Which is Safer and More Effective –Miralax (PEG 3350) or Lactulose?

A recent randomized multicenter study (D Jarzebicka et al. JPGN 2019; 68: 318-24) compared polyethylene glycol 3350 (PEG, aka Miralax) and lactulose for functional constipation in a cohort of 102 patients (12 weeks of treatment and 4 weeks of follow-up).


  • PEG 3350 dosing divided into 2 doses:
    • <8 kg:  5 g/day
    • 8-12 kg: 10 g/day
    • 12-20 kg: 15 g/day
    • >20 kg: 20 g/day
  • Lactulose was given as 1 mL/kg twice a day

A “good clinical outcome” was considered if there were three or more stools per week and an improvement in stool consistency of at least 2 types in the Bristol scale.

Key findings:

  • At week 12, a good clinical outcome was noted in 98% of PEG group compared with 90% in the lactulose group
  • PEG group had increased defecations relative to lactulose: 7.9 vs. 5.7, less stool retention: 7% vs 10%, and fewer hard stools 7% vs 13%
  • Side effects, mainly abdominal pain and bloating, were reported more frequently with lactulose than with PEG: 23 vs 15, P=0.02)

My take: This study showed that PEG 3350 was more effective and had fewer side effects than lactulose for constipation in children between 6 months of age and 6 years.

Related blog posts:

From Civil and Human Rights Museum, Atlanta

PEG 3350 is Not Associated with Elevated Glycol Levels

Everyday parents ask me if Miralax (polyethylene glycol) is safe; this has been driven by social media claims of neurotoxicity and by articles in the NY Times (see prior blog references) indicating that more testing is needed.

A recent study (KC Williams et al. J Pediatr 2018; 195: 148-53) examines one of the areas of concern, whether miralax could result in toxic levels of glycols.  In this study with 9 treated children (ages 6-12 years) and 18 controls, careful study of potentially toxic agents, ethylene glycol (EG), diethylene glycol (DEG), and triethylene glycol (TEG), were measured every 30 minutes for 3 hours after receiving 17 g of PEG 3350.

Key findings:

  • Baseline blood levels of EG (390.51 ng/mmL) and TEG (2.21 ng/mL) did not differ between control and treated groups
  • Baseline DEG levels were lower in the PEG 3350 group (40.12 ng/mL vs 92.83 ng/mL, P=.008)
  • After PEG 3350 dose, EG and TEG levels remained well below toxic levels; DEG levels did not change.  The increases in EG and TEG, which peaked at 90 minutes, were not sustained at levels different from controls.
  • EG peaked at 1032.8 ng/mL. TEG peaked at 35.17 ng/mL
  • The highest levels of EG and DEG were actually identified in control patients. Thus, “all children are exposed routinely and have measureable amounts in the blood.”

With regard to TEG toxicity, in the discussion, the authors note that, based on animal studies, “very large doses of TEG are needed to cause side effects.” Even doses of 4000 mg/kg of TEG daily for 90 days did not result in local or systemic toxicity.  The authors note that TEG concentration in PEG 3350 is “approximately 22.1-30.6 mcg per 17 gram dose of PEG 3350.”

With regard to EG and DEG, “the average EG and DEG content of the PEG samples in this study were a 100 and 800 times less, respectively, than this required 0.2% cutoff” [FDA limit].  The agency of Toxic Substances and Disease Registry profile for EG, has indicated that “EG blood levels greater tan 0.2 mg/mL are needed for acute toxic poisoning.  The average level of EG at the 90-minute peak of 1100 +/- 350 ng/mL was 182 times lower than this level.”  For chronic exposure EG toxicity, the authors estimate that one would need to take “40 capfuls [17 gram each] of PEG 3350 per day for up to a year.”  The EPA also has advisories with regard to EG.  To achieve toxic levels for a 10-kg child, this would necessitate that the child “would have to drink 1 L of water with 50 capfuls (858 g) in 1 day or 15 capfuls (258 g) per day for 10 days.”

An important limitation of this study is that there may be other metabolites that are not measured that could cause neurotoxicity.

My take: This study shows that the theoretical risk of glycol toxicity is highly unlikely.  My advice for miralax usage: (borrowed from expert review): “Generally speaking, if your child has been prescribed PEG 3350 as part of his/her treatment plan, and you feel this medicine provides benefit, you should feel safe continuing PEG 3350. At this time, PEG 3350 appears to be safe based on current medical literature. We recommend discussing any concerns you have about the safety of PEG 3350 with your child’s health care provider. If you would prefer for your child to stop taking PEG 3350, discuss other treatments options with your child’s health care team before stopping PEG 3350 therapy. Although abruptly stopping PEG 3350 is not considered dangerous, it could lead to a relapse/worsening of constipation.”

