Artículo

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Acta Diabetologica

https://doi.org/10.1007/s00592-020-01559-9

ORIGINAL ARTICLE

Metformin use in prediabetes: is earlier intervention better?


Andrew Warrilow1 · Shawn Somerset1 · Kate Pumpa1 · Robert Fleet2

Received: 24 April 2020 / Accepted: 12 June 2020


© Springer-Verlag Italia S.r.l., part of Springer Nature 2020

Abstract
Aim The aim of this study was to investigate the effectiveness of metformin in diabetes prevention in a prediabetic popula-
tion across a range of fasting plasma glucose (FPG) levels at baseline. A secondary aim was to assess the effectiveness of
metformin in preventing diabetes in those participants where impaired fasting glucose (IFG) was relatively more pronounced
as opposed to impaired glucose tolerance (IGT).
Methods Participants randomised to metformin and placebo arms in the Diabetes Prevention Program study were stratified
into cohorts according to level of FPG at baseline. Cumulative incidence of diabetes for the different cohorts was assessed.
Change in FPG, insulin sensitivity, and levels of fasting insulin and proinsulin for the different cohorts were also calculated.
Results The largest reductions in incidence of diabetes and FPG occurred within prediabetic persons with a higher level of
FPG at baseline. Metformin was able to stabilise insulin sensitivity in every stratified sub-cohort except one. Sub-cohorts
which had higher levels of insulin sensitivity at baseline experienced the largest increases in insulin sensitivity. Metformin
reduced the incidence of diabetes by 43% (RR 0.57, CI 0.4–0.9) in those prediabetic persons whose IFG was more pronounced
compared to a 26% (RR 0.74 CI 0.7–0.8) when all participants in the study were included.
Conclusion The largest reductions in both incidence of diabetes and FPG occurred in prediabetic persons with a higher
level of FPG at baseline. Metformin was able to stabilise insulin sensitivity and was more effective in persons with more
pronounced IFG.

Keywords Metformin · Prediabetes · Diabetes prevention · Insulin sensitivity · Insulin resistance

Introduction In that year, 5 million deaths worldwide were attributable


to diabetes and the global healthcare expenditure on peo-
The International Diabetes Federation estimated that glob- ple with diabetes was estimated to be US$850 billion [1].
ally in 2017 there were 451 million people with diabetes A more effective pharmacologically based strategy for the
worldwide, approximately half of whom are undiagnosed. prevention of type 2 diabetes would have significant benefits
globally both for society and the economy.
The Diabetes Prevention Program (DPP) randomly
Managed by Massimo Porta. assigned 3234 nondiabetic persons with elevated fasting and
post-load plasma glucose concentrations to placebo, the drug
* Andrew Warrilow metformin (850 mg twice daily), or a lifestyle-modification
andrew.warrilow@canberra.edu.au
programme. The lifestyle intervention reduced the incidence
Shawn Somerset by 58% (95% CI 48–66%) and metformin by 31% (95% CI
shawn.somerset@canberra.edu.au
17–43%), as compared with placebo; the lifestyle interven-
Kate Pumpa tion was significantly more effective than metformin [2].
kate.pumpa@canberra.edu.au
Prescribed as a first line treatment of the condition after
Robert Fleet diagnosis and after diet and exercise has failed to re-estab-
robert.fleet@anu.edu.au
lish control of blood sugar, there is now growing support
1
University of Canberra, Locked Bag 1, Canberra, ACT​2601, for using metformin in the prediabetic stage [3]. A recent
Australia review discussing next steps for translating diabetes preven-
2
The Australian National University, Canberra, ACT​2600, tion research into real-world interventions concluded that
Australia

