Hormonal versus non‐hormonal contraceptives in women with diabetes mellitus
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
Primary objective:
To investigate whether progestogen‐only contraceptive methods and combined estrogen/progestogen contraceptives, respectively, differ from non‐hormonal contraceptives in terms of efficacy in preventing pregnancy and in their side effects on carbohydrate metabolism, lipid metabolism, and haemostatic function when used by women with diabetes mellitus.
Secondary objectives:
-
To investigate whether there are any differences in terms of efficacy in preventing pregnancy and/or side effects on carbohydrate metabolism, lipid metabolism, and haemostatic function of combined oral contraceptive pills containing <50 µg estrogen as compared to combined oral contraceptive pills containing >= 50 µg estrogen when used by women with diabetes mellitus.
-
To investigate whether there are any differences in terms of efficacy in preventing pregnancy and/or side effects on carbohydrate metabolism, lipid metabolism, and haemostatic function between first‐, second‐ and third‐generation progestogen components in oral contraceptives used by women with diabetes mellitus.
-
To assess serious adverse events (thrombosis, myocard infarct, etc.) associated with the use of hormonal contraceptives by women with diabetes mellitus.
Background
HORMONAL CONTRACEPTIVES AND CARDIOVASCULAR DISEASE
Combined oral contraceptives are one of the most effective types of contraception available. However, hormones in contraceptives could alter plasma lipoprotein metabolism, blood coagulation and endothelial function (Godsland 2000). In addition, plasma insulin levels and tissue insulin resistance might be raised in women using combined oral contraceptives (Godsland 1996). All these changes might possibly be related to the development of cardiovascular disease (Fontbonne 1991, Bass 1993). The association between combined oral contraceptives and cardiovascular disease was found to be dependent on both estrogen and progestogen doses, as reduction in these doses was reported to decrease the incidence of cardiovascular disease (Kay 1982, Lidegaard 1993, Lewis 1999).
DIABETES MELLITUS AND HORMONAL CONTRACEPTIVES
Women with diabetes mellitus already suffer from disturbances in endothelial function, blood coagulation, glucose and lipid metabolism, and are therefore at increased risk of cardiovascular disease. Particularly coronary heart disease and stroke are important causes of mortality in these women (Wong 1991). For women with diabetes mellitus, the risk of cardiovascular disease might be further increased by the use of hormonal contraceptives. Alternatively, a major advantage of hormonal contraceptives is the high efficacy in preventing pregnancy. For women with diabetes mellitus, pregnancy is associated with an increased risk of maternal and infant morbidity and mortality (Pickup 1997). To reduce such risks, strict diabetic control should be achieved before conception, and adequate contraceptive advice is therefore particularly important to these women (Pickup 1997). According to the medical guidelines of the World Health Organisation (WHO 2001), for non‐smoking women with diabetes mellitus, aged younger than 35 years and without further comorbidity, the advantages of using low‐dose oral contraceptives outweigh the disadvantages. In the presence of both microvascular (retino‐, nephro‐, and neuropathies) and macrovascular (coronary artery, cerebrovascular, and peripheral vascular diseases) complications of diabetes mellitus, most progestogen‐only contraceptives are preferred to low‐dose combined oral contraceptives according to WHO guidelines (WHO 2001).
