61)	0.336
Total medium-chain fatty acids		2.76 (2.38–3.33)	2.31 (2.01–3.54)	0.365
Total long-chain fatty acids		26.60 ± 10.68	24.99 ± 11.69	0.658
ω3:ω6 ratio		0.04 (0.03–0.05)	0.04 (0.02–0.05)	0.444

Variables expressed as mean ± SD or median (IQR).

Student’s t-test or Wilcoxon rank-sum test depending on the distribution of the variable.

Milk macronutrients concentrations after cannabis use

There was an interaction between group and timepoint on the concentration of lactose in milk (p = 0.045; Fig. 1 and Supplementary Table S2). In the controls, mean lactose concentration increased from 67.74 ± 1.25 at S3 to 71.54 ± 1.25 g/L at S6 (pairwise comparison, p = 0.004). In contrast, mean lactose concentration in milk in the cases remained stable, with values of 71.17 ± 1.29 and 71.85 ± 1.32 g/L at S3 and S6, respectively (p = 0.938). Protein concentrations were not different by group or timepoint (Supplementary Tables S2 and S3).

FIG. 1.

Estimated means of selected macronutrients in milk produced by women who used cannabis (cases) and women who did not use cannabis (controls). X-axes reflect time after the first cannabis use by cases. p Values represent the significance of group (cases vs controls), timepoint (time after cannabis use by cases), or interaction (group × timepoint) effects on macronutrient concentrations. * indicates significant difference in estimated means (Tukey-adjusted pairwise comparisons; p < 0.05) between groups at the particular timepoint. MUFA, monounsaturated fatty acids.

There was no interaction between group and timepoint for total lipid concentration. However, there was an independent effect of time (p < 0.001) and a trend toward an independent effect of group (p = 0.078) on milk lipid concentration (Fig. 1). Lipid concentration remained stable across timepoints S2, S3, and S4 (mean ± SD: 51.7 ± 3.19, 60 ± 3.27, and 56.0 ± 3.23 mg/mL, respectively; p > 0.100). However, compared to S2, mean lipid concentration decreased at S5 (48.4 ± 3.27 mg/mL; p = 0.032) and at S6 (38.8 ± 3.31 mg/mL), which was lower than S2, S3, and S4 (p = 0.012, <0.001, and <0.001, respectively). Lipid concentration did not differ between S5 and S6 (p = 0.121). Overall mean lipid concentration tended to be lower in cases than controls (47.3 ± 2.97 vs. 54.7 ± 2.95 mg/mL; p = 0.086), with statistically significant differences at S2 and S3 (p = 0.048 and p = 0.037, respectively). However, the total lipids AUC following cannabis use was similar between cases and controls (mean ± SD: 350 ± 146 and 408 ± 159 mg· h/mL, respectively; Student’s t-test: p = 0.248).

There was an independent effect of time on absolute concentrations of all fatty acids (Fig. 1 and Supplementary Table S3), which decreased between S3 and S6 (p < 0.05). There was also an independent group effect on the absolute concentration of 10 of the 39 identified fatty acids [margaric, margaroleic, stearic, C18:1t11, C18:1c11, γ-linolenic, behenic, arachidonic, C22:4ω6, and docosapentaenoic (EPA) acids]. Concentrations of these fatty acids were lower in milk produced by cases than controls (p < 0.05, Fig. 1 and Supplementary Table S3). Absolute concentrations of grouped saturated fatty acids and monounsaturated fatty acids were also lower in milk from cases (mean ± SD: 18.17 ± 1.30 and 16.77 ± 1.41 mg/mL, respectively) compared to controls (21.78 ± 1.21 and 20.66 ± 1.31 mg/mL, respectively) (p = 0.039 and p = 0.043, respectively). Effect sizes of LMEs are summarized in Supplementary Table S4.

Associations among macronutrients and cannabinoid levels in milk from cases

Cannabinoid concentrations (including Δ⁹-THC) in the milk produced by cases in this study have been reported previously.⁶ The relationships between Δ⁹-THC and macronutrient concentrations in samples from cases are summarized in Figure 2 and Supplementary Table S5. Milk Δ⁹-THC and total lipid concentrations were positively correlated [r = 0.55, 95% CI: (0.40, 0.68), p < 0.001], while Δ⁹-THC and lactose concentrations were negatively correlated [r = −0.39, 95% CI: (−0.63, −0.08), p = 0.016]. There were no significant correlations between Δ⁹-THC and relative amounts of other macronutrients in the milk.

