The Short-term Effect of Spinal Manipulation in the Treatment of Infantile Colic: A Randomized Controlled Clinical Trial with a Blinded Observer


The Short-term Effect of Spinal Manipulation in the Treatment of Infantile Colic: A Randomized Controlled Clinical Trial with a Blinded Observer

Jesper M.M. Wiberg, DC, Jan Nordsteen, DC, Niels Nilsson, DC, MD, Phd

Center for Biomechanics,
Odense University, Denmark

OBJECTIVE:   To determine whether there is a short-term effect of spinal manipulation in the treatment of infantile colic.

DESIGN:   A randomized controlled trial.

SETTING:   A private chiropractic practice and the National Health Service’s health visitor nurses in the suburb Ballerup (Copenhagen, Denmark).

SUBJECTS:   Infants seen by the health visitor nurses, who fulfilled the diagnostic criteria for infantile colic.

INTERVENTION:   One group received spinal manipulation for 2 weeks, the other was treated with the drug dimethicone for 2 weeks.

OUTCOME MEASURE:   Changes in daily hours of crying as registered in a colic diary.

RESULTS:   By trial days 4 to 7, hours of crying were reduced by 1 hour in the dimethicone group compared with 2.4 hours in the manipulation group (P = .04). On days 8 through 11, crying was reduced by 1 hour for the dimethicone group, whereas crying in the manipulation group was reduced by 2.7 hours (P = .004). From trial day 5 onward the manipulation group did significantly better that the dimethicone group.

CONCLUSION:   Spinal manipulation is effective in relieving infantile colic.

From the Full-Text Article:


Infantile colic was first described in 1894 as dyspepsia. Since then, different sets of diagnostic criteria have emerged, but there has never been complete agreement about any of these. However, the most accepted definition is: unexplainable and uncontrollable crying in babies from 0 to 3 months old, more than 3 hours a day, more than 3 days a week for 3 weeks or more, usually in the afternoon and evening hours. [1-17]

Several studies have described a difference in the quality of crying in infants with colic compared with normal infants. Infants with colic have more hours of crying per day and more periods of crying per day, [1, 17-20] and the crying in infantile colic has a higher frequency/pitch. [10] Other studies describe motor unrest, typically with flexing of the knees against the abdomen, clenching of the fists, and extension or straightening of the trunk and extremities. Several studies conclude that infants with colic are perfectly normal but cry more than other infants. [21-28] All studies agree that apart from the increased crying, these infants are otherwise healthy, thriving, and have a normal weight gain.

The incidence of infantile colic among newborns is estimated to be between 8% and 49% in different studies. [1-12, 17, 19, 29-35] An average of these estimates would put the incidence of infantile colic at 22.5%, and with a birth rate of 67,731 children in Denmark in 1992 this would mean that some 15,226 newborns would have infantile colic every year in a country with a population of 5 million (Statistisk månedsoversigt 1993:3:12).

Infantile colic is often described as a “benign self-limiting condition” that disappears spontaneously at 3 months of age. But a good observational study has shown that only 47% of infantile colic cases have disappeared at the age of 3 months, a further 41% disappeared before 6 months of age, and the remaining 12% of cases persevered until between the ages of 6 and 12 months. [36]

Although infantile colic may be regarded as a “benign self-limiting condition” that only results in a temporary delay in the development of child, [33] several studies suggest that the effects of infantile colic may be more severe than that. Some studies have concluded that infantile colic can lead to a negative mother-child relationship, which may persist for up to 3 years after the disappearance of symptoms. [31, 32, 37-42] Other studies conclude that these troublesome and screaming infants are in a high-risk group for ill treatment, for possible central nervous system damage, and even death from being shaken (shaken infant syndrome). [10, 21, 43-50]

The cause of infantile colic remains unknown. Many have pointed to increased air in the intestines as the cause. [51-53] But radiologic studies have shown that the amount of air in the intestines of infants with colic is the same as in normal infants. [6, 29, 54] Neither has it been possible to find constrictions in the intestines, and the gastrointestinal transit time in infants with colic is also normal. Others have investigated levels of intestinal hormones, fecal analysis, intestinal microflora, markers of intestinal damage, or occult blood in the stools and hydrogenion concentration in the breath, but none of these studies have convincingly identified abnormalities in infants with colic. [12, 34, 52, 55-59]

Various modes of delivery (cesarean section, vaginal extraction, or vacuum extraction) have also been studied, but no differences could be identified in the incidence of colic in infants delivered by these three modes. [60, 61] Neither does the use of pudendal block, epidural analgesia, general anesthesia, or intravenous oxytocin seem to increase the risk of infantile colic. [62]

Numerous treatments for infantile colic have been advocated, but most of these, such as music and different sound or vibrations (car-simulating devices), have been shown to be without any effect. [3, 18, 63, 64] Treatment with sucrose seems to have a generalized analgesic effect in infants and may therefore also help in infantile colic. [65]