Disclaimer: These blog posts are for educational purposes only. Specific dosing of medications/diets (along with potential adverse effects) should be confirmed by prescribing physician/nutritionist.  This content is not a substitute for medical advice, diagnosis or treatment provided by a qualified healthcare provider. Always seek the advice of your physician or other qualified health provider with any questions you may have regarding a condition.

Related blog posts:

Colonoscopy, Split-dosing bowel preps, and Ottawa scores

Adequate bowel cleansing improves the results of colonoscopy.  Since colonoscopy is a big part of gastroenterology/pediatric gastroenterology, a lot has been written comparing bowel preparations.  A recent study has examined a split-dose polyethylene glycol (PEG) preparation (Gastrointest Endosc 2012; 75: 583-90).

This prospective study from a single center in South Korea examined the effectiveness of a split-dose regimen in 366 patients (18 -65 years).  An interval of 3 to 5 hours between the completion of the last dose of the preparation and the start of the colonoscopy had the highest scores for bowel cleansing.  The authors used the “Ottawa Bowel Preparation Scale” for assessing a quality score.  Favorable odds ratios were noted with optimal preparation interval (OR 1.85), amount of PEG ingested (OR 4.34), and adherence with diet instructions (OR 2.22).

Instructions for preparation in this study:

  • 1. Low-fiber diet for 3 days prior
  • 2. Regular diet for breakfast & lunch day prior; soft diet for dinner, then only clear liquids
  • 3. Two liters of PEG on day prior and then two liters early in morning (5-7 am) on day of procedure –NPO for at least two hours prior to procedure.  While taking PEG, patients were instructed to consume 250 mL every 10 minutes.

Ottawa Scale –score right, mid, and left colon:

  • 0    no liquid
  • 1    minimal liquid –no suctioning required
  • 2    suctioning required to see mucosa
  • 3    wash and suctioning needed
  • 4    solid stool, not washable
  • Plus 0-2 points for overall quantity of fluid.
  • Max score 14 points

In this study, a prolonged interval (> 7 hours) between prep and colonoscopy had increasingly higher scores.  Score for 3-4 hours was 4.25; score for 7-8 hours was 5.20 and score for > 8 hours was 5.92.

As alluded to above, individuals who consumed adequate volume of preparation, had good bowel cleansing.  99% of patients who took at least 75% of PEG (> 3 liters) had a satisfactory preparation.  And, 95% of patients who took less than 75% of PEG had an unsatisfactory preparation.  Compliance with diet instruction was associated with a satisfactory prep in 89% and poor compliance was associated with a poor prep in 81%.

While this study identified these risk factors, it is telling that even in a research study, 141 patients had an unsatisfactory preparation (Ottawa scale 6-14); 225 had a satisfactory preparation (Ottawa scale 0-5).

Additional references –bowel preparations:

  • -JPGN 2011; 53: 71. 2 day prep: Miralax 2 gm/kg, bisacodyl 5mg. n=111.
  • -JPGN 2010; 51: 254. Review of bowel preparations
  • -Gastroenterol 2006; 130: 2240. Oral phospho fleets associated with alarming risk of nephropathy. 21 additional cases identified
  • -J Am Soc Nephrol 2005; 16: 3389-3396
  • -NEJM 2003; 349: 1006. acute phosphate nephropathy after oral phospho fleets in a 71yo woman
Additional references –colonoscopy screening:
  • -Gastroenterol 2010; 138: 73, 27 (ed). Overutilization of colon screening in low risk situations and underutilization in high risk situations in clinical practice.
  • -Clin Gastro & Hep 2009; 7: 1217.  Fewer polyps detected as day progresses at a VA hospital n=477 pts. 27% decline in polyp detection.
  • -NEJM 2009; 361: 1179. Review of screening for colorectal cancer.
  • -Gastroenterol 2009; 137: 792. Use of CT colonography -current appraisal.
  • -Ann Intern Med 2009; 150: 1-8. Says endoscopists miss most cancers on right side & colonosopy reduces cancer by ~60% primarily due to left-sided cancers.  Most, 73%, of colonoscopies not done by GI/colorectal surgery in this study.
  • -Gastroenterol 2008; 134: 1570. Update recommendations from ACS, ACR, US Multi-society task force.
  • -Gastroenterol 2008; 134: 1311. Screening in 40-49 detects similar # of adenomas as in older groups (>50) but fewer advanced Ca.