FIG. 2.

Scatter plots showing the repeated-measures correlations between Δ⁹-THC concentration in human milk and (A) total lipids, and (B) lactose concentrations in milk from cases collected across multiple time points. Each point represents a single observation from a given subject (n = 20), collected at different time points during the study period. Lines represent the common intra-subject correlation trend. Points and lines that correspond to the same subject are illustrated with the same color. Δ⁹-THC, delta-9-tetrahydrocannabinol.

During the study period, 12 cases reported a single instance of cannabis use, while the remaining 8 cases reported multiple uses. Baseline lipid concentrations were not different between participants who used cannabis once and those who used it multiple times (mean ± SD: 49.2 ± 19.7 vs. 40.2 ± 12.9 mg/mL, p = 0.282). However, there was an interaction between usage frequency (once vs. multiple) and timepoint on total lipid concentrations (p = 0.017; Supplementary Fig. S1 and Fig. S2). While total lipids remained similar throughout the study in samples from participants who used multiple times, mean total lipid concentration in milk from participants who used cannabis once decreased from 57.4 ± 5.98, 57.3 ± 6.38, and 57.2 ± 5.66 mg/mL at S2, S3 and S4, respectively, to 34.1 ± 5.66 mg/mL at S6 (p = 0.033, p = 0.055, and p = 0.025, respectively; pairwise comparisons). Concentration at S5 (43.0 ± 5.98 mg/mL) did not differ from other timepoints (p > 0.05). Milk protein and lactose concentrations did not differ by cannabis use frequency on the study period.

Macronutrients concentration in milk and previous patterns of cannabis use

Associations between previous patterns of cannabis use and macronutrient concentrations in milk collected at baseline are presented in Supplementary Table S6. Correlations between total years of cannabis use and milk macronutrient concentrations were calculated for all participants (cases and controls) who reported any lifetime cannabis use (n = 27). No significant correlations were found between total years of cannabis use or the total number of days of cannabis use in the past 30 days and concentration of milk macronutrients (Supplementary Table S6). Additionally, no differences were observed in macronutrient concentrations based on cannabis use frequency during the postpartum period (daily vs. weekly) or primary form of use (marijuana vs. concentrates).

Discussion

The primary objective of this study was to investigate whether maternal cannabis use during lactation is associated with short-term changes in total lipids, lactose, protein, and fatty acid concentrations in milk. Our results show that milk lipid concentrations varied over the study period, regardless of whether the women were using cannabis or not. Total lipid concentration was greater in samples collected 1–4 hours after cannabis use by cases than in those collected 4–12 hours after. Importantly, participants collected a complete breast expression at each timepoint, with the first three samples after initial cannabis use collected within a 4-hour window. Previous research has shown that milk lipid content is influenced by the degree of breast emptying, with higher lipid content in milk expressed from less full breasts.³¹ Furthermore, Hassiotou et al.³² observed that milk lipids peak 30 minutes after a breast expression, which comports with our study’s observed changes over time. As such, we interpret the increase in milk lipid concentration in the first 4 hours after initial cannabis use by cases because of the breast being emptied multiple times over a short period, not an overall impact of other factors such as time of day.

Our data indicate that lipid concentrations were lower in milk collected from cases 1 to 4 hours after cannabis use compared to controls. To our knowledge, only two studies have examined the potential effect of maternal cannabis use on milk nutrient composition. In the study by Josan et al.,⁷ 22 women who reported cannabis use and 18 women who did not provided milk (60 mL) from their daily or weekly pumped milk (collection methods were not standardized). Milk nutrient concentrations were analyzed using the Miris Human Analyzer™ (Miris HMA™). The researchers reported no differences in lipid content between the groups. Conversely, Narayanan et al.¹⁵ analyzed the nutrient content of 165 samples from the Human Milk Research Biorepository at UC San Diego with detectable levels of cannabinoids, 23 samples with no detectable levels of cannabinoids but from self-reported cannabis use participants, and 449 samples from women who self-reported not using cannabis, using a SpectraStar 2400 near infrared instrument. Milk collection methods were not standardized. In line with our results, the researchers found lower lipid concentrations in samples with detectable levels of cannabinoids that were collected through complete breast expression, but not in partially expressed samples—emphasizing the importance of milk collection methods when studying milk composition—particularly lipids. The effect of Δ⁹-THC and other cannabinoids on milk fat content has also been explored in an in vitro study using a murine MEC line.¹⁴ The researchers observed that Δ⁹-THC-treated MEC had reduced expression of genes related to lipid metabolism. Although our results indicate that the same effect may occur in the human mammary gland, further studies using human MEC lines are needed to confirm these findings.

