The side effects of oral contraceptives (OCs) can be minimized by appr
opriate OC selection. Side effects or perceived side effects that mani
fest themselves physically - e.g., weight gain, breakthrough bleeding
(BTB), nausea, headache, breast tenderness, mood swings, acne, and hir
sutism - are the most common causes of premature discontinuation of or
al contraception. The relative androgenicity of the progestin componen
t of combination OCs has become an important differential in selecting
OC formulations. Several studies have indicated that preparations wit
h less androgenic potential can minimize some of the ''physical'' side
effects and adverse metabolic effects traditionally associated with o
ral contraception. Acne and hirsutism, common pre-existing conditions
that are clearly related to the androgenicity of the progestin compone
nt, can be eliminated or improved by use of OCs with low androgenic ac
tivity. Many women perceive that OCs cause weight gain; although weigh
t gain is to some extent androgen related, most studies comparing low-
androgenic OCs with medium- or high-androgenic preparations have found
little or no change in weight regardless of formulation. BTB, which u
sually subsides within a few months, is related to the dose, potency,
and ratio of the estrogen and progestin in the OC formulation. Low-est
rogen-dose OCs (<35 mu g ethinyl estradiol [EE]) containing less andro
genic progestins are associated with bleeding patterns as acceptable a
s older low-estrogen-dose formulations. The same analysis found that s
moking cigarettes promotes BTB in women who use OCs. There is no convi
ncing evidence that the use of one progestin or another is less likely
to cause or exacerbate headache; however, changing preparations somet
imes reduces the incidence. Women with persistent headaches during the
pill-free interval may benefit from a longer cycle of OC treatment. N
ausea and breast tenderness are primarily estrogen-related effects; if
a woman experiences persistent nausea, switching to an OC formulation
containing 20 mu g EE may be appropriate as long as the patient is ca
utioned that BTB is more likely. Mood changes are a common, highly sub
jective complaint whose relationship to OC use is hard to assess. Conc
erns about the potentially deleterious effects of combination OCs on l
ipid/lipoprotein and carbohydrate metabolism have been substantially d
iminished by new epidemiologic findings relative to cardiovascular dis
ease as well as by the development of low- androgenic progestins. Form
ulations containing these progestins lower LDL cholesterol and increas
e HDL cholesterol; they do not affect carbohydrate metabolism as much
as older, more androgenic formulations.