Abstract
Background:
Magnesium sulfate is widely prescribed postpartum for seizure prophylaxis in women with preeclampsia and other hypertensive disorders of pregnancy, yet its potential effects on lactation outcomes remain underexplored.
Objectives:
To evaluate and synthesize the current evidence on how postpartum magnesium sulfate therapy affects lactation outcomes, including breastfeeding initiation, exclusivity, duration, pumping habits, and secretory activation (SA).
Methods:
This integrative review followed PRISMA guidelines and Whittemore and Knafl’s framework. Literature was obtained from five databases without date restrictions. A total of 11 studies met the inclusion criteria. Methodological quality was evaluated using the Joanna Briggs Institute tools and Melnyk and Fineout-Overholt’s hierarchy of evidence.
Results:
Data were synthesized from 11 studies published between 1993 and 2023, encompassing 2,842 participants across diverse hospital settings. Findings indicate that extended postpartum magnesium sulfate administration is associated with delayed breastfeeding initiation, delayed maternal perception of SA, and greater reliance on milk expression. Most researchers did not report maternal side effects or quantitatively measure the frequency of breastfeeding or pumping. Studies reported hospital policies that restricted rooming-in and breastfeeding during magnesium infusion.
Conclusions:
Postpartum magnesium sulfate administration is associated with delays in lactation initiation, missing the evidence-based critical window for frequent early milk removal. However, studies in this review rarely examine maternal side effects or feeding frequency in detail. Future research should use standardized definitions, document both frequency and mode of milk removal, evaluate objective measures of SA, evaluate maternal experience, and hospital policies.
Introduction
Hypertensive disorders of pregnancy (HDP) account for a rising share of maternal and perinatal morbidity and mortality worldwide, with a global prevalence of 3.51 million cases contributing substantially to preventable maternal and neonatal deaths.1–3 HDP is an umbrella term covering a range of conditions characterized by high blood pressure during pregnancy, including gestational hypertension, chronic hypertension with superimposed preeclampsia, preeclampsia, and eclampsia.3,4 In the United States, HDP now complicates more than 15% of pregnancies, with nearly one-third of maternal deaths during delivery hospitalization linked to these disorders. 1 Beyond the peripartum period, women have an associated two- to four-fold higher lifetime risk of cardiovascular disease.1,5–7
The management of severe cases of HDP (e.g., severe preeclampsia) has been transformed by intravenous magnesium sulfate, the most effective therapy demonstrated to prevent and control eclamptic seizures.8–11 Yet, this life-saving intervention inevitably has side effects that tend to worsen with higher doses and more prolonged exposure.12–14 Flushing, nausea, and vomiting are among the most common reactions, affecting about one-quarter of patients at a 1 gram/hour continuous infusion rate and rising to over 70% at a 2 gram/hour rate.12,13 The extended treatment duration further exacerbates these effects, with increased fatigue that directly impairs breastfeeding. 14
Magnesium sulfate therapy overlaps with the critical period of lactation initiation, which hinges on timely and frequent breast stimulation in the first hours after birth. 15 Early milk removal (whether via direct breastfeeding or pumping) is not merely beneficial, but biologically essential, catalyzing the cascade of secretory activation (SA; MOM sodium [Na] concentration ≤ 16 mM) that underpins sustained effective lactation.15–17 Delayed or infrequent attempts in these pivotal hours are strongly associated with delayed SA and lower rates of breastfeeding duration and exclusivity.18,19 The failure to establish effective breastfeeding in this vulnerable population removes critical protective benefits when they are needed most. Breastfeeding can reduce the risk of preeclampsia in future pregnancies and provides protective cardiovascular benefits that may help lessen some long-term consequences of HDP, in addition to the well-known advantages for infant health.20–23 For women with HDP, this risk is acute: both the physiological sequelae of preeclampsia and the logistics of postpartum magnesium therapy may compound barriers to optimal lactation outcomes.