Although we documented a trend toward lower total lipids in milk following cannabis use (compared to controls), we also documented a positive correlation between milk lipid concentration and Δ⁹-THC in the cases. We hypothesize that this relationship is likely attributed to the lipophilic nature of Δ⁹-THC.¹³ It is possible that circulating Δ⁹-THC is physically associated with stored and/or circulating lipids, which become incorporated into milk.

We observed an interaction between group and timepoint on milk lactose concentration, which remained unchanged in milk produced by the cannabis group but increased in milk produced by controls. Additionally, there was a negative correlation between Δ⁹-THC and lactose concentrations in milk. Contrary to these results, Josan et al.⁷ reported greater concentrations of lactose in milk produced by women who used cannabis compared to controls, while Narayanan et al.¹⁵ did not observe an association between cannabinoids and carbohydrates (including human milk oligosaccharides, HMOs) in milk. Lactose is synthesized in the MEC from glucose and galactose, and its levels in milk are relatively stable. Josan et al.¹⁴ reported that treatment of murine MEC with Δ⁹-THC decreased expression of the glucose transporter 1 (GLUT1) gene, the primary glucose transporter in the MEC. This reduction may explain lower levels of lactose in milk produced by women who use cannabis and the negative correlation observed between Δ⁹-THC and lactose levels in their milk. However, additional studies using human MEC lines are needed.

Consistent with findings of Josan et al.,⁷ we observed no differences in milk protein concentrations between the cannabis and control groups, nor a correlation between milk Δ⁹-THC levels and protein content. Narayanan et al.¹⁵ reported greater protein concentrations in milk with detectable levels of cannabinoids compared to milk with no detectable levels. In the in vitro study carried out by Josan et al.,¹⁴ murine MEC treated with cannabinoids had reduced transcription (mRNA levels) of genes involved in milk protein synthesis. Further studies are needed to confirm whether maternal use of cannabis affects milk protein levels in women.

Our results also showed lower concentrations of 10 fatty acids (including stearic and arachidonic acids) in milk produced by cases compared to controls. Cannabinoids can bind to peroxisome proliferator-activated receptors (PPAR) in the MEC,^13,33 a family of nuclear hormone receptors that bind to DNA sequences leading to changes in the transcription of target genes. PPAR expression is increased during pregnancy and lactation³⁴ and plays an important role in lipid metabolism.^34,35 Accordingly, cannabinoids might interfere with mammary lipid and fatty acid synthesis through the activation of these receptors.

Strengths of this study include its repeated-measures design, allowing us to investigate changes in milk composition over time. The case-control design, with participants matched for body mass index, time postpartum, and time of milk collection, allowed us to control for individual variability. Full breast expressions enabled standardized assessment of milk composition, particularly lipid content, and macronutrient analyses were conducted using gold-standard methods previously validated for human milk. Limitations include the lack of dietary control, which can influence the milk fatty acid profile. The self-directed cannabis use resulted in high variability in cannabis exposures, including multiple uses, various methods of consumption, and a wide range of reported Δ⁹-THC concentrations in the products consumed. This factor limited our ability to precisely quantify participants’ cannabis doses. Future research should control for these variables.

Conclusion

Our findings suggest that maternal cannabis use during lactation may acutely influence nutrient composition of human milk, particularly total lipids, lactose, and specific fatty acids. These effects may be a result of the interaction of cannabinoids with the mammary epithelial cell function as well as with the endocrine system. This is the first study providing rigorous evidence of the potential short-term effects of maternal cannabis use on milk composition. More studies are needed to better understand implications, if any, for infant nutrition and health.

Authors’ Contributions

I.C.-N.: Writing—original draft, writing—review and editing, data curation, formal analysis, investigation, and visualization. J.E.W.: Conceptualization, funding acquisition, supervision, and writing—review and editing. L.D.: Investigation and writing—review and editing. H.R.: Investigation and writing—review and editing. A.B.: Investigation and writing—review and editing. D.R.G.: Funding acquisition, investigation, resources, supervision, and writing—review and editing. E.A.H.: Data curation, investigation, and writing—review and editing. B.C.: Investigation and writing—review and editing. C.S.: Investigation and writing—review and editing. C.B.-L.: Conceptualization, funding acquisition, and writing—review and editing. O.B.: Investigation and writing—review and editing. M.A.M.: Conceptualization, resources, funding acquisition, supervision, and writing– review and editing. C.L.M.: Conceptualization, funding acquisition, investigation, project administration, resources, supervision, and writing—review and editing. M.K.M.: Conceptualization, resources, funding acquisition, supervision, and writing—review and editing.

