Cocoa butter lotion for prevention of striae gravidarum: a double-blind, randomised and placebo-controlled trial*

Authors

  • H Osman,

    1. a Department of Health Behaviour and Educationb Department of Family Medicinec Department of Obstetrics and Gynecology and d Department of Dermatology, American University of Beirut Medical Center, Beirut, Lebanone Department of Obstetrics and Gynecology, Nini Women’s Center, Tripoli, Lebanonf Department of Obstetrics and Gynecology, Bahman Hospital, Beirut, Lebanon
    Search for more papers by this author
  • a,b IM Usta,

    1. a Department of Health Behaviour and Educationb Department of Family Medicinec Department of Obstetrics and Gynecology and d Department of Dermatology, American University of Beirut Medical Center, Beirut, Lebanone Department of Obstetrics and Gynecology, Nini Women’s Center, Tripoli, Lebanonf Department of Obstetrics and Gynecology, Bahman Hospital, Beirut, Lebanon
    Search for more papers by this author
  • c N Rubeiz,

    1. a Department of Health Behaviour and Educationb Department of Family Medicinec Department of Obstetrics and Gynecology and d Department of Dermatology, American University of Beirut Medical Center, Beirut, Lebanone Department of Obstetrics and Gynecology, Nini Women’s Center, Tripoli, Lebanonf Department of Obstetrics and Gynecology, Bahman Hospital, Beirut, Lebanon
    Search for more papers by this author
  • d R Abu-Rustum,

    1. a Department of Health Behaviour and Educationb Department of Family Medicinec Department of Obstetrics and Gynecology and d Department of Dermatology, American University of Beirut Medical Center, Beirut, Lebanone Department of Obstetrics and Gynecology, Nini Women’s Center, Tripoli, Lebanonf Department of Obstetrics and Gynecology, Bahman Hospital, Beirut, Lebanon
    Search for more papers by this author
  • e I Charara,

    1. a Department of Health Behaviour and Educationb Department of Family Medicinec Department of Obstetrics and Gynecology and d Department of Dermatology, American University of Beirut Medical Center, Beirut, Lebanone Department of Obstetrics and Gynecology, Nini Women’s Center, Tripoli, Lebanonf Department of Obstetrics and Gynecology, Bahman Hospital, Beirut, Lebanon
    Search for more papers by this author
  • and f AH Nassar c

    Corresponding author
    1. a Department of Health Behaviour and Educationb Department of Family Medicinec Department of Obstetrics and Gynecology and d Department of Dermatology, American University of Beirut Medical Center, Beirut, Lebanone Department of Obstetrics and Gynecology, Nini Women’s Center, Tripoli, Lebanonf Department of Obstetrics and Gynecology, Bahman Hospital, Beirut, Lebanon
    Search for more papers by this author

  • *

    This trial was registered in clinicaltrials.gov on 16 June 2005 under the title ‘Cocoa Butter for Prevention of Stretch Marks’ (registration no. NCT00114660).

Dr AH Nassar, Department of Obstetrics and Gynecology, American University of Beirut Medical Center, PO Box 113-6044/B36, Beirut, Lebanon. Email: an21@aub.edu.lb

Abstract

Objective  To assess whether application of cocoa butter lotion reduces the development of striae gravidarum (SG).

Design  Multicentre, double-blind, randomised and placebo-controlled trial.

Setting  Beirut and Tripoli, Lebanon.

Population  Nulliparous women presenting for prenatal care.

Methods  Nulliparous women presenting in the first trimester were randomly assigned to receive a lotion containing cocoa butter or a placebo lotion. Women were instructed to apply the assigned lotion daily until delivery.

Main outcome measure The development of striae over the abdomen, breasts and thighs postpartum.

Results Of 210 women enrolled, 175 (83%) completed the study. Ninety-one women received the study lotion and 84 received the placebo. There was no difference in the development of SG (45.1% versus 48.8%; P = 0.730) or the severity of SG between cases and controls. The results did not change when presence of stretch marks at enrolment or compliance with the regimen were taken into account.

Conclusion  Topical application of a lotion containing cocoa butter does not appear to reduce the likelihood of developing striae gravidarum.