13
Vol.:(0123456789)
Acta Diabetologica

sufficient evidence exists for a change in the indication of clinics) and 140–199 mg per decilitre (7.8–11.0 mmol per
metformin for preventative use [4]. Based on such conclu- litre) two hours after a 75-g oral glucose load. Some clinics
sions, the Medicines and Healthcare Products Regulatory and individual participants in the original DPP study did not
Agency (MHRA) in the UK has approved the use of met- have Institutional Review Board approval (n = 154) to allow
formin for prevention of diabetes [5]. their results to be added to the NIDDK repository, and these
A 2012 economic analysis found that over 10 years, the results have been excluded from the data (Fig. 1).
cumulative, undiscounted per capita direct medical costs
of the interventions, as implemented during the DPP, were Study design
almost twice as great for lifestyle as metformin [6] and since
publication the cost difference would reasonably be expected For the present study, the effectiveness of metformin com-
to have widened further. While not a substitute for lifestyle- pared to placebo was assessed in five separate analyses:
related interventions such as physical activity and improved
nutrition, a strengthening case also exists for considering 1. Incidence of diabetes in sub-cohorts with varying levels
the use of metformin for the prevention of diabetes on a of FPG at baseline
cost–benefit basis.   Data from the DPP dataset were received from the
Regarding the pathology of diabetes, IFG and IGT rep- NIDDK. A Cox proportional hazards regression analysis
resent intermediate states of abnormal glucose regulation was run using SPSS. Participants were then stratified
that exist between normal glucose homeostasis and diabetes. in order to generate sub-cohorts of as similar a size as
IFG is defined by the American Diabetes Association as an possible: low sub-cohort (FPG of 99–100 mg/dl at base-
elevated fasting plasma glucose (FPG) concentration (≥ 100 line); low–mid sub-cohort (101–105 mg/dl at baseline);
and < 126 mg/dl). IGT is defined by an elevated 2-h plasma mid–high sub-cohort (106–115 mg/dl at baseline); and
glucose concentration (≥ 140 and < 200 mg/dl) after a 75-g high sub-cohort (≥ 116 mg/dl at baseline). Reduction in
glucose load on the oral glucose tolerance test (OGTT) in the the incidence of diabetes between metformin and pla-
presence of an FPG concentration < 126 mg/dl. The various cebo was calculated overall and between sub-cohorts.
pathologies of diabetes and the resultant individualisation 2. Mean difference in FPG in sub-cohorts with varying
of therapy are being increasingly recognised [7]. Those with levels of FPG at baseline
IFG are known to suffer from hepatic insulin resistance [8].   Participants in the DPP study were stratified into 4
It could be hypothesised, therefore, that use of metformin cohorts as described above. Mean difference in FPG
may be more effective in prediabetic persons where IFG is levels between metformin and placebo at 4 years within
more pronounced. the various sub-cohorts was calculated. A Levene test
The optimal timing for preventative treatment with met- was used to determine homogeneity of the differences
formin remains an open question. For those with prediabe- between sub-cohorts. In the event that these were not
tes, fasting blood glucose levels may become elevated with sufficiently homogenous, significance in the difference
increasing rapidity. Therefore, it logically follows that by between sub-cohort results was determined by means
intervening before this unstable period, it is possible that the of a nonparametric Kruskal–Wallis H test. Tukey’s post
onset of frank hyperglycaemia may be even further delayed hoc analysis was used to uncover where differences
[9]. The aims of the present study are to investigate, using between the various sub-cohorts occurred.
the dataset from the DPP study: 1. the effectiveness of met- 3. Mean difference in insulin sensitivity in sub-cohorts with
formin to prevent type 2 diabetes in prediabetic persons with varying levels of FPG at baseline
a lower FPG at baseline; and 2. the effectiveness of met-   Participants in the DPP study were stratified into 4
formin in preventing type 2 diabetes in prediabetic persons cohorts as described above. The degree of insulin sen-
where IFG is more pronounced relative to IGT. sitivity was assessed using the Quantitative Insulin Sen-
sitivity Check Index (QUICKI) [10] as a measure. Dif-
ference in mean insulin sensitivity between metformin
Subjects, materials and methods and placebo at year 4 was assessed. Mean differences in
QUICKI scores between metformin and placebo within
Participants the various sub-cohorts at 4 years were calculated. Sig-
nificance in differences between the sub-cohorts was
Eligibility criteria for the original DPP study included an tested as described above.
age of at least 25 years, a body mass index of 24 or higher 4. Mean difference in fasting proinsulin in sub-cohorts with
(22 or higher in Asians), and a plasma glucose concentration varying levels of FPG at baseline
of 95–125 mg per decilitre (5.3–6.9 mmol per litre) in the   Participants in the DPP study were stratified into 4
fasting state (≤ 125 mg per decilitre in the American Indian cohorts as above. Mean differences in proinsulin levels