Footnotes

Acknowledgments

The authors would like to thank Julia Piaskowski and Harpreet Kaur from the University of Idaho College of Agricultural and Life Sciences Statistical Programs for their advice on the statistical analysis of this article. Additionally, the authors sincerely thank all the women who participated in this study.

Funding Information

This study was supported by the Idaho Agricultural Experiment Station, the State of Washington Initiative Measures 171 and 502, and Washington State University Health Equity Research Center, and the National Institutes of Health, R21HD113959.

Disclosure Statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Supplemental Material

References

Boerner

, Natha

, Baker

, et al. Perceptions of cannabis use and its benefits and risks among breastfeeding mothers. Womens Health Rep (New Rochelle), 2024; 5(1):412–423; doi: 10.1089/whr.2024.0021

Haider

, Karim

, Jayawardhana

, et al. Association between state-level medical marijuana legalization and marijuana use during pregnancy: A population-based study. Am J Addict, 2025; 34(1):75–84; doi: 10.1111/ajad.13651

Marchand

, Masoud

, Govindan

, et al. Birth outcomes of neonates exposed to marijuana in utero. JAMA Netw Open, 2022; 5(1):e2145653; doi: 10.1001/jamanetworkopen.2021.45653

Baker

, Datta

, Rewers-Felkins

, et al. Transfer of inhaled cannabis into human breast milk. Obstet Gynecol, 2018; 131(5):783–788; doi: 10.1097/AOG.0000000000002575

Bertrand

, Hanan

, Honerkamp-Smith

, et al. Marijuana use by breastfeeding mothers and cannabinoid concentrations in breast milk. Pediatrics, 2018; 142(3):e20181076; doi: 10.1542/peds.2018-1076

Holdsworth

, Berim

, Gang

, et al. Human milk cannabinoid concentrations and associations with maternal factors: The Lactation and Cannabis (LAC) study. Breastfeed Med, 2024; 19(7):515–524; doi: 10.1089/bfm.2024.0021

Josan

, Shiplo

, Fusch

, et al. Cannabis use during lactation may alter the composition of human breast milk. Pediatr Res, 2023; 93(7):1959–1968; doi: 10.1038/s41390-022-02315-1

Marchei

, Escuder

, Pallas

, et al. Simultaneous analysis of frequently used licit and illicit psychoactive drugs in breast milk by liquid chromatography tandem mass spectrometry. J Pharm Biomed Anal, 2011; 55(2):309–316; doi: 10.1016/j.jpba.2011.01.028

Moss

, Bushlin

, Kazmierczak

, et al. Cannabis use and measurement of cannabinoids in plasma and breast milk of breastfeeding mothers. Pediatr Res, 2021; 90(4):861–868; doi: 10.1038/s41390-020-01332-2

10.

Perez-Reyes

, Wall

. Presence of Δ9-tetrahydrocannabinol in human milk. N Engl J Med, 1982; 307(13):819–820; doi: 10.1056/NEJM198209233071311

11.

Sempio

, Wymore

, Palmer

, et al. Detection of cannabinoids by LC–MS-MS and ELISA in breast milk. J Anal Toxicol, 2021; 45(7):686–692; doi: 10.1093/jat/bkaa142

12.

Wymore

, Palmer

, Wang

, et al. Persistence of Δ-9-tetrahydrocannabinol in human breast milk. JAMA Pediatr, 2021; 175(6):632–634; doi: 10.1001/jamapediatrics.2020.6098

13.

Castro-Navarro

, McGuire

, Williams

, et al. Maternal cannabis use during lactation and potential effects on human milk composition and production: A narrative review. Adv Nutr, 2024; 15(4):100196; doi: 10.1016/j.advnut.2024.100196

14.

Josan

, Podinic

, Pfaff

, et al. Effect of Delta-9-tetrahydrocannabinol and cannabidiol on milk proteins and lipid levels in HC11 cells. PLoS One, 2022; 17(8):e0272819; doi: 10.1371/journal.pone.0272819

15.