Prior research has shown that women with preeclampsia exhibit lower breastfeeding rates, shorter duration, and delayed SA compared with normotensive counterparts.24–27 Unfortunately, many studies on lactation outcomes do not distinguish between women who received magnesium sulfate and those who did not, making it hard to determine whether lactation challenges are due to both the underlying pathophysiology of HDP and/or side effects of magnesium sulfate or other influences. Additionally, while intrapartum magnesium sulfate can cause neonatal side effects such as hypotonia and feeding difficulties, postpartum administration only minimally elevates milk magnesium levels and shows low oral absorption in breastfeeding infants.28,29
This integrative review was designed to evaluate and synthesize the current evidence on how postpartum magnesium sulfate administration affects lactation outcomes, including breastfeeding initiation, exclusivity, duration, pumping behavior, and SA, in patients with HDP.
Methods
Design
This integrative review was conducted following Whittemore and Knafl’s five-stage framework 30 and is reported in accordance with PRISMA 2020 guidelines. 31 This approach allows for the systematic inclusion and synthesis of both experimental and non-experimental research studies, resulting in a comprehensive assessment of how postpartum magnesium sulfate exposure impacts lactation outcomes.
Data sources and search strategy
A structured search was conducted in May 2025 across five databases: PubMed, CINAHL, Embase, Scopus, and Web of Science. Keywords and MeSH terms were combined using Boolean operators related to “breastfeeding,” “lactation,” “milk expression,” “secretory activation,” and “magnesium sulfate.” (See Supplementary Table S1). No date limits were applied. Reference lists of included studies were manually screened to identify additional sources. The search was performed with librarian assistance and managed through Covidence software.
Eligibility criteria
Studies were included if they met the following criteria:
Conducted with human participants. Reported intravenous magnesium sulfate therapy administered for the prevention or treatment of eclampsia among women diagnosed with HDP (e.g., preeclampsia, eclampsia, gestational or chronic hypertension). Examined at least one lactation outcome, such as initiation, frequency, or duration of breastfeeding, milk expression, pumping habits, or SA. Primary research. Published in a peer-reviewed journal. Available in English.
Screening and selection
A total of 1,184 records were identified, and 590 unique studies remained after duplicates were removed. Title and abstract screening excluded 532 articles that did not meet the inclusion criteria. Fifty-eight full texts were reviewed in detail, resulting in 11 studies included in the final analysis (see Fig. 1, PRISMA flow diagram). Any published article that was not primary research (e.g., review) was analyzed for applicable studies before exclusion. All 11 full-text articles were systematically screened to ensure consistency, using a standardized inclusion/exclusion protocol. 32 Screening was conducted by the primary author (H.H.S.), with discrepancies and uncertainties resolved through discussion and consensus with two additional reviewers (D.L.S. + R.R.S.C.).
Data evaluation
Each included study was assessed for methodological quality using the official Joanna Briggs Institute (JBI) critical appraisal checklist corresponding to its study design (e.g., cohort, RCT, case series). 33 Responses were systematically documented as “Yes,” “No,” “Unclear,” or “Not Applicable,” and recorded in a comparative appraisal table where questions that did not result in a “Yes” response were clarified. Weaknesses identified via these checklists were noted for complete transparency, but no overall numeric summary score was assigned. Levels of evidence were classified according to Melnyk and Fineout-Overholt’s 34 hierarchy.
Data extraction and analysis
Key study characteristics, including design, sample size, maternal diagnosis, lactation-related information, and magnesium sulfate regimen (dosage and duration), were extracted using a standardized review table (Table 1: Article summary). Given the wide variability in magnesium sulfate exposure reporting, studies were categorized in-text by exposure clarity: either “known postpartum duration” or “uncertain/unspecified/none.” Additionally, a figure was created (see Fig. 2: Hours of postpartum magnesium sulfate) to analyze visually postpartum duration differences. Lactation outcomes were synthesized comparatively, with particular emphasis on time to initiation, discharge rates, SA, and available lactation support due to the availability in the studies.

PRISMA flow diagram for the review of magnesium sulfate and breastfeeding outcomes.
Article Summary

Postpartum magnesium sulfate duration.
Results
Study characteristics
Eleven studies published between 1993 and 2023 met all inclusion criteria for final synthesis. Most studies were conducted in the United States (N = 6), with others from India (N = 1), Mexico (N = 1), Pakistan (N = 1), and one location only specified as “Latin America” (N = 1). The included studies comprised three randomized controlled trials, four retrospective cohorts, one cross-sectional study, one case series, and two case reports. Sample sizes ranged from single-case reports41,43 to multicenter cohorts of over 1,100 participants. 45 The majority of studies focused on women with severe preeclampsia; other hypertensive pregnancy disorders were rarely represented.