13
Acta Diabetologica

Fig. 1  Diagram of numerical


breakdown of participants DPP Study Lack of IRB approval
(n=3819) (n=154)

DPP Repository Troglitazone arm


(n=3665) (n=584)

DPP Met/Plcbo/Lfstyl arms Lifestyle arm


(n=3081) (n=1024)

DPP Meormin arm DPP Placebo arm


(n=1027) (n=1030)

between metformin and placebo within the various sub- Results


cohorts at baseline and at 4 years were calculated. Sig-
nificance in differences between sub-cohort was tested Reduction in the incidence of diabetes
as mentioned above. in sub‑cohorts with varying levels of FPG at baseline
5. Prevention of type 2 diabetes in prediabetic persons
where IFG is more pronounced, in sub-cohorts with When all DPP participants were included, there was a 26%
varying levels of FPG at baseline reduction in incidence (RR: 74%, CI 0.65–0.83) over the
life of the DPP study. The cumulative hazard curves for
DPP participants had both IFG and IGT. A cohort of pre- metformin were consistently lower than that for placebo
diabetic persons was created whose IFG was more pronounced (Fig. 2). Reductions in incidence of diabetes for each
by excluding participants with a blood sugar level (BSL) sub-cohort compared to placebo were: low sub-cohort,
150–199 mg/dl 2 h after an oral glucose load at baseline. A 15% (RR 85%, CI 0.6–1.2); low-mid sub-cohort 4% (RR
Cox proportional hazards regression analysis was run using 96%, CI 0.7–1.3); mid-high sub-cohort, 26% (RR 74%, CI
SPSS. Reduction in incidence of diabetes overall and within 0.6–0.9); and high sub-cohort, 34% (RR 66%, CI 0.6–0.8)
each sub-cohort was also calculated.

13
Acta Diabetologica

Fig. 2  Cumulative hazard, inci-


dence of diabetes, metformin
compared to placebo, baseline
to year 4, all participants

Table 1  Change in mean FPG, Mean FPG at baseline (mg/dl) Mean FPG at year 4 (mg/dl) Mean difference at year 4
metformin compared to placebo,
baseline to year 4 Metformin Placebo Metformin Placebo

Low 99.3 99.2 101.2 104.9 3.7 (CI 95% 1.6–5.9)


Low–mid 102.9 103.0 104.9 110.0 6.0 (CI 95% 0.7–11.4)
Mid–high 110.1 109.8 111.6 117.4 5.8 (CI 95% 3.2–8.5)
High 121.1 120.9 121.6 136.1 14.5 (CI 95 − 10.7 to 39.7)

Table 2  Change in mean Mean QUICKI score at Mean QUICKI score at Mean difference in QIUCKI score
QUICKI score, metformin baseline year 4 metformin and placebo at year 4
compared to placebo, baseline
to year 4 Metformin Placebo Metformin Placebo

Low 0.303 0.304 0.302 0.293 0.00819 (CI 0.00817–0.00821)


Low–mid 0.299 0.298 0.304 0.296 0.00792 (CI 0.00791–0.00793)
Mid–high 0.293 0.294 0.297 0.287 0.00973 (CI 0.00973–0.00974)
High 0.285 0.285 0.287 0.283 0.00417 (CI 0.00416–0.00418)

with only mid-high and high sub-cohorts showing signifi- FPG at baseline finished the study with a higher level of FPG.
cance. Sub-cohorts having a higher FPG at baseline had a Although the difference between sub-cohorts was found to be
greater reduction in diabetes incidence compared to sub- significant (mean FPG χ2(7) = 112.498, p < 0.001), post hoc
cohorts having a lower FPG at baseline with the exception analysis revealed there was a significant difference between
of the low sub-cohort. metformin and placebo in the high sub-cohort (baseline
FPG ≥ 116 mg/dl) only, p = 0.009.
Mean difference in FPG in sub‑cohorts with varying
levels of FPG at baseline Mean difference in insulin sensitivity in sub‑cohorts
with varying levels of FPG at baseline
At four years, FPG levels were lower in every metformin sub-
cohort compared to its placebo counterpart (Table 1). Sub- Mean QUICKI scores for the metformin sub-cohorts
cohorts which commenced the study with a higher level of were mostly stable over the 4 years of the study (Table 2).