Introduction

Striae gravidarum (SG), also known as ‘stretch marks’, are a common cosmetic concern for many parturients. They usually appear on the abdomen and breasts and less commonly on the hips, thighs and buttocks. Striae initially appear as slightly raised pink to purple linear bands that eventually mature to become pale, atrophic and finely wrinkled lines. The reported prevalence ranges from 50 to 90% depending on the population studied.1 In a previous study conducted by our group, we found that the prevalence of SG in primigravidae was 61%.2

The aetiology of SG is still unclear but probably involves the interplay between hormonal factors and the dermal support matrix combined with skin stretching.1,3–5 SG have been associated with young maternal age, increased infant birthweight and excessive maternal weight gain during pregnancy.2,6,7 Other factors such as personal history, familial predisposition, race, maternal body mass index (BMI) and skin type are controversial. Chang et al. found family history to be predictive of the development of SG.8 However, in our previous study on SG, although women with a family history of SG were not more likely to develop SG, they were more likely to have severe SG.2 Some authors found white race3 and high BMI7 to be associated with SG, while others found nonwhite race more predictive of SG and prepregnancy BMI to be unrelated.8

SG can be a cause of great concern for many women. Although they are not detrimental to one’s health, they can significantly affect a woman’s self-esteem and image. Many products have been marketed for use in the prevention or treatment of stretch marks; however, there is no evidence to support or refute their efficacy. Despite their psychological and economic burden, SG and their prevention have received little attention in the medical literature.

Cocoa butter, a natural fat derived from cocoa beans (Theobroma cacao), has moisturizing properties and is commonly used in many cultures. Many women believe that the application of cocoa butter to the skin helps prevent the development of SG. Some physicians and midwives advise their patients to use cocoa butter before, during or even after pregnancy for prevention or treatment of SG. However, this practice is not supported by scientific evidence. The objective of this study was to assess the effect of topical application of cocoa butter lotion on the development of SG in nulliparous women compared with a placebo lotion.

Methods

Nulliparous women with singleton pregnancies presenting for prenatal care in the first trimester between November 2004 and December 2005 were invited to participate in the study. Women with known hypersensitivity to cocoa butter or any of the components of the lotion were excluded from the study. The study was approved by the Institutional Review Board of the American University of Beirut and registered with clinicaltrials.gov. Obstetricians were selected based on their patient volume and their willingness to enrol their patients in the study. They were affiliated with four medical centres in different areas of the country, which cater to women from a variety of socio-economic classes. Flyers advertising the study were placed in the offices of participating obstetricians. Obstetricians or their clerical staff approached eligible women to invite them to participate in the study. Women who agreed to participate in the study were given complimentary educational brochures about breastfeeding, newborn care and the postpartum period.

After informed consent, women were randomly assigned to receive the study lotion or the placebo lotion. Randomisation was conducted using a computer-generated random number table by the principal investigator who was not involved in the clinical care of the women. The study lotion was a commercially available lotion containing cocoa butter and tocopheryl acetate (vitamin E). The placebo lotion was made up to look, smell and feel the same as the study lotion but did not contain the active ingredients. The study and placebo lotions were supplied to the researchers by the product manufacturer in identical containers with pump delivery. Women were instructed to apply a thin layer of the lotion to their abdomen (two pumps), breasts and thighs (one pump each) once daily. The application was to be started between 12 and 18 completed weeks of gestation and continued until delivery.

Five assessors were trained to complete the data collection tools and assess the severity of SG based on a scale developed and previously validated by the authors.2 The scale took into consideration the density and width of striae to estimate the surface area of the body part affected. Assessors, researchers, study participants and their physicians were blinded to the lotion assignment. The codes for the study and placebo lotions were opened after the final assessment of the last randomised woman.

An initial assessment of participants was conducted at enrolment. This included the collection of some demographic information, baseline information related to the pregnancy and information about history of SG in mothers and sisters. In addition, the assessors determined the participant’s skin type and conducted a clinical assessment of previous stretch marks. Skin type was determined based on the Fitzpatrick classification, which depends on how often a person burns and how well they tan when exposed to the sun.9

Participants were followed up with monthly telephone calls to assess their compliance with the regimen, encourage women to continue using their lotions and ensure that they were not developing skin reactions. After delivery, women were assessed for the development of SG either at the hospital where they gave birth or during a home visit within 2 weeks of delivery. The primary outcome variable was the proportion of women who developed SG and the secondary outcome variable was the severity of SG by site of involvement.