Participant and infant characteristics
Across the 11 studies included in this review (N = 2,842 women), most participants were diagnosed with severe preeclampsia or preeclampsia with severe features, accounting for over 60.3–98.6% of the study samples.36–40,44,45 A minority of women were diagnosed with superimposed preeclampsia (1.4–19.5%),36,37,40,44,45 and gestational hypertension or pregnancy-induced hypertension.39,41,43 Mushtaq et al. specifically focused on moderate to severe preeclampsia as an inclusion criterion, without stratifying by severity, 42 whereas Burgess et al. reported an equal distribution between early-onset and late-onset preeclampsia, without specifying severity (48.8% and 51.2%, respectively). 35
Maternal age was generally consistent across studies, with most reporting mean ages typically ranging from 24 to 30 years.35–38,40,42,44,45 Smaller studies had individual patient ages ranging from 26 to 35 years.39,41,43 Primiparous patients made up the majority (55–65%) across studies.36–41,43–45 Among studies reporting data, the prevalence of pregestational and gestational diabetes ranged from 2.3% 40 to 15%, 35 except for Cordero et al. (2020), which reported 70% of their sample with diabetes due to its focus on that population. 37 Maternal body mass index was inconsistently reported, generally within the overweight to highly obese range, consistent with the risk profile of hypertensive pregnancies.35,37,39 However, many studies did not specify when weight was measured, limiting reliability.36,38,40–45
Gestational age at delivery clustered around a mean of 36–37 weeks, mainly reflecting late preterm to early term births.34,37,38,42,44,45 Some variability was observed, with some studies reporting means or the majority between 34–36 weeks.35,36,40 Smaller studies involving single cases39,41,43 reported gestations ranging from 37 to 39 weeks.
Cesarean birth rates were consistently elevated compared to national averages, with most studies ranging from 52% to 75%, reflecting the high acuity and management of these hypertensive pregnancies.35–37,40,44,45 Cordero et al. (2021) reported lower cesarean rates with 58% vaginal and 42% cesarean deliveries. 38 Smaller case series reported a combined average cesarean rate of 50%.39,41,43 Among studies that reported neonatal data, neonatal intensive care unit (NICU) admission rates ranged from 44% to 58%.35–39 Most studies did not report on NICU admission data.40–45
Magnesium sulfate regimens
Known postpartum magnesium sulfate duration
In most studies, postpartum magnesium sulfate regimens typically involved a 24-hour continuous infusion after delivery.36–39,42,44,45 Several studies compared these standard protocols with shorter postpartum exposures (6–8 hours) or discontinuation at birth.42,44,45 Haldeman (1993) was unique in having a prolonged postpartum exposure of 48 hours 41 (see Fig. 2: Postpartum magnesium sulfate duration).
Uncertain/Unspecified postpartum magnesium sulfate duration
A minority reported only that magnesium sulfate was used, with limited further information and an unclear or unspecified duration.35,43
Dosage information
Most studies that reported dosage information ranged from 1 gram per hour40,44,45 to 2 grams per hour.39,41 There was also a 4-gram40,44,45 or 6-gram 41 loading dose. Mushtaq et al. (2023) participants received 8 microdrops/minute with unknown volume or grams of the bag 42 (See Table 2: Magnesium dosage). Specific dosage information was not provided in some studies.35–38,43
Postpartum Magnesium Sulfate Dosage Information
Lactation outcomes
Breastfeeding definitions
Some studies have definitions related to lactation.35–38 Cordero defined breastfeeding broadly: in 2012, “successful initiation” meant ≥50% of feedings were human milk (either direct or expressed) at discharge, with 84% meeting this criterion. 36 In contrast, Burgess et al. (2019) defined breastfeeding as women feeding their newborns with their own milk through any modality. 35 Cordero et al. (2020 and 2021) defined breastfeeding initiation as any exclusive or partial breastfeeding during the last 24 hours before hospital discharge, with exclusive being mothers’ own milk by any modality or donor milk, and partial involving formula supplementation.37,38 Most studies did not clarify whether initiation referred to direct breastfeeding or pumping, or whether patients restricted from direct breastfeeding were permitted to pump.