Other references/links:

1. “Take it easy, Doc, you’re boldly going where no man has gone before.”
2. “Find Amelia Earhart yet?”
3. “Can you hear me NOW?”
4. “Are we there yet? Are we there yet? Are we there yet?”
5. “You know, in Arkansas, we’re now legally married.”
6. “Any sign of the trapped miners, Chief?”
7. “You put your left hand in, you take your left hand out. You do the
Hokey Pokey….”
8. “Hey! Now I know how a Muppet feels!”
9. “If your hand doesn’t fit, you must acquit!”
10. “Hey, Doc, let me know if you find my dignity.” and
11. “Could you write me a note for my wife, saying that my head is not, in fact, up there?”

Stimulants for constipation

Overall, 12-19% of Americans are affected by chronic constipation (Am J Gastroenterol 2004; 9: 750-59).  Despite the fact that constipation problems are widespread, the amount of useful research available to guide treatment is quite limited.  Two recent articles do offer some information:

  • Clin Gastroenterol Hepatol 2011; 9: 577-83.
  • Gut 2011; 60: 209-18.

The first reference examined the use of bisacodyl in a randomized, double-blind placebo-controlled study in the UK.  During the 4-week treatment period, patients receiving 10mg/day bisacodyl (n=247) had increased stools, from 1.1 per week to 5.2 per week.  Stool frequency also increased to 1.9 per week in the placebo group (n=121).  All secondary endpoints including constipation-associated symptoms (eg. quality of life indices, physical discomfort) improved significantly compared to placebo.  Average age of patients in this study was 55 years.  The main adverse effect was diarrhea –mainly during the 1st week of therapy.

A selected summary in Gastroenterology (Gastroenterology 2012; 142: 402-404) reviews the first study and makes several useful points:

  • Stimulant laxative use has been hindered by myths & misconceptions along with lack of supporting data.  Most recent studies do not support a role of stimulant use in causing enteric neuropathies, a cathartic colon or increasing the risk of colon cancer
  • Osmotic laxatives have been favored in guidelines but this has not been bolstered by supporting data
  • Only recently have two large randomized controlled studies proven the efficacy and safety of stimulant laxatives over the short-term
  • Long-term prospective studies are not available on the use of stimulant laxatives.

The second reference is a systematic review and meta-analysis of randomized controlled trials (RCTs) of pharmacologic therapy for chronic idiopathic constipation.  Twenty one eligible RCTs were identified: eight laxative studies (n=1411), seven prucalopride studies (n=2639), three lubiprostone (n=610), and three linactolide trials (n=1582).  All of these studies showed treatment was superior to placebo. These studies involved subjects who were mainly adults (>90% older than 16 years).

The results showed benefit from both stimulant and osmotic laxatives.  Overall, the osmotic and stimulant laxative studies showed higher response than the pharmacologic agents like prucalopride, lubiprostone, and linaclotide.  Nevertheless, between 50% and 85% of patients did not fulfill criteria for response to therapy.

Additional references:

  • -J Clin Gastro 2003; 36: 386-389.  Safety of stimulants for long-term use.
  • -Am J Gastro 2005; 100: 232-242.  Myths about constipation.  Stimulants have not been proven to cause a “cathartic colon”
  • -J Pediatr 2009; 154: 258.  Constipation associated w 3-fold increase in health utilization/cost.
  • -Clin Gastro  Hep 2009; 7: 20.  Review of complications assoc c constipation in adults.
  • -Pediatrics 2008; 121: e1334.  Behavioral therapy ineffective in treating childhood constipation.
  • -NEJM 2008; 358: 2332, 2344.  Use of methylnatrexone for opioid-induced constipation & trial of n=620 of prucalopride for severe constipation.  Both drugs were helpful.
  • -Gastroenterology 2004; 126: S33. Review of pediatric incontinence.
  • -J Pediatr 2004; 145: 253-4.  Prevalence of encopresis 15% /constipation 23% in obese children  (n=80).  Questionnaire administered to 80 consecutive obese children.
  • -Gastroenterology 2003; 125: 357.  Longterm constipation followup.  one-third with persistent constipation; 60% better at 1 year.  (tertiary referral group)
  • -Pediatrics 2004; 113: 1753 & e520.  When constipation & toileting difficulties both occur, constipation usually precedes toileting problems