Sample size was calculated based on our previous finding that the incidence of SG in our population was 61%.2 We considered a 20% reduction in this number in favour of the study lotion to be clinically relevant. The sample size calculated required 97 participants in each arm to achieve a power of 80% at a significance level of 0.05.

Compliance was defined as complete if lotion was applied daily and incomplete if applied three to six times per week. Women were considered noncompliant if the lotion was applied two times per week or less. In cases where compliance varied during the course of the study, the poorest compliance was used for analysis. The severity score for each site was rated as 1 for mild, 2 for moderate and 3 for severe SG. When calculating the severity index (SI) for multiple sites, the respective scores were added. Family history SI was based on the addition of the SI for the mother and the sister at the three sites. If a woman did not have a sister who had been pregnant in the past or if she had a sister with no previous SG, the SI was considered zero. In case of a positive history of SG in multiple sisters, the history of the sister with the most severe lesions was used for analysis.

Statistical analysis was performed using the SPSS statistical package version 13 (Chicago, IL, USA). Data were analysed by intention to treat. For categorical variables, comparison was carried out using the chi-square test, two-tailed Fisher’s exact test if the expected cell frequencies were small and the relative risk with 95% confidence interval (CI). Continuous variables were compared using two-tailed Student’s t test. P < 0.05 was considered statistically significant.

Results

A total of 210 women were enrolled in the study. Eight women did not complete the study due to loss of pregnancy, 8 withdrew from the study and 18 were lost to follow up. One woman in the placebo group dropped out of the study due to a possible allergic reaction to the lotion. A total of 175 women (83%) completed the study, 91 of whom received the study lotion and 84 received the placebo lotion (Figure 1).

Figure 1.

Flow chart of recruitment, randomisation and completion status.

Demographic characteristics of both groups are summarised in Table 1. There was no difference in the baseline characteristics of the women or their newborns between groups. At initial assessment, stretch marks were noted in 45.1% of cases compared with 48.8% of controls (P =0.730) (Table 2). More than one site was involved in 13/41 (31.7%) of cases versus 11/41 (26.8%) of controls. No difference in the SI was noted at initial assessment. No differences were noted in patient compliance in both groups where 65.9% of cases were ‘completely’ compliant versus 65.5% of controls (relative risk [RR] 1.0; 95% CI: 0.8–1.3) and only 3.3% of cases were noncompliant versus 1.2% of controls (RR 2.8; 95% CI: 0.4–19.3).

Table 1.  Baseline, demographic and neonatal characteristics
 Cases (n = 91)Controls (n = 84)P
  • *

    25/84 and 29/91 women had sisters who were pregnant and developed SG.

Age (years)26.4 ± 5.025.4 ± 4.20.150
Age ≤2536 (39.6)35 (41.7)0.897
Age ≥357 (7.7)2 (2.4)0.172
Gestational age at enrolment (weeks)12.5 ± 4.212.4 ± 4.10.949
BMI (kg/m2)24.4 ± 3.623.7 ± 3.80.236
BMI ≥ 2531 (34.1)19 (22.6)0.132
Gestational weight gain14.2 ± 5.214.4 ± 4.20.781
Weight gain >14.4 kg38 (41.8)35 (41.7)0.990
Skin type
I04 (4.4)0.234
II18 (21.4)14 (15.4)
III56 (66.7)64 (70.3)
IV9 (10.7)9 (9.9)
Missing1 (1.2)0
Personal history of stretch marks40 (44.0)38 (45.2)0.985
SI0.7 ± 0.90.8 ± 1.40.334
Family history of SG62 (68.1)58 (69.0)0.896
SI*2.4 ± 2.82.0 ± 2.80.478
Smoking3 (3.3)8 (9.5)0.166
Use of other creams13 (14.3)7 (8.3)0.318
Birthweight (g)3293 ± 4323198 ± 3740.123
Macrosomia (>4000 g)3 (3.3)2 (2.4)1.000
% Male46 (50.5)39 (46.1)0.694
Table 2.  Distribution and severity of SG at initial and final assessment
 Cases (n = 91)Controls (n = 84)RR (95% CI)P
Initial assessment of SG
None50 (54.9)43 (51.2)1.1 (0.8–1.4)0.785
1 site28 (30.8)30 (35.7)0.9 (0.6–1.3)
>1 site13 (14.3)11 (13.1)1.1 (0.5–2.3)
SI0.8 ± 1.00.8 ± 1.1 0.633
Final assessment of SG
None19 (20.9)19 (22.6)0.9 (0.5–1.6)0.953
1 site28 (30.8)26 (31.0)1.0 (0.6–1.6)
>1 site44 (48.4)39 (46.4)1.0 (0.8–1.4)
SI2.2 ± 1.82.1 ± 1.8 0.687