13
Acta Diabetologica

Table 3  Change in mean fasting Mean fasting proinsulin at Mean fasting proinsulin at Mean difference at year 4
proinsulin, metformin compared baseline (pM) year 4 (pM)
to placebo, baseline to year 4
Metformin Placebo Metformin Placebo

Low 13.856 13.342 18.857 20.136 1.27922 (CI 1.27920–1.27924)


Low–mid 15.681 15.854 15.732 18.556 2.82341 (CI 2.82340–2.82342)
Mid–high 20.206 19.140 18.938 25.200 6.26250 (CI 6.26249–6.26251)
High 25.889 26.083 23.292 29.604 6.31250 (CI 6.31249–6.31251)

Fig. 3  Cumulative hazard, inci-


dence of diabetes, metformin
compared to placebo, baseline
to year 4, participants with more
pronounced IFG

Sub-cohorts with a lower FPG/higher QUICKI score at showed no significant difference between any metformin
baseline tended to have a larger mean difference in QUICKI sub-cohort and its corresponding placebo equivalent.
score over the 4 years of the study, with the exception of the
mid-high sub-cohort. All metformin cohorts experienced an Reduction in the incidence of diabetes where IFG
increase in QUICKI score/improvement in insulin sensitivity was more pronounced, in sub‑cohorts with varying
with the exception of the low sub-cohort which had a minor levels of FPG at baseline
decrease (a reduction of 0.002 points on the QUICKI score).
Placebo sub-cohorts showed a decrease in QUICKI score When only prediabetic persons with more pronounced
in every sub-cohort. The difference between sub-cohorts IFG were included, the reduction in incidence of diabetes
overall was significant as determined by one-way ANOVA increased to 43% (RR 57%, CI 0.4–0.9) over the period of
(F(7,412) = 4.856, p < 0.001). the DPP study. The cumulative hazard curves showed a
consistently beneficial effect for metformin but with a less
Mean difference in fasting proinsulin in sub‑cohorts steeply sloping curve compared to when all participants were
with varying levels of FPG at baseline included (Fig. 3). Reductions in incidence of diabetes for
each sub-cohort compared to placebo were: low sub-cohort,
Fasting proinsulin was lower in metformin users in the mid- 12% (RR 88%, CI 0.2–3.7); low–mid sub-cohort, 55% (RR
high sub-cohort and the high sub-cohort (Table 3). Fasting 45%, CI 0.1–2.2); mid–high sub-cohort, 44% (RR 56%, CI
proinsulin levels were also lower in the low and low-mid 0.3–1.1); high sub-cohort, 28% (RR 72%, CI 0.4–1.2) with
sub-cohorts, but the differences were not as great. The dif- no sub-cohorts having statistical significance. In contrast to
ference between the sub-cohorts was significant (mean fast- the analysis including all participants, sub-cohorts having
ing proinsulin, χ2(7) = 25.965, p = 0.001). Post hoc analysis a lower FPG at baseline had a greater reduction in diabetes

