Maria F. Gallo, Jennifer Legardy-Williams, Tina Hylton-Kong, Carole Rattray, Athena P. Kourtis, Denise J. Jamieson, Markus J. Steiner

To evaluate initiation of a two-rod, 150-mg levonorgestrel contraceptive implant on women's perceived and observed body weight. We conducted a secondary analysis of data from an open, randomized controlled trial of adult, nonpregnant, human immunodeficiency virus-negative women attending a public clinic in Kingston, Jamaica, who were assigned to initiate implant use either immediately or after a 3-month delay. The primary objective of the parent study was to assess the effect of initiation of the implant on the frequency of condom use. We compared study arms during follow-up using one-sided χ tests for differences in perceived weight gain and loss, one-sided Wilcoxon-Mann-Whitney tests for median gain in measured weight, and logistic regression with generalized estimating equations for risk of gaining greater than 2 kg. From 2012 to 2014, women were assigned to the implant (n=208) or delay arm (n=206). At 3 months, more women in the implant arm (15.3%) reported perceived weight gain than in the control arm (4.3%) (P=.01). Despite differences in perception, the implant and control arms did not differ significantly in median weight gain at 1-month (0.0 kg and 0.0 kg, respectively; P=.44) and 3-month visits (0.5 kg and 0.0 kg, respectively; P=.27). Study arms did not differ in risk of gaining greater than 2 kg (odds ratio 0.9, 95% confidence interval 0.6-1.3). We found no evidence of weight gain from short-term implant use. Through the power of the nocebo effect, the practice of counseling women to expect possible weight gain from initiating implant use could lead them to perceive weight gain even in its absence and contribute to the early discontinuation of this highly effective contraceptive method.