Different medical treatments for infantile colic have also been used. Dicyclomine hydrochloride [5, 15, 66] was often used with some effect until the mid 1980s, when reports about serious side effects stopped this treatment. [67-72] Gripe-water, alcohol, atropine, skopyl, phenobarbital, meperidine, homatropine, and merbentyl have also been used with more or less convincing results. [29, 31, 73-77] Phenobarbital, homatropine, and alcohol have been studied in a double-blind design, which showed no effect compared with a placebo treatment, [75] and most other preparations have had serious side effects. [78-82]

One drug that is still used is dimethicone (Simethicone, Mylicon), and several good controlled studies have shown that this is no better than placebo treatment. [3, 67, 72, 83, 84]

Chiropractors and others have for many years manually treated infants with colic symptoms with apparently good results. [4, 7, 30, 60, 61, 85-90] The background for this has been the purely clinical observation that functional disturbances of the vertebral column may cause symptoms similar to those of infantile colic, and several studies seem to support this clinical observation. [7, 30, 61, 86-92]

In 1985 a retrospective study was undertaken in Denmark on the chiropractic treatment of infantile colic,89 and this was followed in 1989 by a prospective multicenter study. [7] Both studies suggest that there seems to be a positive effect of spinal manipulation for infantile colic, but because none of these studies had a control group, it is impossible to assess whether the effect observed was significantly better than placebo.

This trial is a prospective, randomized controlled clinical trial with a blinded observer for the purpose of studying whether any effect exists beyond placebo from spinal manipulation in the treatment of infantile colic.


When we compare our results with those of the only other prospective trial on spinal manipulation in the treatment of infantile colic, [7] we see that the results are virtually identical for our manipulation group. In the previous trial, mean daily hours with colic were reduced by 66% on day 12 of the trial, and in our trial we saw a reduction of 67% on day 12 in the manipulation group, whereas our dimethicone group only had a reduction in daily hours with colic of 38% by day [12]. These similarities strengthen the conclusion that a positive effect of spinal manipulation exists in the treatment of infantile colic. These strong similarities perhaps also suggest that we may be dealing with an underlying anatomic/physiologic disease mechanism rather than a psychosocial one, but that remains to be clarified.

Initially we planned to have a baseline observation period of 7 days, but soon after the start of the project we had to revise this to 4 days because the families demand for fast action was too strong. If we had upheld the 7-day baseline observation period, it would have been difficult to recruit a sufficient number of participants because parents would have sought help elsewhere. The problem is understandable in retrospect because families of infants with colic are under a tremendous psychologic strain, as anyone with infantile colic experience will appreciate. We do not think that reducing the length of the baseline observation period has any effect on the validity of the results because this was the same for both treatment groups.

We would also have liked a longer follow-up period, but once again families with small infants are busy families (colic or not). It was our impression, that it would not have been possible to keep those participants who still had colic in the trial much longer, whereas those who were helped by the treatment would want to get on with normalizing their family lives. The result of a longer follow-up period would thus have been an unacceptably high number of dropouts. Furthermore, in the previous trial on spinal manipulation in the treatment of infantile colic, they had a follow-up at 4 weeks but found that the reduction in crying was clearly within the first 14 days, and no significant change occurred in the last 2 weeks of follow-up. [7]

The pretrial characteristics of the 2 treatment groups show only 1 significant difference. The age of the infant when colic started was 2.2 weeks in the dimethicone group compared with 1.2 week in the manipulation group (P = .02), and this could potentially be a problem. If infantile colic tends to disappear spontaneously after a given number of weeks, the difference could have meant that infants in the manipulation group might be more likely to experience spontaneous alleviation of symptoms than those in the dimethicone group, and this could be the reason for the observed effect. But because the age of the infant on entry into the trial also has the same 1-week difference in age and because both groups on entry had infantile colic for 3.7 weeks, this could not have been the reason for the observed difference.

If, on the other hand, infantile colic tends to stop spontaneously at a given age, then the 1-week age difference would instead have favored a higher rate of spontaneous disappearance in the dimethicone group, and this was not evident. So even if the difference in age on debut of symptoms was statistically significant, it did not affect the validity of our results.

There were 9 dropouts in the dimethicone group and none in the manipulation group. This could be interpreted as a sign of parents’ bias against dimethicone, but inspection of the data in the last infantile colic profile completed before dropout show clearly that for all where information is available, the dropout was due to a genuine worsening of symptoms and not parents’ bias. By excluding data from the dropouts, we are excluding more severe cases from the dimethicone group, and this has the effect of making that group appear better that it actually was. Thus we are introducing a serious bias against showing an effect of spinal manipulation, and despite this the manipulation group did significantly better.

Spinal manipulation is normally used in the treatment of musculoskeletal disorders, and the results of this trial leave open 2 possible interpretations. Either spinal manipulation is effective in the treatment of the visceral disorder infantile colic or infantile colic is, in fact, a musculoskeletal disorder, and not, as normally assumed, visceral. This study does not address this issue.