Correction to: Diabetologia

https://doi.org/10.1007/s00125-023-05958-9

The authors are publishing this correction to clarify the following points:

• There is a substantial amount of missing data in this trial (>30% for the primary outcome, >50% for the secondary outcomes), meaning that the overall conclusions will need further verification in the future.

• Individuals with missing data were excluded from the analysis, so primary outcome results were based on 46 (strength training group [ST]), 42 (aerobic training group [AER]) and 43 (combined strength and aerobic training group [COMB]) participants. In addition, the analysis assumed that data were missing at random, and no attempt was made to assess whether the ST vs AER superiority conclusion is robust to deviations from this assumption.

• The authors gave the incorrect impression that they compared the change in HbA1c for the ST, AER and COMB groups at 9 months; in fact, their model included data from 3, 6 and 9 months.

• Contrary to CONSORT recommendations, the main focus of the authors’ analyses was changes within randomised groups rather than differences between groups.

• Reference to comparison of baseline characteristics between randomised groups has been removed, as this is considered by CONSORT to be ‘illogical’ and ‘superfluous’ and ‘can mislead’ (reference item 15 [1]).

• The full trial protocol is available at [2].

Specifically, the following changes have been made:

Abstract:

The p value for the decrease in HbA1c levels in the ST group in the intention-to-treat analysis was corrected from p=0.002 to p=0.02.

Methods:

The following sentence was reworded to read: ‘The primary outcome was the absolute change in HbA1c levels within and mean HbA1c across the three groups at 3, 6 and 9 months.’

The following sentence was deleted: ‘Baseline comparison between groups was performed using Welch’s t test or the Mann–Whitney U test if two groups were compared and one-way ANOVA or the Kruskal–Wallis test if three groups were compared.’

Figure 2:

The asterisks indicating significant differences between the ST and AER groups were removed from Fig. 2a and b, as there were no significant differences between these groups at 9 months.

The legend to Fig. 2 was updated to read: ‘The table within each plot shows the results of the pairwise comparisons between groups of the mean HbA1c during follow-up using all data from 3, 6 and 9 months.’

Results:

The following statement was deleted: ‘At baseline, there were no significant differences across the three groups in weight, lean mass, fat mass, or muscle strength (Table 1).’

Table 2:

The samples sizes for the ST, AER and COMB groups for the ITT analysis were changed from 63, 58 and 65 to 46, 42 and 43, respectively.

Discussion:

The discussion was amended to note that the amount of missing data means that the conclusions need further verification, with the phrase ‘and further studies are required to verify our results’ being added to the end of the sentence ‘The follow-up rate was about 45%; therefore, the study was underpowered to obtain conclusive findings.’

The original article has been corrected.