Initiation timing
Despite definitional differences, a consistent pattern emerged: shorter magnesium sulfate exposure was linked to earlier breastfeeding initiation. Across multiple studies, shorter postpartum regimens (6–8 hours) or none (0 hours) resulted in initiation 8–12 hours earlier than standard 24-hour protocols, with mean initiation times ranging from approximately 14 to 26 hours in shorter-duration groups compared to 17 to 37 hours in longer-duration groups.40,42,44 Case reports indicated initiation rates of 30–90 minutes postpartum.39,41 All women in the preeclampsia with severe features (PWSF) group remained in labor and delivery during the first 24 hours postpartum. The main disruption in mother–infant contact after birth was due to the need for immediate transfer of symptomatic prematurely born infants to the NICU (43% of the PWSF and 14% of the preeclampsia without severe features [WOSF] group). 38
Where the intention to breastfeed was documented, initiation rates were high, although the timing varied considerably.36–38 Cordero et al. (2020 and 2021) reported that only 16–18% of mothers started breastfeeding within the first hour, while 40–42% delayed initiation beyond 7 hours.37,38 The 2012 study by Cordero et al. showed that 51% of the total sample-initiated breastfeeding, with 29% of neonates transferred to the Well-Baby Nursery being breastfed at least once on the first day and 26% initiating on the second day or later. 36 Two studies did not specify the timing of initiation.35,43
Breastfeeding rates and duration
Findings on exclusivity and discharge feeding status were mixed, with only half of the studies addressing these aspects. Exclusive breast milk feeding ranged from 26% to 37%, while mixed feeding ranged from 22% to 34%, with lower rates seen in babies admitted to the NICU.36–38 Conversely, Burgess et al. (2019) reported overall breastfeeding discharge rates of 85.7% (n = 210). 35 In smaller studies, exclusive breast milk feeding was observed in 2 out of 3 participants.39,41 Half of the studies did not report breastfeeding discharge rates, and data on breastfeeding after discharge were limited; only some studies provided longer-term outcomes.35,39,41 These results varied, with Burgess reporting breastfeeding at the first postpartum visit at 49.5%, Demirci et al. reporting by 3–4 months, and Haldeman reporting at day 10—all of whom were exclusively breast milk fed. Many studies did not report breastfeeding discharge rates.36,40,42,43,45
Perceived SA
Only two studies reported explicit data on SA. Where reported, perceived onset of copious milk ranged from day 4 39 to day 10 41 postpartum. None of the studies reported objective data, such as sodium and lactose biomarkers, to measure SA. Additionally, none of the studies reported the use of pre- and post-weights to determine milk intake.
Expressed milk and pumping habits
There is significant variation in how milk expression and pumping practices are defined and reported. Haldeman (1993) specifically documented using an electric breast pump on both breasts after each feeding for 10 minutes per session from postpartum day 3 through 7 due to concerns about milk supply. 41 In contrast, Cordero et al. (2012, 2020, 2021) used the term “expressed milk” to include both pumping and possible hand expression, without providing details on the method, timing, or frequency.36–38 At hospital discharge, 33–36% of patients who received magnesium sulfate provided pumped milk to their infants, with 9–10% exclusively feeding pumped milk.37,38 Among NICU neonates (n = 155, 54%) in Cordero et al. (2012), the initiation of expressed milk varied widely: 6 infants on day 1, 24 on day 2, and 44 on or after day 3. 36 Demirci et al. (2018) described postpartum ICU patients engaging in hand expression and pumping, with consistent post-discharge pumping six times per 24 hours. 39 Other studies (Burgess et al., 2019; Gutiérrez-Vela et al., 2021; Mushtaq et al., 2023; Sithara, 2015; Vigil-De Gracia et al., 2017 & 2018) did not provide specific details on pumping or milk expression.35,40,42–44
Lactation support
Lactation support was sparsely reported across studies. Cordero et al. (2020 and 2021) noted that approximately 77–87% of their sample received postpartum lactation consultation when reported.37,38 The 2012 study by Cordero et al. described standard protocols ensuring lactation support access for women planning or undecided about breastfeeding, 36 while individualized support was detailed in case studies.34,39 Most other studies did not provide information on lactation support, including Burgess et al., Gutiérrez-Vela et al., Mushtaq et al., Sithara, and Vigil-De Gracia et al.35,40,42–44
Discussion
This integrative review demonstrates that postpartum magnesium sulfate exposure complicates lactation outcomes for women with HDP, impacting both breastfeeding initiation and discharge rates. The reviewed studies consistently describe a high-risk perinatal population: primarily presenting as women with severe preeclampsia delivering late-preterm to early-term neonates by cesarean, with high NICU admission rates. These clinical factors frequently coincide, reducing early opportunities for milk removal and delaying breastfeeding initiation.
