This poorly reported review concluded that L-carnitine and/or L-acetyl-carnitine may be effective in improving the pregnancy rate and sperm characteristics of infertile men, but further research is required to confirm these results. The conclusions appear to adequately reflect the data, but concerns about the review methodology make it difficult to assess the reliability of the findings.

To evaluate the effectiveness of carnitines for the nutrition treatment of male infertility.

MEDLINE (1950 to 2006), EMBASE (1966 to 2006), the Cochrane CENTRAL Register (1966), CBM (1978 to 2006) and CNKI (1994 to 2006) were searched for studies published in English or Chinese; the search terms were reported.

Blinded and unblinded randomised controlled trials (RCTs) were eligible for inclusion in the review. Some studies included a wash-out period and some appeared to be crossover studies.

Studies that compared L-carnitine (LC) alone, L-acetyl-carnitine (LAC) alone, LC plus LAC, or carnitines combined with other drugs, with a ‘proper’ placebo or other drug treatment were eligible for inclusion. The included studies compared carnitines with placebo, non-steroidal anti-inflammatory drugs (NSAIDs) or VE plus VC (VE and VC not defined). Some studies used cointerventions (including NSAIDs, cinnoxicam and nicetile). The duration of the interventions ranged from 2 to 6 months.

Studies of male patients (aged 18 to 65 years) with infertility (according to World Health Organization criteria) for at least 1 year were eligible. The men had to be having regular sexual intercourse with a gynaecologically normal female partner, who had no identifiable factors for female infertility. In the included studies, the participants were aged from 18 to 65 years and most had oligo-astheno-teratozoospermia.

The primary review outcome was the pregnancy rate. The secondary outcomes were semen analysis (sperm concentration and total and forward sperm motility) and sperm morphology defined using World Health Organization criteria.

One reviewer selected the studies.

Validity was assessed and scored using the Jadad scale, which considers the reporting and handling of randomisation, blinding and the handling of withdrawals. The maximum possible score was 5 points. Studies scoring at least 3 points were defined as high quality. Studies were also assessed for inclusion criteria, baseline comparability and measurement of compliance. The authors did not state how the validity assessment was performed.

The authors did not state how the data were extracted for the review, or how many reviewers performed the data extraction. Authors of studies with unclear or missing information were contacted for additional data. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for dichotomous data, while means with standard deviations were calculated for continuous outcomes.

For studies reporting sufficient data, pooled ORs and weighted mean differences (WMDs) with 95% CIs were calculated. Random-effects models were used when significant heterogeneity (p<0.05) was found. If studies had more than one treatment arm, all comparisons were included in the main meta-analyses.

Statistical heterogeneity was assessed using the chi-squared statistic. A sensitivity analysis was undertaken to examine the influence of study quality, type of carnitine and patient and intervention characteristics, and by excluding some studies. Analyses were repeated for some outcomes, including only one of three possible carnitine/control comparisons reported in a single study.

Nine RCTs (n=862) were included, one of which appeared to be a crossover trial. The sample size ranged from 21 to 325.

The Jadad scores ranged from 1 to 4, with 5 studies scoring at least 3 points. One study reported the randomisation method, seven reported baseline comparability, six reported double-blinding, seven reported drop-outs or side-effects and two reported reasons for drop-outs.

There were no significant differences between LC versus LAC, LC plus LAC versus LC alone, and LC plus LAC versus LAC alone (based on 1 study with 14 or 15 patients per subgroup). Given the absence of any significant effects, the results for all types of carnitines were pooled when assessing the remaining outcomes.

Carnitines were associated with a significant increase in pregnancy rates compared with control (OR 4.10, 95% CI: 2.08, 8.08; based on 7 studies). No statistically significant heterogeneity was detected (p=0.15). One crossover study that reported this outcome was excluded from the analysis but the reasons were not clear. Results from the sensitivity and subgroup analyses were similar.

There was no significant difference between carnitines and control for sperm concentration (WMD 5.69, 95% CI: -4.47, 15.84, p=0.27; based on 3 studies). Statistically significant heterogeneity was detected (p<0.00001). The results were similar after including only one of three possible carnitine/control comparisons from a single study.

Carnitines were associated with a significant increase in total sperm motility (WMD 7.43, 95% CI: 1.72, 13.14, p=0.01; based on 5 studies), forward sperm motility (WMD 11.83, 95% CI: 0.49, 23.16; p=0.04; based on 4 studies) and atypical sperm forms (WMD -5.72, 95% CI: -7.89, -3.56; p<0.00001; based on 2 studies) compared with control. The results for total sperm motility were similar after including only one of three possible carnitine/control comparisons from a single study. The results for forward sperm motility were similar after excluding one low-quality study.

LC and/or LAC may be effective in improving the pregnancy rate and sperm motility in infertile men, but further research is required to confirm these results.

The language in which this review is written is not always clear and some parts are therefore difficult to read and interpret. The review addressed a clear question, using several relevant databases, but restrictions applied to the search strategy might have introduced language and publication bias. The review methods were poorly reported but, where stated, failed to take appropriate precautions to reduce the risk of reviewer error and bias. The validity of the studies and the extent of statistical heterogeneity were assessed, and attempts were made to investigate potential sources of heterogeneity. The authors also examined the influence of including three comparisons with a common control group from one study. However, the assumption of comparable effectiveness of different types or combinations of carnitines was based on only one small study, and thus may not be reliable. Overall, the authors’ conclusions appear to adequately reflect the limitations of the evidence from the small number of included studies, but poor reporting and concerns about the review methodology make it difficult to assess the reliability of the findings.

Practice: The authors did not state any implications for practice.

Research: The authors stated the need for further good-quality large RCTs to compare the effectiveness of different types and combinations of carnitines, and to evaluate the effect of carnitines on sperm motility and morphology. There is also a need to examine the effect of carnitines on the metabolism of the male gamete.

Not stated.

This record is a structured abstract written by CRD reviewers. The original has met a set of quality criteria. Since September 1996 abstracts have been sent to authors for comment. Additional factual information is incorporated into the record. Noted as  [A:....].