Can we rely on delaying the uptake of sugar consumption in the prevention of early childhood caries?

The authors present a well-designed randomised clinical trial to investigate the impact of a tailored sugar prevention programme on the development of early childhood caries (ECC)¹.

The study design is appropriate for the intended research question. It is reassuring that a power calculation was performed to determine sample size although this may have been a post-hoc sample size calculation. Randomisation and allocation to the intervention and control groups were completed effectively through the use of a computerised system and without bias, and the investigator responsible for this was blinded to the recruitment process. There was a large drop-out rate which an intention-to-treat analysis accounted for. However, minimal information was provided on how missing data was dealt with.

The large drop-out rate reduced the sample size. Fortuitously, the drop-out rate was similar for both groups. The authors acknowledge this limitation and attribute this to change in participant address and/or refusal to participate. Alternatively, it could be a disinterest in the study due to the participant burden of monthly phone calls/home visits on parents who are caring for a young child. Engagement with primary dental care or ante-natal team was not been recorded and refusal to participate may also be due to mother and child receiving risk-based prevention from another provider, negating the need for their involvement.

Given the nature of the intervention, dietary counsellors were not blinded. In contrast, data collection on diet (by interviewers) and oral health (by a paediatric dentist) were undertaken by trained, calibrated and blinded individuals. Appropriate inter- and intra-rater scores were provided. Good internal validity to the study was apparent as the follow-up process was robustly reported, reassuring the reader that the methods were undertaken in a consistent and robust manner. Unfortunately, the main outcome was the impact on ECC. ECC is described up to the age of three², thus the short follow-up period limits the full extent of ECC rates in both cohorts. A longer follow-up period of the trial would have been important to fully establish the impact.

The more intensive counselling seems to have been effective at preventing early introduction to dietary and additional sugars to children at both 6-months and 12-months, expressed through the sugar consumption index. The validity of this outcome is however questionable. The use of limited questionnaires identifying only food and drink with added sugar, alongside 24-hour recall, is non-specific and does not provide longitudinal information of the children’s sugar consumption. The intervention group also received advice with a focus on avoiding added sugars. The use of a questionnaire that quantified foods with added sugar to record the sugar consumption index could result in acquiescence bias in the intervention group, with them only reporting the foods that are seen as correct. In addition, the resource and cost required for the intervention counselling has not been accounted for. A simple cost-analysis, or fuller economic evaluation would have supported whether the intervention is cost-effective or not. Having this information would be of benefit to policymakers who consider how best to allocate fixed and scarce resources.

No significant difference was found between the two groups for the occurrence of ECC and caries-affected teeth. The intervention group however did report a lower percentage of those diagnosed with ECC (19.6%) when compared with the control group (14.7%). The intervention group also had a lower maximum number of teeth affected by caries (4) versus the control group (6). This would suggest that an intervention is better than none, but no significant difference was seen between the two. Precision was effectively reported with confidence intervals and no harms or unintended effects of the intervention were made clear by the authors. The programme targeted only diet but did not account for other factors like fluoride use or oral hygiene which contribute to caries disease progression. There was also no record of how the children have been fed. The most common pattern of ECC is that of ‘bottle caries’ which is the introduction of a sugary beverage via a bottle frequently³. This frequent consumption of a sugary drink is often the prime aetiological factor in ECC³ but there are no details of what the children were most commonly eating or drinking. Furthermore, there was no data on the breastfeeding habits of the mothers. Despite varied evidence of the impact of breastfeeding on ECC, breastfeeding up to the first year could have an impact on the occurrence of ECC due to the reported protective effect⁴. No oral hygiene instruction was included in the intervention and both groups presented with similar plaque scores at their 12-month review. With no mention of outside involvement that may be providing further advice, we cannot account for this confounding factor. There has also been no reporting on the brushing habits or use of fluoride toothpaste. Without consideration of these other factors that are known to prevent dental caries, it is not possible to conclude that the programme had no impact on the occurrence of ECC. Accounting for these various confounding factors could have been dealt with in a regression analysis.

While the prevention of sugar intake is a positive, the programme had no impact on ECC and it is unclear whether it is cost-effective. Furthermore, advanced prevention including diet advice can be provided by general dentists and supported by guidance to support the dental practitioner in providing targeted advice^5,6.

The results of this study are perhaps best used in the development of other studies/programmes to target ECC. It highlights that dietary counselling alone is insufficient to influence ECC and suggests that such initiatives should be part of a broader preventive programme. Overall, while the study design was sound, its conclusions are confounded by the lack of consideration of other preventative factors.