13
Acta Diabetologica

incidence, with the exception of the low sub-cohort. Dif- progression to diabetes. The magnitude of the mean differ-
ferences between sub-cohorts were found to be significant: ence was greatest in the low sub-cohort and declined through
mean incidence of diabetes, χ2(7) = 92.736, p < 0.001. Post to the high sub-cohort, with the exception of the mid-high
hoc analysis showed no significant difference between sub-cohort. This could indicate a pattern of improved abil-
any metformin sub-cohort and its corresponding placebo ity to maintain insulin sensitivity in those sub-cohorts in
equivalent. which intervention with metformin occurs earlier. These
results agree with other research using DPP data finding an
association between a higher level of insulin sensitivity and
Discussion lower diabetes risk [16].
Reductions in fasting proinsulin levels were lower in met-
This analysis of the DPP study found that metformin reduced formin users in sub-cohorts with a higher FPG at baseline.
the incidence of diabetes by 26% (RR 74%, CI 0.65–0.83). Fasting proinsulin levels are an accurate indicator of β-cell
As expected, the graph for cumulative hazard shows a diver- exhaustion [17], and these results would suggest this is what
gence between metformin and placebo curves (see Fig. 2). is occurring in these sub-cohorts. 𝛽 -cell mass is normally
The reduction in FPG brought about by metformin was tightly maintained through a balance of β-cell creation and
greater in cohorts which had a higher FPG at baseline. Over- β-cell death through apoptosis [5]. β-cell mass is decreased
all, results on changes in blood glucose in the present study in both obese and lean humans with type 2 diabetes com-
agree with those in the original DPP study. pared with their nondiabetic age- and weight-matched coun-
Levels of blood glucose are directly relevant to progres- terparts [18]. The development and continued progression
sion of patients to hyperglycaemia. Research has suggested of diabetes is a consequence of the failure of β-cells to over-
that chronic hyperglycaemia, independent of hyperlipidae- come insulin resistance, increasing loss of β-cell function.
mia, is toxic for β-cell function, whereas chronic hyperlipi- Under the conditions of glucotoxicity produced as a result of
daemia is deleterious only in the presence of hyperglycaemia this process, β-cell apoptosis is accelerated [19], while the
[11]. Even before the development of hyperglycaemia, the rate of new β-cell formation remains unchanged. Eventually,
aetiology of type 2 diabetes in the large majority of patients the rate of β-cell death overtakes that of creation and β-cell
begins with insulin resistance [12, 13]. A feedback loop mass starts to decline. It is plausible that avoidance of this 3-
controls the interaction between insulin-sensitive cells and to 10-fold increase in the rate of apoptosis could potentially
β-cells. This relationship is hyperbolic, with insulin resist- lead to a restoration of β cell mass [20]. Both higher insu-
ant individuals displaying proportionally greater changes in lin secretion and sensitivity at baseline and improvements
levels of insulin secretion [14]. The progression to type 2 in response to treatment have been associated with lower
diabetes is associated with declining insulin sensitivity and diabetes incidence in other studies conducted on DPP data
β cell function, and these changes occur early in the patho- [16]. Interventions to prevent diabetes should tackle both
genesis of diabetes [15]. pathologies, therefore, and at the earliest stages.
In the present analysis, we assessed changes to insulin Metformin plays a regulatory role in multiple systems
sensitivity by means of the QUICKI score. Mean QUICKI including: regulation of hepatic gluconeogenesis through a
scores for the metformin sub-cohorts were mostly stable variety of mechanisms; reduction of obesity related inflam-
with all sub-cohorts (except the low sub-cohort) showing mation; and various influences on the gastrointestinal tract
an increase over the four years of the study. Placebo sub- and gut microbiota [21]. Those with IFG can suffer from
cohorts, on the other hand, showed a decrease in QUICKI hepatic insulin resistance [8]. The ability of metformin to
score in each sub-cohort indicating a continuing loss of insu- improve insulin sensitivity warranted a separate analysis on
lin sensitivity for those taking placebo over the four years the incidence of diabetes for a cohort with more pronounced
of the study.1 The fact that sub-cohorts with greater insulin IFG. Participants in the DPP study had both diagnosed IFG
sensitivity/higher QUICKI score at baseline generally expe- and IGT, making investigation of these different types of
rienced the largest increases in insulin sensitivity indicates prediabetes in isolation difficult. When a cohort with more
the possibility of evidence in support of earlier intervention pronounced IFG was created by excluding participants
with metformin. having a BSL of > 149 mg/dl 2 h post-OGTT, there was a
Increases in blood sugar levels may be modulated by the 43% reduction in the incidence of diabetes (RR 56%, CI
actions of insulin, and thus, changes in both FPG and insulin 0.38–0.86). These results would suggest that metformin is
levels may create a more accurate picture of the degree of substantially more effective in preventing type 2 diabetes in
prediabetic persons with more pronounced IFG compared to
participants generally.
1
N.B. an increase in QUICKI score denotes an increase in insulin This analysis has demonstrated that metformin is effec-
sensitivity. tive at both lowering FPG, especially in prediabetic persons

