Dieting cost hip bone density; exercise-driven weight loss did not

Losing weight is recommended for a long list of health reasons, but bone is one of the tissues that can quietly come along for the ride. Evidence on whether weight reduction harms bone mineral density has been inconsistent, and the answer plausibly depends on how the weight comes off. That distinction matters because eating less and exercising more place very different mechanical demands on the skeleton. This review set out to pool the randomised trials and separate the effects of calorie restriction from those of exercise training.

The authors conducted a systematic review and meta-analysis of randomised controlled trials in adults, searching MEDLINE and EMBASE comprehensively up to March 2016. Thirty-two trials met the predetermined inclusion criteria. Study-specific mean differences in bone mineral density were pooled using a random effects model, covering total body, hip and lumbar spine sites. Subgroup analysis and meta-regression were used to look for sources of disagreement between studies, including the type of intervention, its duration and the participants' starting body size.

Total body bone mineral density did not differ significantly with weight loss, with a pooled mean difference of 0.007 g/cm2. Hip density fell significantly, by about 0.008 g/cm2, and lumbar spine density fell by about 0.018 g/cm2. Hip losses appeared in interventions running longer than 4 months and were more pronounced in participants who were obese, while significant lumbar spine reductions showed up in calorie restriction lasting longer than 13 months. The split by method was the striking part: calorie restriction alone and calorie restriction combined with exercise both lowered hip bone density, whereas weight loss driven by exercise training without dietary restriction was associated with increased hip BMD.

The analysis helps resolve the inconsistency in earlier reports by showing the answer depends on site and on method: hip and spine respond, total body density does not shift measurably. It supports the long standing idea that mechanical loading protects bone, since the only route to weight loss that spared and even raised hip density involved exercise without dieting. It also flags duration as important, with the clearest losses in longer interventions. Left open are questions about which specific diets or exercise types matter, whether the changes translate into fracture risk, and how much the combination of dieting plus training can offset the loss.

The useful signal here is that the route to weight loss appears to matter for bone at the hip, not just the amount lost. Losing weight by eating less was linked to lower hip and spine density in these trials, while losing it through exercise was not. The absolute changes in density are small, and the review cannot say what they mean for any individual's fracture risk. Still, for anyone planning a long stretch of dieting, particularly beyond a year, this is a reason to be aware that bone is part of the picture and not only fat and muscle.