COMBINED ORAL CONTRACEPTIVES
This review considers the two types of hormonal contraceptives that are available: combined oral contraceptives and progestogen‐only methods. Although low‐dose combined pills, containing estrogens (< 50 µg) and progestogens, are considered to be a good choice for women with diabetes mellitus without micro‐ and macrovascular illness‐related complications, or other risk factors for vascular disease (Speroff 2000; WHO 2001), few randomised controlled trials have been recently conducted in this area. One randomised controlled crossover trial compared a pill containing 50 µg estrogen combined with a first‐generation progestogen (lynestrenol, 2.5 mg) with a preparation containing only 0.5 mg lynestrenol (Radberg 1981, Radberg 1982) in women with diabetes mellitus. This study found that the combined contraceptive pill increased the insulin requirement and plasma triglycerides, without affecting the high‐density lipoprotein lipids. One randomised (Skouby 1986) and one non‐randomised controlled trial (Diab 2000) investigated third‐generation combined oral contraceptives containing gestodene. Both studies found that low‐dose combined oral contraceptives increased fasting blood sugars, plasma triglycerides and very low‐density lipoprotein cholesterol levels in women with diabetes mellitus (Skouby 1986; Diab 2000). In contrast, another non‐randomised controlled clinical trial found no evidence of impaired glycometabolic control or adverse changes in serum levels of lipoproteins (i.e., reduced low‐density lipoprotein lipids; increased levels of triglycerides and lipoprotein A; unchanged levels of total and high‐density lipoprotein lipids) in women with well‐controlled diabetes mellitus using ethinyl estradiol and gestodene (Petersen 1995). However, this study suggested that endothelial function may be disturbed in these women (Petersen 1994). Typically, none of these trials allocated more than 25 women to each condition, and thus may fall short of achieving the statistical power necessary to find a true treatment effect.
PROGESTOGEN‐ONLY METHODS
Progestogen‐only contraceptives have a slightly higher failure rate and give more menstrual‐cycle disturbances than combined oral contraceptives (Hatcher 1998). Progestogen‐only contraceptives can be administered by injection, pill, implant, intrauterine device (IUD) or vaginal ring. In their randomised controlled trial, Radberg et al. (Radberg 1981) found that progestogen‐only pills (containing 0.5 mg lynestrenol) did not influence carbohydrate metabolism and were not associated with adverse changes in serum lipoproteins (i.e., reduced serum triglycerides, cholesterol and phospholipids without affecting high‐density lipoprotein lipids). Diab et al. (Diab 2000) also investigated two progestogen‐only methods: Norplant (implant containing levonorgestrel) and depot medroxyprogesterone acetate (DMPA) administered by injection. From this non‐randomised controlled clinical trial, they concluded that Norplant resulted in minimal adverse metabolic changes (i.e., no change in fasting blood sugar; decreased total, low‐ and high‐density lipoprotein lipids; unchanged triglycerides), although haemostatic function might be affected (prolonged partial thromboplastin time). In contrast, DMPA was associated with an unfavourable outcome as fasting blood sugar, and total and low‐density lipoprotein lipids increased; high‐density lipoprotein lipids decreased while triglycerides remained unchanged. Because of the possible negative effect of DMPA on lipid metabolism, the WHO (WHO 2001) does not usually recommend the use of DMPA in women with diabetes mellitus with micro‐ or macrovascular complications or in women who have had diabetes mellitus for more than 20 years, unless other methods are not available.
Although WHO guidelines (WHO 2001) suggest that hormonal contraceptives might be generally recommended for use by women with diabetes mellitus, there is however limited evidence available. Women with diabetes mellitus do need adequate contraceptive advice not only to prevent unplanned pregnancies but also to avoid co‐morbidity and deterioration of the disease because of the possible side effects of hormonal contraceptives. Therefore, we will conduct a systematic review to examine the efficacy and metabolic influences of progestogen‐only and combined oral contraceptives versus non‐hormonal methods, such as non‐hormonal intrauterine devices and barrier methods, in women with diabetes mellitus.
Objectives
Primary objective:
To investigate whether progestogen‐only contraceptive methods and combined estrogen/progestogen contraceptives, respectively, differ from non‐hormonal contraceptives in terms of efficacy in preventing pregnancy and in their side effects on carbohydrate metabolism, lipid metabolism, and haemostatic function when used by women with diabetes mellitus.
Secondary objectives:
-
To investigate whether there are any differences in terms of efficacy in preventing pregnancy and/or side effects on carbohydrate metabolism, lipid metabolism, and haemostatic function of combined oral contraceptive pills containing <50 µg estrogen as compared to combined oral contraceptive pills containing >= 50 µg estrogen when used by women with diabetes mellitus.
-
To investigate whether there are any differences in terms of efficacy in preventing pregnancy and/or side effects on carbohydrate metabolism, lipid metabolism, and haemostatic function between first‐, second‐ and third‐generation progestogen components in oral contraceptives used by women with diabetes mellitus.