13
Acta Diabetologica

with more pronounced IFG, and has also shown that met- Intervention or Metformin study, the NIDDK Central Repositories, or
formin was capable of maintaining the existing level of insu- the NIDDK.
lin resistance over the four years of the DPP study. Insulin
Funding The authors received no funding from any source.
resistance is a key factor leading to the development of type
2 diabetes. The other key factor implicated in the develop- Availability of data and material The data from the Reduction in the
ment of type 2 diabetes is impairment of β-cell function Incidence of Type 2 Diabetes with Lifestyle Intervention or Metformin
and insulin resistance plays a part here also. As metformin study reported here were supplied by the NIDDK Central Repositories.
addresses insulin resistance, intervention is beneficial at any Due to privacy considerations, the authors do not have consent to pub-
lish the data used for this study.
stage of the progression to type 2 diabetes but may be more
effective when made earlier before impairment of β-cell
Compliance with ethical standards
function.
Additional research is warranted to determine whether Conflict of interest The authors declare no conflict of interest.
the biochemical processes which result concurrently with
initial loss of insulin sensitivity and increased β-cell apop- Consent for publication This manuscript was not prepared in collabo-
tosis could be halted by stabilising insulin sensitivity at the ration with Investigators of the Reduction in the Incidence of Type
2 Diabetes with Lifestyle Intervention or Metformin study and does
early stages. not necessarily reflect the opinions or views of the Reduction in the
Incidence of Type 2 Diabetes with Lifestyle Intervention or Metformin
study, the NIDDK Central Repositories, or the NIDDK. Consent to
publish has been received from the University of Canberra.
Conclusion Ethics approval This study was evaluated as exempt from ethics
approval by the University of Canberra Human Research Ethics Com-
In all but one sub-cohort, metformin was able to stabilise mittee.
insulin sensitivity, which may be a more accurate indica-
tor of progression to diabetes. This compares to a reduction
in insulin sensitivity in all sub-cohorts taking placebo. The
fact that sub-cohorts with greater FPG/insulin sensitivity at References
baseline generally experienced the largest increases in insu-
1. Cho NH et al (2018) IDF Diabetes Atlas: global estimates of dia-
lin sensitivity supports the utility of metformin earlier in the betes prevalence for 2017 and projections for 2045. Diabetes Res
prediabetic phase and perhaps even earlier. Metformin was Clin Pract 138:271–281
more effective in prediabetic persons with more pronounced 2. Group DPPR (2002) Reduction in the incidence of type 2 dia-
betes with lifestyle intervention or metformin. N Engl J Med
IFG, which is consistent with the established metformin
2002(346):393–403
mode of action. 3. Hostalek U, Gwilt M, Hildemann S (2015) Therapeutic use
Both higher insulin secretion and sensitivity at baseline of metformin in prediabetes and diabetes prevention. Drugs
and improvements in response to treatment have been asso- 75(10):1071–1094
4. Cefalu WT et al (2016) Update and next steps for real-world trans-
ciated with lower diabetes incidence in other studies con-
lation of interventions for type 2 diabetes prevention: reflections
ducted on DPP data. In addition, any benefits from taking from a diabetes care editors’ expert forum. Am Diabetes Assoc
metformin are likely to be cumulative. Earlier intervention 5. Merck’s Glucophage® SR Receives Label Extension for Patients
with metformin in prediabetes may be advantageous where at High Risk of Type 2 Diabetes in the UK. [Web Page] 2017
[cited 2019 3rd of May, 2019]; Available from: https:​ //www.prnew​
stabilisation of insulin sensitivity is considered a desirable
swire​.co.uk/news-relea​ses/merck​s-gluco​phage​-sr-recei​ves-label​
therapeutic goal. Further research is warranted to determine -exten​sion-for-patie​nts-at-high-risk-of-type-2-diabe​tes-in-the-
whether the biochemical processes which result concurrently uk-62279​9343.html
with initial loss of insulin sensitivity and increase in the 6. The 10-Year Cost-Effectiveness of Lifestyle Intervention or Met-
formin for Diabetes Prevention (2012) An intent-to-treat analysis
rate of β-cell apoptosis could be halted by stabilising insulin
of the DPP/DPPOS 35(4):723–730
sensitivity at the early stages. 7. Smith RJ et al (2010) Individualizing therapies in Type 2 diabetes
mellitus based on patient characteristics: what we know and what
Acknowledgements Metformin Use in Prediabetes: Is Earlier Inter- we need to know. J Clin Endocrinol Metab 95(4):1566–1574
vention Better? received nonfinancial support from the United States 8. Basu R et al (2013) Pathogenesis of prediabetes: role of the liver in
National Institute of Diabetes and Digestive and Kidney Diseases isolated fasting hyperglycemia and combined fasting and postpran-
(NIDDK). The data from the Reduction in the Incidence of Type 2 dial hyperglycemia. J Clin Endocrinol Metab 98(3):E409–E417
Diabetes with Lifestyle Intervention or Metformin study reported here 9. Tabák AG et al (2009) Trajectories of glycaemia, insulin sensitiv-
were supplied by the NIDDK Central Repositories. This manuscript ity, and insulin secretion before diagnosis of type 2 diabetes: an
was not prepared in collaboration with Investigators of the Reduction analysis from the Whitehall II study. Lancet 373(9682):2215–2221
in the Incidence of Type 2 Diabetes with Lifestyle Intervention or Met- 10. Katz A et al (2000) Quantitative insulin sensitivity check index:
formin study and does not necessarily reflect the opinions or views a simple, accurate method for assessing insulin sensitivity in
of the Reduction in the Incidence of Type 2 Diabetes with Lifestyle humans. J Clin Endocrinol Metab 85(7):2402–2410

