Metformin Effects in Gestational Diabetes
In this study, the patient reported QoL, either assessed as a global score or as specific items across important life domains, was best with metformin as sole treatment and least good with a combination of metformin and insulin. Measured items included key aspects of social functioning such as working life, family life, relationships, and self-confidence. Total treatment satisfaction was also highest in the metformin group. Patients treated with metformin reported that diabetes had the least negative impact on their QoL compared with those taking metformin in combination with insulin. With regard to hypoglycaemia, patients taking metformin had the lowest score and unsurprisingly, those on insulin alone had the highest score for perceived hypoglycaemia. Although not specifically measuring QoL or TTS, our results are comparable to those obtained by the MiG trialists who showed superior treatment acceptability of metformin compared with insulin in GDM patients.
There are several ways in which insulin could have a negative impact on QoL; patients will be worried that their diabetes may be harming their baby and have to quickly master the technique of insulin self-administration and learn to adjust the dose based on their home glucose readings. They may be fearful of hypoglycaemia and concerned about weight gain. None of these considerations apply to metformin.
Nevertheless the most negative impact on QoL in this study was the combination of metformin and insulin rather than insulin alone. Similar results have been found in non-pregnant patients with diabetes where a larger number of treatments resulted in a more negative impact on the QoL. There are several possible reasons that could account for this: firstly QoL is lower when diabetes is less well controlled and this will be the case if insulin needs to be added to metformin. Secondly, combination therapy of insulin and metformin may result in a higher risk of side effects compared with insulin alone. Finally, patient perceptions are important; as one patient reported, 'knowing I needed two medications to treat my diabetes instantly worried me, and even though everything was explained to me, I just always had that constant worry in the back of my mind.'
Metformin was also reported to be a more convenient option for women with GDM in comparison to insulin, or metformin + insulin reflecting its oral administration.
DTSQs is a measure of treatment satisfaction that was designed explicitly to measure issues of importance to patients and therefore any significant differences should be taken seriously. Treatment satisfaction score correlates with the duration of diabetes and perceived blood glucose control. The DTSQs is highly sensitive to major changes in treatment, for example, from tablets to injections or from conventional (more rigid insulin regimen with fixed meal times) to a more flexible insulin dosing (allowing for dietary freedom).
The strengths of this study are that it is multicentre (and hence results more generalisable) with data collected at routine postnatal check-ups. The patient reported outcome measures are well established validated tools to assess QoL and treatment satisfaction in diabetes.
The limitations are that the study was not blinded and treatment was not randomised to metformin or insulin. It is possible therefore that patients treated with insulin had lower QoL at baseline. Although we didn't have the opportunity to assess QoL or TTS before commencing treatment, none of the patients had significant renal or hepatic dysfunction in the post partum period which might have been associated with reduced QoL. Perinatal outcomes were similar in metformin and insulin treated patients.
In summary, this study adds to previous evidence that when tolerated metformin is associated with improved treatment satisfaction and a favourable impact on QoL compared with insulin alone or in combination. Further studies are needed to confirm this finding using a randomized controlled trial design.
Discussion
In this study, the patient reported QoL, either assessed as a global score or as specific items across important life domains, was best with metformin as sole treatment and least good with a combination of metformin and insulin. Measured items included key aspects of social functioning such as working life, family life, relationships, and self-confidence. Total treatment satisfaction was also highest in the metformin group. Patients treated with metformin reported that diabetes had the least negative impact on their QoL compared with those taking metformin in combination with insulin. With regard to hypoglycaemia, patients taking metformin had the lowest score and unsurprisingly, those on insulin alone had the highest score for perceived hypoglycaemia. Although not specifically measuring QoL or TTS, our results are comparable to those obtained by the MiG trialists who showed superior treatment acceptability of metformin compared with insulin in GDM patients.
There are several ways in which insulin could have a negative impact on QoL; patients will be worried that their diabetes may be harming their baby and have to quickly master the technique of insulin self-administration and learn to adjust the dose based on their home glucose readings. They may be fearful of hypoglycaemia and concerned about weight gain. None of these considerations apply to metformin.
Nevertheless the most negative impact on QoL in this study was the combination of metformin and insulin rather than insulin alone. Similar results have been found in non-pregnant patients with diabetes where a larger number of treatments resulted in a more negative impact on the QoL. There are several possible reasons that could account for this: firstly QoL is lower when diabetes is less well controlled and this will be the case if insulin needs to be added to metformin. Secondly, combination therapy of insulin and metformin may result in a higher risk of side effects compared with insulin alone. Finally, patient perceptions are important; as one patient reported, 'knowing I needed two medications to treat my diabetes instantly worried me, and even though everything was explained to me, I just always had that constant worry in the back of my mind.'
Metformin was also reported to be a more convenient option for women with GDM in comparison to insulin, or metformin + insulin reflecting its oral administration.
DTSQs is a measure of treatment satisfaction that was designed explicitly to measure issues of importance to patients and therefore any significant differences should be taken seriously. Treatment satisfaction score correlates with the duration of diabetes and perceived blood glucose control. The DTSQs is highly sensitive to major changes in treatment, for example, from tablets to injections or from conventional (more rigid insulin regimen with fixed meal times) to a more flexible insulin dosing (allowing for dietary freedom).
The strengths of this study are that it is multicentre (and hence results more generalisable) with data collected at routine postnatal check-ups. The patient reported outcome measures are well established validated tools to assess QoL and treatment satisfaction in diabetes.
The limitations are that the study was not blinded and treatment was not randomised to metformin or insulin. It is possible therefore that patients treated with insulin had lower QoL at baseline. Although we didn't have the opportunity to assess QoL or TTS before commencing treatment, none of the patients had significant renal or hepatic dysfunction in the post partum period which might have been associated with reduced QoL. Perinatal outcomes were similar in metformin and insulin treated patients.
In summary, this study adds to previous evidence that when tolerated metformin is associated with improved treatment satisfaction and a favourable impact on QoL compared with insulin alone or in combination. Further studies are needed to confirm this finding using a randomized controlled trial design.