-
To assess serious adverse events (thrombosis, myocard infarct, etc.) associated with the use of hormonal contraceptives by women with diabetes mellitus.
Methods
Criteria for considering studies for this review
Types of studies
All studies using random or quasi‐random patient allocation with a minimum treatment period of 6 months will be eligible. The unit of randomisation might be either women (individual) or health‐care unit (cluster). Except for the evaluation of efficacy in preventing pregnancy, crossover studies will be eligible for inclusion but will be analysed separately from parallel studies.
A study is randomised when it appears that the women (or cluster) followed in the study were definitely or probably assigned prospectively to one or two (or more) alternative forms of health care using random allocation. A study is quasi‐randomised when it appears that the women (or cluster) followed were definitely or possibly assigned prospectively to one of two (or more) alternative forms of health care using some quasi‐random method of allocation (such as alternation, date of birth or case record number).
To asses serious adverse events (secondary objective) cohort studies and case‐controlled studies will be eligible, as recommended in the preliminary guidelines of the non‐randomised methods group (http://www.cochrane.dk/nrsmg/).
Types of participants
All women of fertile age and from all ethnic backgrounds with diabetes mellitus, types I and II, irrespective of the severity of the illness, who desire to use contraception will be eligible for inclusion. To be consistent with changes in classification and diagnostic criteria of diabetes mellitus throughout the years, the diagnosis should have been established using the standard criteria valid at the time the trial was conducted. The intervention must be applied to women seeking contraception. Trials enrolling women receiving contraception for non‐contraceptive purposes, such as acne vulgaris, will be excluded. Studies on women with previous gestational diabetes mellitus and studies on women with impaired glucose intolerance will also be excluded from this review.
Types of interventions
Primary interventions:
1. Any combined oral contraceptive pill compared with any progestogen‐only contraceptive (pill, implant, injection, IUD, vaginal ring) used in women with diabetes mellitus.
2. Any combined oral contraceptive pill compared with any non‐hormonal contraceptive method used in women with diabetes mellitus.
3. Any progestogen‐only contraceptive compared with any non‐hormonal contraceptive method used in women with diabetes mellitus.
Secondary interventions:
1. Any combined oral contraceptive pill containing < 50 µg estrogen compared with any oral contraceptive pill containing >= 50 µg estrogen used in women with diabetes mellitus.
2. Any oral contraceptive containing first‐generation progestogens (generally lynesterol, norethynodrel and norethisterone) compared with any oral contraceptive containing second‐generation progestogens (levonorgestrel, norgestimate, medroxyprogesterone acetate, megestrol acetate, chlormadione and cyproterone acetate) and compared with any oral contraceptive containing third‐generation progestogens (desogestrel and gestodene) used in women with diabetes mellitus.
Types of outcome measures
MAIN OUTCOME MEASURES:
-
contraceptive efficacy (e.g., pregnancy rate)
-
diabetes control/carbohydrate metabolism (e.g., glycated haemoglobin, urinary/fasting plasma glucose)
-
haemostatic function (e.g., clotting factors)
-
lipid metabolism (e.g., cholesterol, triglycerides, lipoprotein A)
-
microalbuminuria
SECONDARY OUTCOME MEASURES
-
continuation rate
-
onset or worsening of microvascular disease (e.g., retinopathy, nephropathy)
-
onset or worsening of macrovasculair complications (e.g., cardiovascular disease)
-
other serious adverse events
Search methods for identification of studies
Electronic databases will be searched using the search strategy outlined below to identify publications that describe randomised or quasi‐randomised controlled and crossover trials comparing contraceptive methods in women with diabetes. The general search strategy for randomised controlled trials (RCT) and controlled clinical trials (CCT) of the Fertility Regulation CRG (Cochrane Library, 2001, issue 3) will be combined with the c (Cochrane Library, 2001, issue 2), and the search for contraceptive agents. Databases that will be searched include MEDLINE, EMBASE, CENTRAL/CCTR, POPLINE, CINAHL, WorldCat, ECO and ArticleFirst. The Science Citation Index will be searched to identify trials that have cited the studies that will be included in the analysis. The British Library Inside will be searched for ongoing trials.
