The Three-Month Trial That Found a Signal
The 2025 randomized trial from Málaga is useful because it tested several calorie-restricted strategies beside a calorie-restricted Mediterranean diet, rather than comparing fasting with no plan at all. Its results give intermittent fasting a more demanding test: can a particular schedule or pattern produce greater short-term weight loss when every group is asked to reduce energy intake? To judge that fairly, we need to follow the trial from its selected participants and shared support to the exact arms, outcomes, and limits of the result.
The study does not settle the broad question by itself. It is one small, three-month comparison with several features changing at once. But it does show what a direct trial can add to a broad review: a closer look at which particular protocols differ from an active dietary alternative, and which do not.
A fairer comparison, still a selected one
The study was a prospective, parallel-arm randomized clinical trial conducted at a university hospital in Málaga, Spain. It enrolled 160 adults with obesity and assigned 32 participants to each of five groups: a Mediterranean-diet control, a ketogenic diet, early time-restricted eating, late time-restricted eating, or modified alternate-day fasting.
Participants were between 18 and 65 years old and had a body-mass index between 30 and 45 kg/m². The group was 70.6% women. These details are not background trivia. They define the population in which the comparison was made. People with diabetes were excluded, as were people with a history of eating disorders, use of medication that could affect weight, chronic kidney disease, several recent or active illnesses, and other conditions specified by the protocol. The findings therefore cannot automatically be transferred to people whose health or circumstances excluded them from the trial.
The design included several features that strengthen an assigned-group comparison. An independent statistician generated the random allocation sequence, allocation was concealed with sealed opaque envelopes, and the statistical analysis was blinded to group assignment. Participants and investigators were not blinded, which is difficult for diets and eating schedules: people know whether they are eating at 8 a.m. or fasting on a particular day, and that knowledge can affect expectations and behavior. The researchers analyzed the data using an intention-to-treat approach, meaning participants were analyzed according to the group to which they were assigned rather than being silently removed because they did not complete the plan.
The intervention was also more structured than ordinary life. Every participant received counseling from expert nutritionists, written support materials, and detailed menus with specified portions. The plan was personalized around an energy deficit of 600 kilocalories per day. Participants were seen in person at one, two, and three months, received telephone follow-up between visits, and could contact the nutritionists through an instant-message application. This support improves the consistency of the study experience, but it also raises a practical question: how much of the result depends on the schedule, and how much on the unusually close assistance used to sustain it?
Randomization helps balance many starting characteristics, but it cannot blind people to the food they are eating or the hours in which they are allowed to eat. Participants know which plan they received, and the trial team knows which counseling materials and replacement meals belong to each arm. Expectations can influence effort, reporting, and persistence even when the scale is measured in the same way. That does not invalidate the comparison; it defines what the comparison estimates.