At final assessment, SG were noted in a similar proportion of cases and controls and the SI was also similar. This continued to hold true when each site was studied individually, breasts, abdomen and thighs (Table 3).

Table 3.  Final assessment of SG by location and severity
 Cases (n = 91)Controls (n = 84)RR (95% CI)P
Abdomen
Absent38 (41.8)32 (38.1)1.1 (0.8–1.6)0.575
Mild21 (23.1)27 (32.1)1.2 (0.6–2.1)
Moderate22 (24.2)16 (19.0)1.0 (0.4–2.4)
Severe10 (11.0)9 (10.7)1.3 (0.7–2.3)
Breasts
Absent57 (62.6)59 (70.2)0.9 (0.7–1.1)0.284
Mild27 (29.7)20 (23.8)1.3 (0.8–2.1)
Moderate6 (6.6)2 (2.4)2.8 (0.7–11.9)
Severe1 (1.1)3 (3.6)0.3 (0.0–2.1)
Thighs
Absent45 (49.5)44 (52.4)0.9 (0.7–1.3)0.690
Mild32 (35.2)25 (29.8)1.2 (0.8–1.8)
Moderate8 (8.8)11 (13.1)0.7 (0.3–1.6)
Severe6 (6.6)4 (4.8)1.4 (0.4–4.5)

Even after grouping women according to the presence or absence of stretch marks at the initial assessment, no differences were noted in the prevalence or severity of SG in cases and controls (Table 4). The results were also comparable when the analysis was restricted to women who were compliant with regimen.

Table 4.  Final assessment of SG by location and severity excluding women with old stretch marks
 CasesControlsRR (95% CI)P
Abdomen
n7872 0.945
Absent38 (48.7)31 (43.1)1.1 (0.8–1.6) 
Mild16 (20.5)20 (27.8)0.7 (0.4–1.3) 
Moderate17 (21.8)15 (20.8)1.1 (0.6–1.9) 
Severe7 (9.0)6 (8.3)1.1 (0.4–3.0) 
SI0.9 ± 1.00.9 ± 1.0 0.855
Breasts
n7367 0.474
Absent55 (75.3)57 (85.1)0.9 (0.8–1.1) 
Mild14 (19.2)7 (10.4)1.8 (0.8–4.3) 
Moderate3 (4.1)1 (1.5)2.8 (0.4–19.1) 
Severe1 (1.4)2 (3.0)0.5 (0.1–3.5) 
SI1.3 ± 0.61.2 ± 0.6 0.342
Thighs
n5751 0.716
Absent42 (73.7)43 (84.3)0.9 (0.7–1.1) 
Mild7 (12.3)4 (7.8)1.6 (0.5–4.9) 
Moderate5 (8.8)2 (3.9)2.2 (0.5–9.9) 
Severe3 (5.3)2 (3.9)1.3 (0.3–6.6) 
SI1.5 ± 0.91.3 ± 0.7 0.222

Discussion

This study confirms that the topical application of cocoa butter in an emollient during pregnancy does not reduce the likelihood of developing SG or their severity, although it does not rule out an effect of an emollient on the development of SG. These findings did not change when compliance with treatment was taken into consideration.