We will also search for reports of suspected adverse effects of hormonal contraceptives in women with diabetes mellitus that will not be reported in the studies that will be retrieved in the aforementioned searches. We will search MEDLINE for cohort studies and case‐control studies using the search strategy outlined below, and will search EMBASE, POPLINE and CINAHL using comparable terms.
The search strategy for this review will include:
#1 Diabetes search
((((((((((((((((("diabetes mellitus"[MESH] OR
"diabet*"[title/abstract word]) OR
(DKA[title/abstract word] OR
IDDM[title/abstract word] OR
DM1[title/abstract word])) OR
(MODY[title/abstract word]OR
DM2[title/abstract word] OR
NIDDM[title/abstract word] OR
IIDM[title/abstract word])) OR
"insulin* secret* dysfunc*"[title/abstract word]) OR
"insulin* resist*"[title/abstract word]) OR
(("impaired glucose tolerance"[title/abstract word] OR
"glucose intoleran*"[title/abstract word] OR
"insulin* resist*"[title/abstract word]) AND
(DM[title/abstract word] OR
DM2[title/abstract word]))) OR
("insulin*depend*"[title/abstract word] OR
"insulindepend*"[title/abstract word] OR
"insulin‐depend*"[title/abstract word])) OR
("non insulin*depend*"[title/abstract word] OR
"noninsulindepend*"[title/abstract word] OR
"noninsulin‐depend*"[title/abstract word] OR
"non insulin‐depend*"[title/abstract word] OR
"noninsulin depend*"[title/abstract word] OR
"non‐insulindepend*"[title/abstract word])) OR
(("typ* 1"[title/abstract word] OR
"typ* I"[title/abstract word]) AND
DM[title/abstract word])) OR
(("typ* 2"[title/abstract word] OR
"typ* II"[title/abstract word]) AND
DM[title/abstract word])) OR
((juvenil*[title/abstract word] OR
child*[title/abstract word] OR
keto*[title/abstract word] OR
labil*[title/abstract word] OR
brittl*[title/abstract word] OR
"early onset"[title/abstract word]) AND
(DM[title/abstract word] OR
DM1[title/abstract word]))) OR
(("keto* prone"[title/abstract word] OR
"autoimmun*"[title/abstract word] OR
"sudden onset"[title/abstract word]) AND
(DM[title/abstract word] OR
DM1[title/abstract word]))) OR
(("keto* resist*"[title/abstract word] OR
"nonketo"[title/abstract word] OR
"non keto"[title/abstract word] OR
"adult* onset"[title/abstract word] OR
"matur* onset"[title/abstract word] OR
"late* onset"[title/abstract word] OR
"slow onset"[title/abstract word] OR
"stabl*"[title/abstract word]) AND
(DM[title/abstract word] OR
DM2[title/abstract word]))) OR
"Insulin Resistance"[MESH]) OR
("insulin* defic*"[title/abstract word] AND
(absolut*[title/abstract word] OR
relativ*[title/abstract word]))) OR
"metabolic* syndrom*"[title/abstract word]) NOT
("Dermatomyositis"[MESH] OR
"Myotonic Dystrophy"[MESH] OR
("Diabetes Insipidus"[MESH] NOT
("Diabetes Mellitus"[MESH] OR
mellitus[title/abstract word]))))
#2 Contraceptives search
(((((((((("Contraceptive Agents, Female"[MESH] OR
(contraceptive[title/abstract word] AND
device*[title/abstract word])) OR
(oral[title/abstract word] AND
contraceptive*[title/abstract word])) OR
(progestagen[title/abstract word] OR
progestogen[title/abstract word] OR
progesteron*[title/abstract word] OR
levonorgestrel[title/abstract word] OR
norethisteron*[title/abstract word] OR
norethindron*[title/abstract word] OR
norgestimat*[title/abstract word] OR
desogestr*[title/abstract word] OR
gestode*[title/abstract word] OR
norgestrel[title/abstract word])) OR
(estrogen*[title/abstract word] OR
estragen*[title/abstract word] OR
oestrogen*[title/abstract word] OR
oestragen*[title/abstract word])) OR
(ethinyl[title/abstract word] AND
(estradiol[title/abstract word] OR
oestradiol[title/abstract word]))) OR
((estrogen*[title/abstract word] OR