Narayanan

, Bertrand

, Waalen

, et al. The effect of cannabis consumption during lactation on the macronutrient concentrations in breast milk. Breastfeed Med, 2025; 20(1):33–41; doi: 10.1089/bfm.2024.0083

16.

Smith

, Schmidt

, Holdsworth

, et al. Cannabis use, decision making, and perceptions of risk among breastfeeding individuals: The Lactation and Cannabis (LAC) study. J Cannabis Res, 2024; 6(1):6; doi: 10.1186/s42238-023-00212-w

17.

Folch

, Lees

, Stanley

GHS

. A simple method for the isolation and purification of total lipides from animal tissues. J Biol Chem, 1957; 226(1):497–509; doi: 10.1016/S0021-9258(18)64849-5

18.

Kumar

, Du Toit

, Kulkarni

, et al. Distinct patterns in human milk microbiota and fatty acid profiles across specific geographic locations. Front Microbiol, 2016; 7:1619; doi: 10.3389/fmicb.2016.01619

19.

Christie

. A simple procedure for rapid transmethylation of glycerolipids and cholesteryl esters. J Lipid Res, 1982; 23(7):1072–1075; doi: 10.1016/S0022-2275(20)38081-0

20.

Polberger

, Lönnerdal

. Simple and rapid macronutrient analysis of human milk for individualized fortification: Basis for improved nutritional management of very-low-birth-weight infants? J Pediatr Gastroenterol Nutr, 1993; 17(3):283–290; doi: 10.1097/00005176-199310000-00009

21.

Harris

, Taylor

, Thielke

, et al. Research Electronic Data Capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform, 2009; 42(2):377–381; doi: 10.1016/j.jbi.2008.08.010

22.

Harris

, Taylor

, Minor

, et al.; REDCap Consortium. The REDCap consortium: Building an international community of software platform partners. J Biomed Inform, 2019; 95:103208; doi: 10.1016/j.jbi.2019.103208

23.

Cuttler

, Spradlin

. Measuring cannabis consumption: Psychometric properties of the daily sessions, frequency, age of onset, and quantity of cannabis use inventory (DFAQ-CU). PLoS One, 2017; 12(5):e0178194; doi: 10.1371/journal.pone.0178194

24.

R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing Team; 2024.

25.

Bates

, Mächler

, Bolker

, et al. Fitting linear mixed-effects models using lme4. J Stat Soft, 2015; 67(1):1–48; doi: 10.18637/jss.v067.i01

26.

Kuznetsova

, Brockhoff

, Christensen

RHB

. lmerTest package: Tests in linear mixed effects models. J Stat Soft, 2017; 82(13):1–26; doi: 10.18637/jss.v082.i13

27.

Hartig

, Lohse

, Leite M de

. DHARMa: Residual diagnostics for hierarchical (Multi-Level/Mixed) regression models. 2024.

28.

Lenth

. Emmeans: Estimated Marginal Means, Aka Least-Squares Means. CRAN; 2025.

29.

Bartoń

. MuMIn: Multi-model inference. CRAN; 2010; doi: 10.32614/CRAN.package.MuMIn

30.

Bakdash

, Marusich

. Repeated measures correlation. Front Psychol, 2017; 8:456; doi: 10.3389/fpsyg.2017.00456

31.

Daly

, Di Rosso

, Owens

, et al. Degree of breast emptying explains changes in the fat content, but not fatty acid composition, of human milk. Exp Physiol, 1993; 78(6):741–755; doi: 10.1113/expphysiol.1993.sp003722

32.

Hassiotou

, Hepworth

, Metzger

, et al. Maternal and infant infections stimulate a rapid leukocyte response in breastmilk. Clin Transl Immunology, 2013; 2(4):e3; doi: 10.1038/cti.2013.1

33.

O’Sullivan

. An update on PPAR activation by cannabinoids. Br J Pharmacol, 2016; 173(12):1899–1910; doi: 10.1111/bph.13497

34.

Bionaz

, Loor

. Gene networks driving bovine milk fat synthesis during the lactation cycle. BMC Genomics, 2008; 9(1):366; doi: 10.1186/1471-2164-9-366

35.

Schoonjans

, Staels

, Auwerx

. The peroxisome proliferator activated receptors (PPARs) and their effects on lipid metabolism and adipocyte differentiation. Biochim Biophys Acta, 1996; 1302(2):93–109; doi: 10.1016/0005-2760(96)00066-5

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.32 MB

0.01 MB

0.10 MB