To date, the only intervention that has been shown to reduce the likelihood of developing SG is a cream containing Centella asiatica extract, alpha-tocopherol and collagen-elastin hydrolysates (41 women received active cream and 39 received placebo).10 Interestingly, the greatest benefit was noted in women who had previously developed stretch marks during puberty, while there seemed to be no apparent benefit in women with no stretch marks at initiation of the trial. Once SG have developed, attempts at treating them with a variety of lasers and topical tretinoin, alpha hydroxyl acids and vitamin C have shown mild improvement at best.11–13

Despite restricting our trial to nulliparas in an attempt to limit it to the development of de novo SG, a high percentage of women (45.1 and 48.8% in cases and controls, respectively) had stretch marks at the enrolment. However, old SG were easy to differentiate from new ones at the postpartum assessment due to a clear difference in colour and texture. Even after running the analysis on the subgroup of women with a priori stretch marks, no differences were observed in the development of SG in the two groups.

Our study also shows that the development of SG is not restricted to the abdomen. A significant proportion of women develop SG over the breasts and thighs. We found the clinical assessment of SG using our standardised scale quite feasible, and it provided more reliable data than the patient self-assessments that have been used in previous studies.

As different skin types have varying likelihood of developing SG, it is possible that response to treatment could be different in women with lighter or darker skin. One limitation of this study is that most participants had Fitzpatrick skin type II and III (86.9%). Our results may therefore not be generalisable to women of all skin types. A similar study with a more heterogeneous population may show different results.

Although we set out to have reliable measures of compliance, we do not feel we were successful in our efforts. Women were asked about compliance once per month during a telephone conversation and again at the time of the final assessment. Reports of compliance varied greatly for each patient and assessors reported that women may have been telling them what they wanted to hear when they asked about compliance. As a result, we do feel that compliance with the regimen may have been a problem. However, our analysis was based on intent to treat, and we believe that use of the lotion in our study population is likely to mirror typical use in the general population.

During the course of the study, political instability in the country led many families to travel. This led us to lose a large number of the study participants. This large loss to follow up limited the final analysis to a lower number of women than originally planned to recruit. A post hoc analysis revealed that our study had 76% power to identify the 20% reduction in the development of SG. However, since there was no trend towards a reduction in the development of SG in the study group, we think it is unlikely that a larger sample size would have shown positive results.

Conclusion

Women will continue to seek ways to prevent the development of SG in pregnancy. Until effective means of prevention of SG are identified, it seems reasonable to counsel women who want to avoid developing SG about the modifiable risk factors that have been linked to the development of SG such as maternal weight gain.2

Our findings do not support the use of cocoa butter lotion for the prevention of SG. Further studies are needed to confirm our findings in other populations and to evaluate the effectiveness of other commonly used products in preventing SG. Future research on SG should include assessments of both the breasts and thighs since these areas appear to be commonly affected. Better characterisation of the epidemiological factors associated with SG would be a valuable addition to the literature.

Contribution to authorship

H.O., N.R. and A.H.N. contributed to the conception and design of the study. H.O., I.M.U., R.A.-R., I.C. and A.H.N. contributed to the collection of data, and H.O., I.M.U. and A.H.N. contributed to analysis and interpretation of the data; H.O., I.M.U., N.R. and A.H.N. contributed to drafting of the manuscript; and all authors contributed to its revision and final approval.

Details of ethics approval

This study was approved by the Institutional Review Board at the American University of Beirut on 24 November 2004 (Protocol no. HO.FM.01a). The study conforms to the US Food and Drug Administration Good Clinical Practice in clinical trials.

Funding

This trial was funded by the Center for Research on Population and Health at the American University of Beirut, Lebanon, with generous support from the Wellcome Trust. Additional support for this study was obtained through the University Research Board (URB) at the American University of Beirut, Beirut, Lebanon. E.T. Browne Drug Company, Inc. provided the study and placebo lotions.

Acknowledgements

The authors acknowledge Dr Oona Campbell for her input and ideas about the study design. They also thank Drs Basma Abdo, Ghassan Khoury, Khalida Bitar, Najat Hazimeh and Rihab Kassem for allowing the researchers access to their patients; and Dr Hazar Kobaya, Farah Barbir, Zeina Assaf, Salam El-Khatib, Rania Beiruti, Sana Masri, Christine Chibel and Dima Sinno for their assistance with data collection and data management.

Ancillary