The maternal side effects of magnesium sulfate, especially at higher doses, often hinder early breastfeeding. Maternal drowsiness and sedation, which become more pronounced with longer or higher-dose treatments, can disrupt breastfeeding and bonding.12–14,46 The incidence of these side effects ranged from 24% (1 g/h) to 71% (2 g/h), demonstrating a clear dose–response relationship.12,13 Burgess et al. noted that women with hypertensive disorders during pregnancy were less likely to breastfeed or pump, often citing illness or medication as reasons for not starting or stopping. 47 Cordero et al. specifically found that among women receiving magnesium sulfate, only 17–18% initiated breastfeeding within the first hour, whereas 40%–42% started it after 7 hours, despite 77–80% intending to breastfeed exclusively. Ultimately, only 30–37% were able to do so at discharge.37,38 These figures are notably lower than the Ohio state average of 51.8% for exclusive breastfeeding. 48
Notably, most lactation studies do not clearly separate the effects of HDP, magnesium sulfate exposure, or other comorbidities. This greatly limits the ability to determine whether lactation difficulties stem from hypertension itself, medication effects, or other confounding factors. Nevertheless, existing research suggests that HDP is linked to shorter breastfeeding duration, delayed SA, and lower latch scores compared with women without hypertension.24–27 Emerging mechanistic evidence also connects HDP to impaired lactation through inflammatory and placental pathways. 49 HDP can cause placental dysfunction, which disrupts hormonal signaling necessary for mammary gland development and secretory differentiation, 49 potentially leading to problems with SA. HDP has been associated with significantly delayed SA and reduced pumping frequency. 50
Lactation support has been shown to have a dose-response relationship with increased rates of exclusive breastfeeding at 6 months. 51 Proactive lactation management can mitigate risks among high-risk mothers, particularly those with NICU infants. For example, among preterm (<34 weeks) mothers (41% preeclamptic, 59% cesarean), pumping in the first 5 days postpartum was strongly associated with improved SA. 52 Each additional hour of pumping daily increased the odds of timely SA by 2.8-fold; 52 similar benefits have been observed for early, regular milk expression outside the NICU setting. 53 Technological advances now enable rapid, biomarker-based clinical assessment of SA.54,55 Validated biomarker protocols using human milk sodium-to-potassium (Na:K) ratios provide objective, standardized measurement of SA timing,54,56 enabling precise identification of at-risk dyads for targeted intervention. Coupled with evidence-based institutional protocols such as the Spatz 10-step method, this represents a key avenue for improving high-risk lactation outcomes.57,58
Elevated NICU admission and cesarean rates are related confounders that increase the risk of lactation issues but may be modifiable through better clinical and institutional practices. NICU admission rates in the reviewed studies ranged from 44% to 56.6%, significantly higher than the national average of 9.8%. 59 Cordero et al. (2012) found that only 4% of NICU neonates received their mother’s own milk on day 1; delayed milk expression directly contributes to later breastfeeding cessation and suboptimal supply, making pumping essential for achieving adequate volume.26,36,60,61 However, Burgess et al. (2019) reported that targeted lactation support within the NICU can mitigate some adverse effects, as evidenced by increased human milk receipt at discharge. 35
Separation of mother and infant due to hospital protocols on magnesium sulfate therapy further restricts opportunities for early initiation and rooming-in. In Cordero et al. (2012), only 78% of neonates in the Well-Baby Nursery saw their mothers within 24 hours, and 22% did so only after discharge from labor and delivery following magnesium sulfate therapy. 36 Postpartum breastfeeding initiation is complicated by hospital protocols that require women on magnesium sulfate to remain on strict bed rest and refrain from breastfeeding until the infusion ends, usually after 24 hours.44,45 Additionally, rooming-in separation may occur because the mother stays on the Labor and Delivery unit or is transferred to the high-dependency unit after birth without her baby, or the baby is transferred to the NICU or nursery.37,42 Such delays reduce opportunities for skin-to-skin contact and are consistently associated with decreased breastfeeding, highlighting the need for policy changes.62,63
In summary, ongoing reliance on institutional policies that restrict breastfeeding during magnesium therapy, frequent mother-infant separation, and inconsistent tracking of feeding and milk expression practices create significant interpretive challenges. Relying on subjective rather than biomarker-confirmed measures of SA further hampers scientific progress, along with short follow-up periods and inconsistent reporting of key lactation behaviors. Importantly, the lack of qualitative studies examining both the side effects of magnesium sulfate and lactation outcomes leaves a substantial gap in understanding.