estragen*[title/abstract word] OR
oestrogen*[title/abstract word] OR
oestragen*[title/abstract word] OR
(ethinyl[title/abstract word] AND
(estradiol[title/abstract word] OR
oestradiol[title/abstract word]))) AND
(low[title/abstract word] AND
dose[title/abstract word]))) OR
(progestagen‐only[title/abstract word] OR
progestogen‐only[title/abstract word] OR
progesteron‐only[title/abstract word] OR
norplant[title/abstract word])) OR
"barrier method*"[title/abstract word]) OR
(IUD*[title/abstract word] OR
IUS*[title/abstract word] OR
(intra‐uterine[title/abstract word] AND
(system[title/abstract word] OR
systems[title/abstract word] OR
device*[title/abstract word])))) NOT
("Menopause"[MESH] OR
"Estrogen Replacement Therapy"[MESH] OR
"Neoplasms"[MESH]))
#3 RCTs search
(((((((((((((((((((((((("randomized controlled trials"[MESH:noexp] OR
"random allocation"[MESH:noexp]) OR
"double‐blind method"[MESH:noexp]) OR
"single‐blind method"[MESH:noexp]) OR
"clinical trials"[MESH]) OR
"placebos"[MESH:noexp]) OR
"research design"[MESH:noexp]) OR
"comparative study"[MESH]) OR
"evaluation studies"[MESH]) OR
"follow‐up studies"[MESH]) OR
"prospective studies"[MESH]) OR
"cross‐over studies"[MESH]) OR
"intervention studies"[MESH]) OR
"randomized controlled trial"[pt]) OR
"controlled clinical trial"[pt]) OR
"clinical trial"[pt]) OR
"clinic* trial*"[title/abstract word]) OR
(((("singl*"[title/abstract word] OR
"doubl*"[title/abstract word]) OR
"tripl*"[title/abstract word]) OR
"trebl*"[title/abstract word]) AND
("blind*"[title/abstract word] OR
"mask*"[title/abstract word]))) OR
"placebo*"[title/abstract word]) OR
"random*"[title/abstract word]) OR
"latin square"[title/abstract word]) OR
"control*"[title/abstract word]) OR
"prospectiv*"[title/abstract word]) OR
"volunteer*"[title/abstract word]) NOT
("animal"[MESH] NOT
"human"[MESH]))
#4 Diabetes Mellitus with Hormonal Contraceptives and RCTs
Search #1 AND #2 AND #3
#5 Adverse events
"probability" [MESH] OR
(risk*[WORD] OR
cohort*[WORD] OR
follow‐up[WORD] OR
predict*[WORD] OR
case‐control*[WORD]) OR
(cause*[WORD] OR
causat*[WORD] OR
causing[WORD] OR
causal*[WORD] OR
etiol*[WORD] OR
aetiol*[WORD]) OR
"case‐control studies"[MESH]
#6 Diabetes Mellitus with Hormonal Contraceptives and Adverse events
Search #1 AND #2 AND #5
This search strategy will be adapted to search the different databases.
-
No language restrictions will be used in the searches.
-
The reference list of all identified studies will be searched for additional, previously unidentified trials.
-
Relevant book chapters and review articles located with the searches or in the reference lists will be searched for all relevant trials.
-
Authors of all potentially or definitely eligible studies will be contacted to find any unidentified published, unpublished or ongoing studies.
-
Attempts will be made to obtain published, unpublished or ongoing trials from pharmaceutical companies marketing contraceptives.
Data collection and analysis
ASSESSMENT OF ELIGIBILITY OF THE STUDY
The titles and abstracts from the literature search will be evaluated by two reviewers (JV and CO). If an abstract or full article is available, a study will be eligible if it includes information on study type (RCTs, CCTs and crossover studies), diabetes mellitus and contraceptives in women. If only a title is available, the full article will be obtained if the title refers to contraceptives and women with diabetes. For all potentially or definitely eligible studies, the full article will be obtained and photocopied. Any disagreement about trial selection will be resolved by discussion or by consulting a third reviewer (HV).