13
Acta Diabetologica

11. Poitout V, Robertson RP (2002) Minireview: secondary β-cell 18. Weir GC, Bonner-Weir S (2004) Five stages of evolving beta-cell
failure in type 2 diabetes—a convergence of glucotoxicity and dysfunction during progression to diabetes. Diabetes 53(suppl
lipotoxicity. Endocrinology 143(2):339–342 3):S16–S21
12. Reaven GM (1988) Role of insulin resistance in human disease. 19. Muoio DM, Newgard CB (2008) Mechanisms of disease: molecu-
Diabetes 37(12):1595–1607 lar and metabolic mechanisms of insulin resistance and [beta]-cell
13. Goldstein BJ (2002) Insulin resistance as the core defect in type failure in type 2 diabetes. Nat Rev Mol Cell Biol 9(3):193
2 diabetes mellitus. Am J Cardiol 90(5):3–10 20. Butler AE et al (2003) β-Cell deficit and increased β-cell apoptosis
14. Kahn S (2003) The relative contributions of insulin resistance and in humans with type 2 diabetes. Diabetes 52(1):102–110
beta-cell dysfunction to the pathophysiology of type 2 diabetes. 21. Foretz M, Guigas B, Viollet B (2019) Understanding the glucoreg-
Diabetologia 46(1):3–19 ulatory mechanisms of metformin in type 2 diabetes mellitus. Nat
15. Weyer C et al (1999) The natural history of insulin secretory Rev Endocrinol 15(10):569–589
dysfunction and insulin resistance in the pathogenesis of type 2
diabetes mellitus. J Clin Investig 104(6):787–794 Publisher’s Note Springer Nature remains neutral with regard to
16. Group, D.P.P.R. (2005) Role of insulin secretion and sensitiv- jurisdictional claims in published maps and institutional affiliations.
ity in the evolution of type 2 diabetes in the diabetes prevention
program: effects of lifestyle intervention and metformin. Diabetes
54(8):2404
17. Pfützner A et al (2004) Fasting intact proinsulin is a highly spe-
cific predictor of insulin resistance in type 2 diabetes. Diabetes
Care 27(3):682–687

13

You might also like