Clinical implications
Based on the findings of this integrative review, magnesium sulfate negatively impacts lactation outcomes in the early postpartum period. Considering the evidence of delayed breastfeeding initiation, disrupted SA, and reduced pumping frequency, clinical teams should prioritize immediate postpartum lactation assessments and provide additional support to these mother-infant pairs. A modified lactation protocol guided by the Spatz 10-step method may be especially helpful in addressing specific challenges such as magnesium side effects, high cesarean section rates, and increased NICU admissions leading to maternal-infant separation. Early and frequent milk removal, whether through direct breastfeeding or pumping, should be actively promoted, with close monitoring to mitigate the effects of therapy-related maternal fatigue or hospital policies on stimulation. Ultimately, individualized feeding plans, education, and institutional support for rooming-in during magnesium therapy can help improve breastfeeding outcomes for this high-risk group.
Future research
To improve care for high-risk mother-infant pairs exposed to HDP and magnesium sulfate, future research should focus on: (1) the use of human milk biomarkers (sodium and lactose) instead of relying on mothers’ subjective reports of “milk coming in,” allowing for precise detection of delayed SA within the critical window. (2) Detailed documentation of milk removal behaviors: Future research must systematically record the frequency, duration, and mode (direct breastfeeding versus pumping) of milk removal during the first 72 hours postpartum, as early and frequent stimulation is biologically vital for SA. (3) Qualitative exploration of maternal experiences: Qualitative studies are necessary to understand how mothers perceive magnesium sulfate side effects (sedation, nausea, weakness), manage institutional policies (such as rooming-in restrictions and lactation support access), and overcome barriers to early milk removal. (4) Examination of modifiable clinical and institutional factors: Research should identify specific hospital policies, lactation support protocols, and magnesium sulfate dosing strategies that can be adjusted to improve lactation outcomes while ensuring maternal safety.
Limitations
Screening for this review was conducted as part of this first author’s qualifying exam for her PhD as per the requirements of the university. The other two authors were consulted for areas of uncertainty or discrepancies. This approach may introduce potential selection bias into the process of identifying studies. Only studies published in English were included, which can limit the comprehensiveness and generalizability of the results. Additionally, relevant unpublished data or studies outside the searched databases might not have been identified, potentially affecting the review’s scope.
Conclusions
Postpartum magnesium sulfate therapy in women with HDP is consistently linked to significant delays in starting lactation, including the beginning of breastfeeding and SA. This disruption is possibly caused by both the medication’s side effects, such as sedation and maternal fatigue, and hospital policies that restrict early infant feeding, rooming-in, and maternal–infant interaction during magnesium treatment. The overall evidence indicates that mothers who undergo extended magnesium therapy face significant obstacles to their lactation journeys, which prevent both mother and child from receiving optimal health benefits.
Authors’ Contributions
H.H.S.: Conceptualization, methodology, literature screening and selection, data extraction, formal analysis, writing—original draft, writing—review and editing, and project administration. D.L.S.: Supervision of PhD student and consultation through the entire process and final editing of the article. R.R.S.C. and D.L.S.: Assessment of eligibility for inclusion or exclusion in cases of uncertainty, methodological evaluation, contributions to organizational clarity through multiple rounds of article review.
Footnotes
Author Disclosure Statement
No competing financial interests exist.
Funding Information
No funding was received for this article.
Supplemental Material
Supplemental Material
References
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.