DEMOGRAPHICS AND POSSIBLE COVARIATES OR CONFOUNDING FACTORS
In addition to the outcomes of interests, the following information will be extracted for all eligible studies
-
Study characteristics (e.g., authors; year of publication; in‐ and exclusion critera; interventions; method of randomisation; allocation concealment; number of participants eligible, randomised and included, respectively; blinding; protocol violations; loss to follow‐up).
-
Age; continuous
-
Parity; nulliparous versus multiparous
-
Social Economical status (SES, Erikson 1983)
-
Ethnicity
-
Previous contraceptive method
-
Smoking; yes/no
-
Diastolic and systolic blood pressure (mmHg)
-
Length of illness (yrs)
-
Severity of illness (White index) (Heineman 1999)
-
Body Mass Index (BMI; kg/m2); continuous
ASSESSMENT OF METHODOLOGICAL QUALITY
All eligible studies will be evaluated whether systematic errors are minimised:
-
Random allocation technique: yes or no
-
Concealment of allocation; Trials will be given a quality score (A: adequate, B: unclear, C: inadequate) as described in the Cochrane Reviewer's Handbook (Clarke and Oxman)
-
Blinding of patients and care providers: adequate or not
-
Blinding of outcome assessors: adequate or not
-
Significant differences in loss to follow‐up and post‐randomisation exclusions
For the quality assessment of the non‐randomised studies, the Newcastle‐Ottawa Scale (Wells NOS) will be used. Two reviewers (JV and CO) will independently conduct assessment of methodological quality after a pilot test of the assessment. Any disagreement will be resolved by discussion or by consulting a third reviewer.
METHOD OF ANALYSIS
First, the analysis will be stratified according to type of intervention (e.g., progestin‐only, combined oral contraceptives), and type of diabetes mellitus. Then trials will be compared on age, BMI, duration of illness, parity, SES, ethnicity, smoking status, blood pressure, and severity of illness to determine possible confounding factors between the studies.
If meta‐analysis is appropriate, Odds Ratios with 95% confidence intervals (95% CI) will be calculated to evaluate differences in dichotomous data between the groups. For continuous data, means or mean differences of change scores with 95% CI will be used to compare different groups. Standardised mean differences will be calculated for different measures of carbohydrate metabolism if appropriate. For comparisons with progestagen‐only contraceptives, contraceptive method will be accounted for. All statistical tests will be two‐tailed, using p<0.05 as the cut‐off value for statistical significance.
HETEROGENEITY
If there are sufficient studies identified, and pooling of data is appropriate, the random‐effect model will be used to allow for differences in treatment effects across studies stratified according to type of intervention and type of diabetes mellitus. The extent of statistical heterogeneity will be evaluated using the Q‐test statistic and the H.
If appropriate, metaregression analysis will be conducted to compare treatment effects of different types of interventions (e.g., progestin‐only methods, combined oral contraceptives). The choice of contraceptive methods may be partially determined by age and parity, and this might explain differences in treatment effects across studies that include different types of contraceptive methods. Furthermore, in diabetes mellitus, duration of illness, smoking, blood pressure and BMI are positively associated with an increased risk of cardiovascular disease (Fontbonne 1991, Bass 1993). Differences in treatment effects across trials and interventions may also be due to the use of restrictive inclusion and exclusion criteria for duration of illness, smoking status, blood pressure, and BMI in some but not all trials (McKee 1999). Therefore, in the presence of significant heterogeneity, the influence of age, parity, duration of illness, smoking, diastolic and systolic blood pressure, and BMI will first be considered (a‐priori analyses).The influence of other factors that might explain differences in treatment effects across trials will be considered as post‐hoc analyses.
SENSITIVITY ANALYSIS
Sensitivity analyses will include comparisons between the different components of methodological quality. If a clinically relevant and statistically significant difference between groups on either of the mentioned demographic and possible confounding factors will be detected, additional sensitivity analysis will be conducted.
